From 5dc4fe1529bb9e1134245db505883364caac3280 Mon Sep 17 00:00:00 2001
From: Itamar Haber <itamar@redislabs.com>
Date: Fri, 2 Dec 2016 18:19:36 +0200
Subject: Verify pairs are provided after subcommands

Fixes https://github.com/antirez/redis/issues/3639
---
 src/t_zset.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/t_zset.c b/src/t_zset.c
index 8d905be02..d36fa30ae 100644
--- a/src/t_zset.c
+++ b/src/t_zset.c
@@ -1521,7 +1521,7 @@ void zaddGenericCommand(client *c, int flags) {
     /* After the options, we expect to have an even number of args, since
      * we expect any number of score-element pairs. */
     elements = c->argc-scoreidx;
-    if (elements % 2) {
+    if (elements % 2 || !elements) {
         addReply(c,shared.syntaxerr);
         return;
     }
-- 
cgit v1.2.1


From b2da5ea773356320af3cda6cfd0f1fc1e7d48ba7 Mon Sep 17 00:00:00 2001
From: oranagra <oran@redislabs.com>
Date: Sat, 24 Dec 2016 17:27:58 +0200
Subject: fix rare assertion in DEBUG DIGEST

getExpire calls dictFind which can do rehashing.
found by calling computeDatasetDigest from serverCron and running the test suite.
---
 src/debug.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/debug.c b/src/debug.c
index b8ad4e511..2fd382fc0 100644
--- a/src/debug.c
+++ b/src/debug.c
@@ -126,7 +126,7 @@ void computeDatasetDigest(unsigned char *final) {
         redisDb *db = server.db+j;
 
         if (dictSize(db->dict) == 0) continue;
-        di = dictGetIterator(db->dict);
+        di = dictGetSafeIterator(db->dict);
 
         /* hash the DB id, so the same dataset moved in a different
          * DB will lead to a different digest */
-- 
cgit v1.2.1


From afaaa9188537b7651b7e42dfe3e9f340af819a0f Mon Sep 17 00:00:00 2001
From: Jan-Erik Rediger <janerik@fnordig.de>
Date: Sun, 8 Jan 2017 18:27:00 +0100
Subject: Initialize help only in repl mode

---
 src/redis-cli.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/src/redis-cli.c b/src/redis-cli.c
index ac4358220..2a9dff712 100644
--- a/src/redis-cli.c
+++ b/src/redis-cli.c
@@ -1273,6 +1273,11 @@ static void repl(void) {
     int argc;
     sds *argv;
 
+    /* Initialize the help and, if possible, use the COMMAND command in order
+     * to retrieve missing entries. */
+    cliInitHelp();
+    cliIntegrateHelp();
+
     config.interactive = 1;
     linenoiseSetMultiLine(1);
     linenoiseSetCompletionCallback(completionCallback);
@@ -2606,11 +2611,6 @@ int main(int argc, char **argv) {
     argc -= firstarg;
     argv += firstarg;
 
-    /* Initialize the help and, if possible, use the COMMAND command in order
-     * to retrieve missing entries. */
-    cliInitHelp();
-    cliIntegrateHelp();
-
     /* Latency mode */
     if (config.latency_mode) {
         if (cliConnect(0) == REDIS_ERR) exit(1);
-- 
cgit v1.2.1


From baa98988215f005b5b57e156c68a5bdc8454a3a5 Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Thu, 12 Jan 2017 09:08:43 +0100
Subject: MEMORY USAGE: support for modules data types.

As a side effect of supporting it, we no longer crash when MEMORY USAGE
is called against a module data type.

Close #3637.
---
 src/module.c            |  2 +-
 src/modules/hellotype.c | 11 +++++++----
 src/object.c            |  8 ++++++++
 3 files changed, 16 insertions(+), 5 deletions(-)

diff --git a/src/module.c b/src/module.c
index a5b3d52ae..1fbc5094f 100644
--- a/src/module.c
+++ b/src/module.c
@@ -2743,8 +2743,8 @@ moduleType *RM_CreateDataType(RedisModuleCtx *ctx, const char *name, int encver,
         moduleTypeLoadFunc rdb_load;
         moduleTypeSaveFunc rdb_save;
         moduleTypeRewriteFunc aof_rewrite;
-        moduleTypeDigestFunc digest;
         moduleTypeMemUsageFunc mem_usage;
+        moduleTypeDigestFunc digest;
         moduleTypeFreeFunc free;
     } *tms = (struct typemethods*) typemethods_ptr;
 
diff --git a/src/modules/hellotype.c b/src/modules/hellotype.c
index 02a5bb477..1aa6c5e63 100644
--- a/src/modules/hellotype.c
+++ b/src/modules/hellotype.c
@@ -226,10 +226,12 @@ void HelloTypeAofRewrite(RedisModuleIO *aof, RedisModuleString *key, void *value
     }
 }
 
-void HelloTypeDigest(RedisModuleDigest *digest, void *value) {
-    REDISMODULE_NOT_USED(digest);
-    REDISMODULE_NOT_USED(value);
-    /* TODO: The DIGEST module interface is yet not implemented. */
+/* The goal of this function is to return the amount of memory used by
+ * the HelloType value. */
+size_t HelloTypeMemUsage(void *value) {
+    struct HelloTypeObject *hto = value;
+    struct HelloTypeNode *node = hto->head;
+    return sizeof(*hto) + sizeof(*node)*hto->len;
 }
 
 void HelloTypeFree(void *value) {
@@ -250,6 +252,7 @@ int RedisModule_OnLoad(RedisModuleCtx *ctx, RedisModuleString **argv, int argc)
         .rdb_load = HelloTypeRdbLoad,
         .rdb_save = HelloTypeRdbSave,
         .aof_rewrite = HelloTypeAofRewrite,
+        .mem_usage = HelloTypeMemUsage,
         .free = HelloTypeFree
     };
 
diff --git a/src/object.c b/src/object.c
index 1ae37c9d1..125c1a595 100644
--- a/src/object.c
+++ b/src/object.c
@@ -786,6 +786,14 @@ size_t objectComputeSize(robj *o, size_t sample_size) {
         } else {
             serverPanic("Unknown hash encoding");
         }
+    } else if (o->type == OBJ_MODULE) {
+        moduleValue *mv = o->ptr;
+        moduleType *mt = mv->type;
+        if (mt->mem_usage != NULL) {
+            asize = mt->mem_usage(mv->value);
+        } else {
+            asize = 0;
+        }
     } else {
         serverPanic("Unknown object type");
     }
-- 
cgit v1.2.1


From 3f79b2f8831bc0fee3c037862aad5a3e9d70cf4b Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Thu, 12 Jan 2017 09:50:40 +0100
Subject: Defrag: don't crash when a module value is encountered.

---
 src/defrag.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/src/defrag.c b/src/defrag.c
index cc7dfdf7f..4a1dcefe4 100644
--- a/src/defrag.c
+++ b/src/defrag.c
@@ -406,6 +406,9 @@ int defragKey(redisDb *db, dictEntry *de) {
         } else {
             serverPanic("Unknown hash encoding");
         }
+    } else if (ob->type == OBJ_MODULE) {
+        /* Currently defragmenting modules private data types
+         * is not supported. */
     } else {
         serverPanic("Unknown object type");
     }
-- 
cgit v1.2.1


From 636c693f447f4c242a1fff5ad987bc6102b3e243 Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Thu, 12 Jan 2017 12:47:46 +0100
Subject: Use const in modules types mem_usage method.

As suggested by @itamarhaber.
---
 src/modules/hellotype.c | 4 ++--
 src/redismodule.h       | 2 +-
 src/server.h            | 2 +-
 3 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/src/modules/hellotype.c b/src/modules/hellotype.c
index 1aa6c5e63..027155d45 100644
--- a/src/modules/hellotype.c
+++ b/src/modules/hellotype.c
@@ -228,8 +228,8 @@ void HelloTypeAofRewrite(RedisModuleIO *aof, RedisModuleString *key, void *value
 
 /* The goal of this function is to return the amount of memory used by
  * the HelloType value. */
-size_t HelloTypeMemUsage(void *value) {
-    struct HelloTypeObject *hto = value;
+size_t HelloTypeMemUsage(const void *value) {
+    const struct HelloTypeObject *hto = value;
     struct HelloTypeNode *node = hto->head;
     return sizeof(*hto) + sizeof(*node)*hto->len;
 }
diff --git a/src/redismodule.h b/src/redismodule.h
index 186e284c0..fcd6aeca3 100644
--- a/src/redismodule.h
+++ b/src/redismodule.h
@@ -91,7 +91,7 @@ typedef int (*RedisModuleCmdFunc) (RedisModuleCtx *ctx, RedisModuleString **argv
 typedef void *(*RedisModuleTypeLoadFunc)(RedisModuleIO *rdb, int encver);
 typedef void (*RedisModuleTypeSaveFunc)(RedisModuleIO *rdb, void *value);
 typedef void (*RedisModuleTypeRewriteFunc)(RedisModuleIO *aof, RedisModuleString *key, void *value);
-typedef size_t (*RedisModuleTypeMemUsageFunc)(void *value);
+typedef size_t (*RedisModuleTypeMemUsageFunc)(const void *value);
 typedef void (*RedisModuleTypeDigestFunc)(RedisModuleDigest *digest, void *value);
 typedef void (*RedisModuleTypeFreeFunc)(void *value);
 
diff --git a/src/server.h b/src/server.h
index bbc121b72..96bbb9ccf 100644
--- a/src/server.h
+++ b/src/server.h
@@ -483,7 +483,7 @@ typedef void *(*moduleTypeLoadFunc)(struct RedisModuleIO *io, int encver);
 typedef void (*moduleTypeSaveFunc)(struct RedisModuleIO *io, void *value);
 typedef void (*moduleTypeRewriteFunc)(struct RedisModuleIO *io, struct redisObject *key, void *value);
 typedef void (*moduleTypeDigestFunc)(struct RedisModuleDigest *digest, void *value);
-typedef size_t (*moduleTypeMemUsageFunc)(void *value);
+typedef size_t (*moduleTypeMemUsageFunc)(const void *value);
 typedef void (*moduleTypeFreeFunc)(void *value);
 
 /* The module type, which is referenced in each value of a given type, defines
-- 
cgit v1.2.1


From 2cd1ae736f7ace5d0620937e4783fd85e9ca7448 Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Fri, 13 Jan 2017 11:47:09 +0100
Subject: Ziplist: remove static from functions, they prevent good crash
 reports.

---
 src/ziplist.c | 28 ++++++++++++++--------------
 1 file changed, 14 insertions(+), 14 deletions(-)

diff --git a/src/ziplist.c b/src/ziplist.c
index a0939f640..f270cdbf9 100644
--- a/src/ziplist.c
+++ b/src/ziplist.c
@@ -187,7 +187,7 @@ typedef struct zlentry {
 void ziplistRepr(unsigned char *zl);
 
 /* Return bytes needed to store integer encoded by 'encoding' */
-static unsigned int zipIntSize(unsigned char encoding) {
+unsigned int zipIntSize(unsigned char encoding) {
     switch(encoding) {
     case ZIP_INT_8B:  return 1;
     case ZIP_INT_16B: return 2;
@@ -202,7 +202,7 @@ static unsigned int zipIntSize(unsigned char encoding) {
 
 /* Encode the length 'rawlen' writing it in 'p'. If p is NULL it just returns
  * the amount of bytes required to encode such a length. */
-static unsigned int zipEncodeLength(unsigned char *p, unsigned char encoding, unsigned int rawlen) {
+unsigned int zipEncodeLength(unsigned char *p, unsigned char encoding, unsigned int rawlen) {
     unsigned char len = 1, buf[5];
 
     if (ZIP_IS_STR(encoding)) {
@@ -266,7 +266,7 @@ static unsigned int zipEncodeLength(unsigned char *p, unsigned char encoding, un
 
 /* Encode the length of the previous entry and write it to "p". Return the
  * number of bytes needed to encode this length if "p" is NULL. */
-static unsigned int zipPrevEncodeLength(unsigned char *p, unsigned int len) {
+unsigned int zipPrevEncodeLength(unsigned char *p, unsigned int len) {
     if (p == NULL) {
         return (len < ZIP_BIGLEN) ? 1 : sizeof(len)+1;
     } else {
@@ -284,7 +284,7 @@ static unsigned int zipPrevEncodeLength(unsigned char *p, unsigned int len) {
 
 /* Encode the length of the previous entry and write it to "p". This only
  * uses the larger encoding (required in __ziplistCascadeUpdate). */
-static void zipPrevEncodeLengthForceLarge(unsigned char *p, unsigned int len) {
+void zipPrevEncodeLengthForceLarge(unsigned char *p, unsigned int len) {
     if (p == NULL) return;
     p[0] = ZIP_BIGLEN;
     memcpy(p+1,&len,sizeof(len));
@@ -316,14 +316,14 @@ static void zipPrevEncodeLengthForceLarge(unsigned char *p, unsigned int len) {
 
 /* Return the difference in number of bytes needed to store the length of the
  * previous element 'len', in the entry pointed to by 'p'. */
-static int zipPrevLenByteDiff(unsigned char *p, unsigned int len) {
+int zipPrevLenByteDiff(unsigned char *p, unsigned int len) {
     unsigned int prevlensize;
     ZIP_DECODE_PREVLENSIZE(p, prevlensize);
     return zipPrevEncodeLength(NULL, len) - prevlensize;
 }
 
 /* Return the total number of bytes used by the entry pointed to by 'p'. */
-static unsigned int zipRawEntryLength(unsigned char *p) {
+unsigned int zipRawEntryLength(unsigned char *p) {
     unsigned int prevlensize, encoding, lensize, len;
     ZIP_DECODE_PREVLENSIZE(p, prevlensize);
     ZIP_DECODE_LENGTH(p + prevlensize, encoding, lensize, len);
@@ -332,7 +332,7 @@ static unsigned int zipRawEntryLength(unsigned char *p) {
 
 /* Check if string pointed to by 'entry' can be encoded as an integer.
  * Stores the integer value in 'v' and its encoding in 'encoding'. */
-static int zipTryEncoding(unsigned char *entry, unsigned int entrylen, long long *v, unsigned char *encoding) {
+int zipTryEncoding(unsigned char *entry, unsigned int entrylen, long long *v, unsigned char *encoding) {
     long long value;
 
     if (entrylen >= 32 || entrylen == 0) return 0;
@@ -359,7 +359,7 @@ static int zipTryEncoding(unsigned char *entry, unsigned int entrylen, long long
 }
 
 /* Store integer 'value' at 'p', encoded as 'encoding' */
-static void zipSaveInteger(unsigned char *p, int64_t value, unsigned char encoding) {
+void zipSaveInteger(unsigned char *p, int64_t value, unsigned char encoding) {
     int16_t i16;
     int32_t i32;
     int64_t i64;
@@ -389,7 +389,7 @@ static void zipSaveInteger(unsigned char *p, int64_t value, unsigned char encodi
 }
 
 /* Read integer encoded as 'encoding' from 'p' */
-static int64_t zipLoadInteger(unsigned char *p, unsigned char encoding) {
+int64_t zipLoadInteger(unsigned char *p, unsigned char encoding) {
     int16_t i16;
     int32_t i32;
     int64_t i64, ret = 0;
@@ -421,7 +421,7 @@ static int64_t zipLoadInteger(unsigned char *p, unsigned char encoding) {
 }
 
 /* Return a struct with all information about an entry. */
-static void zipEntry(unsigned char *p, zlentry *e) {
+void zipEntry(unsigned char *p, zlentry *e) {
 
     ZIP_DECODE_PREVLEN(p, e->prevrawlensize, e->prevrawlen);
     ZIP_DECODE_LENGTH(p + e->prevrawlensize, e->encoding, e->lensize, e->len);
@@ -441,7 +441,7 @@ unsigned char *ziplistNew(void) {
 }
 
 /* Resize the ziplist. */
-static unsigned char *ziplistResize(unsigned char *zl, unsigned int len) {
+unsigned char *ziplistResize(unsigned char *zl, unsigned int len) {
     zl = zrealloc(zl,len);
     ZIPLIST_BYTES(zl) = intrev32ifbe(len);
     zl[len-1] = ZIP_END;
@@ -468,7 +468,7 @@ static unsigned char *ziplistResize(unsigned char *zl, unsigned int len) {
  *
  * The pointer "p" points to the first entry that does NOT need to be
  * updated, i.e. consecutive fields MAY need an update. */
-static unsigned char *__ziplistCascadeUpdate(unsigned char *zl, unsigned char *p) {
+unsigned char *__ziplistCascadeUpdate(unsigned char *zl, unsigned char *p) {
     size_t curlen = intrev32ifbe(ZIPLIST_BYTES(zl)), rawlen, rawlensize;
     size_t offset, noffset, extra;
     unsigned char *np;
@@ -530,7 +530,7 @@ static unsigned char *__ziplistCascadeUpdate(unsigned char *zl, unsigned char *p
 }
 
 /* Delete "num" entries, starting at "p". Returns pointer to the ziplist. */
-static unsigned char *__ziplistDelete(unsigned char *zl, unsigned char *p, unsigned int num) {
+unsigned char *__ziplistDelete(unsigned char *zl, unsigned char *p, unsigned int num) {
     unsigned int i, totlen, deleted = 0;
     size_t offset;
     int nextdiff = 0;
@@ -590,7 +590,7 @@ static unsigned char *__ziplistDelete(unsigned char *zl, unsigned char *p, unsig
 }
 
 /* Insert item at "p". */
-static unsigned char *__ziplistInsert(unsigned char *zl, unsigned char *p, unsigned char *s, unsigned int slen) {
+unsigned char *__ziplistInsert(unsigned char *zl, unsigned char *p, unsigned char *s, unsigned int slen) {
     size_t curlen = intrev32ifbe(ZIPLIST_BYTES(zl)), reqlen;
     unsigned int prevlensize, prevlen = 0;
     size_t offset;
-- 
cgit v1.2.1


From 53b8bf2c89ac9607a8bede2868d60cf8b3cbddd6 Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Wed, 18 Jan 2017 17:05:10 +0100
Subject: serverPanic(): allow printf() alike formatting.

This is of great interest because allows us to print debugging
informations that could be of useful when debugging, like in the
following example:

    serverPanic("Unexpected encoding for object %d, %d",
        obj->type, obj->encoding);
---
 src/debug.c  | 14 ++++++++++++--
 src/server.h |  4 ++--
 2 files changed, 14 insertions(+), 4 deletions(-)

diff --git a/src/debug.c b/src/debug.c
index 0026594d3..a3190f636 100644
--- a/src/debug.c
+++ b/src/debug.c
@@ -266,6 +266,8 @@ void debugCommand(client *c) {
         blen++; addReplyStatus(c,
         "segfault -- Crash the server with sigsegv.");
         blen++; addReplyStatus(c,
+        "panic -- Crash the server simulating a panic.");
+        blen++; addReplyStatus(c,
         "restart  -- Graceful restart: save config, db, restart.");
         blen++; addReplyStatus(c,
         "crash-and-recovery <milliseconds> -- Hard crash and restart after <milliseconds> delay.");
@@ -300,6 +302,8 @@ void debugCommand(client *c) {
         setDeferredMultiBulkLength(c,blenp,blen);
     } else if (!strcasecmp(c->argv[1]->ptr,"segfault")) {
         *((char*)-1) = 'x';
+    } else if (!strcasecmp(c->argv[1]->ptr,"panic")) {
+        serverPanic("DEBUG PANIC called at Unix time %ld", time(NULL));
     } else if (!strcasecmp(c->argv[1]->ptr,"restart") ||
                !strcasecmp(c->argv[1]->ptr,"crash-and-recover"))
     {
@@ -615,11 +619,17 @@ void _serverAssertWithInfo(const client *c, const robj *o, const char *estr, con
     _serverAssert(estr,file,line);
 }
 
-void _serverPanic(const char *msg, const char *file, int line) {
+void _serverPanic(const char *file, int line, const char *msg, ...) {
+    va_list ap;
+    va_start(ap,msg);
+    char fmtmsg[256];
+    vsnprintf(fmtmsg,sizeof(fmtmsg),msg,ap);
+    va_end(ap);
+
     bugReportStart();
     serverLog(LL_WARNING,"------------------------------------------------");
     serverLog(LL_WARNING,"!!! Software Failure. Press left mouse button to continue");
-    serverLog(LL_WARNING,"Guru Meditation: %s #%s:%d",msg,file,line);
+    serverLog(LL_WARNING,"Guru Meditation: %s #%s:%d",fmtmsg,file,line);
 #ifdef HAVE_BACKTRACE
     serverLog(LL_WARNING,"(forcing SIGSEGV in order to print the stack trace)");
 #endif
diff --git a/src/server.h b/src/server.h
index 96bbb9ccf..30d8be849 100644
--- a/src/server.h
+++ b/src/server.h
@@ -435,7 +435,7 @@ typedef long long mstime_t; /* millisecond time type. */
 /* We can print the stacktrace, so our assert is defined this way: */
 #define serverAssertWithInfo(_c,_o,_e) ((_e)?(void)0 : (_serverAssertWithInfo(_c,_o,#_e,__FILE__,__LINE__),_exit(1)))
 #define serverAssert(_e) ((_e)?(void)0 : (_serverAssert(#_e,__FILE__,__LINE__),_exit(1)))
-#define serverPanic(_e) _serverPanic(#_e,__FILE__,__LINE__),_exit(1)
+#define serverPanic(...) _serverPanic(__FILE__,__LINE__,__VA_ARGS__),_exit(1)
 
 /*-----------------------------------------------------------------------------
  * Data types
@@ -1960,7 +1960,7 @@ void *realloc(void *ptr, size_t size) __attribute__ ((deprecated));
 /* Debugging stuff */
 void _serverAssertWithInfo(const client *c, const robj *o, const char *estr, const char *file, int line);
 void _serverAssert(const char *estr, const char *file, int line);
-void _serverPanic(const char *msg, const char *file, int line);
+void _serverPanic(const char *file, int line, const char *msg, ...);
 void bugReportStart(void);
 void serverLogObjectDebugInfo(const robj *o);
 void sigsegvHandler(int sig, siginfo_t *info, void *secret);
-- 
cgit v1.2.1


From 17ac46ea78fc3bed81b764d096ec8fb38e3ae3c9 Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Wed, 18 Jan 2017 17:11:57 +0100
Subject: Add panic() into redisassert.h.

This header file is for libs, like ziplist.c, that we want to leave
almost separted from the core. The panic() calls will be easy to delete
in order to use such files outside, but the debugging info we gain are
very valuable compared to simple assertions where it is not possible to
print debugging info.
---
 src/redisassert.h | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/src/redisassert.h b/src/redisassert.h
index 34e95268a..c9b78327c 100644
--- a/src/redisassert.h
+++ b/src/redisassert.h
@@ -41,7 +41,9 @@
 #include <unistd.h> /* for _exit() */
 
 #define assert(_e) ((_e)?(void)0 : (_serverAssert(#_e,__FILE__,__LINE__),_exit(1)))
+#define panic(...) _serverPanic(__FILE__,__LINE__,__VA_ARGS__),_exit(1)
 
 void _serverAssert(char *estr, char *file, int line);
+void _serverPanic(const char *file, int line, const char *msg, ...);
 
 #endif
-- 
cgit v1.2.1


From 3ec1a001fbc11b54c0231f9c58384eb53ca3802e Mon Sep 17 00:00:00 2001
From: miter <Miterv@163.com>
Date: Thu, 26 Jan 2017 21:36:26 +0900
Subject: Change switch statment to if statment

---
 src/object.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/src/object.c b/src/object.c
index 125c1a595..90df9e38b 100644
--- a/src/object.c
+++ b/src/object.c
@@ -246,11 +246,9 @@ void freeStringObject(robj *o) {
 }
 
 void freeListObject(robj *o) {
-    switch (o->encoding) {
-    case OBJ_ENCODING_QUICKLIST:
+    if (o->encoding == OBJ_ENCODING_QUICKLIST) {
         quicklistRelease(o->ptr);
-        break;
-    default:
+    } else {
         serverPanic("Unknown list encoding type");
     }
 }
-- 
cgit v1.2.1


From 3c9b817217b03d0377bed3857f0159a473711490 Mon Sep 17 00:00:00 2001
From: Jan-Erik Rediger <janerik@fnordig.de>
Date: Fri, 27 Jan 2017 16:23:15 +0100
Subject: Don't divide by zero

Previously Redis crashed on `MEMORY DOCTOR` when it has no slaves attached.

Fixes #3783
---
 src/object.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/object.c b/src/object.c
index 90df9e38b..08c9ad956 100644
--- a/src/object.c
+++ b/src/object.c
@@ -951,7 +951,7 @@ sds getMemoryDoctorReport(void) {
         }
 
         /* Slaves using more than 10 MB each? */
-        if (mh->clients_slaves / numslaves > (1024*1024*10)) {
+        if (numslaves > 0 && mh->clients_slaves / numslaves > (1024*1024*10)) {
             big_slave_buf = 1;
             num_reports++;
         }
-- 
cgit v1.2.1


From 27e29f4fe61d822eb23d948bcb72db76c4c887e5 Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Mon, 30 Jan 2017 09:58:34 +0100
Subject: Jemalloc updated to 4.4.0.

The original jemalloc source tree was modified to:

1. Remove the configure error that prevents nested builds.
2. Insert the Redis private Jemalloc API in order to allow the
Redis fragmentation function to work.
---
 deps/jemalloc/.appveyor.yml                        |   28 +
 deps/jemalloc/.gitignore                           |   16 +
 deps/jemalloc/.travis.yml                          |   29 +
 deps/jemalloc/COPYING                              |    4 +-
 deps/jemalloc/ChangeLog                            |  220 +++
 deps/jemalloc/INSTALL                              |   26 +
 deps/jemalloc/Makefile.in                          |  128 +-
 deps/jemalloc/README                               |    2 +-
 deps/jemalloc/VERSION                              |    2 +-
 deps/jemalloc/bin/jeprof.in                        |  131 +-
 deps/jemalloc/build-aux/config.guess               | 1462 ++++++++++++++
 deps/jemalloc/build-aux/config.sub                 | 1825 +++++++++++++++++
 deps/jemalloc/build-aux/install-sh                 |  250 +++
 deps/jemalloc/config.guess                         | 1420 --------------
 deps/jemalloc/config.sub                           | 1797 -----------------
 deps/jemalloc/configure                            | 1567 +++++++++++++--
 deps/jemalloc/configure.ac                         |  436 ++++-
 deps/jemalloc/doc/html.xsl.in                      |    1 +
 deps/jemalloc/doc/jemalloc.3                       |  859 ++++----
 deps/jemalloc/doc/jemalloc.html                    | 1507 ++++++---------
 deps/jemalloc/doc/jemalloc.xml.in                  |  544 ++++--
 deps/jemalloc/doc/stylesheet.xsl                   |    7 +-
 deps/jemalloc/include/jemalloc/internal/arena.h    |  582 ++++--
 deps/jemalloc/include/jemalloc/internal/assert.h   |   45 +
 deps/jemalloc/include/jemalloc/internal/atomic.h   |    4 +-
 deps/jemalloc/include/jemalloc/internal/base.h     |   11 +-
 deps/jemalloc/include/jemalloc/internal/bitmap.h   |   76 +-
 deps/jemalloc/include/jemalloc/internal/chunk.h    |   38 +-
 .../jemalloc/include/jemalloc/internal/chunk_dss.h |   10 +-
 .../include/jemalloc/internal/chunk_mmap.h         |    4 +-
 deps/jemalloc/include/jemalloc/internal/ckh.h      |    6 +-
 deps/jemalloc/include/jemalloc/internal/ctl.h      |   29 +-
 deps/jemalloc/include/jemalloc/internal/extent.h   |   43 +-
 deps/jemalloc/include/jemalloc/internal/hash.h     |   33 +-
 deps/jemalloc/include/jemalloc/internal/huge.h     |   21 +-
 .../jemalloc/internal/jemalloc_internal.h.in       |  440 +++--
 .../jemalloc/internal/jemalloc_internal_decls.h    |   11 +
 .../jemalloc/internal/jemalloc_internal_defs.h.in  |   73 +-
 deps/jemalloc/include/jemalloc/internal/mb.h       |   10 +-
 deps/jemalloc/include/jemalloc/internal/mutex.h    |   62 +-
 deps/jemalloc/include/jemalloc/internal/nstime.h   |   48 +
 deps/jemalloc/include/jemalloc/internal/pages.h    |    7 +-
 deps/jemalloc/include/jemalloc/internal/ph.h       |  345 ++++
 .../include/jemalloc/internal/private_symbols.txt  |  252 ++-
 deps/jemalloc/include/jemalloc/internal/prng.h     |  195 +-
 deps/jemalloc/include/jemalloc/internal/prof.h     |   86 +-
 deps/jemalloc/include/jemalloc/internal/rb.h       |  208 +-
 deps/jemalloc/include/jemalloc/internal/rtree.h    |  160 +-
 .../include/jemalloc/internal/size_classes.sh      |   50 +-
 .../include/jemalloc/internal/smoothstep.h         |  246 +++
 .../include/jemalloc/internal/smoothstep.sh        |  115 ++
 deps/jemalloc/include/jemalloc/internal/spin.h     |   51 +
 deps/jemalloc/include/jemalloc/internal/stats.h    |   14 +
 deps/jemalloc/include/jemalloc/internal/tcache.h   |  159 +-
 deps/jemalloc/include/jemalloc/internal/ticker.h   |   75 +
 deps/jemalloc/include/jemalloc/internal/tsd.h      |  164 +-
 deps/jemalloc/include/jemalloc/internal/util.h     |  214 +-
 deps/jemalloc/include/jemalloc/internal/valgrind.h |   40 +-
 deps/jemalloc/include/jemalloc/internal/witness.h  |  266 +++
 deps/jemalloc/include/jemalloc/jemalloc_defs.h.in  |    8 +
 .../jemalloc/include/jemalloc/jemalloc_macros.h.in |   61 +-
 deps/jemalloc/include/msvc_compat/strings.h        |   30 +
 deps/jemalloc/include/msvc_compat/windows_extra.h  |   22 +-
 deps/jemalloc/install-sh                           |  250 ---
 deps/jemalloc/jemalloc.pc.in                       |    2 +-
 deps/jemalloc/msvc/ReadMe.txt                      |   24 +
 deps/jemalloc/msvc/jemalloc_vc2015.sln             |   63 +
 .../msvc/projects/vc2015/jemalloc/jemalloc.vcxproj |  402 ++++
 .../vc2015/jemalloc/jemalloc.vcxproj.filters       |  272 +++
 .../projects/vc2015/test_threads/test_threads.cpp  |   89 +
 .../projects/vc2015/test_threads/test_threads.h    |    3 +
 .../vc2015/test_threads/test_threads.vcxproj       |  327 ++++
 .../test_threads/test_threads.vcxproj.filters      |   26 +
 .../vc2015/test_threads/test_threads_main.cpp      |   12 +
 deps/jemalloc/src/arena.c                          | 2041 +++++++++++++-------
 deps/jemalloc/src/base.c                           |   73 +-
 deps/jemalloc/src/bitmap.c                         |   59 +-
 deps/jemalloc/src/chunk.c                          |  426 ++--
 deps/jemalloc/src/chunk_dss.c                      |  182 +-
 deps/jemalloc/src/chunk_mmap.c                     |   18 +-
 deps/jemalloc/src/ckh.c                            |   43 +-
 deps/jemalloc/src/ctl.c                            |  761 +++++---
 deps/jemalloc/src/extent.c                         |   70 +-
 deps/jemalloc/src/huge.c                           |  238 ++-
 deps/jemalloc/src/jemalloc.c                       | 1492 ++++++++------
 deps/jemalloc/src/mutex.c                          |   23 +-
 deps/jemalloc/src/nstime.c                         |  194 ++
 deps/jemalloc/src/pages.c                          |  177 +-
 deps/jemalloc/src/prng.c                           |    2 +
 deps/jemalloc/src/prof.c                           |  664 ++++---
 deps/jemalloc/src/quarantine.c                     |   50 +-
 deps/jemalloc/src/rtree.c                          |    9 +-
 deps/jemalloc/src/spin.c                           |    2 +
 deps/jemalloc/src/stats.c                          | 1220 ++++++++----
 deps/jemalloc/src/tcache.c                         |  170 +-
 deps/jemalloc/src/ticker.c                         |    2 +
 deps/jemalloc/src/tsd.c                            |   28 +-
 deps/jemalloc/src/util.c                           |   42 +-
 deps/jemalloc/src/witness.c                        |  136 ++
 deps/jemalloc/src/zone.c                           |  198 +-
 deps/jemalloc/test/include/test/jemalloc_test.h.in |   80 +-
 deps/jemalloc/test/include/test/mtx.h              |    2 +
 deps/jemalloc/test/include/test/test.h             |    4 +
 deps/jemalloc/test/include/test/timer.h            |   19 +-
 deps/jemalloc/test/integration/MALLOCX_ARENA.c     |    4 +-
 deps/jemalloc/test/integration/aligned_alloc.c     |   20 +-
 deps/jemalloc/test/integration/allocated.c         |   17 +-
 deps/jemalloc/test/integration/chunk.c             |   98 +-
 deps/jemalloc/test/integration/mallocx.c           |  106 +-
 deps/jemalloc/test/integration/overflow.c          |    8 +-
 deps/jemalloc/test/integration/posix_memalign.c    |   20 +-
 deps/jemalloc/test/integration/rallocx.c           |   86 +-
 deps/jemalloc/test/integration/sdallocx.c          |    4 +-
 deps/jemalloc/test/integration/thread_arena.c      |   10 +-
 .../test/integration/thread_tcache_enabled.c       |   39 +-
 deps/jemalloc/test/integration/xallocx.c           |  120 +-
 deps/jemalloc/test/src/mtx.c                       |    7 +
 deps/jemalloc/test/src/test.c                      |   56 +-
 deps/jemalloc/test/src/timer.c                     |   45 +-
 deps/jemalloc/test/stress/microbench.c             |    3 +-
 deps/jemalloc/test/unit/a0.c                       |   19 +
 deps/jemalloc/test/unit/arena_reset.c              |  159 ++
 deps/jemalloc/test/unit/bitmap.c                   |   26 +-
 deps/jemalloc/test/unit/ckh.c                      |    8 +-
 deps/jemalloc/test/unit/decay.c                    |  374 ++++
 deps/jemalloc/test/unit/fork.c                     |   64 +
 deps/jemalloc/test/unit/hash.c                     |   36 +-
 deps/jemalloc/test/unit/junk.c                     |   17 +-
 deps/jemalloc/test/unit/junk_alloc.c               |    2 +-
 deps/jemalloc/test/unit/junk_free.c                |    2 +-
 deps/jemalloc/test/unit/mallctl.c                  |  319 ++-
 deps/jemalloc/test/unit/math.c                     |    4 +
 deps/jemalloc/test/unit/nstime.c                   |  227 +++
 deps/jemalloc/test/unit/pack.c                     |  206 ++
 deps/jemalloc/test/unit/pages.c                    |   27 +
 deps/jemalloc/test/unit/ph.c                       |  290 +++
 deps/jemalloc/test/unit/prng.c                     |  263 +++
 deps/jemalloc/test/unit/prof_accum.c               |    5 +-
 deps/jemalloc/test/unit/prof_active.c              |    5 +-
 deps/jemalloc/test/unit/prof_gdump.c               |   13 +-
 deps/jemalloc/test/unit/prof_idump.c               |    5 +-
 deps/jemalloc/test/unit/prof_reset.c               |   16 +-
 deps/jemalloc/test/unit/prof_thread_name.c         |   22 +-
 deps/jemalloc/test/unit/rb.c                       |   60 +-
 deps/jemalloc/test/unit/run_quantize.c             |  149 ++
 deps/jemalloc/test/unit/size_classes.c             |  105 +-
 deps/jemalloc/test/unit/smoothstep.c               |  106 +
 deps/jemalloc/test/unit/stats.c                    |  241 +--
 deps/jemalloc/test/unit/ticker.c                   |   76 +
 deps/jemalloc/test/unit/tsd.c                      |   13 +-
 deps/jemalloc/test/unit/util.c                     |   89 +-
 deps/jemalloc/test/unit/witness.c                  |  278 +++
 deps/jemalloc/test/unit/zero.c                     |   16 +-
 153 files changed, 20619 insertions(+), 9736 deletions(-)
 create mode 100644 deps/jemalloc/.appveyor.yml
 create mode 100644 deps/jemalloc/.travis.yml
 create mode 100755 deps/jemalloc/build-aux/config.guess
 create mode 100755 deps/jemalloc/build-aux/config.sub
 create mode 100755 deps/jemalloc/build-aux/install-sh
 delete mode 100755 deps/jemalloc/config.guess
 delete mode 100755 deps/jemalloc/config.sub
 create mode 100644 deps/jemalloc/include/jemalloc/internal/assert.h
 create mode 100644 deps/jemalloc/include/jemalloc/internal/nstime.h
 create mode 100644 deps/jemalloc/include/jemalloc/internal/ph.h
 create mode 100644 deps/jemalloc/include/jemalloc/internal/smoothstep.h
 create mode 100755 deps/jemalloc/include/jemalloc/internal/smoothstep.sh
 create mode 100644 deps/jemalloc/include/jemalloc/internal/spin.h
 create mode 100644 deps/jemalloc/include/jemalloc/internal/ticker.h
 create mode 100644 deps/jemalloc/include/jemalloc/internal/witness.h
 delete mode 100755 deps/jemalloc/install-sh
 create mode 100644 deps/jemalloc/msvc/ReadMe.txt
 create mode 100644 deps/jemalloc/msvc/jemalloc_vc2015.sln
 create mode 100644 deps/jemalloc/msvc/projects/vc2015/jemalloc/jemalloc.vcxproj
 create mode 100644 deps/jemalloc/msvc/projects/vc2015/jemalloc/jemalloc.vcxproj.filters
 create mode 100755 deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.cpp
 create mode 100644 deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.h
 create mode 100644 deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.vcxproj
 create mode 100644 deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.vcxproj.filters
 create mode 100644 deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads_main.cpp
 create mode 100644 deps/jemalloc/src/nstime.c
 create mode 100644 deps/jemalloc/src/prng.c
 create mode 100644 deps/jemalloc/src/spin.c
 mode change 100644 => 100755 deps/jemalloc/src/stats.c
 mode change 100644 => 100755 deps/jemalloc/src/tcache.c
 create mode 100644 deps/jemalloc/src/ticker.c
 mode change 100644 => 100755 deps/jemalloc/src/util.c
 create mode 100644 deps/jemalloc/src/witness.c
 mode change 100644 => 100755 deps/jemalloc/test/integration/MALLOCX_ARENA.c
 mode change 100644 => 100755 deps/jemalloc/test/integration/allocated.c
 mode change 100644 => 100755 deps/jemalloc/test/integration/mallocx.c
 mode change 100644 => 100755 deps/jemalloc/test/integration/overflow.c
 mode change 100644 => 100755 deps/jemalloc/test/integration/rallocx.c
 mode change 100644 => 100755 deps/jemalloc/test/integration/thread_arena.c
 mode change 100644 => 100755 deps/jemalloc/test/integration/thread_tcache_enabled.c
 mode change 100644 => 100755 deps/jemalloc/test/integration/xallocx.c
 create mode 100644 deps/jemalloc/test/unit/a0.c
 create mode 100755 deps/jemalloc/test/unit/arena_reset.c
 create mode 100755 deps/jemalloc/test/unit/decay.c
 create mode 100644 deps/jemalloc/test/unit/fork.c
 mode change 100644 => 100755 deps/jemalloc/test/unit/mallctl.c
 create mode 100644 deps/jemalloc/test/unit/nstime.c
 create mode 100644 deps/jemalloc/test/unit/pack.c
 create mode 100644 deps/jemalloc/test/unit/pages.c
 create mode 100644 deps/jemalloc/test/unit/ph.c
 create mode 100644 deps/jemalloc/test/unit/prng.c
 mode change 100644 => 100755 deps/jemalloc/test/unit/prof_accum.c
 mode change 100644 => 100755 deps/jemalloc/test/unit/prof_active.c
 mode change 100644 => 100755 deps/jemalloc/test/unit/prof_gdump.c
 mode change 100644 => 100755 deps/jemalloc/test/unit/prof_idump.c
 mode change 100644 => 100755 deps/jemalloc/test/unit/prof_reset.c
 mode change 100644 => 100755 deps/jemalloc/test/unit/prof_thread_name.c
 create mode 100644 deps/jemalloc/test/unit/run_quantize.c
 mode change 100644 => 100755 deps/jemalloc/test/unit/size_classes.c
 create mode 100644 deps/jemalloc/test/unit/smoothstep.c
 mode change 100644 => 100755 deps/jemalloc/test/unit/stats.c
 create mode 100644 deps/jemalloc/test/unit/ticker.c
 create mode 100644 deps/jemalloc/test/unit/witness.c

diff --git a/deps/jemalloc/.appveyor.yml b/deps/jemalloc/.appveyor.yml
new file mode 100644
index 000000000..ddd5c5711
--- /dev/null
+++ b/deps/jemalloc/.appveyor.yml
@@ -0,0 +1,28 @@
+version: '{build}'
+
+environment:
+  matrix:
+  - MSYSTEM: MINGW64
+    CPU: x86_64
+    MSVC: amd64
+  - MSYSTEM: MINGW32
+    CPU: i686
+    MSVC: x86
+  - MSYSTEM: MINGW64
+    CPU: x86_64
+  - MSYSTEM: MINGW32
+    CPU: i686
+
+install:
+  - set PATH=c:\msys64\%MSYSTEM%\bin;c:\msys64\usr\bin;%PATH%
+  - if defined MSVC call "c:\Program Files (x86)\Microsoft Visual Studio 14.0\VC\vcvarsall.bat" %MSVC%
+  - if defined MSVC pacman --noconfirm -Rsc mingw-w64-%CPU%-gcc gcc
+  - pacman --noconfirm -Suy mingw-w64-%CPU%-make
+
+build_script:
+  - bash -c "autoconf"
+  - bash -c "./configure"
+  - mingw32-make -j3
+  - file lib/jemalloc.dll
+  - mingw32-make -j3 tests
+  - mingw32-make -k check
diff --git a/deps/jemalloc/.gitignore b/deps/jemalloc/.gitignore
index d0e393619..08278d087 100644
--- a/deps/jemalloc/.gitignore
+++ b/deps/jemalloc/.gitignore
@@ -73,3 +73,19 @@ test/include/test/jemalloc_test_defs.h
 /test/unit/*.out
 
 /VERSION
+
+*.pdb
+*.sdf
+*.opendb
+*.opensdf
+*.cachefile
+*.suo
+*.user
+*.sln.docstates
+*.tmp
+/msvc/Win32/
+/msvc/x64/
+/msvc/projects/*/*/Debug*/
+/msvc/projects/*/*/Release*/
+/msvc/projects/*/*/Win32/
+/msvc/projects/*/*/x64/
diff --git a/deps/jemalloc/.travis.yml b/deps/jemalloc/.travis.yml
new file mode 100644
index 000000000..1fed4f8e6
--- /dev/null
+++ b/deps/jemalloc/.travis.yml
@@ -0,0 +1,29 @@
+language: c
+
+matrix:
+  include:
+    - os: linux
+      compiler: gcc
+    - os: linux
+      compiler: gcc
+      env:
+        - EXTRA_FLAGS=-m32
+      addons:
+        apt:
+          packages:
+          - gcc-multilib
+    - os: osx
+      compiler: clang
+    - os: osx
+      compiler: clang
+      env:
+        - EXTRA_FLAGS=-m32
+
+before_script:
+  - autoconf
+  - ./configure${EXTRA_FLAGS:+ CC="$CC $EXTRA_FLAGS"}
+  - make -j3
+  - make -j3 tests
+
+script:
+  - make check
diff --git a/deps/jemalloc/COPYING b/deps/jemalloc/COPYING
index 611968cda..104b1f8b0 100644
--- a/deps/jemalloc/COPYING
+++ b/deps/jemalloc/COPYING
@@ -1,10 +1,10 @@
 Unless otherwise specified, files in the jemalloc source distribution are
 subject to the following license:
 --------------------------------------------------------------------------------
-Copyright (C) 2002-2015 Jason Evans <jasone@canonware.com>.
+Copyright (C) 2002-2016 Jason Evans <jasone@canonware.com>.
 All rights reserved.
 Copyright (C) 2007-2012 Mozilla Foundation.  All rights reserved.
-Copyright (C) 2009-2015 Facebook, Inc.  All rights reserved.
+Copyright (C) 2009-2016 Facebook, Inc.  All rights reserved.
 
 Redistribution and use in source and binary forms, with or without
 modification, are permitted provided that the following conditions are met:
diff --git a/deps/jemalloc/ChangeLog b/deps/jemalloc/ChangeLog
index e3b0a5190..f75edd933 100644
--- a/deps/jemalloc/ChangeLog
+++ b/deps/jemalloc/ChangeLog
@@ -4,6 +4,226 @@ brevity.  Much more detail can be found in the git revision history:
 
     https://github.com/jemalloc/jemalloc
 
+* 4.4.0 (December 3, 2016)
+
+  New features:
+  - Add configure support for *-*-linux-android.  (@cferris1000, @jasone)
+  - Add the --disable-syscall configure option, for use on systems that place
+    security-motivated limitations on syscall(2).  (@jasone)
+  - Add support for Debian GNU/kFreeBSD.  (@thesam)
+
+  Optimizations:
+  - Add extent serial numbers and use them where appropriate as a sort key that
+    is higher priority than address, so that the allocation policy prefers older
+    extents.  This tends to improve locality (decrease fragmentation) when
+    memory grows downward.  (@jasone)
+  - Refactor madvise(2) configuration so that MADV_FREE is detected and utilized
+    on Linux 4.5 and newer.  (@jasone)
+  - Mark partially purged arena chunks as non-huge-page.  This improves
+    interaction with Linux's transparent huge page functionality.  (@jasone)
+
+  Bug fixes:
+  - Fix size class computations for edge conditions involving extremely large
+    allocations.  This regression was first released in 4.0.0.  (@jasone,
+    @ingvarha)
+  - Remove overly restrictive assertions related to the cactive statistic.  This
+    regression was first released in 4.1.0.  (@jasone)
+  - Implement a more reliable detection scheme for os_unfair_lock on macOS.
+    (@jszakmeister)
+
+* 4.3.1 (November 7, 2016)
+
+  Bug fixes:
+  - Fix a severe virtual memory leak.  This regression was first released in
+    4.3.0.  (@interwq, @jasone)
+  - Refactor atomic and prng APIs to restore support for 32-bit platforms that
+    use pre-C11 toolchains, e.g. FreeBSD's mips.  (@jasone)
+
+* 4.3.0 (November 4, 2016)
+
+  This is the first release that passes the test suite for multiple Windows
+  configurations, thanks in large part to @glandium setting up continuous
+  integration via AppVeyor (and Travis CI for Linux and OS X).
+
+  New features:
+  - Add "J" (JSON) support to malloc_stats_print().  (@jasone)
+  - Add Cray compiler support.  (@ronawho)
+
+  Optimizations:
+  - Add/use adaptive spinning for bootstrapping and radix tree node
+    initialization.  (@jasone)
+
+  Bug fixes:
+  - Fix large allocation to search starting in the optimal size class heap,
+    which can substantially reduce virtual memory churn and fragmentation.  This
+    regression was first released in 4.0.0.  (@mjp41, @jasone)
+  - Fix stats.arenas.<i>.nthreads accounting.  (@interwq)
+  - Fix and simplify decay-based purging.  (@jasone)
+  - Make DSS (sbrk(2)-related) operations lockless, which resolves potential
+    deadlocks during thread exit.  (@jasone)
+  - Fix over-sized allocation of radix tree leaf nodes.  (@mjp41, @ogaun,
+    @jasone)
+  - Fix over-sized allocation of arena_t (plus associated stats) data
+    structures.  (@jasone, @interwq)
+  - Fix EXTRA_CFLAGS to not affect configuration.  (@jasone)
+  - Fix a Valgrind integration bug.  (@ronawho)
+  - Disallow 0x5a junk filling when running in Valgrind.  (@jasone)
+  - Fix a file descriptor leak on Linux.  This regression was first released in
+    4.2.0.  (@vsarunas, @jasone)
+  - Fix static linking of jemalloc with glibc.  (@djwatson)
+  - Use syscall(2) rather than {open,read,close}(2) during boot on Linux.  This
+    works around other libraries' system call wrappers performing reentrant
+    allocation.  (@kspinka, @Whissi, @jasone)
+  - Fix OS X default zone replacement to work with OS X 10.12.  (@glandium,
+    @jasone)
+  - Fix cached memory management to avoid needless commit/decommit operations
+    during purging, which resolves permanent virtual memory map fragmentation
+    issues on Windows.  (@mjp41, @jasone)
+  - Fix TSD fetches to avoid (recursive) allocation.  This is relevant to
+    non-TLS and Windows configurations.  (@jasone)
+  - Fix malloc_conf overriding to work on Windows.  (@jasone)
+  - Forcibly disable lazy-lock on Windows (was forcibly *enabled*).  (@jasone)
+
+* 4.2.1 (June 8, 2016)
+
+  Bug fixes:
+  - Fix bootstrapping issues for configurations that require allocation during
+    tsd initialization (e.g. --disable-tls).  (@cferris1000, @jasone)
+  - Fix gettimeofday() version of nstime_update().  (@ronawho)
+  - Fix Valgrind regressions in calloc() and chunk_alloc_wrapper().  (@ronawho)
+  - Fix potential VM map fragmentation regression.  (@jasone)
+  - Fix opt_zero-triggered in-place huge reallocation zeroing.  (@jasone)
+  - Fix heap profiling context leaks in reallocation edge cases.  (@jasone)
+
+* 4.2.0 (May 12, 2016)
+
+  New features:
+  - Add the arena.<i>.reset mallctl, which makes it possible to discard all of
+    an arena's allocations in a single operation.  (@jasone)
+  - Add the stats.retained and stats.arenas.<i>.retained statistics.  (@jasone)
+  - Add the --with-version configure option.  (@jasone)
+  - Support --with-lg-page values larger than actual page size.  (@jasone)
+
+  Optimizations:
+  - Use pairing heaps rather than red-black trees for various hot data
+    structures.  (@djwatson, @jasone)
+  - Streamline fast paths of rtree operations.  (@jasone)
+  - Optimize the fast paths of calloc() and [m,d,sd]allocx().  (@jasone)
+  - Decommit unused virtual memory if the OS does not overcommit.  (@jasone)
+  - Specify MAP_NORESERVE on Linux if [heuristic] overcommit is active, in order
+    to avoid unfortunate interactions during fork(2).  (@jasone)
+
+  Bug fixes:
+  - Fix chunk accounting related to triggering gdump profiles.  (@jasone)
+  - Link against librt for clock_gettime(2) if glibc < 2.17.  (@jasone)
+  - Scale leak report summary according to sampling probability.  (@jasone)
+
+* 4.1.1 (May 3, 2016)
+
+  This bugfix release resolves a variety of mostly minor issues, though the
+  bitmap fix is critical for 64-bit Windows.
+
+  Bug fixes:
+  - Fix the linear scan version of bitmap_sfu() to shift by the proper amount
+    even when sizeof(long) is not the same as sizeof(void *), as on 64-bit
+    Windows.  (@jasone)
+  - Fix hashing functions to avoid unaligned memory accesses (and resulting
+    crashes).  This is relevant at least to some ARM-based platforms.
+    (@rkmisra)
+  - Fix fork()-related lock rank ordering reversals.  These reversals were
+    unlikely to cause deadlocks in practice except when heap profiling was
+    enabled and active.  (@jasone)
+  - Fix various chunk leaks in OOM code paths.  (@jasone)
+  - Fix malloc_stats_print() to print opt.narenas correctly.  (@jasone)
+  - Fix MSVC-specific build/test issues.  (@rustyx, @yuslepukhin)
+  - Fix a variety of test failures that were due to test fragility rather than
+    core bugs.  (@jasone)
+
+* 4.1.0 (February 28, 2016)
+
+  This release is primarily about optimizations, but it also incorporates a lot
+  of portability-motivated refactoring and enhancements.  Many people worked on
+  this release, to an extent that even with the omission here of minor changes
+  (see git revision history), and of the people who reported and diagnosed
+  issues, so much of the work was contributed that starting with this release,
+  changes are annotated with author credits to help reflect the collaborative
+  effort involved.
+
+  New features:
+  - Implement decay-based unused dirty page purging, a major optimization with
+    mallctl API impact.  This is an alternative to the existing ratio-based
+    unused dirty page purging, and is intended to eventually become the sole
+    purging mechanism.  New mallctls:
+    + opt.purge
+    + opt.decay_time
+    + arena.<i>.decay
+    + arena.<i>.decay_time
+    + arenas.decay_time
+    + stats.arenas.<i>.decay_time
+    (@jasone, @cevans87)
+  - Add --with-malloc-conf, which makes it possible to embed a default
+    options string during configuration.  This was motivated by the desire to
+    specify --with-malloc-conf=purge:decay , since the default must remain
+    purge:ratio until the 5.0.0 release.  (@jasone)
+  - Add MS Visual Studio 2015 support.  (@rustyx, @yuslepukhin)
+  - Make *allocx() size class overflow behavior defined.  The maximum
+    size class is now less than PTRDIFF_MAX to protect applications against
+    numerical overflow, and all allocation functions are guaranteed to indicate
+    errors rather than potentially crashing if the request size exceeds the
+    maximum size class.  (@jasone)
+  - jeprof:
+    + Add raw heap profile support.  (@jasone)
+    + Add --retain and --exclude for backtrace symbol filtering.  (@jasone)
+
+  Optimizations:
+  - Optimize the fast path to combine various bootstrapping and configuration
+    checks and execute more streamlined code in the common case.  (@interwq)
+  - Use linear scan for small bitmaps (used for small object tracking).  In
+    addition to speeding up bitmap operations on 64-bit systems, this reduces
+    allocator metadata overhead by approximately 0.2%.  (@djwatson)
+  - Separate arena_avail trees, which substantially speeds up run tree
+    operations.  (@djwatson)
+  - Use memoization (boot-time-computed table) for run quantization.  Separate
+    arena_avail trees reduced the importance of this optimization.  (@jasone)
+  - Attempt mmap-based in-place huge reallocation.  This can dramatically speed
+    up incremental huge reallocation.  (@jasone)
+
+  Incompatible changes:
+  - Make opt.narenas unsigned rather than size_t.  (@jasone)
+
+  Bug fixes:
+  - Fix stats.cactive accounting regression.  (@rustyx, @jasone)
+  - Handle unaligned keys in hash().  This caused problems for some ARM systems.
+    (@jasone, @cferris1000)
+  - Refactor arenas array.  In addition to fixing a fork-related deadlock, this
+    makes arena lookups faster and simpler.  (@jasone)
+  - Move retained memory allocation out of the default chunk allocation
+    function, to a location that gets executed even if the application installs
+    a custom chunk allocation function.  This resolves a virtual memory leak.
+    (@buchgr)
+  - Fix a potential tsd cleanup leak.  (@cferris1000, @jasone)
+  - Fix run quantization.  In practice this bug had no impact unless
+    applications requested memory with alignment exceeding one page.
+    (@jasone, @djwatson)
+  - Fix LinuxThreads-specific bootstrapping deadlock.  (Cosmin Paraschiv)
+  - jeprof:
+    + Don't discard curl options if timeout is not defined.  (@djwatson)
+    + Detect failed profile fetches.  (@djwatson)
+  - Fix stats.arenas.<i>.{dss,lg_dirty_mult,decay_time,pactive,pdirty} for
+    --disable-stats case.  (@jasone)
+
+* 4.0.4 (October 24, 2015)
+
+  This bugfix release fixes another xallocx() regression.  No other regressions
+  have come to light in over a month, so this is likely a good starting point
+  for people who prefer to wait for "dot one" releases with all the major issues
+  shaken out.
+
+  Bug fixes:
+  - Fix xallocx(..., MALLOCX_ZERO to zero the last full trailing page of large
+    allocations that have been randomly assigned an offset of 0 when
+    --enable-cache-oblivious configure option is enabled.
+
 * 4.0.3 (September 24, 2015)
 
   This bugfix release continues the trend of xallocx() and heap profiling fixes.
diff --git a/deps/jemalloc/INSTALL b/deps/jemalloc/INSTALL
index 8d3968745..cce3ed711 100644
--- a/deps/jemalloc/INSTALL
+++ b/deps/jemalloc/INSTALL
@@ -35,6 +35,10 @@ any of the following arguments (not a definitive list) to 'configure':
     will cause files to be installed into /usr/local/include, /usr/local/lib,
     and /usr/local/man.
 
+--with-version=<major>.<minor>.<bugfix>-<nrev>-g<gid>
+    Use the specified version string rather than trying to generate one (if in
+    a git repository) or use existing the VERSION file (if present).
+
 --with-rpath=<colon-separated-rpath>
     Embed one or more library paths, so that libjemalloc can find the libraries
     it is linked to.  This works only on ELF-based systems.
@@ -84,6 +88,14 @@ any of the following arguments (not a definitive list) to 'configure':
     versions of jemalloc can coexist in the same installation directory.  For
     example, libjemalloc.so.0 becomes libjemalloc<suffix>.so.0.
 
+--with-malloc-conf=<malloc_conf>
+    Embed <malloc_conf> as a run-time options string that is processed prior to
+    the malloc_conf global variable, the /etc/malloc.conf symlink, and the
+    MALLOC_CONF environment variable.  For example, to change the default chunk
+    size to 256 KiB:
+
+      --with-malloc-conf=lg_chunk:18
+
 --disable-cc-silence
     Disable code that silences non-useful compiler warnings.  This is mainly
     useful during development when auditing the set of warnings that are being
@@ -194,6 +206,11 @@ any of the following arguments (not a definitive list) to 'configure':
     most extreme case increases physical memory usage for the 16 KiB size class
     to 20 KiB.
 
+--disable-syscall
+    Disable use of syscall(2) rather than {open,read,write,close}(2).  This is
+    intended as a workaround for systems that place security limitations on
+    syscall(2).
+
 --with-xslroot=<path>
     Specify where to find DocBook XSL stylesheets when building the
     documentation.
@@ -315,6 +332,15 @@ LDFLAGS="?"
 PATH="?"
     'configure' uses this to find programs.
 
+In some cases it may be necessary to work around configuration results that do
+not match reality.  For example, Linux 4.5 added support for the MADV_FREE flag
+to madvise(2), which can cause problems if building on a host with MADV_FREE
+support and deploying to a target without.  To work around this, use a cache
+file to override the relevant configuration variable defined in configure.ac,
+e.g.:
+
+    echo "je_cv_madv_free=no" > config.cache && ./configure -C
+
 === Advanced compilation =======================================================
 
 To build only parts of jemalloc, use the following targets:
diff --git a/deps/jemalloc/Makefile.in b/deps/jemalloc/Makefile.in
index 1ac6f2926..c70536391 100644
--- a/deps/jemalloc/Makefile.in
+++ b/deps/jemalloc/Makefile.in
@@ -24,11 +24,11 @@ abs_objroot := @abs_objroot@
 
 # Build parameters.
 CPPFLAGS := @CPPFLAGS@ -I$(srcroot)include -I$(objroot)include
-CFLAGS := @CFLAGS@
+EXTRA_CFLAGS := @EXTRA_CFLAGS@
+CFLAGS := @CFLAGS@ $(EXTRA_CFLAGS)
 LDFLAGS := @LDFLAGS@
 EXTRA_LDFLAGS := @EXTRA_LDFLAGS@
 LIBS := @LIBS@
-TESTLIBS := @TESTLIBS@
 RPATH_EXTRA := @RPATH_EXTRA@
 SO := @so@
 IMPORTLIB := @importlib@
@@ -53,15 +53,19 @@ enable_prof := @enable_prof@
 enable_valgrind := @enable_valgrind@
 enable_zone_allocator := @enable_zone_allocator@
 MALLOC_CONF := @JEMALLOC_CPREFIX@MALLOC_CONF
+link_whole_archive := @link_whole_archive@
 DSO_LDFLAGS = @DSO_LDFLAGS@
 SOREV = @SOREV@
 PIC_CFLAGS = @PIC_CFLAGS@
 CTARGET = @CTARGET@
 LDTARGET = @LDTARGET@
+TEST_LD_MODE = @TEST_LD_MODE@
 MKLIB = @MKLIB@
 AR = @AR@
 ARFLAGS = @ARFLAGS@
 CC_MM = @CC_MM@
+LM := @LM@
+INSTALL = @INSTALL@
 
 ifeq (macho, $(ABI))
 TEST_LIBRARY_PATH := DYLD_FALLBACK_LIBRARY_PATH="$(objroot)lib"
@@ -78,15 +82,34 @@ LIBJEMALLOC := $(LIBPREFIX)jemalloc$(install_suffix)
 # Lists of files.
 BINS := $(objroot)bin/jemalloc-config $(objroot)bin/jemalloc.sh $(objroot)bin/jeprof
 C_HDRS := $(objroot)include/jemalloc/jemalloc$(install_suffix).h
-C_SRCS := $(srcroot)src/jemalloc.c $(srcroot)src/arena.c \
-	$(srcroot)src/atomic.c $(srcroot)src/base.c $(srcroot)src/bitmap.c \
-	$(srcroot)src/chunk.c $(srcroot)src/chunk_dss.c \
-	$(srcroot)src/chunk_mmap.c $(srcroot)src/ckh.c $(srcroot)src/ctl.c \
-	$(srcroot)src/extent.c $(srcroot)src/hash.c $(srcroot)src/huge.c \
-	$(srcroot)src/mb.c $(srcroot)src/mutex.c $(srcroot)src/pages.c \
-	$(srcroot)src/prof.c $(srcroot)src/quarantine.c $(srcroot)src/rtree.c \
-	$(srcroot)src/stats.c $(srcroot)src/tcache.c $(srcroot)src/util.c \
-	$(srcroot)src/tsd.c
+C_SRCS := $(srcroot)src/jemalloc.c \
+	$(srcroot)src/arena.c \
+	$(srcroot)src/atomic.c \
+	$(srcroot)src/base.c \
+	$(srcroot)src/bitmap.c \
+	$(srcroot)src/chunk.c \
+	$(srcroot)src/chunk_dss.c \
+	$(srcroot)src/chunk_mmap.c \
+	$(srcroot)src/ckh.c \
+	$(srcroot)src/ctl.c \
+	$(srcroot)src/extent.c \
+	$(srcroot)src/hash.c \
+	$(srcroot)src/huge.c \
+	$(srcroot)src/mb.c \
+	$(srcroot)src/mutex.c \
+	$(srcroot)src/nstime.c \
+	$(srcroot)src/pages.c \
+	$(srcroot)src/prng.c \
+	$(srcroot)src/prof.c \
+	$(srcroot)src/quarantine.c \
+	$(srcroot)src/rtree.c \
+	$(srcroot)src/stats.c \
+	$(srcroot)src/spin.c \
+	$(srcroot)src/tcache.c \
+	$(srcroot)src/ticker.c \
+	$(srcroot)src/tsd.c \
+	$(srcroot)src/util.c \
+	$(srcroot)src/witness.c
 ifeq ($(enable_valgrind), 1)
 C_SRCS += $(srcroot)src/valgrind.c
 endif
@@ -105,6 +128,11 @@ DSOS := $(objroot)lib/$(LIBJEMALLOC).$(SOREV)
 ifneq ($(SOREV),$(SO))
 DSOS += $(objroot)lib/$(LIBJEMALLOC).$(SO)
 endif
+ifeq (1, $(link_whole_archive))
+LJEMALLOC := -Wl,--whole-archive -L$(objroot)lib -l$(LIBJEMALLOC) -Wl,--no-whole-archive
+else
+LJEMALLOC := $(objroot)lib/$(LIBJEMALLOC).$(IMPORTLIB)
+endif
 PC := $(objroot)jemalloc.pc
 MAN3 := $(objroot)doc/jemalloc$(install_suffix).3
 DOCS_XML := $(objroot)doc/jemalloc$(install_suffix).xml
@@ -116,10 +144,19 @@ C_TESTLIB_SRCS := $(srcroot)test/src/btalloc.c $(srcroot)test/src/btalloc_0.c \
 	$(srcroot)test/src/mtx.c $(srcroot)test/src/mq.c \
 	$(srcroot)test/src/SFMT.c $(srcroot)test/src/test.c \
 	$(srcroot)test/src/thd.c $(srcroot)test/src/timer.c
-C_UTIL_INTEGRATION_SRCS := $(srcroot)src/util.c
-TESTS_UNIT := $(srcroot)test/unit/atomic.c \
+ifeq (1, $(link_whole_archive))
+C_UTIL_INTEGRATION_SRCS :=
+else
+C_UTIL_INTEGRATION_SRCS := $(srcroot)src/nstime.c $(srcroot)src/util.c
+endif
+TESTS_UNIT := \
+	$(srcroot)test/unit/a0.c \
+	$(srcroot)test/unit/arena_reset.c \
+	$(srcroot)test/unit/atomic.c \
 	$(srcroot)test/unit/bitmap.c \
 	$(srcroot)test/unit/ckh.c \
+	$(srcroot)test/unit/decay.c \
+	$(srcroot)test/unit/fork.c \
 	$(srcroot)test/unit/hash.c \
 	$(srcroot)test/unit/junk.c \
 	$(srcroot)test/unit/junk_alloc.c \
@@ -129,6 +166,10 @@ TESTS_UNIT := $(srcroot)test/unit/atomic.c \
 	$(srcroot)test/unit/math.c \
 	$(srcroot)test/unit/mq.c \
 	$(srcroot)test/unit/mtx.c \
+	$(srcroot)test/unit/pack.c \
+	$(srcroot)test/unit/pages.c \
+	$(srcroot)test/unit/ph.c \
+	$(srcroot)test/unit/prng.c \
 	$(srcroot)test/unit/prof_accum.c \
 	$(srcroot)test/unit/prof_active.c \
 	$(srcroot)test/unit/prof_gdump.c \
@@ -140,11 +181,16 @@ TESTS_UNIT := $(srcroot)test/unit/atomic.c \
 	$(srcroot)test/unit/quarantine.c \
 	$(srcroot)test/unit/rb.c \
 	$(srcroot)test/unit/rtree.c \
+	$(srcroot)test/unit/run_quantize.c \
 	$(srcroot)test/unit/SFMT.c \
 	$(srcroot)test/unit/size_classes.c \
+	$(srcroot)test/unit/smoothstep.c \
 	$(srcroot)test/unit/stats.c \
+	$(srcroot)test/unit/ticker.c \
+	$(srcroot)test/unit/nstime.c \
 	$(srcroot)test/unit/tsd.c \
 	$(srcroot)test/unit/util.c \
+	$(srcroot)test/unit/witness.c \
 	$(srcroot)test/unit/zero.c
 TESTS_INTEGRATION := $(srcroot)test/integration/aligned_alloc.c \
 	$(srcroot)test/integration/allocated.c \
@@ -266,69 +312,69 @@ $(STATIC_LIBS):
 
 $(objroot)test/unit/%$(EXE): $(objroot)test/unit/%.$(O) $(TESTS_UNIT_LINK_OBJS) $(C_JET_OBJS) $(C_TESTLIB_UNIT_OBJS)
 	@mkdir -p $(@D)
-	$(CC) $(LDTARGET) $(filter %.$(O),$^) $(call RPATH,$(objroot)lib) $(LDFLAGS) $(filter-out -lm,$(LIBS)) -lm $(TESTLIBS) $(EXTRA_LDFLAGS)
+	$(CC) $(LDTARGET) $(filter %.$(O),$^) $(call RPATH,$(objroot)lib) $(LDFLAGS) $(filter-out -lm,$(LIBS)) $(LM) $(EXTRA_LDFLAGS)
 
 $(objroot)test/integration/%$(EXE): $(objroot)test/integration/%.$(O) $(C_TESTLIB_INTEGRATION_OBJS) $(C_UTIL_INTEGRATION_OBJS) $(objroot)lib/$(LIBJEMALLOC).$(IMPORTLIB)
 	@mkdir -p $(@D)
-	$(CC) $(LDTARGET) $(filter %.$(O),$^) $(call RPATH,$(objroot)lib) $(objroot)lib/$(LIBJEMALLOC).$(IMPORTLIB) $(LDFLAGS) $(filter-out -lm,$(filter -lpthread,$(LIBS))) -lm $(TESTLIBS) $(EXTRA_LDFLAGS)
+	$(CC) $(TEST_LD_MODE) $(LDTARGET) $(filter %.$(O),$^) $(call RPATH,$(objroot)lib) $(LJEMALLOC) $(LDFLAGS) $(filter-out -lm,$(filter -lrt -lpthread,$(LIBS))) $(LM) $(EXTRA_LDFLAGS)
 
 $(objroot)test/stress/%$(EXE): $(objroot)test/stress/%.$(O) $(C_JET_OBJS) $(C_TESTLIB_STRESS_OBJS) $(objroot)lib/$(LIBJEMALLOC).$(IMPORTLIB)
 	@mkdir -p $(@D)
-	$(CC) $(LDTARGET) $(filter %.$(O),$^) $(call RPATH,$(objroot)lib) $(objroot)lib/$(LIBJEMALLOC).$(IMPORTLIB) $(LDFLAGS) $(filter-out -lm,$(LIBS)) -lm $(TESTLIBS) $(EXTRA_LDFLAGS)
+	$(CC) $(TEST_LD_MODE) $(LDTARGET) $(filter %.$(O),$^) $(call RPATH,$(objroot)lib) $(objroot)lib/$(LIBJEMALLOC).$(IMPORTLIB) $(LDFLAGS) $(filter-out -lm,$(LIBS)) $(LM) $(EXTRA_LDFLAGS)
 
 build_lib_shared: $(DSOS)
 build_lib_static: $(STATIC_LIBS)
 build_lib: build_lib_shared build_lib_static
 
 install_bin:
-	install -d $(BINDIR)
+	$(INSTALL) -d $(BINDIR)
 	@for b in $(BINS); do \
-	echo "install -m 755 $$b $(BINDIR)"; \
-	install -m 755 $$b $(BINDIR); \
+	echo "$(INSTALL) -m 755 $$b $(BINDIR)"; \
+	$(INSTALL) -m 755 $$b $(BINDIR); \
 done
 
 install_include:
-	install -d $(INCLUDEDIR)/jemalloc
+	$(INSTALL) -d $(INCLUDEDIR)/jemalloc
 	@for h in $(C_HDRS); do \
-	echo "install -m 644 $$h $(INCLUDEDIR)/jemalloc"; \
-	install -m 644 $$h $(INCLUDEDIR)/jemalloc; \
+	echo "$(INSTALL) -m 644 $$h $(INCLUDEDIR)/jemalloc"; \
+	$(INSTALL) -m 644 $$h $(INCLUDEDIR)/jemalloc; \
 done
 
 install_lib_shared: $(DSOS)
-	install -d $(LIBDIR)
-	install -m 755 $(objroot)lib/$(LIBJEMALLOC).$(SOREV) $(LIBDIR)
+	$(INSTALL) -d $(LIBDIR)
+	$(INSTALL) -m 755 $(objroot)lib/$(LIBJEMALLOC).$(SOREV) $(LIBDIR)
 ifneq ($(SOREV),$(SO))
 	ln -sf $(LIBJEMALLOC).$(SOREV) $(LIBDIR)/$(LIBJEMALLOC).$(SO)
 endif
 
 install_lib_static: $(STATIC_LIBS)
-	install -d $(LIBDIR)
+	$(INSTALL) -d $(LIBDIR)
 	@for l in $(STATIC_LIBS); do \
-	echo "install -m 755 $$l $(LIBDIR)"; \
-	install -m 755 $$l $(LIBDIR); \
+	echo "$(INSTALL) -m 755 $$l $(LIBDIR)"; \
+	$(INSTALL) -m 755 $$l $(LIBDIR); \
 done
 
 install_lib_pc: $(PC)
-	install -d $(LIBDIR)/pkgconfig
+	$(INSTALL) -d $(LIBDIR)/pkgconfig
 	@for l in $(PC); do \
-	echo "install -m 644 $$l $(LIBDIR)/pkgconfig"; \
-	install -m 644 $$l $(LIBDIR)/pkgconfig; \
+	echo "$(INSTALL) -m 644 $$l $(LIBDIR)/pkgconfig"; \
+	$(INSTALL) -m 644 $$l $(LIBDIR)/pkgconfig; \
 done
 
 install_lib: install_lib_shared install_lib_static install_lib_pc
 
 install_doc_html:
-	install -d $(DATADIR)/doc/jemalloc$(install_suffix)
+	$(INSTALL) -d $(DATADIR)/doc/jemalloc$(install_suffix)
 	@for d in $(DOCS_HTML); do \
-	echo "install -m 644 $$d $(DATADIR)/doc/jemalloc$(install_suffix)"; \
-	install -m 644 $$d $(DATADIR)/doc/jemalloc$(install_suffix); \
+	echo "$(INSTALL) -m 644 $$d $(DATADIR)/doc/jemalloc$(install_suffix)"; \
+	$(INSTALL) -m 644 $$d $(DATADIR)/doc/jemalloc$(install_suffix); \
 done
 
 install_doc_man:
-	install -d $(MANDIR)/man3
+	$(INSTALL) -d $(MANDIR)/man3
 	@for d in $(DOCS_MAN3); do \
-	echo "install -m 644 $$d $(MANDIR)/man3"; \
-	install -m 644 $$d $(MANDIR)/man3; \
+	echo "$(INSTALL) -m 644 $$d $(MANDIR)/man3"; \
+	$(INSTALL) -m 644 $$d $(MANDIR)/man3; \
 done
 
 install_doc: install_doc_html install_doc_man
@@ -349,18 +395,22 @@ stress_dir:
 check_dir: check_unit_dir check_integration_dir
 
 check_unit: tests_unit check_unit_dir
-	$(SHELL) $(objroot)test/test.sh $(TESTS_UNIT:$(srcroot)%.c=$(objroot)%)
+	$(MALLOC_CONF)="purge:ratio" $(SHELL) $(objroot)test/test.sh $(TESTS_UNIT:$(srcroot)%.c=$(objroot)%)
+	$(MALLOC_CONF)="purge:decay" $(SHELL) $(objroot)test/test.sh $(TESTS_UNIT:$(srcroot)%.c=$(objroot)%)
 check_integration_prof: tests_integration check_integration_dir
 ifeq ($(enable_prof), 1)
 	$(MALLOC_CONF)="prof:true" $(SHELL) $(objroot)test/test.sh $(TESTS_INTEGRATION:$(srcroot)%.c=$(objroot)%)
 	$(MALLOC_CONF)="prof:true,prof_active:false" $(SHELL) $(objroot)test/test.sh $(TESTS_INTEGRATION:$(srcroot)%.c=$(objroot)%)
 endif
+check_integration_decay: tests_integration check_integration_dir
+	$(MALLOC_CONF)="purge:decay,decay_time:-1" $(SHELL) $(objroot)test/test.sh $(TESTS_INTEGRATION:$(srcroot)%.c=$(objroot)%)
+	$(MALLOC_CONF)="purge:decay,decay_time:0" $(SHELL) $(objroot)test/test.sh $(TESTS_INTEGRATION:$(srcroot)%.c=$(objroot)%)
+	$(MALLOC_CONF)="purge:decay" $(SHELL) $(objroot)test/test.sh $(TESTS_INTEGRATION:$(srcroot)%.c=$(objroot)%)
 check_integration: tests_integration check_integration_dir
 	$(SHELL) $(objroot)test/test.sh $(TESTS_INTEGRATION:$(srcroot)%.c=$(objroot)%)
 stress: tests_stress stress_dir
 	$(SHELL) $(objroot)test/test.sh $(TESTS_STRESS:$(srcroot)%.c=$(objroot)%)
-check: tests check_dir check_integration_prof
-	$(SHELL) $(objroot)test/test.sh $(TESTS_UNIT:$(srcroot)%.c=$(objroot)%) $(TESTS_INTEGRATION:$(srcroot)%.c=$(objroot)%)
+check: check_unit check_integration check_integration_decay check_integration_prof
 
 ifeq ($(enable_code_coverage), 1)
 coverage_unit: check_unit
diff --git a/deps/jemalloc/README b/deps/jemalloc/README
index 9b268f422..5ff24a9ef 100644
--- a/deps/jemalloc/README
+++ b/deps/jemalloc/README
@@ -17,4 +17,4 @@ jemalloc.
 
 The ChangeLog file contains a brief summary of changes for each release.
 
-URL: http://www.canonware.com/jemalloc/
+URL: http://jemalloc.net/
diff --git a/deps/jemalloc/VERSION b/deps/jemalloc/VERSION
index f1f9f1c61..810bd6d4c 100644
--- a/deps/jemalloc/VERSION
+++ b/deps/jemalloc/VERSION
@@ -1 +1 @@
-4.0.3-0-ge9192eacf8935e29fc62fddc2701f7942b1cc02c
+4.4.0-0-gf1f76357313e7dcad7262f17a48ff0a2e005fcdc
diff --git a/deps/jemalloc/bin/jeprof.in b/deps/jemalloc/bin/jeprof.in
index e7178078a..42087fcec 100644
--- a/deps/jemalloc/bin/jeprof.in
+++ b/deps/jemalloc/bin/jeprof.in
@@ -95,7 +95,7 @@ my @EVINCE = ("evince");    # could also be xpdf or perhaps acroread
 my @KCACHEGRIND = ("kcachegrind");
 my @PS2PDF = ("ps2pdf");
 # These are used for dynamic profiles
-my @URL_FETCHER = ("curl", "-s");
+my @URL_FETCHER = ("curl", "-s", "--fail");
 
 # These are the web pages that servers need to support for dynamic profiles
 my $HEAP_PAGE = "/pprof/heap";
@@ -223,12 +223,14 @@ Call-graph Options:
    --nodefraction=<f>  Hide nodes below <f>*total [default=.005]
    --edgefraction=<f>  Hide edges below <f>*total [default=.001]
    --maxdegree=<n>     Max incoming/outgoing edges per node [default=8]
-   --focus=<regexp>    Focus on nodes matching <regexp>
+   --focus=<regexp>    Focus on backtraces with nodes matching <regexp>
    --thread=<n>        Show profile for thread <n>
-   --ignore=<regexp>   Ignore nodes matching <regexp>
+   --ignore=<regexp>   Ignore backtraces with nodes matching <regexp>
    --scale=<n>         Set GV scaling [default=0]
    --heapcheck         Make nodes with non-0 object counts
                        (i.e. direct leak generators) more visible
+   --retain=<regexp>   Retain only nodes that match <regexp>
+   --exclude=<regexp>  Exclude all nodes that match <regexp>
 
 Miscellaneous:
    --tools=<prefix or binary:fullpath>[,...]   \$PATH for object tool pathnames
@@ -339,6 +341,8 @@ sub Init() {
   $main::opt_ignore = '';
   $main::opt_scale = 0;
   $main::opt_heapcheck = 0;
+  $main::opt_retain = '';
+  $main::opt_exclude = '';
   $main::opt_seconds = 30;
   $main::opt_lib = "";
 
@@ -410,6 +414,8 @@ sub Init() {
              "ignore=s"       => \$main::opt_ignore,
              "scale=i"        => \$main::opt_scale,
              "heapcheck"      => \$main::opt_heapcheck,
+             "retain=s"       => \$main::opt_retain,
+             "exclude=s"      => \$main::opt_exclude,
              "inuse_space!"   => \$main::opt_inuse_space,
              "inuse_objects!" => \$main::opt_inuse_objects,
              "alloc_space!"   => \$main::opt_alloc_space,
@@ -1160,8 +1166,21 @@ sub PrintSymbolizedProfile {
   }
   print '---', "\n";
 
-  $PROFILE_PAGE =~ m,[^/]+$,;    # matches everything after the last slash
-  my $profile_marker = $&;
+  my $profile_marker;
+  if ($main::profile_type eq 'heap') {
+    $HEAP_PAGE =~ m,[^/]+$,;    # matches everything after the last slash
+    $profile_marker = $&;
+  } elsif ($main::profile_type eq 'growth') {
+    $GROWTH_PAGE =~ m,[^/]+$,;    # matches everything after the last slash
+    $profile_marker = $&;
+  } elsif ($main::profile_type eq 'contention') {
+    $CONTENTION_PAGE =~ m,[^/]+$,;    # matches everything after the last slash
+    $profile_marker = $&;
+  } else { # elsif ($main::profile_type eq 'cpu')
+    $PROFILE_PAGE =~ m,[^/]+$,;    # matches everything after the last slash
+    $profile_marker = $&;
+  }
+
   print '--- ', $profile_marker, "\n";
   if (defined($main::collected_profile)) {
     # if used with remote fetch, simply dump the collected profile to output.
@@ -1171,6 +1190,12 @@ sub PrintSymbolizedProfile {
     }
     close(SRC);
   } else {
+    # --raw/http: For everything to work correctly for non-remote profiles, we
+    # would need to extend PrintProfileData() to handle all possible profile
+    # types, re-enable the code that is currently disabled in ReadCPUProfile()
+    # and FixCallerAddresses(), and remove the remote profile dumping code in
+    # the block above.
+    die "--raw/http: jeprof can only dump remote profiles for --raw\n";
     # dump a cpu-format profile to standard out
     PrintProfileData($profile);
   }
@@ -2821,6 +2846,43 @@ sub ExtractCalls {
   return $calls;
 }
 
+sub FilterFrames {
+  my $symbols = shift;
+  my $profile = shift;
+
+  if ($main::opt_retain eq '' && $main::opt_exclude eq '') {
+    return $profile;
+  }
+
+  my $result = {};
+  foreach my $k (keys(%{$profile})) {
+    my $count = $profile->{$k};
+    my @addrs = split(/\n/, $k);
+    my @path = ();
+    foreach my $a (@addrs) {
+      my $sym;
+      if (exists($symbols->{$a})) {
+        $sym = $symbols->{$a}->[0];
+      } else {
+        $sym = $a;
+      }
+      if ($main::opt_retain ne '' && $sym !~ m/$main::opt_retain/) {
+        next;
+      }
+      if ($main::opt_exclude ne '' && $sym =~ m/$main::opt_exclude/) {
+        next;
+      }
+      push(@path, $a);
+    }
+    if (scalar(@path) > 0) {
+      my $reduced_path = join("\n", @path);
+      AddEntry($result, $reduced_path, $count);
+    }
+  }
+
+  return $result;
+}
+
 sub RemoveUninterestingFrames {
   my $symbols = shift;
   my $profile = shift;
@@ -2965,6 +3027,9 @@ sub RemoveUninterestingFrames {
     my $reduced_path = join("\n", @path);
     AddEntry($result, $reduced_path, $count);
   }
+
+  $result = FilterFrames($symbols, $result);
+
   return $result;
 }
 
@@ -3274,7 +3339,7 @@ sub ResolveRedirectionForCurl {
 # Add a timeout flat to URL_FETCHER.  Returns a new list.
 sub AddFetchTimeout {
   my $timeout = shift;
-  my @fetcher = shift;
+  my @fetcher = @_;
   if (defined($timeout)) {
     if (join(" ", @fetcher) =~ m/\bcurl -s/) {
       push(@fetcher, "--max-time", sprintf("%d", $timeout));
@@ -3320,6 +3385,27 @@ sub ReadSymbols {
   return $map;
 }
 
+sub URLEncode {
+  my $str = shift;
+  $str =~ s/([^A-Za-z0-9\-_.!~*'()])/ sprintf "%%%02x", ord $1 /eg;
+  return $str;
+}
+
+sub AppendSymbolFilterParams {
+  my $url = shift;
+  my @params = ();
+  if ($main::opt_retain ne '') {
+    push(@params, sprintf("retain=%s", URLEncode($main::opt_retain)));
+  }
+  if ($main::opt_exclude ne '') {
+    push(@params, sprintf("exclude=%s", URLEncode($main::opt_exclude)));
+  }
+  if (scalar @params > 0) {
+    $url = sprintf("%s?%s", $url, join("&", @params));
+  }
+  return $url;
+}
+
 # Fetches and processes symbols to prepare them for use in the profile output
 # code.  If the optional 'symbol_map' arg is not given, fetches symbols from
 # $SYMBOL_PAGE for all PC values found in profile.  Otherwise, the raw symbols
@@ -3344,9 +3430,11 @@ sub FetchSymbols {
     my $command_line;
     if (join(" ", @URL_FETCHER) =~ m/\bcurl -s/) {
       $url = ResolveRedirectionForCurl($url);
+      $url = AppendSymbolFilterParams($url);
       $command_line = ShellEscape(@URL_FETCHER, "-d", "\@$main::tmpfile_sym",
                                   $url);
     } else {
+      $url = AppendSymbolFilterParams($url);
       $command_line = (ShellEscape(@URL_FETCHER, "--post", $url)
                        . " < " . ShellEscape($main::tmpfile_sym));
     }
@@ -3427,12 +3515,22 @@ sub FetchDynamicProfile {
       }
       $url .= sprintf("seconds=%d", $main::opt_seconds);
       $fetch_timeout = $main::opt_seconds * 1.01 + 60;
+      # Set $profile_type for consumption by PrintSymbolizedProfile.
+      $main::profile_type = 'cpu';
     } else {
       # For non-CPU profiles, we add a type-extension to
       # the target profile file name.
       my $suffix = $path;
       $suffix =~ s,/,.,g;
       $profile_file .= $suffix;
+      # Set $profile_type for consumption by PrintSymbolizedProfile.
+      if ($path =~ m/$HEAP_PAGE/) {
+        $main::profile_type = 'heap';
+      } elsif ($path =~ m/$GROWTH_PAGE/) {
+        $main::profile_type = 'growth';
+      } elsif ($path =~ m/$CONTENTION_PAGE/) {
+        $main::profile_type = 'contention';
+      }
     }
 
     my $profile_dir = $ENV{"JEPROF_TMPDIR"} || ($ENV{HOME} . "/jeprof");
@@ -3730,6 +3828,8 @@ sub ReadProfile {
   my $symbol_marker = $&;
   $PROFILE_PAGE =~ m,[^/]+$,;    # matches everything after the last slash
   my $profile_marker = $&;
+  $HEAP_PAGE =~ m,[^/]+$,;    # matches everything after the last slash
+  my $heap_marker = $&;
 
   # Look at first line to see if it is a heap or a CPU profile.
   # CPU profile may start with no header at all, and just binary data
@@ -3756,7 +3856,13 @@ sub ReadProfile {
     $header = ReadProfileHeader(*PROFILE) || "";
   }
 
+  if ($header =~ m/^--- *($heap_marker|$growth_marker)/o) {
+    # Skip "--- ..." line for profile types that have their own headers.
+    $header = ReadProfileHeader(*PROFILE) || "";
+  }
+
   $main::profile_type = '';
+
   if ($header =~ m/^heap profile:.*$growth_marker/o) {
     $main::profile_type = 'growth';
     $result =  ReadHeapProfile($prog, *PROFILE, $header);
@@ -3808,9 +3914,9 @@ sub ReadProfile {
 # independent implementation.
 sub FixCallerAddresses {
   my $stack = shift;
-  if ($main::use_symbolized_profile) {
-    return $stack;
-  } else {
+  # --raw/http: Always subtract one from pc's, because PrintSymbolizedProfile()
+  # dumps unadjusted profiles.
+  {
     $stack =~ /(\s)/;
     my $delimiter = $1;
     my @addrs = split(' ', $stack);
@@ -3878,12 +3984,7 @@ sub ReadCPUProfile {
     for (my $j = 0; $j < $d; $j++) {
       my $pc = $slots->get($i+$j);
       # Subtract one from caller pc so we map back to call instr.
-      # However, don't do this if we're reading a symbolized profile
-      # file, in which case the subtract-one was done when the file
-      # was written.
-      if ($j > 0 && !$main::use_symbolized_profile) {
-        $pc--;
-      }
+      $pc--;
       $pc = sprintf("%0*x", $address_length, $pc);
       $pcs->{$pc} = 1;
       push @k, $pc;
diff --git a/deps/jemalloc/build-aux/config.guess b/deps/jemalloc/build-aux/config.guess
new file mode 100755
index 000000000..2e9ad7fe8
--- /dev/null
+++ b/deps/jemalloc/build-aux/config.guess
@@ -0,0 +1,1462 @@
+#! /bin/sh
+# Attempt to guess a canonical system name.
+#   Copyright 1992-2016 Free Software Foundation, Inc.
+
+timestamp='2016-10-02'
+
+# This file is free software; you can redistribute it and/or modify it
+# under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful, but
+# WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+# General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, see <http://www.gnu.org/licenses/>.
+#
+# As a special exception to the GNU General Public License, if you
+# distribute this file as part of a program that contains a
+# configuration script generated by Autoconf, you may include it under
+# the same distribution terms that you use for the rest of that
+# program.  This Exception is an additional permission under section 7
+# of the GNU General Public License, version 3 ("GPLv3").
+#
+# Originally written by Per Bothner; maintained since 2000 by Ben Elliston.
+#
+# You can get the latest version of this script from:
+# http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.guess
+#
+# Please send patches to <config-patches@gnu.org>.
+
+
+me=`echo "$0" | sed -e 's,.*/,,'`
+
+usage="\
+Usage: $0 [OPTION]
+
+Output the configuration name of the system \`$me' is run on.
+
+Operation modes:
+  -h, --help         print this help, then exit
+  -t, --time-stamp   print date of last modification, then exit
+  -v, --version      print version number, then exit
+
+Report bugs and patches to <config-patches@gnu.org>."
+
+version="\
+GNU config.guess ($timestamp)
+
+Originally written by Per Bothner.
+Copyright 1992-2016 Free Software Foundation, Inc.
+
+This is free software; see the source for copying conditions.  There is NO
+warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE."
+
+help="
+Try \`$me --help' for more information."
+
+# Parse command line
+while test $# -gt 0 ; do
+  case $1 in
+    --time-stamp | --time* | -t )
+       echo "$timestamp" ; exit ;;
+    --version | -v )
+       echo "$version" ; exit ;;
+    --help | --h* | -h )
+       echo "$usage"; exit ;;
+    -- )     # Stop option processing
+       shift; break ;;
+    - )	# Use stdin as input.
+       break ;;
+    -* )
+       echo "$me: invalid option $1$help" >&2
+       exit 1 ;;
+    * )
+       break ;;
+  esac
+done
+
+if test $# != 0; then
+  echo "$me: too many arguments$help" >&2
+  exit 1
+fi
+
+trap 'exit 1' 1 2 15
+
+# CC_FOR_BUILD -- compiler used by this script. Note that the use of a
+# compiler to aid in system detection is discouraged as it requires
+# temporary files to be created and, as you can see below, it is a
+# headache to deal with in a portable fashion.
+
+# Historically, `CC_FOR_BUILD' used to be named `HOST_CC'. We still
+# use `HOST_CC' if defined, but it is deprecated.
+
+# Portable tmp directory creation inspired by the Autoconf team.
+
+set_cc_for_build='
+trap "exitcode=\$?; (rm -f \$tmpfiles 2>/dev/null; rmdir \$tmp 2>/dev/null) && exit \$exitcode" 0 ;
+trap "rm -f \$tmpfiles 2>/dev/null; rmdir \$tmp 2>/dev/null; exit 1" 1 2 13 15 ;
+: ${TMPDIR=/tmp} ;
+ { tmp=`(umask 077 && mktemp -d "$TMPDIR/cgXXXXXX") 2>/dev/null` && test -n "$tmp" && test -d "$tmp" ; } ||
+ { test -n "$RANDOM" && tmp=$TMPDIR/cg$$-$RANDOM && (umask 077 && mkdir $tmp) ; } ||
+ { tmp=$TMPDIR/cg-$$ && (umask 077 && mkdir $tmp) && echo "Warning: creating insecure temp directory" >&2 ; } ||
+ { echo "$me: cannot create a temporary directory in $TMPDIR" >&2 ; exit 1 ; } ;
+dummy=$tmp/dummy ;
+tmpfiles="$dummy.c $dummy.o $dummy.rel $dummy" ;
+case $CC_FOR_BUILD,$HOST_CC,$CC in
+ ,,)    echo "int x;" > $dummy.c ;
+	for c in cc gcc c89 c99 ; do
+	  if ($c -c -o $dummy.o $dummy.c) >/dev/null 2>&1 ; then
+	     CC_FOR_BUILD="$c"; break ;
+	  fi ;
+	done ;
+	if test x"$CC_FOR_BUILD" = x ; then
+	  CC_FOR_BUILD=no_compiler_found ;
+	fi
+	;;
+ ,,*)   CC_FOR_BUILD=$CC ;;
+ ,*,*)  CC_FOR_BUILD=$HOST_CC ;;
+esac ; set_cc_for_build= ;'
+
+# This is needed to find uname on a Pyramid OSx when run in the BSD universe.
+# (ghazi@noc.rutgers.edu 1994-08-24)
+if (test -f /.attbin/uname) >/dev/null 2>&1 ; then
+	PATH=$PATH:/.attbin ; export PATH
+fi
+
+UNAME_MACHINE=`(uname -m) 2>/dev/null` || UNAME_MACHINE=unknown
+UNAME_RELEASE=`(uname -r) 2>/dev/null` || UNAME_RELEASE=unknown
+UNAME_SYSTEM=`(uname -s) 2>/dev/null`  || UNAME_SYSTEM=unknown
+UNAME_VERSION=`(uname -v) 2>/dev/null` || UNAME_VERSION=unknown
+
+case "${UNAME_SYSTEM}" in
+Linux|GNU|GNU/*)
+	# If the system lacks a compiler, then just pick glibc.
+	# We could probably try harder.
+	LIBC=gnu
+
+	eval $set_cc_for_build
+	cat <<-EOF > $dummy.c
+	#include <features.h>
+	#if defined(__UCLIBC__)
+	LIBC=uclibc
+	#elif defined(__dietlibc__)
+	LIBC=dietlibc
+	#else
+	LIBC=gnu
+	#endif
+	EOF
+	eval `$CC_FOR_BUILD -E $dummy.c 2>/dev/null | grep '^LIBC' | sed 's, ,,g'`
+	;;
+esac
+
+# Note: order is significant - the case branches are not exclusive.
+
+case "${UNAME_MACHINE}:${UNAME_SYSTEM}:${UNAME_RELEASE}:${UNAME_VERSION}" in
+    *:NetBSD:*:*)
+	# NetBSD (nbsd) targets should (where applicable) match one or
+	# more of the tuples: *-*-netbsdelf*, *-*-netbsdaout*,
+	# *-*-netbsdecoff* and *-*-netbsd*.  For targets that recently
+	# switched to ELF, *-*-netbsd* would select the old
+	# object file format.  This provides both forward
+	# compatibility and a consistent mechanism for selecting the
+	# object file format.
+	#
+	# Note: NetBSD doesn't particularly care about the vendor
+	# portion of the name.  We always set it to "unknown".
+	sysctl="sysctl -n hw.machine_arch"
+	UNAME_MACHINE_ARCH=`(uname -p 2>/dev/null || \
+	    /sbin/$sysctl 2>/dev/null || \
+	    /usr/sbin/$sysctl 2>/dev/null || \
+	    echo unknown)`
+	case "${UNAME_MACHINE_ARCH}" in
+	    armeb) machine=armeb-unknown ;;
+	    arm*) machine=arm-unknown ;;
+	    sh3el) machine=shl-unknown ;;
+	    sh3eb) machine=sh-unknown ;;
+	    sh5el) machine=sh5le-unknown ;;
+	    earmv*)
+		arch=`echo ${UNAME_MACHINE_ARCH} | sed -e 's,^e\(armv[0-9]\).*$,\1,'`
+		endian=`echo ${UNAME_MACHINE_ARCH} | sed -ne 's,^.*\(eb\)$,\1,p'`
+		machine=${arch}${endian}-unknown
+		;;
+	    *) machine=${UNAME_MACHINE_ARCH}-unknown ;;
+	esac
+	# The Operating System including object format, if it has switched
+	# to ELF recently (or will in the future) and ABI.
+	case "${UNAME_MACHINE_ARCH}" in
+	    earm*)
+		os=netbsdelf
+		;;
+	    arm*|i386|m68k|ns32k|sh3*|sparc|vax)
+		eval $set_cc_for_build
+		if echo __ELF__ | $CC_FOR_BUILD -E - 2>/dev/null \
+			| grep -q __ELF__
+		then
+		    # Once all utilities can be ECOFF (netbsdecoff) or a.out (netbsdaout).
+		    # Return netbsd for either.  FIX?
+		    os=netbsd
+		else
+		    os=netbsdelf
+		fi
+		;;
+	    *)
+		os=netbsd
+		;;
+	esac
+	# Determine ABI tags.
+	case "${UNAME_MACHINE_ARCH}" in
+	    earm*)
+		expr='s/^earmv[0-9]/-eabi/;s/eb$//'
+		abi=`echo ${UNAME_MACHINE_ARCH} | sed -e "$expr"`
+		;;
+	esac
+	# The OS release
+	# Debian GNU/NetBSD machines have a different userland, and
+	# thus, need a distinct triplet. However, they do not need
+	# kernel version information, so it can be replaced with a
+	# suitable tag, in the style of linux-gnu.
+	case "${UNAME_VERSION}" in
+	    Debian*)
+		release='-gnu'
+		;;
+	    *)
+		release=`echo ${UNAME_RELEASE} | sed -e 's/[-_].*//' | cut -d. -f1,2`
+		;;
+	esac
+	# Since CPU_TYPE-MANUFACTURER-KERNEL-OPERATING_SYSTEM:
+	# contains redundant information, the shorter form:
+	# CPU_TYPE-MANUFACTURER-OPERATING_SYSTEM is used.
+	echo "${machine}-${os}${release}${abi}"
+	exit ;;
+    *:Bitrig:*:*)
+	UNAME_MACHINE_ARCH=`arch | sed 's/Bitrig.//'`
+	echo ${UNAME_MACHINE_ARCH}-unknown-bitrig${UNAME_RELEASE}
+	exit ;;
+    *:OpenBSD:*:*)
+	UNAME_MACHINE_ARCH=`arch | sed 's/OpenBSD.//'`
+	echo ${UNAME_MACHINE_ARCH}-unknown-openbsd${UNAME_RELEASE}
+	exit ;;
+    *:LibertyBSD:*:*)
+	UNAME_MACHINE_ARCH=`arch | sed 's/^.*BSD\.//'`
+	echo ${UNAME_MACHINE_ARCH}-unknown-libertybsd${UNAME_RELEASE}
+	exit ;;
+    *:ekkoBSD:*:*)
+	echo ${UNAME_MACHINE}-unknown-ekkobsd${UNAME_RELEASE}
+	exit ;;
+    *:SolidBSD:*:*)
+	echo ${UNAME_MACHINE}-unknown-solidbsd${UNAME_RELEASE}
+	exit ;;
+    macppc:MirBSD:*:*)
+	echo powerpc-unknown-mirbsd${UNAME_RELEASE}
+	exit ;;
+    *:MirBSD:*:*)
+	echo ${UNAME_MACHINE}-unknown-mirbsd${UNAME_RELEASE}
+	exit ;;
+    *:Sortix:*:*)
+	echo ${UNAME_MACHINE}-unknown-sortix
+	exit ;;
+    alpha:OSF1:*:*)
+	case $UNAME_RELEASE in
+	*4.0)
+		UNAME_RELEASE=`/usr/sbin/sizer -v | awk '{print $3}'`
+		;;
+	*5.*)
+		UNAME_RELEASE=`/usr/sbin/sizer -v | awk '{print $4}'`
+		;;
+	esac
+	# According to Compaq, /usr/sbin/psrinfo has been available on
+	# OSF/1 and Tru64 systems produced since 1995.  I hope that
+	# covers most systems running today.  This code pipes the CPU
+	# types through head -n 1, so we only detect the type of CPU 0.
+	ALPHA_CPU_TYPE=`/usr/sbin/psrinfo -v | sed -n -e 's/^  The alpha \(.*\) processor.*$/\1/p' | head -n 1`
+	case "$ALPHA_CPU_TYPE" in
+	    "EV4 (21064)")
+		UNAME_MACHINE=alpha ;;
+	    "EV4.5 (21064)")
+		UNAME_MACHINE=alpha ;;
+	    "LCA4 (21066/21068)")
+		UNAME_MACHINE=alpha ;;
+	    "EV5 (21164)")
+		UNAME_MACHINE=alphaev5 ;;
+	    "EV5.6 (21164A)")
+		UNAME_MACHINE=alphaev56 ;;
+	    "EV5.6 (21164PC)")
+		UNAME_MACHINE=alphapca56 ;;
+	    "EV5.7 (21164PC)")
+		UNAME_MACHINE=alphapca57 ;;
+	    "EV6 (21264)")
+		UNAME_MACHINE=alphaev6 ;;
+	    "EV6.7 (21264A)")
+		UNAME_MACHINE=alphaev67 ;;
+	    "EV6.8CB (21264C)")
+		UNAME_MACHINE=alphaev68 ;;
+	    "EV6.8AL (21264B)")
+		UNAME_MACHINE=alphaev68 ;;
+	    "EV6.8CX (21264D)")
+		UNAME_MACHINE=alphaev68 ;;
+	    "EV6.9A (21264/EV69A)")
+		UNAME_MACHINE=alphaev69 ;;
+	    "EV7 (21364)")
+		UNAME_MACHINE=alphaev7 ;;
+	    "EV7.9 (21364A)")
+		UNAME_MACHINE=alphaev79 ;;
+	esac
+	# A Pn.n version is a patched version.
+	# A Vn.n version is a released version.
+	# A Tn.n version is a released field test version.
+	# A Xn.n version is an unreleased experimental baselevel.
+	# 1.2 uses "1.2" for uname -r.
+	echo ${UNAME_MACHINE}-dec-osf`echo ${UNAME_RELEASE} | sed -e 's/^[PVTX]//' | tr ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz`
+	# Reset EXIT trap before exiting to avoid spurious non-zero exit code.
+	exitcode=$?
+	trap '' 0
+	exit $exitcode ;;
+    Alpha\ *:Windows_NT*:*)
+	# How do we know it's Interix rather than the generic POSIX subsystem?
+	# Should we change UNAME_MACHINE based on the output of uname instead
+	# of the specific Alpha model?
+	echo alpha-pc-interix
+	exit ;;
+    21064:Windows_NT:50:3)
+	echo alpha-dec-winnt3.5
+	exit ;;
+    Amiga*:UNIX_System_V:4.0:*)
+	echo m68k-unknown-sysv4
+	exit ;;
+    *:[Aa]miga[Oo][Ss]:*:*)
+	echo ${UNAME_MACHINE}-unknown-amigaos
+	exit ;;
+    *:[Mm]orph[Oo][Ss]:*:*)
+	echo ${UNAME_MACHINE}-unknown-morphos
+	exit ;;
+    *:OS/390:*:*)
+	echo i370-ibm-openedition
+	exit ;;
+    *:z/VM:*:*)
+	echo s390-ibm-zvmoe
+	exit ;;
+    *:OS400:*:*)
+	echo powerpc-ibm-os400
+	exit ;;
+    arm:RISC*:1.[012]*:*|arm:riscix:1.[012]*:*)
+	echo arm-acorn-riscix${UNAME_RELEASE}
+	exit ;;
+    arm*:riscos:*:*|arm*:RISCOS:*:*)
+	echo arm-unknown-riscos
+	exit ;;
+    SR2?01:HI-UX/MPP:*:* | SR8000:HI-UX/MPP:*:*)
+	echo hppa1.1-hitachi-hiuxmpp
+	exit ;;
+    Pyramid*:OSx*:*:* | MIS*:OSx*:*:* | MIS*:SMP_DC-OSx*:*:*)
+	# akee@wpdis03.wpafb.af.mil (Earle F. Ake) contributed MIS and NILE.
+	if test "`(/bin/universe) 2>/dev/null`" = att ; then
+		echo pyramid-pyramid-sysv3
+	else
+		echo pyramid-pyramid-bsd
+	fi
+	exit ;;
+    NILE*:*:*:dcosx)
+	echo pyramid-pyramid-svr4
+	exit ;;
+    DRS?6000:unix:4.0:6*)
+	echo sparc-icl-nx6
+	exit ;;
+    DRS?6000:UNIX_SV:4.2*:7* | DRS?6000:isis:4.2*:7*)
+	case `/usr/bin/uname -p` in
+	    sparc) echo sparc-icl-nx7; exit ;;
+	esac ;;
+    s390x:SunOS:*:*)
+	echo ${UNAME_MACHINE}-ibm-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'`
+	exit ;;
+    sun4H:SunOS:5.*:*)
+	echo sparc-hal-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'`
+	exit ;;
+    sun4*:SunOS:5.*:* | tadpole*:SunOS:5.*:*)
+	echo sparc-sun-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'`
+	exit ;;
+    i86pc:AuroraUX:5.*:* | i86xen:AuroraUX:5.*:*)
+	echo i386-pc-auroraux${UNAME_RELEASE}
+	exit ;;
+    i86pc:SunOS:5.*:* | i86xen:SunOS:5.*:*)
+	eval $set_cc_for_build
+	SUN_ARCH=i386
+	# If there is a compiler, see if it is configured for 64-bit objects.
+	# Note that the Sun cc does not turn __LP64__ into 1 like gcc does.
+	# This test works for both compilers.
+	if [ "$CC_FOR_BUILD" != no_compiler_found ]; then
+	    if (echo '#ifdef __amd64'; echo IS_64BIT_ARCH; echo '#endif') | \
+		(CCOPTS="" $CC_FOR_BUILD -E - 2>/dev/null) | \
+		grep IS_64BIT_ARCH >/dev/null
+	    then
+		SUN_ARCH=x86_64
+	    fi
+	fi
+	echo ${SUN_ARCH}-pc-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'`
+	exit ;;
+    sun4*:SunOS:6*:*)
+	# According to config.sub, this is the proper way to canonicalize
+	# SunOS6.  Hard to guess exactly what SunOS6 will be like, but
+	# it's likely to be more like Solaris than SunOS4.
+	echo sparc-sun-solaris3`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'`
+	exit ;;
+    sun4*:SunOS:*:*)
+	case "`/usr/bin/arch -k`" in
+	    Series*|S4*)
+		UNAME_RELEASE=`uname -v`
+		;;
+	esac
+	# Japanese Language versions have a version number like `4.1.3-JL'.
+	echo sparc-sun-sunos`echo ${UNAME_RELEASE}|sed -e 's/-/_/'`
+	exit ;;
+    sun3*:SunOS:*:*)
+	echo m68k-sun-sunos${UNAME_RELEASE}
+	exit ;;
+    sun*:*:4.2BSD:*)
+	UNAME_RELEASE=`(sed 1q /etc/motd | awk '{print substr($5,1,3)}') 2>/dev/null`
+	test "x${UNAME_RELEASE}" = x && UNAME_RELEASE=3
+	case "`/bin/arch`" in
+	    sun3)
+		echo m68k-sun-sunos${UNAME_RELEASE}
+		;;
+	    sun4)
+		echo sparc-sun-sunos${UNAME_RELEASE}
+		;;
+	esac
+	exit ;;
+    aushp:SunOS:*:*)
+	echo sparc-auspex-sunos${UNAME_RELEASE}
+	exit ;;
+    # The situation for MiNT is a little confusing.  The machine name
+    # can be virtually everything (everything which is not
+    # "atarist" or "atariste" at least should have a processor
+    # > m68000).  The system name ranges from "MiNT" over "FreeMiNT"
+    # to the lowercase version "mint" (or "freemint").  Finally
+    # the system name "TOS" denotes a system which is actually not
+    # MiNT.  But MiNT is downward compatible to TOS, so this should
+    # be no problem.
+    atarist[e]:*MiNT:*:* | atarist[e]:*mint:*:* | atarist[e]:*TOS:*:*)
+	echo m68k-atari-mint${UNAME_RELEASE}
+	exit ;;
+    atari*:*MiNT:*:* | atari*:*mint:*:* | atarist[e]:*TOS:*:*)
+	echo m68k-atari-mint${UNAME_RELEASE}
+	exit ;;
+    *falcon*:*MiNT:*:* | *falcon*:*mint:*:* | *falcon*:*TOS:*:*)
+	echo m68k-atari-mint${UNAME_RELEASE}
+	exit ;;
+    milan*:*MiNT:*:* | milan*:*mint:*:* | *milan*:*TOS:*:*)
+	echo m68k-milan-mint${UNAME_RELEASE}
+	exit ;;
+    hades*:*MiNT:*:* | hades*:*mint:*:* | *hades*:*TOS:*:*)
+	echo m68k-hades-mint${UNAME_RELEASE}
+	exit ;;
+    *:*MiNT:*:* | *:*mint:*:* | *:*TOS:*:*)
+	echo m68k-unknown-mint${UNAME_RELEASE}
+	exit ;;
+    m68k:machten:*:*)
+	echo m68k-apple-machten${UNAME_RELEASE}
+	exit ;;
+    powerpc:machten:*:*)
+	echo powerpc-apple-machten${UNAME_RELEASE}
+	exit ;;
+    RISC*:Mach:*:*)
+	echo mips-dec-mach_bsd4.3
+	exit ;;
+    RISC*:ULTRIX:*:*)
+	echo mips-dec-ultrix${UNAME_RELEASE}
+	exit ;;
+    VAX*:ULTRIX*:*:*)
+	echo vax-dec-ultrix${UNAME_RELEASE}
+	exit ;;
+    2020:CLIX:*:* | 2430:CLIX:*:*)
+	echo clipper-intergraph-clix${UNAME_RELEASE}
+	exit ;;
+    mips:*:*:UMIPS | mips:*:*:RISCos)
+	eval $set_cc_for_build
+	sed 's/^	//' << EOF >$dummy.c
+#ifdef __cplusplus
+#include <stdio.h>  /* for printf() prototype */
+	int main (int argc, char *argv[]) {
+#else
+	int main (argc, argv) int argc; char *argv[]; {
+#endif
+	#if defined (host_mips) && defined (MIPSEB)
+	#if defined (SYSTYPE_SYSV)
+	  printf ("mips-mips-riscos%ssysv\n", argv[1]); exit (0);
+	#endif
+	#if defined (SYSTYPE_SVR4)
+	  printf ("mips-mips-riscos%ssvr4\n", argv[1]); exit (0);
+	#endif
+	#if defined (SYSTYPE_BSD43) || defined(SYSTYPE_BSD)
+	  printf ("mips-mips-riscos%sbsd\n", argv[1]); exit (0);
+	#endif
+	#endif
+	  exit (-1);
+	}
+EOF
+	$CC_FOR_BUILD -o $dummy $dummy.c &&
+	  dummyarg=`echo "${UNAME_RELEASE}" | sed -n 's/\([0-9]*\).*/\1/p'` &&
+	  SYSTEM_NAME=`$dummy $dummyarg` &&
+	    { echo "$SYSTEM_NAME"; exit; }
+	echo mips-mips-riscos${UNAME_RELEASE}
+	exit ;;
+    Motorola:PowerMAX_OS:*:*)
+	echo powerpc-motorola-powermax
+	exit ;;
+    Motorola:*:4.3:PL8-*)
+	echo powerpc-harris-powermax
+	exit ;;
+    Night_Hawk:*:*:PowerMAX_OS | Synergy:PowerMAX_OS:*:*)
+	echo powerpc-harris-powermax
+	exit ;;
+    Night_Hawk:Power_UNIX:*:*)
+	echo powerpc-harris-powerunix
+	exit ;;
+    m88k:CX/UX:7*:*)
+	echo m88k-harris-cxux7
+	exit ;;
+    m88k:*:4*:R4*)
+	echo m88k-motorola-sysv4
+	exit ;;
+    m88k:*:3*:R3*)
+	echo m88k-motorola-sysv3
+	exit ;;
+    AViiON:dgux:*:*)
+	# DG/UX returns AViiON for all architectures
+	UNAME_PROCESSOR=`/usr/bin/uname -p`
+	if [ $UNAME_PROCESSOR = mc88100 ] || [ $UNAME_PROCESSOR = mc88110 ]
+	then
+	    if [ ${TARGET_BINARY_INTERFACE}x = m88kdguxelfx ] || \
+	       [ ${TARGET_BINARY_INTERFACE}x = x ]
+	    then
+		echo m88k-dg-dgux${UNAME_RELEASE}
+	    else
+		echo m88k-dg-dguxbcs${UNAME_RELEASE}
+	    fi
+	else
+	    echo i586-dg-dgux${UNAME_RELEASE}
+	fi
+	exit ;;
+    M88*:DolphinOS:*:*)	# DolphinOS (SVR3)
+	echo m88k-dolphin-sysv3
+	exit ;;
+    M88*:*:R3*:*)
+	# Delta 88k system running SVR3
+	echo m88k-motorola-sysv3
+	exit ;;
+    XD88*:*:*:*) # Tektronix XD88 system running UTekV (SVR3)
+	echo m88k-tektronix-sysv3
+	exit ;;
+    Tek43[0-9][0-9]:UTek:*:*) # Tektronix 4300 system running UTek (BSD)
+	echo m68k-tektronix-bsd
+	exit ;;
+    *:IRIX*:*:*)
+	echo mips-sgi-irix`echo ${UNAME_RELEASE}|sed -e 's/-/_/g'`
+	exit ;;
+    ????????:AIX?:[12].1:2)   # AIX 2.2.1 or AIX 2.1.1 is RT/PC AIX.
+	echo romp-ibm-aix     # uname -m gives an 8 hex-code CPU id
+	exit ;;               # Note that: echo "'`uname -s`'" gives 'AIX '
+    i*86:AIX:*:*)
+	echo i386-ibm-aix
+	exit ;;
+    ia64:AIX:*:*)
+	if [ -x /usr/bin/oslevel ] ; then
+		IBM_REV=`/usr/bin/oslevel`
+	else
+		IBM_REV=${UNAME_VERSION}.${UNAME_RELEASE}
+	fi
+	echo ${UNAME_MACHINE}-ibm-aix${IBM_REV}
+	exit ;;
+    *:AIX:2:3)
+	if grep bos325 /usr/include/stdio.h >/dev/null 2>&1; then
+		eval $set_cc_for_build
+		sed 's/^		//' << EOF >$dummy.c
+		#include <sys/systemcfg.h>
+
+		main()
+			{
+			if (!__power_pc())
+				exit(1);
+			puts("powerpc-ibm-aix3.2.5");
+			exit(0);
+			}
+EOF
+		if $CC_FOR_BUILD -o $dummy $dummy.c && SYSTEM_NAME=`$dummy`
+		then
+			echo "$SYSTEM_NAME"
+		else
+			echo rs6000-ibm-aix3.2.5
+		fi
+	elif grep bos324 /usr/include/stdio.h >/dev/null 2>&1; then
+		echo rs6000-ibm-aix3.2.4
+	else
+		echo rs6000-ibm-aix3.2
+	fi
+	exit ;;
+    *:AIX:*:[4567])
+	IBM_CPU_ID=`/usr/sbin/lsdev -C -c processor -S available | sed 1q | awk '{ print $1 }'`
+	if /usr/sbin/lsattr -El ${IBM_CPU_ID} | grep ' POWER' >/dev/null 2>&1; then
+		IBM_ARCH=rs6000
+	else
+		IBM_ARCH=powerpc
+	fi
+	if [ -x /usr/bin/lslpp ] ; then
+		IBM_REV=`/usr/bin/lslpp -Lqc bos.rte.libc |
+			   awk -F: '{ print $3 }' | sed s/[0-9]*$/0/`
+	else
+		IBM_REV=${UNAME_VERSION}.${UNAME_RELEASE}
+	fi
+	echo ${IBM_ARCH}-ibm-aix${IBM_REV}
+	exit ;;
+    *:AIX:*:*)
+	echo rs6000-ibm-aix
+	exit ;;
+    ibmrt:4.4BSD:*|romp-ibm:BSD:*)
+	echo romp-ibm-bsd4.4
+	exit ;;
+    ibmrt:*BSD:*|romp-ibm:BSD:*)            # covers RT/PC BSD and
+	echo romp-ibm-bsd${UNAME_RELEASE}   # 4.3 with uname added to
+	exit ;;                             # report: romp-ibm BSD 4.3
+    *:BOSX:*:*)
+	echo rs6000-bull-bosx
+	exit ;;
+    DPX/2?00:B.O.S.:*:*)
+	echo m68k-bull-sysv3
+	exit ;;
+    9000/[34]??:4.3bsd:1.*:*)
+	echo m68k-hp-bsd
+	exit ;;
+    hp300:4.4BSD:*:* | 9000/[34]??:4.3bsd:2.*:*)
+	echo m68k-hp-bsd4.4
+	exit ;;
+    9000/[34678]??:HP-UX:*:*)
+	HPUX_REV=`echo ${UNAME_RELEASE}|sed -e 's/[^.]*.[0B]*//'`
+	case "${UNAME_MACHINE}" in
+	    9000/31? )            HP_ARCH=m68000 ;;
+	    9000/[34]?? )         HP_ARCH=m68k ;;
+	    9000/[678][0-9][0-9])
+		if [ -x /usr/bin/getconf ]; then
+		    sc_cpu_version=`/usr/bin/getconf SC_CPU_VERSION 2>/dev/null`
+		    sc_kernel_bits=`/usr/bin/getconf SC_KERNEL_BITS 2>/dev/null`
+		    case "${sc_cpu_version}" in
+		      523) HP_ARCH=hppa1.0 ;; # CPU_PA_RISC1_0
+		      528) HP_ARCH=hppa1.1 ;; # CPU_PA_RISC1_1
+		      532)                      # CPU_PA_RISC2_0
+			case "${sc_kernel_bits}" in
+			  32) HP_ARCH=hppa2.0n ;;
+			  64) HP_ARCH=hppa2.0w ;;
+			  '') HP_ARCH=hppa2.0 ;;   # HP-UX 10.20
+			esac ;;
+		    esac
+		fi
+		if [ "${HP_ARCH}" = "" ]; then
+		    eval $set_cc_for_build
+		    sed 's/^		//' << EOF >$dummy.c
+
+		#define _HPUX_SOURCE
+		#include <stdlib.h>
+		#include <unistd.h>
+
+		int main ()
+		{
+		#if defined(_SC_KERNEL_BITS)
+		    long bits = sysconf(_SC_KERNEL_BITS);
+		#endif
+		    long cpu  = sysconf (_SC_CPU_VERSION);
+
+		    switch (cpu)
+			{
+			case CPU_PA_RISC1_0: puts ("hppa1.0"); break;
+			case CPU_PA_RISC1_1: puts ("hppa1.1"); break;
+			case CPU_PA_RISC2_0:
+		#if defined(_SC_KERNEL_BITS)
+			    switch (bits)
+				{
+				case 64: puts ("hppa2.0w"); break;
+				case 32: puts ("hppa2.0n"); break;
+				default: puts ("hppa2.0"); break;
+				} break;
+		#else  /* !defined(_SC_KERNEL_BITS) */
+			    puts ("hppa2.0"); break;
+		#endif
+			default: puts ("hppa1.0"); break;
+			}
+		    exit (0);
+		}
+EOF
+		    (CCOPTS="" $CC_FOR_BUILD -o $dummy $dummy.c 2>/dev/null) && HP_ARCH=`$dummy`
+		    test -z "$HP_ARCH" && HP_ARCH=hppa
+		fi ;;
+	esac
+	if [ ${HP_ARCH} = hppa2.0w ]
+	then
+	    eval $set_cc_for_build
+
+	    # hppa2.0w-hp-hpux* has a 64-bit kernel and a compiler generating
+	    # 32-bit code.  hppa64-hp-hpux* has the same kernel and a compiler
+	    # generating 64-bit code.  GNU and HP use different nomenclature:
+	    #
+	    # $ CC_FOR_BUILD=cc ./config.guess
+	    # => hppa2.0w-hp-hpux11.23
+	    # $ CC_FOR_BUILD="cc +DA2.0w" ./config.guess
+	    # => hppa64-hp-hpux11.23
+
+	    if echo __LP64__ | (CCOPTS="" $CC_FOR_BUILD -E - 2>/dev/null) |
+		grep -q __LP64__
+	    then
+		HP_ARCH=hppa2.0w
+	    else
+		HP_ARCH=hppa64
+	    fi
+	fi
+	echo ${HP_ARCH}-hp-hpux${HPUX_REV}
+	exit ;;
+    ia64:HP-UX:*:*)
+	HPUX_REV=`echo ${UNAME_RELEASE}|sed -e 's/[^.]*.[0B]*//'`
+	echo ia64-hp-hpux${HPUX_REV}
+	exit ;;
+    3050*:HI-UX:*:*)
+	eval $set_cc_for_build
+	sed 's/^	//' << EOF >$dummy.c
+	#include <unistd.h>
+	int
+	main ()
+	{
+	  long cpu = sysconf (_SC_CPU_VERSION);
+	  /* The order matters, because CPU_IS_HP_MC68K erroneously returns
+	     true for CPU_PA_RISC1_0.  CPU_IS_PA_RISC returns correct
+	     results, however.  */
+	  if (CPU_IS_PA_RISC (cpu))
+	    {
+	      switch (cpu)
+		{
+		  case CPU_PA_RISC1_0: puts ("hppa1.0-hitachi-hiuxwe2"); break;
+		  case CPU_PA_RISC1_1: puts ("hppa1.1-hitachi-hiuxwe2"); break;
+		  case CPU_PA_RISC2_0: puts ("hppa2.0-hitachi-hiuxwe2"); break;
+		  default: puts ("hppa-hitachi-hiuxwe2"); break;
+		}
+	    }
+	  else if (CPU_IS_HP_MC68K (cpu))
+	    puts ("m68k-hitachi-hiuxwe2");
+	  else puts ("unknown-hitachi-hiuxwe2");
+	  exit (0);
+	}
+EOF
+	$CC_FOR_BUILD -o $dummy $dummy.c && SYSTEM_NAME=`$dummy` &&
+		{ echo "$SYSTEM_NAME"; exit; }
+	echo unknown-hitachi-hiuxwe2
+	exit ;;
+    9000/7??:4.3bsd:*:* | 9000/8?[79]:4.3bsd:*:* )
+	echo hppa1.1-hp-bsd
+	exit ;;
+    9000/8??:4.3bsd:*:*)
+	echo hppa1.0-hp-bsd
+	exit ;;
+    *9??*:MPE/iX:*:* | *3000*:MPE/iX:*:*)
+	echo hppa1.0-hp-mpeix
+	exit ;;
+    hp7??:OSF1:*:* | hp8?[79]:OSF1:*:* )
+	echo hppa1.1-hp-osf
+	exit ;;
+    hp8??:OSF1:*:*)
+	echo hppa1.0-hp-osf
+	exit ;;
+    i*86:OSF1:*:*)
+	if [ -x /usr/sbin/sysversion ] ; then
+	    echo ${UNAME_MACHINE}-unknown-osf1mk
+	else
+	    echo ${UNAME_MACHINE}-unknown-osf1
+	fi
+	exit ;;
+    parisc*:Lites*:*:*)
+	echo hppa1.1-hp-lites
+	exit ;;
+    C1*:ConvexOS:*:* | convex:ConvexOS:C1*:*)
+	echo c1-convex-bsd
+	exit ;;
+    C2*:ConvexOS:*:* | convex:ConvexOS:C2*:*)
+	if getsysinfo -f scalar_acc
+	then echo c32-convex-bsd
+	else echo c2-convex-bsd
+	fi
+	exit ;;
+    C34*:ConvexOS:*:* | convex:ConvexOS:C34*:*)
+	echo c34-convex-bsd
+	exit ;;
+    C38*:ConvexOS:*:* | convex:ConvexOS:C38*:*)
+	echo c38-convex-bsd
+	exit ;;
+    C4*:ConvexOS:*:* | convex:ConvexOS:C4*:*)
+	echo c4-convex-bsd
+	exit ;;
+    CRAY*Y-MP:*:*:*)
+	echo ymp-cray-unicos${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/'
+	exit ;;
+    CRAY*[A-Z]90:*:*:*)
+	echo ${UNAME_MACHINE}-cray-unicos${UNAME_RELEASE} \
+	| sed -e 's/CRAY.*\([A-Z]90\)/\1/' \
+	      -e y/ABCDEFGHIJKLMNOPQRSTUVWXYZ/abcdefghijklmnopqrstuvwxyz/ \
+	      -e 's/\.[^.]*$/.X/'
+	exit ;;
+    CRAY*TS:*:*:*)
+	echo t90-cray-unicos${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/'
+	exit ;;
+    CRAY*T3E:*:*:*)
+	echo alphaev5-cray-unicosmk${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/'
+	exit ;;
+    CRAY*SV1:*:*:*)
+	echo sv1-cray-unicos${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/'
+	exit ;;
+    *:UNICOS/mp:*:*)
+	echo craynv-cray-unicosmp${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/'
+	exit ;;
+    F30[01]:UNIX_System_V:*:* | F700:UNIX_System_V:*:*)
+	FUJITSU_PROC=`uname -m | tr ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz`
+	FUJITSU_SYS=`uname -p | tr ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz | sed -e 's/\///'`
+	FUJITSU_REL=`echo ${UNAME_RELEASE} | sed -e 's/ /_/'`
+	echo "${FUJITSU_PROC}-fujitsu-${FUJITSU_SYS}${FUJITSU_REL}"
+	exit ;;
+    5000:UNIX_System_V:4.*:*)
+	FUJITSU_SYS=`uname -p | tr ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz | sed -e 's/\///'`
+	FUJITSU_REL=`echo ${UNAME_RELEASE} | tr ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz | sed -e 's/ /_/'`
+	echo "sparc-fujitsu-${FUJITSU_SYS}${FUJITSU_REL}"
+	exit ;;
+    i*86:BSD/386:*:* | i*86:BSD/OS:*:* | *:Ascend\ Embedded/OS:*:*)
+	echo ${UNAME_MACHINE}-pc-bsdi${UNAME_RELEASE}
+	exit ;;
+    sparc*:BSD/OS:*:*)
+	echo sparc-unknown-bsdi${UNAME_RELEASE}
+	exit ;;
+    *:BSD/OS:*:*)
+	echo ${UNAME_MACHINE}-unknown-bsdi${UNAME_RELEASE}
+	exit ;;
+    *:FreeBSD:*:*)
+	UNAME_PROCESSOR=`/usr/bin/uname -p`
+	case ${UNAME_PROCESSOR} in
+	    amd64)
+		echo x86_64-unknown-freebsd`echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'` ;;
+	    *)
+		echo ${UNAME_PROCESSOR}-unknown-freebsd`echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'` ;;
+	esac
+	exit ;;
+    i*:CYGWIN*:*)
+	echo ${UNAME_MACHINE}-pc-cygwin
+	exit ;;
+    *:MINGW64*:*)
+	echo ${UNAME_MACHINE}-pc-mingw64
+	exit ;;
+    *:MINGW*:*)
+	echo ${UNAME_MACHINE}-pc-mingw32
+	exit ;;
+    *:MSYS*:*)
+	echo ${UNAME_MACHINE}-pc-msys
+	exit ;;
+    i*:windows32*:*)
+	# uname -m includes "-pc" on this system.
+	echo ${UNAME_MACHINE}-mingw32
+	exit ;;
+    i*:PW*:*)
+	echo ${UNAME_MACHINE}-pc-pw32
+	exit ;;
+    *:Interix*:*)
+	case ${UNAME_MACHINE} in
+	    x86)
+		echo i586-pc-interix${UNAME_RELEASE}
+		exit ;;
+	    authenticamd | genuineintel | EM64T)
+		echo x86_64-unknown-interix${UNAME_RELEASE}
+		exit ;;
+	    IA64)
+		echo ia64-unknown-interix${UNAME_RELEASE}
+		exit ;;
+	esac ;;
+    [345]86:Windows_95:* | [345]86:Windows_98:* | [345]86:Windows_NT:*)
+	echo i${UNAME_MACHINE}-pc-mks
+	exit ;;
+    8664:Windows_NT:*)
+	echo x86_64-pc-mks
+	exit ;;
+    i*:Windows_NT*:* | Pentium*:Windows_NT*:*)
+	# How do we know it's Interix rather than the generic POSIX subsystem?
+	# It also conflicts with pre-2.0 versions of AT&T UWIN. Should we
+	# UNAME_MACHINE based on the output of uname instead of i386?
+	echo i586-pc-interix
+	exit ;;
+    i*:UWIN*:*)
+	echo ${UNAME_MACHINE}-pc-uwin
+	exit ;;
+    amd64:CYGWIN*:*:* | x86_64:CYGWIN*:*:*)
+	echo x86_64-unknown-cygwin
+	exit ;;
+    p*:CYGWIN*:*)
+	echo powerpcle-unknown-cygwin
+	exit ;;
+    prep*:SunOS:5.*:*)
+	echo powerpcle-unknown-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'`
+	exit ;;
+    *:GNU:*:*)
+	# the GNU system
+	echo `echo ${UNAME_MACHINE}|sed -e 's,[-/].*$,,'`-unknown-${LIBC}`echo ${UNAME_RELEASE}|sed -e 's,/.*$,,'`
+	exit ;;
+    *:GNU/*:*:*)
+	# other systems with GNU libc and userland
+	echo ${UNAME_MACHINE}-unknown-`echo ${UNAME_SYSTEM} | sed 's,^[^/]*/,,' | tr "[:upper:]" "[:lower:]"``echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'`-${LIBC}
+	exit ;;
+    i*86:Minix:*:*)
+	echo ${UNAME_MACHINE}-pc-minix
+	exit ;;
+    aarch64:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    aarch64_be:Linux:*:*)
+	UNAME_MACHINE=aarch64_be
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    alpha:Linux:*:*)
+	case `sed -n '/^cpu model/s/^.*: \(.*\)/\1/p' < /proc/cpuinfo` in
+	  EV5)   UNAME_MACHINE=alphaev5 ;;
+	  EV56)  UNAME_MACHINE=alphaev56 ;;
+	  PCA56) UNAME_MACHINE=alphapca56 ;;
+	  PCA57) UNAME_MACHINE=alphapca56 ;;
+	  EV6)   UNAME_MACHINE=alphaev6 ;;
+	  EV67)  UNAME_MACHINE=alphaev67 ;;
+	  EV68*) UNAME_MACHINE=alphaev68 ;;
+	esac
+	objdump --private-headers /bin/sh | grep -q ld.so.1
+	if test "$?" = 0 ; then LIBC=gnulibc1 ; fi
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    arc:Linux:*:* | arceb:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    arm*:Linux:*:*)
+	eval $set_cc_for_build
+	if echo __ARM_EABI__ | $CC_FOR_BUILD -E - 2>/dev/null \
+	    | grep -q __ARM_EABI__
+	then
+	    echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	else
+	    if echo __ARM_PCS_VFP | $CC_FOR_BUILD -E - 2>/dev/null \
+		| grep -q __ARM_PCS_VFP
+	    then
+		echo ${UNAME_MACHINE}-unknown-linux-${LIBC}eabi
+	    else
+		echo ${UNAME_MACHINE}-unknown-linux-${LIBC}eabihf
+	    fi
+	fi
+	exit ;;
+    avr32*:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    cris:Linux:*:*)
+	echo ${UNAME_MACHINE}-axis-linux-${LIBC}
+	exit ;;
+    crisv32:Linux:*:*)
+	echo ${UNAME_MACHINE}-axis-linux-${LIBC}
+	exit ;;
+    e2k:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    frv:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    hexagon:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    i*86:Linux:*:*)
+	echo ${UNAME_MACHINE}-pc-linux-${LIBC}
+	exit ;;
+    ia64:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    k1om:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    m32r*:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    m68*:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    mips:Linux:*:* | mips64:Linux:*:*)
+	eval $set_cc_for_build
+	sed 's/^	//' << EOF >$dummy.c
+	#undef CPU
+	#undef ${UNAME_MACHINE}
+	#undef ${UNAME_MACHINE}el
+	#if defined(__MIPSEL__) || defined(__MIPSEL) || defined(_MIPSEL) || defined(MIPSEL)
+	CPU=${UNAME_MACHINE}el
+	#else
+	#if defined(__MIPSEB__) || defined(__MIPSEB) || defined(_MIPSEB) || defined(MIPSEB)
+	CPU=${UNAME_MACHINE}
+	#else
+	CPU=
+	#endif
+	#endif
+EOF
+	eval `$CC_FOR_BUILD -E $dummy.c 2>/dev/null | grep '^CPU'`
+	test x"${CPU}" != x && { echo "${CPU}-unknown-linux-${LIBC}"; exit; }
+	;;
+    mips64el:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    openrisc*:Linux:*:*)
+	echo or1k-unknown-linux-${LIBC}
+	exit ;;
+    or32:Linux:*:* | or1k*:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    padre:Linux:*:*)
+	echo sparc-unknown-linux-${LIBC}
+	exit ;;
+    parisc64:Linux:*:* | hppa64:Linux:*:*)
+	echo hppa64-unknown-linux-${LIBC}
+	exit ;;
+    parisc:Linux:*:* | hppa:Linux:*:*)
+	# Look for CPU level
+	case `grep '^cpu[^a-z]*:' /proc/cpuinfo 2>/dev/null | cut -d' ' -f2` in
+	  PA7*) echo hppa1.1-unknown-linux-${LIBC} ;;
+	  PA8*) echo hppa2.0-unknown-linux-${LIBC} ;;
+	  *)    echo hppa-unknown-linux-${LIBC} ;;
+	esac
+	exit ;;
+    ppc64:Linux:*:*)
+	echo powerpc64-unknown-linux-${LIBC}
+	exit ;;
+    ppc:Linux:*:*)
+	echo powerpc-unknown-linux-${LIBC}
+	exit ;;
+    ppc64le:Linux:*:*)
+	echo powerpc64le-unknown-linux-${LIBC}
+	exit ;;
+    ppcle:Linux:*:*)
+	echo powerpcle-unknown-linux-${LIBC}
+	exit ;;
+    riscv32:Linux:*:* | riscv64:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    s390:Linux:*:* | s390x:Linux:*:*)
+	echo ${UNAME_MACHINE}-ibm-linux-${LIBC}
+	exit ;;
+    sh64*:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    sh*:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    sparc:Linux:*:* | sparc64:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    tile*:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    vax:Linux:*:*)
+	echo ${UNAME_MACHINE}-dec-linux-${LIBC}
+	exit ;;
+    x86_64:Linux:*:*)
+	echo ${UNAME_MACHINE}-pc-linux-${LIBC}
+	exit ;;
+    xtensa*:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    i*86:DYNIX/ptx:4*:*)
+	# ptx 4.0 does uname -s correctly, with DYNIX/ptx in there.
+	# earlier versions are messed up and put the nodename in both
+	# sysname and nodename.
+	echo i386-sequent-sysv4
+	exit ;;
+    i*86:UNIX_SV:4.2MP:2.*)
+	# Unixware is an offshoot of SVR4, but it has its own version
+	# number series starting with 2...
+	# I am not positive that other SVR4 systems won't match this,
+	# I just have to hope.  -- rms.
+	# Use sysv4.2uw... so that sysv4* matches it.
+	echo ${UNAME_MACHINE}-pc-sysv4.2uw${UNAME_VERSION}
+	exit ;;
+    i*86:OS/2:*:*)
+	# If we were able to find `uname', then EMX Unix compatibility
+	# is probably installed.
+	echo ${UNAME_MACHINE}-pc-os2-emx
+	exit ;;
+    i*86:XTS-300:*:STOP)
+	echo ${UNAME_MACHINE}-unknown-stop
+	exit ;;
+    i*86:atheos:*:*)
+	echo ${UNAME_MACHINE}-unknown-atheos
+	exit ;;
+    i*86:syllable:*:*)
+	echo ${UNAME_MACHINE}-pc-syllable
+	exit ;;
+    i*86:LynxOS:2.*:* | i*86:LynxOS:3.[01]*:* | i*86:LynxOS:4.[02]*:*)
+	echo i386-unknown-lynxos${UNAME_RELEASE}
+	exit ;;
+    i*86:*DOS:*:*)
+	echo ${UNAME_MACHINE}-pc-msdosdjgpp
+	exit ;;
+    i*86:*:4.*:* | i*86:SYSTEM_V:4.*:*)
+	UNAME_REL=`echo ${UNAME_RELEASE} | sed 's/\/MP$//'`
+	if grep Novell /usr/include/link.h >/dev/null 2>/dev/null; then
+		echo ${UNAME_MACHINE}-univel-sysv${UNAME_REL}
+	else
+		echo ${UNAME_MACHINE}-pc-sysv${UNAME_REL}
+	fi
+	exit ;;
+    i*86:*:5:[678]*)
+	# UnixWare 7.x, OpenUNIX and OpenServer 6.
+	case `/bin/uname -X | grep "^Machine"` in
+	    *486*)	     UNAME_MACHINE=i486 ;;
+	    *Pentium)	     UNAME_MACHINE=i586 ;;
+	    *Pent*|*Celeron) UNAME_MACHINE=i686 ;;
+	esac
+	echo ${UNAME_MACHINE}-unknown-sysv${UNAME_RELEASE}${UNAME_SYSTEM}${UNAME_VERSION}
+	exit ;;
+    i*86:*:3.2:*)
+	if test -f /usr/options/cb.name; then
+		UNAME_REL=`sed -n 's/.*Version //p' </usr/options/cb.name`
+		echo ${UNAME_MACHINE}-pc-isc$UNAME_REL
+	elif /bin/uname -X 2>/dev/null >/dev/null ; then
+		UNAME_REL=`(/bin/uname -X|grep Release|sed -e 's/.*= //')`
+		(/bin/uname -X|grep i80486 >/dev/null) && UNAME_MACHINE=i486
+		(/bin/uname -X|grep '^Machine.*Pentium' >/dev/null) \
+			&& UNAME_MACHINE=i586
+		(/bin/uname -X|grep '^Machine.*Pent *II' >/dev/null) \
+			&& UNAME_MACHINE=i686
+		(/bin/uname -X|grep '^Machine.*Pentium Pro' >/dev/null) \
+			&& UNAME_MACHINE=i686
+		echo ${UNAME_MACHINE}-pc-sco$UNAME_REL
+	else
+		echo ${UNAME_MACHINE}-pc-sysv32
+	fi
+	exit ;;
+    pc:*:*:*)
+	# Left here for compatibility:
+	# uname -m prints for DJGPP always 'pc', but it prints nothing about
+	# the processor, so we play safe by assuming i586.
+	# Note: whatever this is, it MUST be the same as what config.sub
+	# prints for the "djgpp" host, or else GDB configure will decide that
+	# this is a cross-build.
+	echo i586-pc-msdosdjgpp
+	exit ;;
+    Intel:Mach:3*:*)
+	echo i386-pc-mach3
+	exit ;;
+    paragon:*:*:*)
+	echo i860-intel-osf1
+	exit ;;
+    i860:*:4.*:*) # i860-SVR4
+	if grep Stardent /usr/include/sys/uadmin.h >/dev/null 2>&1 ; then
+	  echo i860-stardent-sysv${UNAME_RELEASE} # Stardent Vistra i860-SVR4
+	else # Add other i860-SVR4 vendors below as they are discovered.
+	  echo i860-unknown-sysv${UNAME_RELEASE}  # Unknown i860-SVR4
+	fi
+	exit ;;
+    mini*:CTIX:SYS*5:*)
+	# "miniframe"
+	echo m68010-convergent-sysv
+	exit ;;
+    mc68k:UNIX:SYSTEM5:3.51m)
+	echo m68k-convergent-sysv
+	exit ;;
+    M680?0:D-NIX:5.3:*)
+	echo m68k-diab-dnix
+	exit ;;
+    M68*:*:R3V[5678]*:*)
+	test -r /sysV68 && { echo 'm68k-motorola-sysv'; exit; } ;;
+    3[345]??:*:4.0:3.0 | 3[34]??A:*:4.0:3.0 | 3[34]??,*:*:4.0:3.0 | 3[34]??/*:*:4.0:3.0 | 4400:*:4.0:3.0 | 4850:*:4.0:3.0 | SKA40:*:4.0:3.0 | SDS2:*:4.0:3.0 | SHG2:*:4.0:3.0 | S7501*:*:4.0:3.0)
+	OS_REL=''
+	test -r /etc/.relid \
+	&& OS_REL=.`sed -n 's/[^ ]* [^ ]* \([0-9][0-9]\).*/\1/p' < /etc/.relid`
+	/bin/uname -p 2>/dev/null | grep 86 >/dev/null \
+	  && { echo i486-ncr-sysv4.3${OS_REL}; exit; }
+	/bin/uname -p 2>/dev/null | /bin/grep entium >/dev/null \
+	  && { echo i586-ncr-sysv4.3${OS_REL}; exit; } ;;
+    3[34]??:*:4.0:* | 3[34]??,*:*:4.0:*)
+	/bin/uname -p 2>/dev/null | grep 86 >/dev/null \
+	  && { echo i486-ncr-sysv4; exit; } ;;
+    NCR*:*:4.2:* | MPRAS*:*:4.2:*)
+	OS_REL='.3'
+	test -r /etc/.relid \
+	    && OS_REL=.`sed -n 's/[^ ]* [^ ]* \([0-9][0-9]\).*/\1/p' < /etc/.relid`
+	/bin/uname -p 2>/dev/null | grep 86 >/dev/null \
+	    && { echo i486-ncr-sysv4.3${OS_REL}; exit; }
+	/bin/uname -p 2>/dev/null | /bin/grep entium >/dev/null \
+	    && { echo i586-ncr-sysv4.3${OS_REL}; exit; }
+	/bin/uname -p 2>/dev/null | /bin/grep pteron >/dev/null \
+	    && { echo i586-ncr-sysv4.3${OS_REL}; exit; } ;;
+    m68*:LynxOS:2.*:* | m68*:LynxOS:3.0*:*)
+	echo m68k-unknown-lynxos${UNAME_RELEASE}
+	exit ;;
+    mc68030:UNIX_System_V:4.*:*)
+	echo m68k-atari-sysv4
+	exit ;;
+    TSUNAMI:LynxOS:2.*:*)
+	echo sparc-unknown-lynxos${UNAME_RELEASE}
+	exit ;;
+    rs6000:LynxOS:2.*:*)
+	echo rs6000-unknown-lynxos${UNAME_RELEASE}
+	exit ;;
+    PowerPC:LynxOS:2.*:* | PowerPC:LynxOS:3.[01]*:* | PowerPC:LynxOS:4.[02]*:*)
+	echo powerpc-unknown-lynxos${UNAME_RELEASE}
+	exit ;;
+    SM[BE]S:UNIX_SV:*:*)
+	echo mips-dde-sysv${UNAME_RELEASE}
+	exit ;;
+    RM*:ReliantUNIX-*:*:*)
+	echo mips-sni-sysv4
+	exit ;;
+    RM*:SINIX-*:*:*)
+	echo mips-sni-sysv4
+	exit ;;
+    *:SINIX-*:*:*)
+	if uname -p 2>/dev/null >/dev/null ; then
+		UNAME_MACHINE=`(uname -p) 2>/dev/null`
+		echo ${UNAME_MACHINE}-sni-sysv4
+	else
+		echo ns32k-sni-sysv
+	fi
+	exit ;;
+    PENTIUM:*:4.0*:*)	# Unisys `ClearPath HMP IX 4000' SVR4/MP effort
+			# says <Richard.M.Bartel@ccMail.Census.GOV>
+	echo i586-unisys-sysv4
+	exit ;;
+    *:UNIX_System_V:4*:FTX*)
+	# From Gerald Hewes <hewes@openmarket.com>.
+	# How about differentiating between stratus architectures? -djm
+	echo hppa1.1-stratus-sysv4
+	exit ;;
+    *:*:*:FTX*)
+	# From seanf@swdc.stratus.com.
+	echo i860-stratus-sysv4
+	exit ;;
+    i*86:VOS:*:*)
+	# From Paul.Green@stratus.com.
+	echo ${UNAME_MACHINE}-stratus-vos
+	exit ;;
+    *:VOS:*:*)
+	# From Paul.Green@stratus.com.
+	echo hppa1.1-stratus-vos
+	exit ;;
+    mc68*:A/UX:*:*)
+	echo m68k-apple-aux${UNAME_RELEASE}
+	exit ;;
+    news*:NEWS-OS:6*:*)
+	echo mips-sony-newsos6
+	exit ;;
+    R[34]000:*System_V*:*:* | R4000:UNIX_SYSV:*:* | R*000:UNIX_SV:*:*)
+	if [ -d /usr/nec ]; then
+		echo mips-nec-sysv${UNAME_RELEASE}
+	else
+		echo mips-unknown-sysv${UNAME_RELEASE}
+	fi
+	exit ;;
+    BeBox:BeOS:*:*)	# BeOS running on hardware made by Be, PPC only.
+	echo powerpc-be-beos
+	exit ;;
+    BeMac:BeOS:*:*)	# BeOS running on Mac or Mac clone, PPC only.
+	echo powerpc-apple-beos
+	exit ;;
+    BePC:BeOS:*:*)	# BeOS running on Intel PC compatible.
+	echo i586-pc-beos
+	exit ;;
+    BePC:Haiku:*:*)	# Haiku running on Intel PC compatible.
+	echo i586-pc-haiku
+	exit ;;
+    x86_64:Haiku:*:*)
+	echo x86_64-unknown-haiku
+	exit ;;
+    SX-4:SUPER-UX:*:*)
+	echo sx4-nec-superux${UNAME_RELEASE}
+	exit ;;
+    SX-5:SUPER-UX:*:*)
+	echo sx5-nec-superux${UNAME_RELEASE}
+	exit ;;
+    SX-6:SUPER-UX:*:*)
+	echo sx6-nec-superux${UNAME_RELEASE}
+	exit ;;
+    SX-7:SUPER-UX:*:*)
+	echo sx7-nec-superux${UNAME_RELEASE}
+	exit ;;
+    SX-8:SUPER-UX:*:*)
+	echo sx8-nec-superux${UNAME_RELEASE}
+	exit ;;
+    SX-8R:SUPER-UX:*:*)
+	echo sx8r-nec-superux${UNAME_RELEASE}
+	exit ;;
+    SX-ACE:SUPER-UX:*:*)
+	echo sxace-nec-superux${UNAME_RELEASE}
+	exit ;;
+    Power*:Rhapsody:*:*)
+	echo powerpc-apple-rhapsody${UNAME_RELEASE}
+	exit ;;
+    *:Rhapsody:*:*)
+	echo ${UNAME_MACHINE}-apple-rhapsody${UNAME_RELEASE}
+	exit ;;
+    *:Darwin:*:*)
+	UNAME_PROCESSOR=`uname -p` || UNAME_PROCESSOR=unknown
+	eval $set_cc_for_build
+	if test "$UNAME_PROCESSOR" = unknown ; then
+	    UNAME_PROCESSOR=powerpc
+	fi
+	if test `echo "$UNAME_RELEASE" | sed -e 's/\..*//'` -le 10 ; then
+	    if [ "$CC_FOR_BUILD" != no_compiler_found ]; then
+		if (echo '#ifdef __LP64__'; echo IS_64BIT_ARCH; echo '#endif') | \
+		    (CCOPTS="" $CC_FOR_BUILD -E - 2>/dev/null) | \
+		    grep IS_64BIT_ARCH >/dev/null
+		then
+		    case $UNAME_PROCESSOR in
+			i386) UNAME_PROCESSOR=x86_64 ;;
+			powerpc) UNAME_PROCESSOR=powerpc64 ;;
+		    esac
+		fi
+	    fi
+	elif test "$UNAME_PROCESSOR" = i386 ; then
+	    # Avoid executing cc on OS X 10.9, as it ships with a stub
+	    # that puts up a graphical alert prompting to install
+	    # developer tools.  Any system running Mac OS X 10.7 or
+	    # later (Darwin 11 and later) is required to have a 64-bit
+	    # processor. This is not true of the ARM version of Darwin
+	    # that Apple uses in portable devices.
+	    UNAME_PROCESSOR=x86_64
+	fi
+	echo ${UNAME_PROCESSOR}-apple-darwin${UNAME_RELEASE}
+	exit ;;
+    *:procnto*:*:* | *:QNX:[0123456789]*:*)
+	UNAME_PROCESSOR=`uname -p`
+	if test "$UNAME_PROCESSOR" = x86; then
+		UNAME_PROCESSOR=i386
+		UNAME_MACHINE=pc
+	fi
+	echo ${UNAME_PROCESSOR}-${UNAME_MACHINE}-nto-qnx${UNAME_RELEASE}
+	exit ;;
+    *:QNX:*:4*)
+	echo i386-pc-qnx
+	exit ;;
+    NEO-?:NONSTOP_KERNEL:*:*)
+	echo neo-tandem-nsk${UNAME_RELEASE}
+	exit ;;
+    NSE-*:NONSTOP_KERNEL:*:*)
+	echo nse-tandem-nsk${UNAME_RELEASE}
+	exit ;;
+    NSR-?:NONSTOP_KERNEL:*:*)
+	echo nsr-tandem-nsk${UNAME_RELEASE}
+	exit ;;
+    *:NonStop-UX:*:*)
+	echo mips-compaq-nonstopux
+	exit ;;
+    BS2000:POSIX*:*:*)
+	echo bs2000-siemens-sysv
+	exit ;;
+    DS/*:UNIX_System_V:*:*)
+	echo ${UNAME_MACHINE}-${UNAME_SYSTEM}-${UNAME_RELEASE}
+	exit ;;
+    *:Plan9:*:*)
+	# "uname -m" is not consistent, so use $cputype instead. 386
+	# is converted to i386 for consistency with other x86
+	# operating systems.
+	if test "$cputype" = 386; then
+	    UNAME_MACHINE=i386
+	else
+	    UNAME_MACHINE="$cputype"
+	fi
+	echo ${UNAME_MACHINE}-unknown-plan9
+	exit ;;
+    *:TOPS-10:*:*)
+	echo pdp10-unknown-tops10
+	exit ;;
+    *:TENEX:*:*)
+	echo pdp10-unknown-tenex
+	exit ;;
+    KS10:TOPS-20:*:* | KL10:TOPS-20:*:* | TYPE4:TOPS-20:*:*)
+	echo pdp10-dec-tops20
+	exit ;;
+    XKL-1:TOPS-20:*:* | TYPE5:TOPS-20:*:*)
+	echo pdp10-xkl-tops20
+	exit ;;
+    *:TOPS-20:*:*)
+	echo pdp10-unknown-tops20
+	exit ;;
+    *:ITS:*:*)
+	echo pdp10-unknown-its
+	exit ;;
+    SEI:*:*:SEIUX)
+	echo mips-sei-seiux${UNAME_RELEASE}
+	exit ;;
+    *:DragonFly:*:*)
+	echo ${UNAME_MACHINE}-unknown-dragonfly`echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'`
+	exit ;;
+    *:*VMS:*:*)
+	UNAME_MACHINE=`(uname -p) 2>/dev/null`
+	case "${UNAME_MACHINE}" in
+	    A*) echo alpha-dec-vms ; exit ;;
+	    I*) echo ia64-dec-vms ; exit ;;
+	    V*) echo vax-dec-vms ; exit ;;
+	esac ;;
+    *:XENIX:*:SysV)
+	echo i386-pc-xenix
+	exit ;;
+    i*86:skyos:*:*)
+	echo ${UNAME_MACHINE}-pc-skyos`echo ${UNAME_RELEASE} | sed -e 's/ .*$//'`
+	exit ;;
+    i*86:rdos:*:*)
+	echo ${UNAME_MACHINE}-pc-rdos
+	exit ;;
+    i*86:AROS:*:*)
+	echo ${UNAME_MACHINE}-pc-aros
+	exit ;;
+    x86_64:VMkernel:*:*)
+	echo ${UNAME_MACHINE}-unknown-esx
+	exit ;;
+    amd64:Isilon\ OneFS:*:*)
+	echo x86_64-unknown-onefs
+	exit ;;
+esac
+
+cat >&2 <<EOF
+$0: unable to guess system type
+
+This script (version $timestamp), has failed to recognize the
+operating system you are using. If your script is old, overwrite
+config.guess and config.sub with the latest versions from:
+
+  http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.guess
+and
+  http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.sub
+
+If $0 has already been updated, send the following data and any
+information you think might be pertinent to config-patches@gnu.org to
+provide the necessary information to handle your system.
+
+config.guess timestamp = $timestamp
+
+uname -m = `(uname -m) 2>/dev/null || echo unknown`
+uname -r = `(uname -r) 2>/dev/null || echo unknown`
+uname -s = `(uname -s) 2>/dev/null || echo unknown`
+uname -v = `(uname -v) 2>/dev/null || echo unknown`
+
+/usr/bin/uname -p = `(/usr/bin/uname -p) 2>/dev/null`
+/bin/uname -X     = `(/bin/uname -X) 2>/dev/null`
+
+hostinfo               = `(hostinfo) 2>/dev/null`
+/bin/universe          = `(/bin/universe) 2>/dev/null`
+/usr/bin/arch -k       = `(/usr/bin/arch -k) 2>/dev/null`
+/bin/arch              = `(/bin/arch) 2>/dev/null`
+/usr/bin/oslevel       = `(/usr/bin/oslevel) 2>/dev/null`
+/usr/convex/getsysinfo = `(/usr/convex/getsysinfo) 2>/dev/null`
+
+UNAME_MACHINE = ${UNAME_MACHINE}
+UNAME_RELEASE = ${UNAME_RELEASE}
+UNAME_SYSTEM  = ${UNAME_SYSTEM}
+UNAME_VERSION = ${UNAME_VERSION}
+EOF
+
+exit 1
+
+# Local variables:
+# eval: (add-hook 'write-file-hooks 'time-stamp)
+# time-stamp-start: "timestamp='"
+# time-stamp-format: "%:y-%02m-%02d"
+# time-stamp-end: "'"
+# End:
diff --git a/deps/jemalloc/build-aux/config.sub b/deps/jemalloc/build-aux/config.sub
new file mode 100755
index 000000000..dd2ca93c6
--- /dev/null
+++ b/deps/jemalloc/build-aux/config.sub
@@ -0,0 +1,1825 @@
+#! /bin/sh
+# Configuration validation subroutine script.
+#   Copyright 1992-2016 Free Software Foundation, Inc.
+
+timestamp='2016-11-04'
+
+# This file is free software; you can redistribute it and/or modify it
+# under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful, but
+# WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+# General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, see <http://www.gnu.org/licenses/>.
+#
+# As a special exception to the GNU General Public License, if you
+# distribute this file as part of a program that contains a
+# configuration script generated by Autoconf, you may include it under
+# the same distribution terms that you use for the rest of that
+# program.  This Exception is an additional permission under section 7
+# of the GNU General Public License, version 3 ("GPLv3").
+
+
+# Please send patches to <config-patches@gnu.org>.
+#
+# Configuration subroutine to validate and canonicalize a configuration type.
+# Supply the specified configuration type as an argument.
+# If it is invalid, we print an error message on stderr and exit with code 1.
+# Otherwise, we print the canonical config type on stdout and succeed.
+
+# You can get the latest version of this script from:
+# http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.sub
+
+# This file is supposed to be the same for all GNU packages
+# and recognize all the CPU types, system types and aliases
+# that are meaningful with *any* GNU software.
+# Each package is responsible for reporting which valid configurations
+# it does not support.  The user should be able to distinguish
+# a failure to support a valid configuration from a meaningless
+# configuration.
+
+# The goal of this file is to map all the various variations of a given
+# machine specification into a single specification in the form:
+#	CPU_TYPE-MANUFACTURER-OPERATING_SYSTEM
+# or in some cases, the newer four-part form:
+#	CPU_TYPE-MANUFACTURER-KERNEL-OPERATING_SYSTEM
+# It is wrong to echo any other type of specification.
+
+me=`echo "$0" | sed -e 's,.*/,,'`
+
+usage="\
+Usage: $0 [OPTION] CPU-MFR-OPSYS or ALIAS
+
+Canonicalize a configuration name.
+
+Operation modes:
+  -h, --help         print this help, then exit
+  -t, --time-stamp   print date of last modification, then exit
+  -v, --version      print version number, then exit
+
+Report bugs and patches to <config-patches@gnu.org>."
+
+version="\
+GNU config.sub ($timestamp)
+
+Copyright 1992-2016 Free Software Foundation, Inc.
+
+This is free software; see the source for copying conditions.  There is NO
+warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE."
+
+help="
+Try \`$me --help' for more information."
+
+# Parse command line
+while test $# -gt 0 ; do
+  case $1 in
+    --time-stamp | --time* | -t )
+       echo "$timestamp" ; exit ;;
+    --version | -v )
+       echo "$version" ; exit ;;
+    --help | --h* | -h )
+       echo "$usage"; exit ;;
+    -- )     # Stop option processing
+       shift; break ;;
+    - )	# Use stdin as input.
+       break ;;
+    -* )
+       echo "$me: invalid option $1$help"
+       exit 1 ;;
+
+    *local*)
+       # First pass through any local machine types.
+       echo $1
+       exit ;;
+
+    * )
+       break ;;
+  esac
+done
+
+case $# in
+ 0) echo "$me: missing argument$help" >&2
+    exit 1;;
+ 1) ;;
+ *) echo "$me: too many arguments$help" >&2
+    exit 1;;
+esac
+
+# Separate what the user gave into CPU-COMPANY and OS or KERNEL-OS (if any).
+# Here we must recognize all the valid KERNEL-OS combinations.
+maybe_os=`echo $1 | sed 's/^\(.*\)-\([^-]*-[^-]*\)$/\2/'`
+case $maybe_os in
+  nto-qnx* | linux-gnu* | linux-android* | linux-dietlibc | linux-newlib* | \
+  linux-musl* | linux-uclibc* | uclinux-uclibc* | uclinux-gnu* | kfreebsd*-gnu* | \
+  knetbsd*-gnu* | netbsd*-gnu* | netbsd*-eabi* | \
+  kopensolaris*-gnu* | cloudabi*-eabi* | \
+  storm-chaos* | os2-emx* | rtmk-nova*)
+    os=-$maybe_os
+    basic_machine=`echo $1 | sed 's/^\(.*\)-\([^-]*-[^-]*\)$/\1/'`
+    ;;
+  android-linux)
+    os=-linux-android
+    basic_machine=`echo $1 | sed 's/^\(.*\)-\([^-]*-[^-]*\)$/\1/'`-unknown
+    ;;
+  *)
+    basic_machine=`echo $1 | sed 's/-[^-]*$//'`
+    if [ $basic_machine != $1 ]
+    then os=`echo $1 | sed 's/.*-/-/'`
+    else os=; fi
+    ;;
+esac
+
+### Let's recognize common machines as not being operating systems so
+### that things like config.sub decstation-3100 work.  We also
+### recognize some manufacturers as not being operating systems, so we
+### can provide default operating systems below.
+case $os in
+	-sun*os*)
+		# Prevent following clause from handling this invalid input.
+		;;
+	-dec* | -mips* | -sequent* | -encore* | -pc532* | -sgi* | -sony* | \
+	-att* | -7300* | -3300* | -delta* | -motorola* | -sun[234]* | \
+	-unicom* | -ibm* | -next | -hp | -isi* | -apollo | -altos* | \
+	-convergent* | -ncr* | -news | -32* | -3600* | -3100* | -hitachi* |\
+	-c[123]* | -convex* | -sun | -crds | -omron* | -dg | -ultra | -tti* | \
+	-harris | -dolphin | -highlevel | -gould | -cbm | -ns | -masscomp | \
+	-apple | -axis | -knuth | -cray | -microblaze*)
+		os=
+		basic_machine=$1
+		;;
+	-bluegene*)
+		os=-cnk
+		;;
+	-sim | -cisco | -oki | -wec | -winbond)
+		os=
+		basic_machine=$1
+		;;
+	-scout)
+		;;
+	-wrs)
+		os=-vxworks
+		basic_machine=$1
+		;;
+	-chorusos*)
+		os=-chorusos
+		basic_machine=$1
+		;;
+	-chorusrdb)
+		os=-chorusrdb
+		basic_machine=$1
+		;;
+	-hiux*)
+		os=-hiuxwe2
+		;;
+	-sco6)
+		os=-sco5v6
+		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
+		;;
+	-sco5)
+		os=-sco3.2v5
+		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
+		;;
+	-sco4)
+		os=-sco3.2v4
+		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
+		;;
+	-sco3.2.[4-9]*)
+		os=`echo $os | sed -e 's/sco3.2./sco3.2v/'`
+		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
+		;;
+	-sco3.2v[4-9]*)
+		# Don't forget version if it is 3.2v4 or newer.
+		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
+		;;
+	-sco5v6*)
+		# Don't forget version if it is 3.2v4 or newer.
+		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
+		;;
+	-sco*)
+		os=-sco3.2v2
+		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
+		;;
+	-udk*)
+		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
+		;;
+	-isc)
+		os=-isc2.2
+		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
+		;;
+	-clix*)
+		basic_machine=clipper-intergraph
+		;;
+	-isc*)
+		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
+		;;
+	-lynx*178)
+		os=-lynxos178
+		;;
+	-lynx*5)
+		os=-lynxos5
+		;;
+	-lynx*)
+		os=-lynxos
+		;;
+	-ptx*)
+		basic_machine=`echo $1 | sed -e 's/86-.*/86-sequent/'`
+		;;
+	-windowsnt*)
+		os=`echo $os | sed -e 's/windowsnt/winnt/'`
+		;;
+	-psos*)
+		os=-psos
+		;;
+	-mint | -mint[0-9]*)
+		basic_machine=m68k-atari
+		os=-mint
+		;;
+esac
+
+# Decode aliases for certain CPU-COMPANY combinations.
+case $basic_machine in
+	# Recognize the basic CPU types without company name.
+	# Some are omitted here because they have special meanings below.
+	1750a | 580 \
+	| a29k \
+	| aarch64 | aarch64_be \
+	| alpha | alphaev[4-8] | alphaev56 | alphaev6[78] | alphapca5[67] \
+	| alpha64 | alpha64ev[4-8] | alpha64ev56 | alpha64ev6[78] | alpha64pca5[67] \
+	| am33_2.0 \
+	| arc | arceb \
+	| arm | arm[bl]e | arme[lb] | armv[2-8] | armv[3-8][lb] | armv7[arm] \
+	| avr | avr32 \
+	| ba \
+	| be32 | be64 \
+	| bfin \
+	| c4x | c8051 | clipper \
+	| d10v | d30v | dlx | dsp16xx \
+	| e2k | epiphany \
+	| fido | fr30 | frv | ft32 \
+	| h8300 | h8500 | hppa | hppa1.[01] | hppa2.0 | hppa2.0[nw] | hppa64 \
+	| hexagon \
+	| i370 | i860 | i960 | ia64 \
+	| ip2k | iq2000 \
+	| k1om \
+	| le32 | le64 \
+	| lm32 \
+	| m32c | m32r | m32rle | m68000 | m68k | m88k \
+	| maxq | mb | microblaze | microblazeel | mcore | mep | metag \
+	| mips | mipsbe | mipseb | mipsel | mipsle \
+	| mips16 \
+	| mips64 | mips64el \
+	| mips64octeon | mips64octeonel \
+	| mips64orion | mips64orionel \
+	| mips64r5900 | mips64r5900el \
+	| mips64vr | mips64vrel \
+	| mips64vr4100 | mips64vr4100el \
+	| mips64vr4300 | mips64vr4300el \
+	| mips64vr5000 | mips64vr5000el \
+	| mips64vr5900 | mips64vr5900el \
+	| mipsisa32 | mipsisa32el \
+	| mipsisa32r2 | mipsisa32r2el \
+	| mipsisa32r6 | mipsisa32r6el \
+	| mipsisa64 | mipsisa64el \
+	| mipsisa64r2 | mipsisa64r2el \
+	| mipsisa64r6 | mipsisa64r6el \
+	| mipsisa64sb1 | mipsisa64sb1el \
+	| mipsisa64sr71k | mipsisa64sr71kel \
+	| mipsr5900 | mipsr5900el \
+	| mipstx39 | mipstx39el \
+	| mn10200 | mn10300 \
+	| moxie \
+	| mt \
+	| msp430 \
+	| nds32 | nds32le | nds32be \
+	| nios | nios2 | nios2eb | nios2el \
+	| ns16k | ns32k \
+	| open8 | or1k | or1knd | or32 \
+	| pdp10 | pdp11 | pj | pjl \
+	| powerpc | powerpc64 | powerpc64le | powerpcle \
+	| pru \
+	| pyramid \
+	| riscv32 | riscv64 \
+	| rl78 | rx \
+	| score \
+	| sh | sh[1234] | sh[24]a | sh[24]aeb | sh[23]e | sh[234]eb | sheb | shbe | shle | sh[1234]le | sh3ele \
+	| sh64 | sh64le \
+	| sparc | sparc64 | sparc64b | sparc64v | sparc86x | sparclet | sparclite \
+	| sparcv8 | sparcv9 | sparcv9b | sparcv9v \
+	| spu \
+	| tahoe | tic4x | tic54x | tic55x | tic6x | tic80 | tron \
+	| ubicom32 \
+	| v850 | v850e | v850e1 | v850e2 | v850es | v850e2v3 \
+	| visium \
+	| we32k \
+	| x86 | xc16x | xstormy16 | xtensa \
+	| z8k | z80)
+		basic_machine=$basic_machine-unknown
+		;;
+	c54x)
+		basic_machine=tic54x-unknown
+		;;
+	c55x)
+		basic_machine=tic55x-unknown
+		;;
+	c6x)
+		basic_machine=tic6x-unknown
+		;;
+	leon|leon[3-9])
+		basic_machine=sparc-$basic_machine
+		;;
+	m6811 | m68hc11 | m6812 | m68hc12 | m68hcs12x | nvptx | picochip)
+		basic_machine=$basic_machine-unknown
+		os=-none
+		;;
+	m88110 | m680[12346]0 | m683?2 | m68360 | m5200 | v70 | w65 | z8k)
+		;;
+	ms1)
+		basic_machine=mt-unknown
+		;;
+
+	strongarm | thumb | xscale)
+		basic_machine=arm-unknown
+		;;
+	xgate)
+		basic_machine=$basic_machine-unknown
+		os=-none
+		;;
+	xscaleeb)
+		basic_machine=armeb-unknown
+		;;
+
+	xscaleel)
+		basic_machine=armel-unknown
+		;;
+
+	# We use `pc' rather than `unknown'
+	# because (1) that's what they normally are, and
+	# (2) the word "unknown" tends to confuse beginning users.
+	i*86 | x86_64)
+	  basic_machine=$basic_machine-pc
+	  ;;
+	# Object if more than one company name word.
+	*-*-*)
+		echo Invalid configuration \`$1\': machine \`$basic_machine\' not recognized 1>&2
+		exit 1
+		;;
+	# Recognize the basic CPU types with company name.
+	580-* \
+	| a29k-* \
+	| aarch64-* | aarch64_be-* \
+	| alpha-* | alphaev[4-8]-* | alphaev56-* | alphaev6[78]-* \
+	| alpha64-* | alpha64ev[4-8]-* | alpha64ev56-* | alpha64ev6[78]-* \
+	| alphapca5[67]-* | alpha64pca5[67]-* | arc-* | arceb-* \
+	| arm-*  | armbe-* | armle-* | armeb-* | armv*-* \
+	| avr-* | avr32-* \
+	| ba-* \
+	| be32-* | be64-* \
+	| bfin-* | bs2000-* \
+	| c[123]* | c30-* | [cjt]90-* | c4x-* \
+	| c8051-* | clipper-* | craynv-* | cydra-* \
+	| d10v-* | d30v-* | dlx-* \
+	| e2k-* | elxsi-* \
+	| f30[01]-* | f700-* | fido-* | fr30-* | frv-* | fx80-* \
+	| h8300-* | h8500-* \
+	| hppa-* | hppa1.[01]-* | hppa2.0-* | hppa2.0[nw]-* | hppa64-* \
+	| hexagon-* \
+	| i*86-* | i860-* | i960-* | ia64-* \
+	| ip2k-* | iq2000-* \
+	| k1om-* \
+	| le32-* | le64-* \
+	| lm32-* \
+	| m32c-* | m32r-* | m32rle-* \
+	| m68000-* | m680[012346]0-* | m68360-* | m683?2-* | m68k-* \
+	| m88110-* | m88k-* | maxq-* | mcore-* | metag-* \
+	| microblaze-* | microblazeel-* \
+	| mips-* | mipsbe-* | mipseb-* | mipsel-* | mipsle-* \
+	| mips16-* \
+	| mips64-* | mips64el-* \
+	| mips64octeon-* | mips64octeonel-* \
+	| mips64orion-* | mips64orionel-* \
+	| mips64r5900-* | mips64r5900el-* \
+	| mips64vr-* | mips64vrel-* \
+	| mips64vr4100-* | mips64vr4100el-* \
+	| mips64vr4300-* | mips64vr4300el-* \
+	| mips64vr5000-* | mips64vr5000el-* \
+	| mips64vr5900-* | mips64vr5900el-* \
+	| mipsisa32-* | mipsisa32el-* \
+	| mipsisa32r2-* | mipsisa32r2el-* \
+	| mipsisa32r6-* | mipsisa32r6el-* \
+	| mipsisa64-* | mipsisa64el-* \
+	| mipsisa64r2-* | mipsisa64r2el-* \
+	| mipsisa64r6-* | mipsisa64r6el-* \
+	| mipsisa64sb1-* | mipsisa64sb1el-* \
+	| mipsisa64sr71k-* | mipsisa64sr71kel-* \
+	| mipsr5900-* | mipsr5900el-* \
+	| mipstx39-* | mipstx39el-* \
+	| mmix-* \
+	| mt-* \
+	| msp430-* \
+	| nds32-* | nds32le-* | nds32be-* \
+	| nios-* | nios2-* | nios2eb-* | nios2el-* \
+	| none-* | np1-* | ns16k-* | ns32k-* \
+	| open8-* \
+	| or1k*-* \
+	| orion-* \
+	| pdp10-* | pdp11-* | pj-* | pjl-* | pn-* | power-* \
+	| powerpc-* | powerpc64-* | powerpc64le-* | powerpcle-* \
+	| pru-* \
+	| pyramid-* \
+	| riscv32-* | riscv64-* \
+	| rl78-* | romp-* | rs6000-* | rx-* \
+	| sh-* | sh[1234]-* | sh[24]a-* | sh[24]aeb-* | sh[23]e-* | sh[34]eb-* | sheb-* | shbe-* \
+	| shle-* | sh[1234]le-* | sh3ele-* | sh64-* | sh64le-* \
+	| sparc-* | sparc64-* | sparc64b-* | sparc64v-* | sparc86x-* | sparclet-* \
+	| sparclite-* \
+	| sparcv8-* | sparcv9-* | sparcv9b-* | sparcv9v-* | sv1-* | sx*-* \
+	| tahoe-* \
+	| tic30-* | tic4x-* | tic54x-* | tic55x-* | tic6x-* | tic80-* \
+	| tile*-* \
+	| tron-* \
+	| ubicom32-* \
+	| v850-* | v850e-* | v850e1-* | v850es-* | v850e2-* | v850e2v3-* \
+	| vax-* \
+	| visium-* \
+	| we32k-* \
+	| x86-* | x86_64-* | xc16x-* | xps100-* \
+	| xstormy16-* | xtensa*-* \
+	| ymp-* \
+	| z8k-* | z80-*)
+		;;
+	# Recognize the basic CPU types without company name, with glob match.
+	xtensa*)
+		basic_machine=$basic_machine-unknown
+		;;
+	# Recognize the various machine names and aliases which stand
+	# for a CPU type and a company and sometimes even an OS.
+	386bsd)
+		basic_machine=i386-unknown
+		os=-bsd
+		;;
+	3b1 | 7300 | 7300-att | att-7300 | pc7300 | safari | unixpc)
+		basic_machine=m68000-att
+		;;
+	3b*)
+		basic_machine=we32k-att
+		;;
+	a29khif)
+		basic_machine=a29k-amd
+		os=-udi
+		;;
+	abacus)
+		basic_machine=abacus-unknown
+		;;
+	adobe68k)
+		basic_machine=m68010-adobe
+		os=-scout
+		;;
+	alliant | fx80)
+		basic_machine=fx80-alliant
+		;;
+	altos | altos3068)
+		basic_machine=m68k-altos
+		;;
+	am29k)
+		basic_machine=a29k-none
+		os=-bsd
+		;;
+	amd64)
+		basic_machine=x86_64-pc
+		;;
+	amd64-*)
+		basic_machine=x86_64-`echo $basic_machine | sed 's/^[^-]*-//'`
+		;;
+	amdahl)
+		basic_machine=580-amdahl
+		os=-sysv
+		;;
+	amiga | amiga-*)
+		basic_machine=m68k-unknown
+		;;
+	amigaos | amigados)
+		basic_machine=m68k-unknown
+		os=-amigaos
+		;;
+	amigaunix | amix)
+		basic_machine=m68k-unknown
+		os=-sysv4
+		;;
+	apollo68)
+		basic_machine=m68k-apollo
+		os=-sysv
+		;;
+	apollo68bsd)
+		basic_machine=m68k-apollo
+		os=-bsd
+		;;
+	aros)
+		basic_machine=i386-pc
+		os=-aros
+		;;
+	asmjs)
+		basic_machine=asmjs-unknown
+		;;
+	aux)
+		basic_machine=m68k-apple
+		os=-aux
+		;;
+	balance)
+		basic_machine=ns32k-sequent
+		os=-dynix
+		;;
+	blackfin)
+		basic_machine=bfin-unknown
+		os=-linux
+		;;
+	blackfin-*)
+		basic_machine=bfin-`echo $basic_machine | sed 's/^[^-]*-//'`
+		os=-linux
+		;;
+	bluegene*)
+		basic_machine=powerpc-ibm
+		os=-cnk
+		;;
+	c54x-*)
+		basic_machine=tic54x-`echo $basic_machine | sed 's/^[^-]*-//'`
+		;;
+	c55x-*)
+		basic_machine=tic55x-`echo $basic_machine | sed 's/^[^-]*-//'`
+		;;
+	c6x-*)
+		basic_machine=tic6x-`echo $basic_machine | sed 's/^[^-]*-//'`
+		;;
+	c90)
+		basic_machine=c90-cray
+		os=-unicos
+		;;
+	cegcc)
+		basic_machine=arm-unknown
+		os=-cegcc
+		;;
+	convex-c1)
+		basic_machine=c1-convex
+		os=-bsd
+		;;
+	convex-c2)
+		basic_machine=c2-convex
+		os=-bsd
+		;;
+	convex-c32)
+		basic_machine=c32-convex
+		os=-bsd
+		;;
+	convex-c34)
+		basic_machine=c34-convex
+		os=-bsd
+		;;
+	convex-c38)
+		basic_machine=c38-convex
+		os=-bsd
+		;;
+	cray | j90)
+		basic_machine=j90-cray
+		os=-unicos
+		;;
+	craynv)
+		basic_machine=craynv-cray
+		os=-unicosmp
+		;;
+	cr16 | cr16-*)
+		basic_machine=cr16-unknown
+		os=-elf
+		;;
+	crds | unos)
+		basic_machine=m68k-crds
+		;;
+	crisv32 | crisv32-* | etraxfs*)
+		basic_machine=crisv32-axis
+		;;
+	cris | cris-* | etrax*)
+		basic_machine=cris-axis
+		;;
+	crx)
+		basic_machine=crx-unknown
+		os=-elf
+		;;
+	da30 | da30-*)
+		basic_machine=m68k-da30
+		;;
+	decstation | decstation-3100 | pmax | pmax-* | pmin | dec3100 | decstatn)
+		basic_machine=mips-dec
+		;;
+	decsystem10* | dec10*)
+		basic_machine=pdp10-dec
+		os=-tops10
+		;;
+	decsystem20* | dec20*)
+		basic_machine=pdp10-dec
+		os=-tops20
+		;;
+	delta | 3300 | motorola-3300 | motorola-delta \
+	      | 3300-motorola | delta-motorola)
+		basic_machine=m68k-motorola
+		;;
+	delta88)
+		basic_machine=m88k-motorola
+		os=-sysv3
+		;;
+	dicos)
+		basic_machine=i686-pc
+		os=-dicos
+		;;
+	djgpp)
+		basic_machine=i586-pc
+		os=-msdosdjgpp
+		;;
+	dpx20 | dpx20-*)
+		basic_machine=rs6000-bull
+		os=-bosx
+		;;
+	dpx2* | dpx2*-bull)
+		basic_machine=m68k-bull
+		os=-sysv3
+		;;
+	e500v[12])
+		basic_machine=powerpc-unknown
+		os=$os"spe"
+		;;
+	e500v[12]-*)
+		basic_machine=powerpc-`echo $basic_machine | sed 's/^[^-]*-//'`
+		os=$os"spe"
+		;;
+	ebmon29k)
+		basic_machine=a29k-amd
+		os=-ebmon
+		;;
+	elxsi)
+		basic_machine=elxsi-elxsi
+		os=-bsd
+		;;
+	encore | umax | mmax)
+		basic_machine=ns32k-encore
+		;;
+	es1800 | OSE68k | ose68k | ose | OSE)
+		basic_machine=m68k-ericsson
+		os=-ose
+		;;
+	fx2800)
+		basic_machine=i860-alliant
+		;;
+	genix)
+		basic_machine=ns32k-ns
+		;;
+	gmicro)
+		basic_machine=tron-gmicro
+		os=-sysv
+		;;
+	go32)
+		basic_machine=i386-pc
+		os=-go32
+		;;
+	h3050r* | hiux*)
+		basic_machine=hppa1.1-hitachi
+		os=-hiuxwe2
+		;;
+	h8300hms)
+		basic_machine=h8300-hitachi
+		os=-hms
+		;;
+	h8300xray)
+		basic_machine=h8300-hitachi
+		os=-xray
+		;;
+	h8500hms)
+		basic_machine=h8500-hitachi
+		os=-hms
+		;;
+	harris)
+		basic_machine=m88k-harris
+		os=-sysv3
+		;;
+	hp300-*)
+		basic_machine=m68k-hp
+		;;
+	hp300bsd)
+		basic_machine=m68k-hp
+		os=-bsd
+		;;
+	hp300hpux)
+		basic_machine=m68k-hp
+		os=-hpux
+		;;
+	hp3k9[0-9][0-9] | hp9[0-9][0-9])
+		basic_machine=hppa1.0-hp
+		;;
+	hp9k2[0-9][0-9] | hp9k31[0-9])
+		basic_machine=m68000-hp
+		;;
+	hp9k3[2-9][0-9])
+		basic_machine=m68k-hp
+		;;
+	hp9k6[0-9][0-9] | hp6[0-9][0-9])
+		basic_machine=hppa1.0-hp
+		;;
+	hp9k7[0-79][0-9] | hp7[0-79][0-9])
+		basic_machine=hppa1.1-hp
+		;;
+	hp9k78[0-9] | hp78[0-9])
+		# FIXME: really hppa2.0-hp
+		basic_machine=hppa1.1-hp
+		;;
+	hp9k8[67]1 | hp8[67]1 | hp9k80[24] | hp80[24] | hp9k8[78]9 | hp8[78]9 | hp9k893 | hp893)
+		# FIXME: really hppa2.0-hp
+		basic_machine=hppa1.1-hp
+		;;
+	hp9k8[0-9][13679] | hp8[0-9][13679])
+		basic_machine=hppa1.1-hp
+		;;
+	hp9k8[0-9][0-9] | hp8[0-9][0-9])
+		basic_machine=hppa1.0-hp
+		;;
+	hppa-next)
+		os=-nextstep3
+		;;
+	hppaosf)
+		basic_machine=hppa1.1-hp
+		os=-osf
+		;;
+	hppro)
+		basic_machine=hppa1.1-hp
+		os=-proelf
+		;;
+	i370-ibm* | ibm*)
+		basic_machine=i370-ibm
+		;;
+	i*86v32)
+		basic_machine=`echo $1 | sed -e 's/86.*/86-pc/'`
+		os=-sysv32
+		;;
+	i*86v4*)
+		basic_machine=`echo $1 | sed -e 's/86.*/86-pc/'`
+		os=-sysv4
+		;;
+	i*86v)
+		basic_machine=`echo $1 | sed -e 's/86.*/86-pc/'`
+		os=-sysv
+		;;
+	i*86sol2)
+		basic_machine=`echo $1 | sed -e 's/86.*/86-pc/'`
+		os=-solaris2
+		;;
+	i386mach)
+		basic_machine=i386-mach
+		os=-mach
+		;;
+	i386-vsta | vsta)
+		basic_machine=i386-unknown
+		os=-vsta
+		;;
+	iris | iris4d)
+		basic_machine=mips-sgi
+		case $os in
+		    -irix*)
+			;;
+		    *)
+			os=-irix4
+			;;
+		esac
+		;;
+	isi68 | isi)
+		basic_machine=m68k-isi
+		os=-sysv
+		;;
+	leon-*|leon[3-9]-*)
+		basic_machine=sparc-`echo $basic_machine | sed 's/-.*//'`
+		;;
+	m68knommu)
+		basic_machine=m68k-unknown
+		os=-linux
+		;;
+	m68knommu-*)
+		basic_machine=m68k-`echo $basic_machine | sed 's/^[^-]*-//'`
+		os=-linux
+		;;
+	m88k-omron*)
+		basic_machine=m88k-omron
+		;;
+	magnum | m3230)
+		basic_machine=mips-mips
+		os=-sysv
+		;;
+	merlin)
+		basic_machine=ns32k-utek
+		os=-sysv
+		;;
+	microblaze*)
+		basic_machine=microblaze-xilinx
+		;;
+	mingw64)
+		basic_machine=x86_64-pc
+		os=-mingw64
+		;;
+	mingw32)
+		basic_machine=i686-pc
+		os=-mingw32
+		;;
+	mingw32ce)
+		basic_machine=arm-unknown
+		os=-mingw32ce
+		;;
+	miniframe)
+		basic_machine=m68000-convergent
+		;;
+	*mint | -mint[0-9]* | *MiNT | *MiNT[0-9]*)
+		basic_machine=m68k-atari
+		os=-mint
+		;;
+	mips3*-*)
+		basic_machine=`echo $basic_machine | sed -e 's/mips3/mips64/'`
+		;;
+	mips3*)
+		basic_machine=`echo $basic_machine | sed -e 's/mips3/mips64/'`-unknown
+		;;
+	monitor)
+		basic_machine=m68k-rom68k
+		os=-coff
+		;;
+	morphos)
+		basic_machine=powerpc-unknown
+		os=-morphos
+		;;
+	moxiebox)
+		basic_machine=moxie-unknown
+		os=-moxiebox
+		;;
+	msdos)
+		basic_machine=i386-pc
+		os=-msdos
+		;;
+	ms1-*)
+		basic_machine=`echo $basic_machine | sed -e 's/ms1-/mt-/'`
+		;;
+	msys)
+		basic_machine=i686-pc
+		os=-msys
+		;;
+	mvs)
+		basic_machine=i370-ibm
+		os=-mvs
+		;;
+	nacl)
+		basic_machine=le32-unknown
+		os=-nacl
+		;;
+	ncr3000)
+		basic_machine=i486-ncr
+		os=-sysv4
+		;;
+	netbsd386)
+		basic_machine=i386-unknown
+		os=-netbsd
+		;;
+	netwinder)
+		basic_machine=armv4l-rebel
+		os=-linux
+		;;
+	news | news700 | news800 | news900)
+		basic_machine=m68k-sony
+		os=-newsos
+		;;
+	news1000)
+		basic_machine=m68030-sony
+		os=-newsos
+		;;
+	news-3600 | risc-news)
+		basic_machine=mips-sony
+		os=-newsos
+		;;
+	necv70)
+		basic_machine=v70-nec
+		os=-sysv
+		;;
+	next | m*-next )
+		basic_machine=m68k-next
+		case $os in
+		    -nextstep* )
+			;;
+		    -ns2*)
+		      os=-nextstep2
+			;;
+		    *)
+		      os=-nextstep3
+			;;
+		esac
+		;;
+	nh3000)
+		basic_machine=m68k-harris
+		os=-cxux
+		;;
+	nh[45]000)
+		basic_machine=m88k-harris
+		os=-cxux
+		;;
+	nindy960)
+		basic_machine=i960-intel
+		os=-nindy
+		;;
+	mon960)
+		basic_machine=i960-intel
+		os=-mon960
+		;;
+	nonstopux)
+		basic_machine=mips-compaq
+		os=-nonstopux
+		;;
+	np1)
+		basic_machine=np1-gould
+		;;
+	neo-tandem)
+		basic_machine=neo-tandem
+		;;
+	nse-tandem)
+		basic_machine=nse-tandem
+		;;
+	nsr-tandem)
+		basic_machine=nsr-tandem
+		;;
+	op50n-* | op60c-*)
+		basic_machine=hppa1.1-oki
+		os=-proelf
+		;;
+	openrisc | openrisc-*)
+		basic_machine=or32-unknown
+		;;
+	os400)
+		basic_machine=powerpc-ibm
+		os=-os400
+		;;
+	OSE68000 | ose68000)
+		basic_machine=m68000-ericsson
+		os=-ose
+		;;
+	os68k)
+		basic_machine=m68k-none
+		os=-os68k
+		;;
+	pa-hitachi)
+		basic_machine=hppa1.1-hitachi
+		os=-hiuxwe2
+		;;
+	paragon)
+		basic_machine=i860-intel
+		os=-osf
+		;;
+	parisc)
+		basic_machine=hppa-unknown
+		os=-linux
+		;;
+	parisc-*)
+		basic_machine=hppa-`echo $basic_machine | sed 's/^[^-]*-//'`
+		os=-linux
+		;;
+	pbd)
+		basic_machine=sparc-tti
+		;;
+	pbb)
+		basic_machine=m68k-tti
+		;;
+	pc532 | pc532-*)
+		basic_machine=ns32k-pc532
+		;;
+	pc98)
+		basic_machine=i386-pc
+		;;
+	pc98-*)
+		basic_machine=i386-`echo $basic_machine | sed 's/^[^-]*-//'`
+		;;
+	pentium | p5 | k5 | k6 | nexgen | viac3)
+		basic_machine=i586-pc
+		;;
+	pentiumpro | p6 | 6x86 | athlon | athlon_*)
+		basic_machine=i686-pc
+		;;
+	pentiumii | pentium2 | pentiumiii | pentium3)
+		basic_machine=i686-pc
+		;;
+	pentium4)
+		basic_machine=i786-pc
+		;;
+	pentium-* | p5-* | k5-* | k6-* | nexgen-* | viac3-*)
+		basic_machine=i586-`echo $basic_machine | sed 's/^[^-]*-//'`
+		;;
+	pentiumpro-* | p6-* | 6x86-* | athlon-*)
+		basic_machine=i686-`echo $basic_machine | sed 's/^[^-]*-//'`
+		;;
+	pentiumii-* | pentium2-* | pentiumiii-* | pentium3-*)
+		basic_machine=i686-`echo $basic_machine | sed 's/^[^-]*-//'`
+		;;
+	pentium4-*)
+		basic_machine=i786-`echo $basic_machine | sed 's/^[^-]*-//'`
+		;;
+	pn)
+		basic_machine=pn-gould
+		;;
+	power)	basic_machine=power-ibm
+		;;
+	ppc | ppcbe)	basic_machine=powerpc-unknown
+		;;
+	ppc-* | ppcbe-*)
+		basic_machine=powerpc-`echo $basic_machine | sed 's/^[^-]*-//'`
+		;;
+	ppcle | powerpclittle)
+		basic_machine=powerpcle-unknown
+		;;
+	ppcle-* | powerpclittle-*)
+		basic_machine=powerpcle-`echo $basic_machine | sed 's/^[^-]*-//'`
+		;;
+	ppc64)	basic_machine=powerpc64-unknown
+		;;
+	ppc64-*) basic_machine=powerpc64-`echo $basic_machine | sed 's/^[^-]*-//'`
+		;;
+	ppc64le | powerpc64little)
+		basic_machine=powerpc64le-unknown
+		;;
+	ppc64le-* | powerpc64little-*)
+		basic_machine=powerpc64le-`echo $basic_machine | sed 's/^[^-]*-//'`
+		;;
+	ps2)
+		basic_machine=i386-ibm
+		;;
+	pw32)
+		basic_machine=i586-unknown
+		os=-pw32
+		;;
+	rdos | rdos64)
+		basic_machine=x86_64-pc
+		os=-rdos
+		;;
+	rdos32)
+		basic_machine=i386-pc
+		os=-rdos
+		;;
+	rom68k)
+		basic_machine=m68k-rom68k
+		os=-coff
+		;;
+	rm[46]00)
+		basic_machine=mips-siemens
+		;;
+	rtpc | rtpc-*)
+		basic_machine=romp-ibm
+		;;
+	s390 | s390-*)
+		basic_machine=s390-ibm
+		;;
+	s390x | s390x-*)
+		basic_machine=s390x-ibm
+		;;
+	sa29200)
+		basic_machine=a29k-amd
+		os=-udi
+		;;
+	sb1)
+		basic_machine=mipsisa64sb1-unknown
+		;;
+	sb1el)
+		basic_machine=mipsisa64sb1el-unknown
+		;;
+	sde)
+		basic_machine=mipsisa32-sde
+		os=-elf
+		;;
+	sei)
+		basic_machine=mips-sei
+		os=-seiux
+		;;
+	sequent)
+		basic_machine=i386-sequent
+		;;
+	sh)
+		basic_machine=sh-hitachi
+		os=-hms
+		;;
+	sh5el)
+		basic_machine=sh5le-unknown
+		;;
+	sh64)
+		basic_machine=sh64-unknown
+		;;
+	sparclite-wrs | simso-wrs)
+		basic_machine=sparclite-wrs
+		os=-vxworks
+		;;
+	sps7)
+		basic_machine=m68k-bull
+		os=-sysv2
+		;;
+	spur)
+		basic_machine=spur-unknown
+		;;
+	st2000)
+		basic_machine=m68k-tandem
+		;;
+	stratus)
+		basic_machine=i860-stratus
+		os=-sysv4
+		;;
+	strongarm-* | thumb-*)
+		basic_machine=arm-`echo $basic_machine | sed 's/^[^-]*-//'`
+		;;
+	sun2)
+		basic_machine=m68000-sun
+		;;
+	sun2os3)
+		basic_machine=m68000-sun
+		os=-sunos3
+		;;
+	sun2os4)
+		basic_machine=m68000-sun
+		os=-sunos4
+		;;
+	sun3os3)
+		basic_machine=m68k-sun
+		os=-sunos3
+		;;
+	sun3os4)
+		basic_machine=m68k-sun
+		os=-sunos4
+		;;
+	sun4os3)
+		basic_machine=sparc-sun
+		os=-sunos3
+		;;
+	sun4os4)
+		basic_machine=sparc-sun
+		os=-sunos4
+		;;
+	sun4sol2)
+		basic_machine=sparc-sun
+		os=-solaris2
+		;;
+	sun3 | sun3-*)
+		basic_machine=m68k-sun
+		;;
+	sun4)
+		basic_machine=sparc-sun
+		;;
+	sun386 | sun386i | roadrunner)
+		basic_machine=i386-sun
+		;;
+	sv1)
+		basic_machine=sv1-cray
+		os=-unicos
+		;;
+	symmetry)
+		basic_machine=i386-sequent
+		os=-dynix
+		;;
+	t3e)
+		basic_machine=alphaev5-cray
+		os=-unicos
+		;;
+	t90)
+		basic_machine=t90-cray
+		os=-unicos
+		;;
+	tile*)
+		basic_machine=$basic_machine-unknown
+		os=-linux-gnu
+		;;
+	tx39)
+		basic_machine=mipstx39-unknown
+		;;
+	tx39el)
+		basic_machine=mipstx39el-unknown
+		;;
+	toad1)
+		basic_machine=pdp10-xkl
+		os=-tops20
+		;;
+	tower | tower-32)
+		basic_machine=m68k-ncr
+		;;
+	tpf)
+		basic_machine=s390x-ibm
+		os=-tpf
+		;;
+	udi29k)
+		basic_machine=a29k-amd
+		os=-udi
+		;;
+	ultra3)
+		basic_machine=a29k-nyu
+		os=-sym1
+		;;
+	v810 | necv810)
+		basic_machine=v810-nec
+		os=-none
+		;;
+	vaxv)
+		basic_machine=vax-dec
+		os=-sysv
+		;;
+	vms)
+		basic_machine=vax-dec
+		os=-vms
+		;;
+	vpp*|vx|vx-*)
+		basic_machine=f301-fujitsu
+		;;
+	vxworks960)
+		basic_machine=i960-wrs
+		os=-vxworks
+		;;
+	vxworks68)
+		basic_machine=m68k-wrs
+		os=-vxworks
+		;;
+	vxworks29k)
+		basic_machine=a29k-wrs
+		os=-vxworks
+		;;
+	w65*)
+		basic_machine=w65-wdc
+		os=-none
+		;;
+	w89k-*)
+		basic_machine=hppa1.1-winbond
+		os=-proelf
+		;;
+	xbox)
+		basic_machine=i686-pc
+		os=-mingw32
+		;;
+	xps | xps100)
+		basic_machine=xps100-honeywell
+		;;
+	xscale-* | xscalee[bl]-*)
+		basic_machine=`echo $basic_machine | sed 's/^xscale/arm/'`
+		;;
+	ymp)
+		basic_machine=ymp-cray
+		os=-unicos
+		;;
+	z8k-*-coff)
+		basic_machine=z8k-unknown
+		os=-sim
+		;;
+	z80-*-coff)
+		basic_machine=z80-unknown
+		os=-sim
+		;;
+	none)
+		basic_machine=none-none
+		os=-none
+		;;
+
+# Here we handle the default manufacturer of certain CPU types.  It is in
+# some cases the only manufacturer, in others, it is the most popular.
+	w89k)
+		basic_machine=hppa1.1-winbond
+		;;
+	op50n)
+		basic_machine=hppa1.1-oki
+		;;
+	op60c)
+		basic_machine=hppa1.1-oki
+		;;
+	romp)
+		basic_machine=romp-ibm
+		;;
+	mmix)
+		basic_machine=mmix-knuth
+		;;
+	rs6000)
+		basic_machine=rs6000-ibm
+		;;
+	vax)
+		basic_machine=vax-dec
+		;;
+	pdp10)
+		# there are many clones, so DEC is not a safe bet
+		basic_machine=pdp10-unknown
+		;;
+	pdp11)
+		basic_machine=pdp11-dec
+		;;
+	we32k)
+		basic_machine=we32k-att
+		;;
+	sh[1234] | sh[24]a | sh[24]aeb | sh[34]eb | sh[1234]le | sh[23]ele)
+		basic_machine=sh-unknown
+		;;
+	sparc | sparcv8 | sparcv9 | sparcv9b | sparcv9v)
+		basic_machine=sparc-sun
+		;;
+	cydra)
+		basic_machine=cydra-cydrome
+		;;
+	orion)
+		basic_machine=orion-highlevel
+		;;
+	orion105)
+		basic_machine=clipper-highlevel
+		;;
+	mac | mpw | mac-mpw)
+		basic_machine=m68k-apple
+		;;
+	pmac | pmac-mpw)
+		basic_machine=powerpc-apple
+		;;
+	*-unknown)
+		# Make sure to match an already-canonicalized machine name.
+		;;
+	*)
+		echo Invalid configuration \`$1\': machine \`$basic_machine\' not recognized 1>&2
+		exit 1
+		;;
+esac
+
+# Here we canonicalize certain aliases for manufacturers.
+case $basic_machine in
+	*-digital*)
+		basic_machine=`echo $basic_machine | sed 's/digital.*/dec/'`
+		;;
+	*-commodore*)
+		basic_machine=`echo $basic_machine | sed 's/commodore.*/cbm/'`
+		;;
+	*)
+		;;
+esac
+
+# Decode manufacturer-specific aliases for certain operating systems.
+
+if [ x"$os" != x"" ]
+then
+case $os in
+	# First match some system type aliases
+	# that might get confused with valid system types.
+	# -solaris* is a basic system type, with this one exception.
+	-auroraux)
+		os=-auroraux
+		;;
+	-solaris1 | -solaris1.*)
+		os=`echo $os | sed -e 's|solaris1|sunos4|'`
+		;;
+	-solaris)
+		os=-solaris2
+		;;
+	-svr4*)
+		os=-sysv4
+		;;
+	-unixware*)
+		os=-sysv4.2uw
+		;;
+	-gnu/linux*)
+		os=`echo $os | sed -e 's|gnu/linux|linux-gnu|'`
+		;;
+	# First accept the basic system types.
+	# The portable systems comes first.
+	# Each alternative MUST END IN A *, to match a version number.
+	# -sysv* is not here because it comes later, after sysvr4.
+	-gnu* | -bsd* | -mach* | -minix* | -genix* | -ultrix* | -irix* \
+	      | -*vms* | -sco* | -esix* | -isc* | -aix* | -cnk* | -sunos | -sunos[34]*\
+	      | -hpux* | -unos* | -osf* | -luna* | -dgux* | -auroraux* | -solaris* \
+	      | -sym* | -kopensolaris* | -plan9* \
+	      | -amigaos* | -amigados* | -msdos* | -newsos* | -unicos* | -aof* \
+	      | -aos* | -aros* | -cloudabi* | -sortix* \
+	      | -nindy* | -vxsim* | -vxworks* | -ebmon* | -hms* | -mvs* \
+	      | -clix* | -riscos* | -uniplus* | -iris* | -rtu* | -xenix* \
+	      | -hiux* | -386bsd* | -knetbsd* | -mirbsd* | -netbsd* \
+	      | -bitrig* | -openbsd* | -solidbsd* | -libertybsd* \
+	      | -ekkobsd* | -kfreebsd* | -freebsd* | -riscix* | -lynxos* \
+	      | -bosx* | -nextstep* | -cxux* | -aout* | -elf* | -oabi* \
+	      | -ptx* | -coff* | -ecoff* | -winnt* | -domain* | -vsta* \
+	      | -udi* | -eabi* | -lites* | -ieee* | -go32* | -aux* \
+	      | -chorusos* | -chorusrdb* | -cegcc* \
+	      | -cygwin* | -msys* | -pe* | -psos* | -moss* | -proelf* | -rtems* \
+	      | -midipix* | -mingw32* | -mingw64* | -linux-gnu* | -linux-android* \
+	      | -linux-newlib* | -linux-musl* | -linux-uclibc* \
+	      | -uxpv* | -beos* | -mpeix* | -udk* | -moxiebox* \
+	      | -interix* | -uwin* | -mks* | -rhapsody* | -darwin* | -opened* \
+	      | -openstep* | -oskit* | -conix* | -pw32* | -nonstopux* \
+	      | -storm-chaos* | -tops10* | -tenex* | -tops20* | -its* \
+	      | -os2* | -vos* | -palmos* | -uclinux* | -nucleus* \
+	      | -morphos* | -superux* | -rtmk* | -rtmk-nova* | -windiss* \
+	      | -powermax* | -dnix* | -nx6 | -nx7 | -sei* | -dragonfly* \
+	      | -skyos* | -haiku* | -rdos* | -toppers* | -drops* | -es* \
+	      | -onefs* | -tirtos* | -phoenix* | -fuchsia*)
+	# Remember, each alternative MUST END IN *, to match a version number.
+		;;
+	-qnx*)
+		case $basic_machine in
+		    x86-* | i*86-*)
+			;;
+		    *)
+			os=-nto$os
+			;;
+		esac
+		;;
+	-nto-qnx*)
+		;;
+	-nto*)
+		os=`echo $os | sed -e 's|nto|nto-qnx|'`
+		;;
+	-sim | -es1800* | -hms* | -xray | -os68k* | -none* | -v88r* \
+	      | -windows* | -osx | -abug | -netware* | -os9* | -beos* | -haiku* \
+	      | -macos* | -mpw* | -magic* | -mmixware* | -mon960* | -lnews*)
+		;;
+	-mac*)
+		os=`echo $os | sed -e 's|mac|macos|'`
+		;;
+	-linux-dietlibc)
+		os=-linux-dietlibc
+		;;
+	-linux*)
+		os=`echo $os | sed -e 's|linux|linux-gnu|'`
+		;;
+	-sunos5*)
+		os=`echo $os | sed -e 's|sunos5|solaris2|'`
+		;;
+	-sunos6*)
+		os=`echo $os | sed -e 's|sunos6|solaris3|'`
+		;;
+	-opened*)
+		os=-openedition
+		;;
+	-os400*)
+		os=-os400
+		;;
+	-wince*)
+		os=-wince
+		;;
+	-osfrose*)
+		os=-osfrose
+		;;
+	-osf*)
+		os=-osf
+		;;
+	-utek*)
+		os=-bsd
+		;;
+	-dynix*)
+		os=-bsd
+		;;
+	-acis*)
+		os=-aos
+		;;
+	-atheos*)
+		os=-atheos
+		;;
+	-syllable*)
+		os=-syllable
+		;;
+	-386bsd)
+		os=-bsd
+		;;
+	-ctix* | -uts*)
+		os=-sysv
+		;;
+	-nova*)
+		os=-rtmk-nova
+		;;
+	-ns2 )
+		os=-nextstep2
+		;;
+	-nsk*)
+		os=-nsk
+		;;
+	# Preserve the version number of sinix5.
+	-sinix5.*)
+		os=`echo $os | sed -e 's|sinix|sysv|'`
+		;;
+	-sinix*)
+		os=-sysv4
+		;;
+	-tpf*)
+		os=-tpf
+		;;
+	-triton*)
+		os=-sysv3
+		;;
+	-oss*)
+		os=-sysv3
+		;;
+	-svr4)
+		os=-sysv4
+		;;
+	-svr3)
+		os=-sysv3
+		;;
+	-sysvr4)
+		os=-sysv4
+		;;
+	# This must come after -sysvr4.
+	-sysv*)
+		;;
+	-ose*)
+		os=-ose
+		;;
+	-es1800*)
+		os=-ose
+		;;
+	-xenix)
+		os=-xenix
+		;;
+	-*mint | -mint[0-9]* | -*MiNT | -MiNT[0-9]*)
+		os=-mint
+		;;
+	-aros*)
+		os=-aros
+		;;
+	-zvmoe)
+		os=-zvmoe
+		;;
+	-dicos*)
+		os=-dicos
+		;;
+	-nacl*)
+		;;
+	-ios)
+		;;
+	-none)
+		;;
+	*)
+		# Get rid of the `-' at the beginning of $os.
+		os=`echo $os | sed 's/[^-]*-//'`
+		echo Invalid configuration \`$1\': system \`$os\' not recognized 1>&2
+		exit 1
+		;;
+esac
+else
+
+# Here we handle the default operating systems that come with various machines.
+# The value should be what the vendor currently ships out the door with their
+# machine or put another way, the most popular os provided with the machine.
+
+# Note that if you're going to try to match "-MANUFACTURER" here (say,
+# "-sun"), then you have to tell the case statement up towards the top
+# that MANUFACTURER isn't an operating system.  Otherwise, code above
+# will signal an error saying that MANUFACTURER isn't an operating
+# system, and we'll never get to this point.
+
+case $basic_machine in
+	score-*)
+		os=-elf
+		;;
+	spu-*)
+		os=-elf
+		;;
+	*-acorn)
+		os=-riscix1.2
+		;;
+	arm*-rebel)
+		os=-linux
+		;;
+	arm*-semi)
+		os=-aout
+		;;
+	c4x-* | tic4x-*)
+		os=-coff
+		;;
+	c8051-*)
+		os=-elf
+		;;
+	hexagon-*)
+		os=-elf
+		;;
+	tic54x-*)
+		os=-coff
+		;;
+	tic55x-*)
+		os=-coff
+		;;
+	tic6x-*)
+		os=-coff
+		;;
+	# This must come before the *-dec entry.
+	pdp10-*)
+		os=-tops20
+		;;
+	pdp11-*)
+		os=-none
+		;;
+	*-dec | vax-*)
+		os=-ultrix4.2
+		;;
+	m68*-apollo)
+		os=-domain
+		;;
+	i386-sun)
+		os=-sunos4.0.2
+		;;
+	m68000-sun)
+		os=-sunos3
+		;;
+	m68*-cisco)
+		os=-aout
+		;;
+	mep-*)
+		os=-elf
+		;;
+	mips*-cisco)
+		os=-elf
+		;;
+	mips*-*)
+		os=-elf
+		;;
+	or32-*)
+		os=-coff
+		;;
+	*-tti)	# must be before sparc entry or we get the wrong os.
+		os=-sysv3
+		;;
+	sparc-* | *-sun)
+		os=-sunos4.1.1
+		;;
+	*-be)
+		os=-beos
+		;;
+	*-haiku)
+		os=-haiku
+		;;
+	*-ibm)
+		os=-aix
+		;;
+	*-knuth)
+		os=-mmixware
+		;;
+	*-wec)
+		os=-proelf
+		;;
+	*-winbond)
+		os=-proelf
+		;;
+	*-oki)
+		os=-proelf
+		;;
+	*-hp)
+		os=-hpux
+		;;
+	*-hitachi)
+		os=-hiux
+		;;
+	i860-* | *-att | *-ncr | *-altos | *-motorola | *-convergent)
+		os=-sysv
+		;;
+	*-cbm)
+		os=-amigaos
+		;;
+	*-dg)
+		os=-dgux
+		;;
+	*-dolphin)
+		os=-sysv3
+		;;
+	m68k-ccur)
+		os=-rtu
+		;;
+	m88k-omron*)
+		os=-luna
+		;;
+	*-next )
+		os=-nextstep
+		;;
+	*-sequent)
+		os=-ptx
+		;;
+	*-crds)
+		os=-unos
+		;;
+	*-ns)
+		os=-genix
+		;;
+	i370-*)
+		os=-mvs
+		;;
+	*-next)
+		os=-nextstep3
+		;;
+	*-gould)
+		os=-sysv
+		;;
+	*-highlevel)
+		os=-bsd
+		;;
+	*-encore)
+		os=-bsd
+		;;
+	*-sgi)
+		os=-irix
+		;;
+	*-siemens)
+		os=-sysv4
+		;;
+	*-masscomp)
+		os=-rtu
+		;;
+	f30[01]-fujitsu | f700-fujitsu)
+		os=-uxpv
+		;;
+	*-rom68k)
+		os=-coff
+		;;
+	*-*bug)
+		os=-coff
+		;;
+	*-apple)
+		os=-macos
+		;;
+	*-atari*)
+		os=-mint
+		;;
+	*)
+		os=-none
+		;;
+esac
+fi
+
+# Here we handle the case where we know the os, and the CPU type, but not the
+# manufacturer.  We pick the logical manufacturer.
+vendor=unknown
+case $basic_machine in
+	*-unknown)
+		case $os in
+			-riscix*)
+				vendor=acorn
+				;;
+			-sunos*)
+				vendor=sun
+				;;
+			-cnk*|-aix*)
+				vendor=ibm
+				;;
+			-beos*)
+				vendor=be
+				;;
+			-hpux*)
+				vendor=hp
+				;;
+			-mpeix*)
+				vendor=hp
+				;;
+			-hiux*)
+				vendor=hitachi
+				;;
+			-unos*)
+				vendor=crds
+				;;
+			-dgux*)
+				vendor=dg
+				;;
+			-luna*)
+				vendor=omron
+				;;
+			-genix*)
+				vendor=ns
+				;;
+			-mvs* | -opened*)
+				vendor=ibm
+				;;
+			-os400*)
+				vendor=ibm
+				;;
+			-ptx*)
+				vendor=sequent
+				;;
+			-tpf*)
+				vendor=ibm
+				;;
+			-vxsim* | -vxworks* | -windiss*)
+				vendor=wrs
+				;;
+			-aux*)
+				vendor=apple
+				;;
+			-hms*)
+				vendor=hitachi
+				;;
+			-mpw* | -macos*)
+				vendor=apple
+				;;
+			-*mint | -mint[0-9]* | -*MiNT | -MiNT[0-9]*)
+				vendor=atari
+				;;
+			-vos*)
+				vendor=stratus
+				;;
+		esac
+		basic_machine=`echo $basic_machine | sed "s/unknown/$vendor/"`
+		;;
+esac
+
+echo $basic_machine$os
+exit
+
+# Local variables:
+# eval: (add-hook 'write-file-hooks 'time-stamp)
+# time-stamp-start: "timestamp='"
+# time-stamp-format: "%:y-%02m-%02d"
+# time-stamp-end: "'"
+# End:
diff --git a/deps/jemalloc/build-aux/install-sh b/deps/jemalloc/build-aux/install-sh
new file mode 100755
index 000000000..ebc66913e
--- /dev/null
+++ b/deps/jemalloc/build-aux/install-sh
@@ -0,0 +1,250 @@
+#! /bin/sh
+#
+# install - install a program, script, or datafile
+# This comes from X11R5 (mit/util/scripts/install.sh).
+#
+# Copyright 1991 by the Massachusetts Institute of Technology
+#
+# Permission to use, copy, modify, distribute, and sell this software and its
+# documentation for any purpose is hereby granted without fee, provided that
+# the above copyright notice appear in all copies and that both that
+# copyright notice and this permission notice appear in supporting
+# documentation, and that the name of M.I.T. not be used in advertising or
+# publicity pertaining to distribution of the software without specific,
+# written prior permission.  M.I.T. makes no representations about the
+# suitability of this software for any purpose.  It is provided "as is"
+# without express or implied warranty.
+#
+# Calling this script install-sh is preferred over install.sh, to prevent
+# `make' implicit rules from creating a file called install from it
+# when there is no Makefile.
+#
+# This script is compatible with the BSD install script, but was written
+# from scratch.  It can only install one file at a time, a restriction
+# shared with many OS's install programs.
+
+
+# set DOITPROG to echo to test this script
+
+# Don't use :- since 4.3BSD and earlier shells don't like it.
+doit="${DOITPROG-}"
+
+
+# put in absolute paths if you don't have them in your path; or use env. vars.
+
+mvprog="${MVPROG-mv}"
+cpprog="${CPPROG-cp}"
+chmodprog="${CHMODPROG-chmod}"
+chownprog="${CHOWNPROG-chown}"
+chgrpprog="${CHGRPPROG-chgrp}"
+stripprog="${STRIPPROG-strip}"
+rmprog="${RMPROG-rm}"
+mkdirprog="${MKDIRPROG-mkdir}"
+
+transformbasename=""
+transform_arg=""
+instcmd="$mvprog"
+chmodcmd="$chmodprog 0755"
+chowncmd=""
+chgrpcmd=""
+stripcmd=""
+rmcmd="$rmprog -f"
+mvcmd="$mvprog"
+src=""
+dst=""
+dir_arg=""
+
+while [ x"$1" != x ]; do
+    case $1 in
+	-c) instcmd="$cpprog"
+	    shift
+	    continue;;
+
+	-d) dir_arg=true
+	    shift
+	    continue;;
+
+	-m) chmodcmd="$chmodprog $2"
+	    shift
+	    shift
+	    continue;;
+
+	-o) chowncmd="$chownprog $2"
+	    shift
+	    shift
+	    continue;;
+
+	-g) chgrpcmd="$chgrpprog $2"
+	    shift
+	    shift
+	    continue;;
+
+	-s) stripcmd="$stripprog"
+	    shift
+	    continue;;
+
+	-t=*) transformarg=`echo $1 | sed 's/-t=//'`
+	    shift
+	    continue;;
+
+	-b=*) transformbasename=`echo $1 | sed 's/-b=//'`
+	    shift
+	    continue;;
+
+	*)  if [ x"$src" = x ]
+	    then
+		src=$1
+	    else
+		# this colon is to work around a 386BSD /bin/sh bug
+		:
+		dst=$1
+	    fi
+	    shift
+	    continue;;
+    esac
+done
+
+if [ x"$src" = x ]
+then
+	echo "install:	no input file specified"
+	exit 1
+else
+	true
+fi
+
+if [ x"$dir_arg" != x ]; then
+	dst=$src
+	src=""
+	
+	if [ -d $dst ]; then
+		instcmd=:
+	else
+		instcmd=mkdir
+	fi
+else
+
+# Waiting for this to be detected by the "$instcmd $src $dsttmp" command
+# might cause directories to be created, which would be especially bad 
+# if $src (and thus $dsttmp) contains '*'.
+
+	if [ -f $src -o -d $src ]
+	then
+		true
+	else
+		echo "install:  $src does not exist"
+		exit 1
+	fi
+	
+	if [ x"$dst" = x ]
+	then
+		echo "install:	no destination specified"
+		exit 1
+	else
+		true
+	fi
+
+# If destination is a directory, append the input filename; if your system
+# does not like double slashes in filenames, you may need to add some logic
+
+	if [ -d $dst ]
+	then
+		dst="$dst"/`basename $src`
+	else
+		true
+	fi
+fi
+
+## this sed command emulates the dirname command
+dstdir=`echo $dst | sed -e 's,[^/]*$,,;s,/$,,;s,^$,.,'`
+
+# Make sure that the destination directory exists.
+#  this part is taken from Noah Friedman's mkinstalldirs script
+
+# Skip lots of stat calls in the usual case.
+if [ ! -d "$dstdir" ]; then
+defaultIFS='	
+'
+IFS="${IFS-${defaultIFS}}"
+
+oIFS="${IFS}"
+# Some sh's can't handle IFS=/ for some reason.
+IFS='%'
+set - `echo ${dstdir} | sed -e 's@/@%@g' -e 's@^%@/@'`
+IFS="${oIFS}"
+
+pathcomp=''
+
+while [ $# -ne 0 ] ; do
+	pathcomp="${pathcomp}${1}"
+	shift
+
+	if [ ! -d "${pathcomp}" ] ;
+        then
+		$mkdirprog "${pathcomp}"
+	else
+		true
+	fi
+
+	pathcomp="${pathcomp}/"
+done
+fi
+
+if [ x"$dir_arg" != x ]
+then
+	$doit $instcmd $dst &&
+
+	if [ x"$chowncmd" != x ]; then $doit $chowncmd $dst; else true ; fi &&
+	if [ x"$chgrpcmd" != x ]; then $doit $chgrpcmd $dst; else true ; fi &&
+	if [ x"$stripcmd" != x ]; then $doit $stripcmd $dst; else true ; fi &&
+	if [ x"$chmodcmd" != x ]; then $doit $chmodcmd $dst; else true ; fi
+else
+
+# If we're going to rename the final executable, determine the name now.
+
+	if [ x"$transformarg" = x ] 
+	then
+		dstfile=`basename $dst`
+	else
+		dstfile=`basename $dst $transformbasename | 
+			sed $transformarg`$transformbasename
+	fi
+
+# don't allow the sed command to completely eliminate the filename
+
+	if [ x"$dstfile" = x ] 
+	then
+		dstfile=`basename $dst`
+	else
+		true
+	fi
+
+# Make a temp file name in the proper directory.
+
+	dsttmp=$dstdir/#inst.$$#
+
+# Move or copy the file name to the temp name
+
+	$doit $instcmd $src $dsttmp &&
+
+	trap "rm -f ${dsttmp}" 0 &&
+
+# and set any options; do chmod last to preserve setuid bits
+
+# If any of these fail, we abort the whole thing.  If we want to
+# ignore errors from any of these, just make sure not to ignore
+# errors from the above "$doit $instcmd $src $dsttmp" command.
+
+	if [ x"$chowncmd" != x ]; then $doit $chowncmd $dsttmp; else true;fi &&
+	if [ x"$chgrpcmd" != x ]; then $doit $chgrpcmd $dsttmp; else true;fi &&
+	if [ x"$stripcmd" != x ]; then $doit $stripcmd $dsttmp; else true;fi &&
+	if [ x"$chmodcmd" != x ]; then $doit $chmodcmd $dsttmp; else true;fi &&
+
+# Now rename the file to the real destination.
+
+	$doit $rmcmd -f $dstdir/$dstfile &&
+	$doit $mvcmd $dsttmp $dstdir/$dstfile 
+
+fi &&
+
+
+exit 0
diff --git a/deps/jemalloc/config.guess b/deps/jemalloc/config.guess
deleted file mode 100755
index 1f5c50c0d..000000000
--- a/deps/jemalloc/config.guess
+++ /dev/null
@@ -1,1420 +0,0 @@
-#! /bin/sh
-# Attempt to guess a canonical system name.
-#   Copyright 1992-2014 Free Software Foundation, Inc.
-
-timestamp='2014-03-23'
-
-# This file is free software; you can redistribute it and/or modify it
-# under the terms of the GNU General Public License as published by
-# the Free Software Foundation; either version 3 of the License, or
-# (at your option) any later version.
-#
-# This program is distributed in the hope that it will be useful, but
-# WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
-# General Public License for more details.
-#
-# You should have received a copy of the GNU General Public License
-# along with this program; if not, see <http://www.gnu.org/licenses/>.
-#
-# As a special exception to the GNU General Public License, if you
-# distribute this file as part of a program that contains a
-# configuration script generated by Autoconf, you may include it under
-# the same distribution terms that you use for the rest of that
-# program.  This Exception is an additional permission under section 7
-# of the GNU General Public License, version 3 ("GPLv3").
-#
-# Originally written by Per Bothner.
-#
-# You can get the latest version of this script from:
-# http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.guess;hb=HEAD
-#
-# Please send patches with a ChangeLog entry to config-patches@gnu.org.
-
-
-me=`echo "$0" | sed -e 's,.*/,,'`
-
-usage="\
-Usage: $0 [OPTION]
-
-Output the configuration name of the system \`$me' is run on.
-
-Operation modes:
-  -h, --help         print this help, then exit
-  -t, --time-stamp   print date of last modification, then exit
-  -v, --version      print version number, then exit
-
-Report bugs and patches to <config-patches@gnu.org>."
-
-version="\
-GNU config.guess ($timestamp)
-
-Originally written by Per Bothner.
-Copyright 1992-2014 Free Software Foundation, Inc.
-
-This is free software; see the source for copying conditions.  There is NO
-warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE."
-
-help="
-Try \`$me --help' for more information."
-
-# Parse command line
-while test $# -gt 0 ; do
-  case $1 in
-    --time-stamp | --time* | -t )
-       echo "$timestamp" ; exit ;;
-    --version | -v )
-       echo "$version" ; exit ;;
-    --help | --h* | -h )
-       echo "$usage"; exit ;;
-    -- )     # Stop option processing
-       shift; break ;;
-    - )	# Use stdin as input.
-       break ;;
-    -* )
-       echo "$me: invalid option $1$help" >&2
-       exit 1 ;;
-    * )
-       break ;;
-  esac
-done
-
-if test $# != 0; then
-  echo "$me: too many arguments$help" >&2
-  exit 1
-fi
-
-trap 'exit 1' 1 2 15
-
-# CC_FOR_BUILD -- compiler used by this script. Note that the use of a
-# compiler to aid in system detection is discouraged as it requires
-# temporary files to be created and, as you can see below, it is a
-# headache to deal with in a portable fashion.
-
-# Historically, `CC_FOR_BUILD' used to be named `HOST_CC'. We still
-# use `HOST_CC' if defined, but it is deprecated.
-
-# Portable tmp directory creation inspired by the Autoconf team.
-
-set_cc_for_build='
-trap "exitcode=\$?; (rm -f \$tmpfiles 2>/dev/null; rmdir \$tmp 2>/dev/null) && exit \$exitcode" 0 ;
-trap "rm -f \$tmpfiles 2>/dev/null; rmdir \$tmp 2>/dev/null; exit 1" 1 2 13 15 ;
-: ${TMPDIR=/tmp} ;
- { tmp=`(umask 077 && mktemp -d "$TMPDIR/cgXXXXXX") 2>/dev/null` && test -n "$tmp" && test -d "$tmp" ; } ||
- { test -n "$RANDOM" && tmp=$TMPDIR/cg$$-$RANDOM && (umask 077 && mkdir $tmp) ; } ||
- { tmp=$TMPDIR/cg-$$ && (umask 077 && mkdir $tmp) && echo "Warning: creating insecure temp directory" >&2 ; } ||
- { echo "$me: cannot create a temporary directory in $TMPDIR" >&2 ; exit 1 ; } ;
-dummy=$tmp/dummy ;
-tmpfiles="$dummy.c $dummy.o $dummy.rel $dummy" ;
-case $CC_FOR_BUILD,$HOST_CC,$CC in
- ,,)    echo "int x;" > $dummy.c ;
-	for c in cc gcc c89 c99 ; do
-	  if ($c -c -o $dummy.o $dummy.c) >/dev/null 2>&1 ; then
-	     CC_FOR_BUILD="$c"; break ;
-	  fi ;
-	done ;
-	if test x"$CC_FOR_BUILD" = x ; then
-	  CC_FOR_BUILD=no_compiler_found ;
-	fi
-	;;
- ,,*)   CC_FOR_BUILD=$CC ;;
- ,*,*)  CC_FOR_BUILD=$HOST_CC ;;
-esac ; set_cc_for_build= ;'
-
-# This is needed to find uname on a Pyramid OSx when run in the BSD universe.
-# (ghazi@noc.rutgers.edu 1994-08-24)
-if (test -f /.attbin/uname) >/dev/null 2>&1 ; then
-	PATH=$PATH:/.attbin ; export PATH
-fi
-
-UNAME_MACHINE=`(uname -m) 2>/dev/null` || UNAME_MACHINE=unknown
-UNAME_RELEASE=`(uname -r) 2>/dev/null` || UNAME_RELEASE=unknown
-UNAME_SYSTEM=`(uname -s) 2>/dev/null`  || UNAME_SYSTEM=unknown
-UNAME_VERSION=`(uname -v) 2>/dev/null` || UNAME_VERSION=unknown
-
-case "${UNAME_SYSTEM}" in
-Linux|GNU|GNU/*)
-	# If the system lacks a compiler, then just pick glibc.
-	# We could probably try harder.
-	LIBC=gnu
-
-	eval $set_cc_for_build
-	cat <<-EOF > $dummy.c
-	#include <features.h>
-	#if defined(__UCLIBC__)
-	LIBC=uclibc
-	#elif defined(__dietlibc__)
-	LIBC=dietlibc
-	#else
-	LIBC=gnu
-	#endif
-	EOF
-	eval `$CC_FOR_BUILD -E $dummy.c 2>/dev/null | grep '^LIBC' | sed 's, ,,g'`
-	;;
-esac
-
-# Note: order is significant - the case branches are not exclusive.
-
-case "${UNAME_MACHINE}:${UNAME_SYSTEM}:${UNAME_RELEASE}:${UNAME_VERSION}" in
-    *:NetBSD:*:*)
-	# NetBSD (nbsd) targets should (where applicable) match one or
-	# more of the tuples: *-*-netbsdelf*, *-*-netbsdaout*,
-	# *-*-netbsdecoff* and *-*-netbsd*.  For targets that recently
-	# switched to ELF, *-*-netbsd* would select the old
-	# object file format.  This provides both forward
-	# compatibility and a consistent mechanism for selecting the
-	# object file format.
-	#
-	# Note: NetBSD doesn't particularly care about the vendor
-	# portion of the name.  We always set it to "unknown".
-	sysctl="sysctl -n hw.machine_arch"
-	UNAME_MACHINE_ARCH=`(/sbin/$sysctl 2>/dev/null || \
-	    /usr/sbin/$sysctl 2>/dev/null || echo unknown)`
-	case "${UNAME_MACHINE_ARCH}" in
-	    armeb) machine=armeb-unknown ;;
-	    arm*) machine=arm-unknown ;;
-	    sh3el) machine=shl-unknown ;;
-	    sh3eb) machine=sh-unknown ;;
-	    sh5el) machine=sh5le-unknown ;;
-	    *) machine=${UNAME_MACHINE_ARCH}-unknown ;;
-	esac
-	# The Operating System including object format, if it has switched
-	# to ELF recently, or will in the future.
-	case "${UNAME_MACHINE_ARCH}" in
-	    arm*|i386|m68k|ns32k|sh3*|sparc|vax)
-		eval $set_cc_for_build
-		if echo __ELF__ | $CC_FOR_BUILD -E - 2>/dev/null \
-			| grep -q __ELF__
-		then
-		    # Once all utilities can be ECOFF (netbsdecoff) or a.out (netbsdaout).
-		    # Return netbsd for either.  FIX?
-		    os=netbsd
-		else
-		    os=netbsdelf
-		fi
-		;;
-	    *)
-		os=netbsd
-		;;
-	esac
-	# The OS release
-	# Debian GNU/NetBSD machines have a different userland, and
-	# thus, need a distinct triplet. However, they do not need
-	# kernel version information, so it can be replaced with a
-	# suitable tag, in the style of linux-gnu.
-	case "${UNAME_VERSION}" in
-	    Debian*)
-		release='-gnu'
-		;;
-	    *)
-		release=`echo ${UNAME_RELEASE}|sed -e 's/[-_].*/\./'`
-		;;
-	esac
-	# Since CPU_TYPE-MANUFACTURER-KERNEL-OPERATING_SYSTEM:
-	# contains redundant information, the shorter form:
-	# CPU_TYPE-MANUFACTURER-OPERATING_SYSTEM is used.
-	echo "${machine}-${os}${release}"
-	exit ;;
-    *:Bitrig:*:*)
-	UNAME_MACHINE_ARCH=`arch | sed 's/Bitrig.//'`
-	echo ${UNAME_MACHINE_ARCH}-unknown-bitrig${UNAME_RELEASE}
-	exit ;;
-    *:OpenBSD:*:*)
-	UNAME_MACHINE_ARCH=`arch | sed 's/OpenBSD.//'`
-	echo ${UNAME_MACHINE_ARCH}-unknown-openbsd${UNAME_RELEASE}
-	exit ;;
-    *:ekkoBSD:*:*)
-	echo ${UNAME_MACHINE}-unknown-ekkobsd${UNAME_RELEASE}
-	exit ;;
-    *:SolidBSD:*:*)
-	echo ${UNAME_MACHINE}-unknown-solidbsd${UNAME_RELEASE}
-	exit ;;
-    macppc:MirBSD:*:*)
-	echo powerpc-unknown-mirbsd${UNAME_RELEASE}
-	exit ;;
-    *:MirBSD:*:*)
-	echo ${UNAME_MACHINE}-unknown-mirbsd${UNAME_RELEASE}
-	exit ;;
-    alpha:OSF1:*:*)
-	case $UNAME_RELEASE in
-	*4.0)
-		UNAME_RELEASE=`/usr/sbin/sizer -v | awk '{print $3}'`
-		;;
-	*5.*)
-		UNAME_RELEASE=`/usr/sbin/sizer -v | awk '{print $4}'`
-		;;
-	esac
-	# According to Compaq, /usr/sbin/psrinfo has been available on
-	# OSF/1 and Tru64 systems produced since 1995.  I hope that
-	# covers most systems running today.  This code pipes the CPU
-	# types through head -n 1, so we only detect the type of CPU 0.
-	ALPHA_CPU_TYPE=`/usr/sbin/psrinfo -v | sed -n -e 's/^  The alpha \(.*\) processor.*$/\1/p' | head -n 1`
-	case "$ALPHA_CPU_TYPE" in
-	    "EV4 (21064)")
-		UNAME_MACHINE="alpha" ;;
-	    "EV4.5 (21064)")
-		UNAME_MACHINE="alpha" ;;
-	    "LCA4 (21066/21068)")
-		UNAME_MACHINE="alpha" ;;
-	    "EV5 (21164)")
-		UNAME_MACHINE="alphaev5" ;;
-	    "EV5.6 (21164A)")
-		UNAME_MACHINE="alphaev56" ;;
-	    "EV5.6 (21164PC)")
-		UNAME_MACHINE="alphapca56" ;;
-	    "EV5.7 (21164PC)")
-		UNAME_MACHINE="alphapca57" ;;
-	    "EV6 (21264)")
-		UNAME_MACHINE="alphaev6" ;;
-	    "EV6.7 (21264A)")
-		UNAME_MACHINE="alphaev67" ;;
-	    "EV6.8CB (21264C)")
-		UNAME_MACHINE="alphaev68" ;;
-	    "EV6.8AL (21264B)")
-		UNAME_MACHINE="alphaev68" ;;
-	    "EV6.8CX (21264D)")
-		UNAME_MACHINE="alphaev68" ;;
-	    "EV6.9A (21264/EV69A)")
-		UNAME_MACHINE="alphaev69" ;;
-	    "EV7 (21364)")
-		UNAME_MACHINE="alphaev7" ;;
-	    "EV7.9 (21364A)")
-		UNAME_MACHINE="alphaev79" ;;
-	esac
-	# A Pn.n version is a patched version.
-	# A Vn.n version is a released version.
-	# A Tn.n version is a released field test version.
-	# A Xn.n version is an unreleased experimental baselevel.
-	# 1.2 uses "1.2" for uname -r.
-	echo ${UNAME_MACHINE}-dec-osf`echo ${UNAME_RELEASE} | sed -e 's/^[PVTX]//' | tr 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' 'abcdefghijklmnopqrstuvwxyz'`
-	# Reset EXIT trap before exiting to avoid spurious non-zero exit code.
-	exitcode=$?
-	trap '' 0
-	exit $exitcode ;;
-    Alpha\ *:Windows_NT*:*)
-	# How do we know it's Interix rather than the generic POSIX subsystem?
-	# Should we change UNAME_MACHINE based on the output of uname instead
-	# of the specific Alpha model?
-	echo alpha-pc-interix
-	exit ;;
-    21064:Windows_NT:50:3)
-	echo alpha-dec-winnt3.5
-	exit ;;
-    Amiga*:UNIX_System_V:4.0:*)
-	echo m68k-unknown-sysv4
-	exit ;;
-    *:[Aa]miga[Oo][Ss]:*:*)
-	echo ${UNAME_MACHINE}-unknown-amigaos
-	exit ;;
-    *:[Mm]orph[Oo][Ss]:*:*)
-	echo ${UNAME_MACHINE}-unknown-morphos
-	exit ;;
-    *:OS/390:*:*)
-	echo i370-ibm-openedition
-	exit ;;
-    *:z/VM:*:*)
-	echo s390-ibm-zvmoe
-	exit ;;
-    *:OS400:*:*)
-	echo powerpc-ibm-os400
-	exit ;;
-    arm:RISC*:1.[012]*:*|arm:riscix:1.[012]*:*)
-	echo arm-acorn-riscix${UNAME_RELEASE}
-	exit ;;
-    arm*:riscos:*:*|arm*:RISCOS:*:*)
-	echo arm-unknown-riscos
-	exit ;;
-    SR2?01:HI-UX/MPP:*:* | SR8000:HI-UX/MPP:*:*)
-	echo hppa1.1-hitachi-hiuxmpp
-	exit ;;
-    Pyramid*:OSx*:*:* | MIS*:OSx*:*:* | MIS*:SMP_DC-OSx*:*:*)
-	# akee@wpdis03.wpafb.af.mil (Earle F. Ake) contributed MIS and NILE.
-	if test "`(/bin/universe) 2>/dev/null`" = att ; then
-		echo pyramid-pyramid-sysv3
-	else
-		echo pyramid-pyramid-bsd
-	fi
-	exit ;;
-    NILE*:*:*:dcosx)
-	echo pyramid-pyramid-svr4
-	exit ;;
-    DRS?6000:unix:4.0:6*)
-	echo sparc-icl-nx6
-	exit ;;
-    DRS?6000:UNIX_SV:4.2*:7* | DRS?6000:isis:4.2*:7*)
-	case `/usr/bin/uname -p` in
-	    sparc) echo sparc-icl-nx7; exit ;;
-	esac ;;
-    s390x:SunOS:*:*)
-	echo ${UNAME_MACHINE}-ibm-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'`
-	exit ;;
-    sun4H:SunOS:5.*:*)
-	echo sparc-hal-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'`
-	exit ;;
-    sun4*:SunOS:5.*:* | tadpole*:SunOS:5.*:*)
-	echo sparc-sun-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'`
-	exit ;;
-    i86pc:AuroraUX:5.*:* | i86xen:AuroraUX:5.*:*)
-	echo i386-pc-auroraux${UNAME_RELEASE}
-	exit ;;
-    i86pc:SunOS:5.*:* | i86xen:SunOS:5.*:*)
-	eval $set_cc_for_build
-	SUN_ARCH="i386"
-	# If there is a compiler, see if it is configured for 64-bit objects.
-	# Note that the Sun cc does not turn __LP64__ into 1 like gcc does.
-	# This test works for both compilers.
-	if [ "$CC_FOR_BUILD" != 'no_compiler_found' ]; then
-	    if (echo '#ifdef __amd64'; echo IS_64BIT_ARCH; echo '#endif') | \
-		(CCOPTS= $CC_FOR_BUILD -E - 2>/dev/null) | \
-		grep IS_64BIT_ARCH >/dev/null
-	    then
-		SUN_ARCH="x86_64"
-	    fi
-	fi
-	echo ${SUN_ARCH}-pc-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'`
-	exit ;;
-    sun4*:SunOS:6*:*)
-	# According to config.sub, this is the proper way to canonicalize
-	# SunOS6.  Hard to guess exactly what SunOS6 will be like, but
-	# it's likely to be more like Solaris than SunOS4.
-	echo sparc-sun-solaris3`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'`
-	exit ;;
-    sun4*:SunOS:*:*)
-	case "`/usr/bin/arch -k`" in
-	    Series*|S4*)
-		UNAME_RELEASE=`uname -v`
-		;;
-	esac
-	# Japanese Language versions have a version number like `4.1.3-JL'.
-	echo sparc-sun-sunos`echo ${UNAME_RELEASE}|sed -e 's/-/_/'`
-	exit ;;
-    sun3*:SunOS:*:*)
-	echo m68k-sun-sunos${UNAME_RELEASE}
-	exit ;;
-    sun*:*:4.2BSD:*)
-	UNAME_RELEASE=`(sed 1q /etc/motd | awk '{print substr($5,1,3)}') 2>/dev/null`
-	test "x${UNAME_RELEASE}" = "x" && UNAME_RELEASE=3
-	case "`/bin/arch`" in
-	    sun3)
-		echo m68k-sun-sunos${UNAME_RELEASE}
-		;;
-	    sun4)
-		echo sparc-sun-sunos${UNAME_RELEASE}
-		;;
-	esac
-	exit ;;
-    aushp:SunOS:*:*)
-	echo sparc-auspex-sunos${UNAME_RELEASE}
-	exit ;;
-    # The situation for MiNT is a little confusing.  The machine name
-    # can be virtually everything (everything which is not
-    # "atarist" or "atariste" at least should have a processor
-    # > m68000).  The system name ranges from "MiNT" over "FreeMiNT"
-    # to the lowercase version "mint" (or "freemint").  Finally
-    # the system name "TOS" denotes a system which is actually not
-    # MiNT.  But MiNT is downward compatible to TOS, so this should
-    # be no problem.
-    atarist[e]:*MiNT:*:* | atarist[e]:*mint:*:* | atarist[e]:*TOS:*:*)
-	echo m68k-atari-mint${UNAME_RELEASE}
-	exit ;;
-    atari*:*MiNT:*:* | atari*:*mint:*:* | atarist[e]:*TOS:*:*)
-	echo m68k-atari-mint${UNAME_RELEASE}
-	exit ;;
-    *falcon*:*MiNT:*:* | *falcon*:*mint:*:* | *falcon*:*TOS:*:*)
-	echo m68k-atari-mint${UNAME_RELEASE}
-	exit ;;
-    milan*:*MiNT:*:* | milan*:*mint:*:* | *milan*:*TOS:*:*)
-	echo m68k-milan-mint${UNAME_RELEASE}
-	exit ;;
-    hades*:*MiNT:*:* | hades*:*mint:*:* | *hades*:*TOS:*:*)
-	echo m68k-hades-mint${UNAME_RELEASE}
-	exit ;;
-    *:*MiNT:*:* | *:*mint:*:* | *:*TOS:*:*)
-	echo m68k-unknown-mint${UNAME_RELEASE}
-	exit ;;
-    m68k:machten:*:*)
-	echo m68k-apple-machten${UNAME_RELEASE}
-	exit ;;
-    powerpc:machten:*:*)
-	echo powerpc-apple-machten${UNAME_RELEASE}
-	exit ;;
-    RISC*:Mach:*:*)
-	echo mips-dec-mach_bsd4.3
-	exit ;;
-    RISC*:ULTRIX:*:*)
-	echo mips-dec-ultrix${UNAME_RELEASE}
-	exit ;;
-    VAX*:ULTRIX*:*:*)
-	echo vax-dec-ultrix${UNAME_RELEASE}
-	exit ;;
-    2020:CLIX:*:* | 2430:CLIX:*:*)
-	echo clipper-intergraph-clix${UNAME_RELEASE}
-	exit ;;
-    mips:*:*:UMIPS | mips:*:*:RISCos)
-	eval $set_cc_for_build
-	sed 's/^	//' << EOF >$dummy.c
-#ifdef __cplusplus
-#include <stdio.h>  /* for printf() prototype */
-	int main (int argc, char *argv[]) {
-#else
-	int main (argc, argv) int argc; char *argv[]; {
-#endif
-	#if defined (host_mips) && defined (MIPSEB)
-	#if defined (SYSTYPE_SYSV)
-	  printf ("mips-mips-riscos%ssysv\n", argv[1]); exit (0);
-	#endif
-	#if defined (SYSTYPE_SVR4)
-	  printf ("mips-mips-riscos%ssvr4\n", argv[1]); exit (0);
-	#endif
-	#if defined (SYSTYPE_BSD43) || defined(SYSTYPE_BSD)
-	  printf ("mips-mips-riscos%sbsd\n", argv[1]); exit (0);
-	#endif
-	#endif
-	  exit (-1);
-	}
-EOF
-	$CC_FOR_BUILD -o $dummy $dummy.c &&
-	  dummyarg=`echo "${UNAME_RELEASE}" | sed -n 's/\([0-9]*\).*/\1/p'` &&
-	  SYSTEM_NAME=`$dummy $dummyarg` &&
-	    { echo "$SYSTEM_NAME"; exit; }
-	echo mips-mips-riscos${UNAME_RELEASE}
-	exit ;;
-    Motorola:PowerMAX_OS:*:*)
-	echo powerpc-motorola-powermax
-	exit ;;
-    Motorola:*:4.3:PL8-*)
-	echo powerpc-harris-powermax
-	exit ;;
-    Night_Hawk:*:*:PowerMAX_OS | Synergy:PowerMAX_OS:*:*)
-	echo powerpc-harris-powermax
-	exit ;;
-    Night_Hawk:Power_UNIX:*:*)
-	echo powerpc-harris-powerunix
-	exit ;;
-    m88k:CX/UX:7*:*)
-	echo m88k-harris-cxux7
-	exit ;;
-    m88k:*:4*:R4*)
-	echo m88k-motorola-sysv4
-	exit ;;
-    m88k:*:3*:R3*)
-	echo m88k-motorola-sysv3
-	exit ;;
-    AViiON:dgux:*:*)
-	# DG/UX returns AViiON for all architectures
-	UNAME_PROCESSOR=`/usr/bin/uname -p`
-	if [ $UNAME_PROCESSOR = mc88100 ] || [ $UNAME_PROCESSOR = mc88110 ]
-	then
-	    if [ ${TARGET_BINARY_INTERFACE}x = m88kdguxelfx ] || \
-	       [ ${TARGET_BINARY_INTERFACE}x = x ]
-	    then
-		echo m88k-dg-dgux${UNAME_RELEASE}
-	    else
-		echo m88k-dg-dguxbcs${UNAME_RELEASE}
-	    fi
-	else
-	    echo i586-dg-dgux${UNAME_RELEASE}
-	fi
-	exit ;;
-    M88*:DolphinOS:*:*)	# DolphinOS (SVR3)
-	echo m88k-dolphin-sysv3
-	exit ;;
-    M88*:*:R3*:*)
-	# Delta 88k system running SVR3
-	echo m88k-motorola-sysv3
-	exit ;;
-    XD88*:*:*:*) # Tektronix XD88 system running UTekV (SVR3)
-	echo m88k-tektronix-sysv3
-	exit ;;
-    Tek43[0-9][0-9]:UTek:*:*) # Tektronix 4300 system running UTek (BSD)
-	echo m68k-tektronix-bsd
-	exit ;;
-    *:IRIX*:*:*)
-	echo mips-sgi-irix`echo ${UNAME_RELEASE}|sed -e 's/-/_/g'`
-	exit ;;
-    ????????:AIX?:[12].1:2)   # AIX 2.2.1 or AIX 2.1.1 is RT/PC AIX.
-	echo romp-ibm-aix     # uname -m gives an 8 hex-code CPU id
-	exit ;;               # Note that: echo "'`uname -s`'" gives 'AIX '
-    i*86:AIX:*:*)
-	echo i386-ibm-aix
-	exit ;;
-    ia64:AIX:*:*)
-	if [ -x /usr/bin/oslevel ] ; then
-		IBM_REV=`/usr/bin/oslevel`
-	else
-		IBM_REV=${UNAME_VERSION}.${UNAME_RELEASE}
-	fi
-	echo ${UNAME_MACHINE}-ibm-aix${IBM_REV}
-	exit ;;
-    *:AIX:2:3)
-	if grep bos325 /usr/include/stdio.h >/dev/null 2>&1; then
-		eval $set_cc_for_build
-		sed 's/^		//' << EOF >$dummy.c
-		#include <sys/systemcfg.h>
-
-		main()
-			{
-			if (!__power_pc())
-				exit(1);
-			puts("powerpc-ibm-aix3.2.5");
-			exit(0);
-			}
-EOF
-		if $CC_FOR_BUILD -o $dummy $dummy.c && SYSTEM_NAME=`$dummy`
-		then
-			echo "$SYSTEM_NAME"
-		else
-			echo rs6000-ibm-aix3.2.5
-		fi
-	elif grep bos324 /usr/include/stdio.h >/dev/null 2>&1; then
-		echo rs6000-ibm-aix3.2.4
-	else
-		echo rs6000-ibm-aix3.2
-	fi
-	exit ;;
-    *:AIX:*:[4567])
-	IBM_CPU_ID=`/usr/sbin/lsdev -C -c processor -S available | sed 1q | awk '{ print $1 }'`
-	if /usr/sbin/lsattr -El ${IBM_CPU_ID} | grep ' POWER' >/dev/null 2>&1; then
-		IBM_ARCH=rs6000
-	else
-		IBM_ARCH=powerpc
-	fi
-	if [ -x /usr/bin/oslevel ] ; then
-		IBM_REV=`/usr/bin/oslevel`
-	else
-		IBM_REV=${UNAME_VERSION}.${UNAME_RELEASE}
-	fi
-	echo ${IBM_ARCH}-ibm-aix${IBM_REV}
-	exit ;;
-    *:AIX:*:*)
-	echo rs6000-ibm-aix
-	exit ;;
-    ibmrt:4.4BSD:*|romp-ibm:BSD:*)
-	echo romp-ibm-bsd4.4
-	exit ;;
-    ibmrt:*BSD:*|romp-ibm:BSD:*)            # covers RT/PC BSD and
-	echo romp-ibm-bsd${UNAME_RELEASE}   # 4.3 with uname added to
-	exit ;;                             # report: romp-ibm BSD 4.3
-    *:BOSX:*:*)
-	echo rs6000-bull-bosx
-	exit ;;
-    DPX/2?00:B.O.S.:*:*)
-	echo m68k-bull-sysv3
-	exit ;;
-    9000/[34]??:4.3bsd:1.*:*)
-	echo m68k-hp-bsd
-	exit ;;
-    hp300:4.4BSD:*:* | 9000/[34]??:4.3bsd:2.*:*)
-	echo m68k-hp-bsd4.4
-	exit ;;
-    9000/[34678]??:HP-UX:*:*)
-	HPUX_REV=`echo ${UNAME_RELEASE}|sed -e 's/[^.]*.[0B]*//'`
-	case "${UNAME_MACHINE}" in
-	    9000/31? )            HP_ARCH=m68000 ;;
-	    9000/[34]?? )         HP_ARCH=m68k ;;
-	    9000/[678][0-9][0-9])
-		if [ -x /usr/bin/getconf ]; then
-		    sc_cpu_version=`/usr/bin/getconf SC_CPU_VERSION 2>/dev/null`
-		    sc_kernel_bits=`/usr/bin/getconf SC_KERNEL_BITS 2>/dev/null`
-		    case "${sc_cpu_version}" in
-		      523) HP_ARCH="hppa1.0" ;; # CPU_PA_RISC1_0
-		      528) HP_ARCH="hppa1.1" ;; # CPU_PA_RISC1_1
-		      532)                      # CPU_PA_RISC2_0
-			case "${sc_kernel_bits}" in
-			  32) HP_ARCH="hppa2.0n" ;;
-			  64) HP_ARCH="hppa2.0w" ;;
-			  '') HP_ARCH="hppa2.0" ;;   # HP-UX 10.20
-			esac ;;
-		    esac
-		fi
-		if [ "${HP_ARCH}" = "" ]; then
-		    eval $set_cc_for_build
-		    sed 's/^		//' << EOF >$dummy.c
-
-		#define _HPUX_SOURCE
-		#include <stdlib.h>
-		#include <unistd.h>
-
-		int main ()
-		{
-		#if defined(_SC_KERNEL_BITS)
-		    long bits = sysconf(_SC_KERNEL_BITS);
-		#endif
-		    long cpu  = sysconf (_SC_CPU_VERSION);
-
-		    switch (cpu)
-			{
-			case CPU_PA_RISC1_0: puts ("hppa1.0"); break;
-			case CPU_PA_RISC1_1: puts ("hppa1.1"); break;
-			case CPU_PA_RISC2_0:
-		#if defined(_SC_KERNEL_BITS)
-			    switch (bits)
-				{
-				case 64: puts ("hppa2.0w"); break;
-				case 32: puts ("hppa2.0n"); break;
-				default: puts ("hppa2.0"); break;
-				} break;
-		#else  /* !defined(_SC_KERNEL_BITS) */
-			    puts ("hppa2.0"); break;
-		#endif
-			default: puts ("hppa1.0"); break;
-			}
-		    exit (0);
-		}
-EOF
-		    (CCOPTS= $CC_FOR_BUILD -o $dummy $dummy.c 2>/dev/null) && HP_ARCH=`$dummy`
-		    test -z "$HP_ARCH" && HP_ARCH=hppa
-		fi ;;
-	esac
-	if [ ${HP_ARCH} = "hppa2.0w" ]
-	then
-	    eval $set_cc_for_build
-
-	    # hppa2.0w-hp-hpux* has a 64-bit kernel and a compiler generating
-	    # 32-bit code.  hppa64-hp-hpux* has the same kernel and a compiler
-	    # generating 64-bit code.  GNU and HP use different nomenclature:
-	    #
-	    # $ CC_FOR_BUILD=cc ./config.guess
-	    # => hppa2.0w-hp-hpux11.23
-	    # $ CC_FOR_BUILD="cc +DA2.0w" ./config.guess
-	    # => hppa64-hp-hpux11.23
-
-	    if echo __LP64__ | (CCOPTS= $CC_FOR_BUILD -E - 2>/dev/null) |
-		grep -q __LP64__
-	    then
-		HP_ARCH="hppa2.0w"
-	    else
-		HP_ARCH="hppa64"
-	    fi
-	fi
-	echo ${HP_ARCH}-hp-hpux${HPUX_REV}
-	exit ;;
-    ia64:HP-UX:*:*)
-	HPUX_REV=`echo ${UNAME_RELEASE}|sed -e 's/[^.]*.[0B]*//'`
-	echo ia64-hp-hpux${HPUX_REV}
-	exit ;;
-    3050*:HI-UX:*:*)
-	eval $set_cc_for_build
-	sed 's/^	//' << EOF >$dummy.c
-	#include <unistd.h>
-	int
-	main ()
-	{
-	  long cpu = sysconf (_SC_CPU_VERSION);
-	  /* The order matters, because CPU_IS_HP_MC68K erroneously returns
-	     true for CPU_PA_RISC1_0.  CPU_IS_PA_RISC returns correct
-	     results, however.  */
-	  if (CPU_IS_PA_RISC (cpu))
-	    {
-	      switch (cpu)
-		{
-		  case CPU_PA_RISC1_0: puts ("hppa1.0-hitachi-hiuxwe2"); break;
-		  case CPU_PA_RISC1_1: puts ("hppa1.1-hitachi-hiuxwe2"); break;
-		  case CPU_PA_RISC2_0: puts ("hppa2.0-hitachi-hiuxwe2"); break;
-		  default: puts ("hppa-hitachi-hiuxwe2"); break;
-		}
-	    }
-	  else if (CPU_IS_HP_MC68K (cpu))
-	    puts ("m68k-hitachi-hiuxwe2");
-	  else puts ("unknown-hitachi-hiuxwe2");
-	  exit (0);
-	}
-EOF
-	$CC_FOR_BUILD -o $dummy $dummy.c && SYSTEM_NAME=`$dummy` &&
-		{ echo "$SYSTEM_NAME"; exit; }
-	echo unknown-hitachi-hiuxwe2
-	exit ;;
-    9000/7??:4.3bsd:*:* | 9000/8?[79]:4.3bsd:*:* )
-	echo hppa1.1-hp-bsd
-	exit ;;
-    9000/8??:4.3bsd:*:*)
-	echo hppa1.0-hp-bsd
-	exit ;;
-    *9??*:MPE/iX:*:* | *3000*:MPE/iX:*:*)
-	echo hppa1.0-hp-mpeix
-	exit ;;
-    hp7??:OSF1:*:* | hp8?[79]:OSF1:*:* )
-	echo hppa1.1-hp-osf
-	exit ;;
-    hp8??:OSF1:*:*)
-	echo hppa1.0-hp-osf
-	exit ;;
-    i*86:OSF1:*:*)
-	if [ -x /usr/sbin/sysversion ] ; then
-	    echo ${UNAME_MACHINE}-unknown-osf1mk
-	else
-	    echo ${UNAME_MACHINE}-unknown-osf1
-	fi
-	exit ;;
-    parisc*:Lites*:*:*)
-	echo hppa1.1-hp-lites
-	exit ;;
-    C1*:ConvexOS:*:* | convex:ConvexOS:C1*:*)
-	echo c1-convex-bsd
-	exit ;;
-    C2*:ConvexOS:*:* | convex:ConvexOS:C2*:*)
-	if getsysinfo -f scalar_acc
-	then echo c32-convex-bsd
-	else echo c2-convex-bsd
-	fi
-	exit ;;
-    C34*:ConvexOS:*:* | convex:ConvexOS:C34*:*)
-	echo c34-convex-bsd
-	exit ;;
-    C38*:ConvexOS:*:* | convex:ConvexOS:C38*:*)
-	echo c38-convex-bsd
-	exit ;;
-    C4*:ConvexOS:*:* | convex:ConvexOS:C4*:*)
-	echo c4-convex-bsd
-	exit ;;
-    CRAY*Y-MP:*:*:*)
-	echo ymp-cray-unicos${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/'
-	exit ;;
-    CRAY*[A-Z]90:*:*:*)
-	echo ${UNAME_MACHINE}-cray-unicos${UNAME_RELEASE} \
-	| sed -e 's/CRAY.*\([A-Z]90\)/\1/' \
-	      -e y/ABCDEFGHIJKLMNOPQRSTUVWXYZ/abcdefghijklmnopqrstuvwxyz/ \
-	      -e 's/\.[^.]*$/.X/'
-	exit ;;
-    CRAY*TS:*:*:*)
-	echo t90-cray-unicos${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/'
-	exit ;;
-    CRAY*T3E:*:*:*)
-	echo alphaev5-cray-unicosmk${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/'
-	exit ;;
-    CRAY*SV1:*:*:*)
-	echo sv1-cray-unicos${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/'
-	exit ;;
-    *:UNICOS/mp:*:*)
-	echo craynv-cray-unicosmp${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/'
-	exit ;;
-    F30[01]:UNIX_System_V:*:* | F700:UNIX_System_V:*:*)
-	FUJITSU_PROC=`uname -m | tr 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' 'abcdefghijklmnopqrstuvwxyz'`
-	FUJITSU_SYS=`uname -p | tr 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' 'abcdefghijklmnopqrstuvwxyz' | sed -e 's/\///'`
-	FUJITSU_REL=`echo ${UNAME_RELEASE} | sed -e 's/ /_/'`
-	echo "${FUJITSU_PROC}-fujitsu-${FUJITSU_SYS}${FUJITSU_REL}"
-	exit ;;
-    5000:UNIX_System_V:4.*:*)
-	FUJITSU_SYS=`uname -p | tr 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' 'abcdefghijklmnopqrstuvwxyz' | sed -e 's/\///'`
-	FUJITSU_REL=`echo ${UNAME_RELEASE} | tr 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' 'abcdefghijklmnopqrstuvwxyz' | sed -e 's/ /_/'`
-	echo "sparc-fujitsu-${FUJITSU_SYS}${FUJITSU_REL}"
-	exit ;;
-    i*86:BSD/386:*:* | i*86:BSD/OS:*:* | *:Ascend\ Embedded/OS:*:*)
-	echo ${UNAME_MACHINE}-pc-bsdi${UNAME_RELEASE}
-	exit ;;
-    sparc*:BSD/OS:*:*)
-	echo sparc-unknown-bsdi${UNAME_RELEASE}
-	exit ;;
-    *:BSD/OS:*:*)
-	echo ${UNAME_MACHINE}-unknown-bsdi${UNAME_RELEASE}
-	exit ;;
-    *:FreeBSD:*:*)
-	UNAME_PROCESSOR=`/usr/bin/uname -p`
-	case ${UNAME_PROCESSOR} in
-	    amd64)
-		echo x86_64-unknown-freebsd`echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'` ;;
-	    *)
-		echo ${UNAME_PROCESSOR}-unknown-freebsd`echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'` ;;
-	esac
-	exit ;;
-    i*:CYGWIN*:*)
-	echo ${UNAME_MACHINE}-pc-cygwin
-	exit ;;
-    *:MINGW64*:*)
-	echo ${UNAME_MACHINE}-pc-mingw64
-	exit ;;
-    *:MINGW*:*)
-	echo ${UNAME_MACHINE}-pc-mingw32
-	exit ;;
-    *:MSYS*:*)
-	echo ${UNAME_MACHINE}-pc-msys
-	exit ;;
-    i*:windows32*:*)
-	# uname -m includes "-pc" on this system.
-	echo ${UNAME_MACHINE}-mingw32
-	exit ;;
-    i*:PW*:*)
-	echo ${UNAME_MACHINE}-pc-pw32
-	exit ;;
-    *:Interix*:*)
-	case ${UNAME_MACHINE} in
-	    x86)
-		echo i586-pc-interix${UNAME_RELEASE}
-		exit ;;
-	    authenticamd | genuineintel | EM64T)
-		echo x86_64-unknown-interix${UNAME_RELEASE}
-		exit ;;
-	    IA64)
-		echo ia64-unknown-interix${UNAME_RELEASE}
-		exit ;;
-	esac ;;
-    [345]86:Windows_95:* | [345]86:Windows_98:* | [345]86:Windows_NT:*)
-	echo i${UNAME_MACHINE}-pc-mks
-	exit ;;
-    8664:Windows_NT:*)
-	echo x86_64-pc-mks
-	exit ;;
-    i*:Windows_NT*:* | Pentium*:Windows_NT*:*)
-	# How do we know it's Interix rather than the generic POSIX subsystem?
-	# It also conflicts with pre-2.0 versions of AT&T UWIN. Should we
-	# UNAME_MACHINE based on the output of uname instead of i386?
-	echo i586-pc-interix
-	exit ;;
-    i*:UWIN*:*)
-	echo ${UNAME_MACHINE}-pc-uwin
-	exit ;;
-    amd64:CYGWIN*:*:* | x86_64:CYGWIN*:*:*)
-	echo x86_64-unknown-cygwin
-	exit ;;
-    p*:CYGWIN*:*)
-	echo powerpcle-unknown-cygwin
-	exit ;;
-    prep*:SunOS:5.*:*)
-	echo powerpcle-unknown-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'`
-	exit ;;
-    *:GNU:*:*)
-	# the GNU system
-	echo `echo ${UNAME_MACHINE}|sed -e 's,[-/].*$,,'`-unknown-${LIBC}`echo ${UNAME_RELEASE}|sed -e 's,/.*$,,'`
-	exit ;;
-    *:GNU/*:*:*)
-	# other systems with GNU libc and userland
-	echo ${UNAME_MACHINE}-unknown-`echo ${UNAME_SYSTEM} | sed 's,^[^/]*/,,' | tr '[A-Z]' '[a-z]'``echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'`-${LIBC}
-	exit ;;
-    i*86:Minix:*:*)
-	echo ${UNAME_MACHINE}-pc-minix
-	exit ;;
-    aarch64:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
-	exit ;;
-    aarch64_be:Linux:*:*)
-	UNAME_MACHINE=aarch64_be
-	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
-	exit ;;
-    alpha:Linux:*:*)
-	case `sed -n '/^cpu model/s/^.*: \(.*\)/\1/p' < /proc/cpuinfo` in
-	  EV5)   UNAME_MACHINE=alphaev5 ;;
-	  EV56)  UNAME_MACHINE=alphaev56 ;;
-	  PCA56) UNAME_MACHINE=alphapca56 ;;
-	  PCA57) UNAME_MACHINE=alphapca56 ;;
-	  EV6)   UNAME_MACHINE=alphaev6 ;;
-	  EV67)  UNAME_MACHINE=alphaev67 ;;
-	  EV68*) UNAME_MACHINE=alphaev68 ;;
-	esac
-	objdump --private-headers /bin/sh | grep -q ld.so.1
-	if test "$?" = 0 ; then LIBC="gnulibc1" ; fi
-	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
-	exit ;;
-    arc:Linux:*:* | arceb:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
-	exit ;;
-    arm*:Linux:*:*)
-	eval $set_cc_for_build
-	if echo __ARM_EABI__ | $CC_FOR_BUILD -E - 2>/dev/null \
-	    | grep -q __ARM_EABI__
-	then
-	    echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
-	else
-	    if echo __ARM_PCS_VFP | $CC_FOR_BUILD -E - 2>/dev/null \
-		| grep -q __ARM_PCS_VFP
-	    then
-		echo ${UNAME_MACHINE}-unknown-linux-${LIBC}eabi
-	    else
-		echo ${UNAME_MACHINE}-unknown-linux-${LIBC}eabihf
-	    fi
-	fi
-	exit ;;
-    avr32*:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
-	exit ;;
-    cris:Linux:*:*)
-	echo ${UNAME_MACHINE}-axis-linux-${LIBC}
-	exit ;;
-    crisv32:Linux:*:*)
-	echo ${UNAME_MACHINE}-axis-linux-${LIBC}
-	exit ;;
-    frv:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
-	exit ;;
-    hexagon:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
-	exit ;;
-    i*86:Linux:*:*)
-	echo ${UNAME_MACHINE}-pc-linux-${LIBC}
-	exit ;;
-    ia64:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
-	exit ;;
-    m32r*:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
-	exit ;;
-    m68*:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
-	exit ;;
-    mips:Linux:*:* | mips64:Linux:*:*)
-	eval $set_cc_for_build
-	sed 's/^	//' << EOF >$dummy.c
-	#undef CPU
-	#undef ${UNAME_MACHINE}
-	#undef ${UNAME_MACHINE}el
-	#if defined(__MIPSEL__) || defined(__MIPSEL) || defined(_MIPSEL) || defined(MIPSEL)
-	CPU=${UNAME_MACHINE}el
-	#else
-	#if defined(__MIPSEB__) || defined(__MIPSEB) || defined(_MIPSEB) || defined(MIPSEB)
-	CPU=${UNAME_MACHINE}
-	#else
-	CPU=
-	#endif
-	#endif
-EOF
-	eval `$CC_FOR_BUILD -E $dummy.c 2>/dev/null | grep '^CPU'`
-	test x"${CPU}" != x && { echo "${CPU}-unknown-linux-${LIBC}"; exit; }
-	;;
-    openrisc*:Linux:*:*)
-	echo or1k-unknown-linux-${LIBC}
-	exit ;;
-    or32:Linux:*:* | or1k*:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
-	exit ;;
-    padre:Linux:*:*)
-	echo sparc-unknown-linux-${LIBC}
-	exit ;;
-    parisc64:Linux:*:* | hppa64:Linux:*:*)
-	echo hppa64-unknown-linux-${LIBC}
-	exit ;;
-    parisc:Linux:*:* | hppa:Linux:*:*)
-	# Look for CPU level
-	case `grep '^cpu[^a-z]*:' /proc/cpuinfo 2>/dev/null | cut -d' ' -f2` in
-	  PA7*) echo hppa1.1-unknown-linux-${LIBC} ;;
-	  PA8*) echo hppa2.0-unknown-linux-${LIBC} ;;
-	  *)    echo hppa-unknown-linux-${LIBC} ;;
-	esac
-	exit ;;
-    ppc64:Linux:*:*)
-	echo powerpc64-unknown-linux-${LIBC}
-	exit ;;
-    ppc:Linux:*:*)
-	echo powerpc-unknown-linux-${LIBC}
-	exit ;;
-    ppc64le:Linux:*:*)
-	echo powerpc64le-unknown-linux-${LIBC}
-	exit ;;
-    ppcle:Linux:*:*)
-	echo powerpcle-unknown-linux-${LIBC}
-	exit ;;
-    s390:Linux:*:* | s390x:Linux:*:*)
-	echo ${UNAME_MACHINE}-ibm-linux-${LIBC}
-	exit ;;
-    sh64*:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
-	exit ;;
-    sh*:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
-	exit ;;
-    sparc:Linux:*:* | sparc64:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
-	exit ;;
-    tile*:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
-	exit ;;
-    vax:Linux:*:*)
-	echo ${UNAME_MACHINE}-dec-linux-${LIBC}
-	exit ;;
-    x86_64:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
-	exit ;;
-    xtensa*:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
-	exit ;;
-    i*86:DYNIX/ptx:4*:*)
-	# ptx 4.0 does uname -s correctly, with DYNIX/ptx in there.
-	# earlier versions are messed up and put the nodename in both
-	# sysname and nodename.
-	echo i386-sequent-sysv4
-	exit ;;
-    i*86:UNIX_SV:4.2MP:2.*)
-	# Unixware is an offshoot of SVR4, but it has its own version
-	# number series starting with 2...
-	# I am not positive that other SVR4 systems won't match this,
-	# I just have to hope.  -- rms.
-	# Use sysv4.2uw... so that sysv4* matches it.
-	echo ${UNAME_MACHINE}-pc-sysv4.2uw${UNAME_VERSION}
-	exit ;;
-    i*86:OS/2:*:*)
-	# If we were able to find `uname', then EMX Unix compatibility
-	# is probably installed.
-	echo ${UNAME_MACHINE}-pc-os2-emx
-	exit ;;
-    i*86:XTS-300:*:STOP)
-	echo ${UNAME_MACHINE}-unknown-stop
-	exit ;;
-    i*86:atheos:*:*)
-	echo ${UNAME_MACHINE}-unknown-atheos
-	exit ;;
-    i*86:syllable:*:*)
-	echo ${UNAME_MACHINE}-pc-syllable
-	exit ;;
-    i*86:LynxOS:2.*:* | i*86:LynxOS:3.[01]*:* | i*86:LynxOS:4.[02]*:*)
-	echo i386-unknown-lynxos${UNAME_RELEASE}
-	exit ;;
-    i*86:*DOS:*:*)
-	echo ${UNAME_MACHINE}-pc-msdosdjgpp
-	exit ;;
-    i*86:*:4.*:* | i*86:SYSTEM_V:4.*:*)
-	UNAME_REL=`echo ${UNAME_RELEASE} | sed 's/\/MP$//'`
-	if grep Novell /usr/include/link.h >/dev/null 2>/dev/null; then
-		echo ${UNAME_MACHINE}-univel-sysv${UNAME_REL}
-	else
-		echo ${UNAME_MACHINE}-pc-sysv${UNAME_REL}
-	fi
-	exit ;;
-    i*86:*:5:[678]*)
-	# UnixWare 7.x, OpenUNIX and OpenServer 6.
-	case `/bin/uname -X | grep "^Machine"` in
-	    *486*)	     UNAME_MACHINE=i486 ;;
-	    *Pentium)	     UNAME_MACHINE=i586 ;;
-	    *Pent*|*Celeron) UNAME_MACHINE=i686 ;;
-	esac
-	echo ${UNAME_MACHINE}-unknown-sysv${UNAME_RELEASE}${UNAME_SYSTEM}${UNAME_VERSION}
-	exit ;;
-    i*86:*:3.2:*)
-	if test -f /usr/options/cb.name; then
-		UNAME_REL=`sed -n 's/.*Version //p' </usr/options/cb.name`
-		echo ${UNAME_MACHINE}-pc-isc$UNAME_REL
-	elif /bin/uname -X 2>/dev/null >/dev/null ; then
-		UNAME_REL=`(/bin/uname -X|grep Release|sed -e 's/.*= //')`
-		(/bin/uname -X|grep i80486 >/dev/null) && UNAME_MACHINE=i486
-		(/bin/uname -X|grep '^Machine.*Pentium' >/dev/null) \
-			&& UNAME_MACHINE=i586
-		(/bin/uname -X|grep '^Machine.*Pent *II' >/dev/null) \
-			&& UNAME_MACHINE=i686
-		(/bin/uname -X|grep '^Machine.*Pentium Pro' >/dev/null) \
-			&& UNAME_MACHINE=i686
-		echo ${UNAME_MACHINE}-pc-sco$UNAME_REL
-	else
-		echo ${UNAME_MACHINE}-pc-sysv32
-	fi
-	exit ;;
-    pc:*:*:*)
-	# Left here for compatibility:
-	# uname -m prints for DJGPP always 'pc', but it prints nothing about
-	# the processor, so we play safe by assuming i586.
-	# Note: whatever this is, it MUST be the same as what config.sub
-	# prints for the "djgpp" host, or else GDB configury will decide that
-	# this is a cross-build.
-	echo i586-pc-msdosdjgpp
-	exit ;;
-    Intel:Mach:3*:*)
-	echo i386-pc-mach3
-	exit ;;
-    paragon:*:*:*)
-	echo i860-intel-osf1
-	exit ;;
-    i860:*:4.*:*) # i860-SVR4
-	if grep Stardent /usr/include/sys/uadmin.h >/dev/null 2>&1 ; then
-	  echo i860-stardent-sysv${UNAME_RELEASE} # Stardent Vistra i860-SVR4
-	else # Add other i860-SVR4 vendors below as they are discovered.
-	  echo i860-unknown-sysv${UNAME_RELEASE}  # Unknown i860-SVR4
-	fi
-	exit ;;
-    mini*:CTIX:SYS*5:*)
-	# "miniframe"
-	echo m68010-convergent-sysv
-	exit ;;
-    mc68k:UNIX:SYSTEM5:3.51m)
-	echo m68k-convergent-sysv
-	exit ;;
-    M680?0:D-NIX:5.3:*)
-	echo m68k-diab-dnix
-	exit ;;
-    M68*:*:R3V[5678]*:*)
-	test -r /sysV68 && { echo 'm68k-motorola-sysv'; exit; } ;;
-    3[345]??:*:4.0:3.0 | 3[34]??A:*:4.0:3.0 | 3[34]??,*:*:4.0:3.0 | 3[34]??/*:*:4.0:3.0 | 4400:*:4.0:3.0 | 4850:*:4.0:3.0 | SKA40:*:4.0:3.0 | SDS2:*:4.0:3.0 | SHG2:*:4.0:3.0 | S7501*:*:4.0:3.0)
-	OS_REL=''
-	test -r /etc/.relid \
-	&& OS_REL=.`sed -n 's/[^ ]* [^ ]* \([0-9][0-9]\).*/\1/p' < /etc/.relid`
-	/bin/uname -p 2>/dev/null | grep 86 >/dev/null \
-	  && { echo i486-ncr-sysv4.3${OS_REL}; exit; }
-	/bin/uname -p 2>/dev/null | /bin/grep entium >/dev/null \
-	  && { echo i586-ncr-sysv4.3${OS_REL}; exit; } ;;
-    3[34]??:*:4.0:* | 3[34]??,*:*:4.0:*)
-	/bin/uname -p 2>/dev/null | grep 86 >/dev/null \
-	  && { echo i486-ncr-sysv4; exit; } ;;
-    NCR*:*:4.2:* | MPRAS*:*:4.2:*)
-	OS_REL='.3'
-	test -r /etc/.relid \
-	    && OS_REL=.`sed -n 's/[^ ]* [^ ]* \([0-9][0-9]\).*/\1/p' < /etc/.relid`
-	/bin/uname -p 2>/dev/null | grep 86 >/dev/null \
-	    && { echo i486-ncr-sysv4.3${OS_REL}; exit; }
-	/bin/uname -p 2>/dev/null | /bin/grep entium >/dev/null \
-	    && { echo i586-ncr-sysv4.3${OS_REL}; exit; }
-	/bin/uname -p 2>/dev/null | /bin/grep pteron >/dev/null \
-	    && { echo i586-ncr-sysv4.3${OS_REL}; exit; } ;;
-    m68*:LynxOS:2.*:* | m68*:LynxOS:3.0*:*)
-	echo m68k-unknown-lynxos${UNAME_RELEASE}
-	exit ;;
-    mc68030:UNIX_System_V:4.*:*)
-	echo m68k-atari-sysv4
-	exit ;;
-    TSUNAMI:LynxOS:2.*:*)
-	echo sparc-unknown-lynxos${UNAME_RELEASE}
-	exit ;;
-    rs6000:LynxOS:2.*:*)
-	echo rs6000-unknown-lynxos${UNAME_RELEASE}
-	exit ;;
-    PowerPC:LynxOS:2.*:* | PowerPC:LynxOS:3.[01]*:* | PowerPC:LynxOS:4.[02]*:*)
-	echo powerpc-unknown-lynxos${UNAME_RELEASE}
-	exit ;;
-    SM[BE]S:UNIX_SV:*:*)
-	echo mips-dde-sysv${UNAME_RELEASE}
-	exit ;;
-    RM*:ReliantUNIX-*:*:*)
-	echo mips-sni-sysv4
-	exit ;;
-    RM*:SINIX-*:*:*)
-	echo mips-sni-sysv4
-	exit ;;
-    *:SINIX-*:*:*)
-	if uname -p 2>/dev/null >/dev/null ; then
-		UNAME_MACHINE=`(uname -p) 2>/dev/null`
-		echo ${UNAME_MACHINE}-sni-sysv4
-	else
-		echo ns32k-sni-sysv
-	fi
-	exit ;;
-    PENTIUM:*:4.0*:*)	# Unisys `ClearPath HMP IX 4000' SVR4/MP effort
-			# says <Richard.M.Bartel@ccMail.Census.GOV>
-	echo i586-unisys-sysv4
-	exit ;;
-    *:UNIX_System_V:4*:FTX*)
-	# From Gerald Hewes <hewes@openmarket.com>.
-	# How about differentiating between stratus architectures? -djm
-	echo hppa1.1-stratus-sysv4
-	exit ;;
-    *:*:*:FTX*)
-	# From seanf@swdc.stratus.com.
-	echo i860-stratus-sysv4
-	exit ;;
-    i*86:VOS:*:*)
-	# From Paul.Green@stratus.com.
-	echo ${UNAME_MACHINE}-stratus-vos
-	exit ;;
-    *:VOS:*:*)
-	# From Paul.Green@stratus.com.
-	echo hppa1.1-stratus-vos
-	exit ;;
-    mc68*:A/UX:*:*)
-	echo m68k-apple-aux${UNAME_RELEASE}
-	exit ;;
-    news*:NEWS-OS:6*:*)
-	echo mips-sony-newsos6
-	exit ;;
-    R[34]000:*System_V*:*:* | R4000:UNIX_SYSV:*:* | R*000:UNIX_SV:*:*)
-	if [ -d /usr/nec ]; then
-		echo mips-nec-sysv${UNAME_RELEASE}
-	else
-		echo mips-unknown-sysv${UNAME_RELEASE}
-	fi
-	exit ;;
-    BeBox:BeOS:*:*)	# BeOS running on hardware made by Be, PPC only.
-	echo powerpc-be-beos
-	exit ;;
-    BeMac:BeOS:*:*)	# BeOS running on Mac or Mac clone, PPC only.
-	echo powerpc-apple-beos
-	exit ;;
-    BePC:BeOS:*:*)	# BeOS running on Intel PC compatible.
-	echo i586-pc-beos
-	exit ;;
-    BePC:Haiku:*:*)	# Haiku running on Intel PC compatible.
-	echo i586-pc-haiku
-	exit ;;
-    x86_64:Haiku:*:*)
-	echo x86_64-unknown-haiku
-	exit ;;
-    SX-4:SUPER-UX:*:*)
-	echo sx4-nec-superux${UNAME_RELEASE}
-	exit ;;
-    SX-5:SUPER-UX:*:*)
-	echo sx5-nec-superux${UNAME_RELEASE}
-	exit ;;
-    SX-6:SUPER-UX:*:*)
-	echo sx6-nec-superux${UNAME_RELEASE}
-	exit ;;
-    SX-7:SUPER-UX:*:*)
-	echo sx7-nec-superux${UNAME_RELEASE}
-	exit ;;
-    SX-8:SUPER-UX:*:*)
-	echo sx8-nec-superux${UNAME_RELEASE}
-	exit ;;
-    SX-8R:SUPER-UX:*:*)
-	echo sx8r-nec-superux${UNAME_RELEASE}
-	exit ;;
-    Power*:Rhapsody:*:*)
-	echo powerpc-apple-rhapsody${UNAME_RELEASE}
-	exit ;;
-    *:Rhapsody:*:*)
-	echo ${UNAME_MACHINE}-apple-rhapsody${UNAME_RELEASE}
-	exit ;;
-    *:Darwin:*:*)
-	UNAME_PROCESSOR=`uname -p` || UNAME_PROCESSOR=unknown
-	eval $set_cc_for_build
-	if test "$UNAME_PROCESSOR" = unknown ; then
-	    UNAME_PROCESSOR=powerpc
-	fi
-	if test `echo "$UNAME_RELEASE" | sed -e 's/\..*//'` -le 10 ; then
-	    if [ "$CC_FOR_BUILD" != 'no_compiler_found' ]; then
-		if (echo '#ifdef __LP64__'; echo IS_64BIT_ARCH; echo '#endif') | \
-		    (CCOPTS= $CC_FOR_BUILD -E - 2>/dev/null) | \
-		    grep IS_64BIT_ARCH >/dev/null
-		then
-		    case $UNAME_PROCESSOR in
-			i386) UNAME_PROCESSOR=x86_64 ;;
-			powerpc) UNAME_PROCESSOR=powerpc64 ;;
-		    esac
-		fi
-	    fi
-	elif test "$UNAME_PROCESSOR" = i386 ; then
-	    # Avoid executing cc on OS X 10.9, as it ships with a stub
-	    # that puts up a graphical alert prompting to install
-	    # developer tools.  Any system running Mac OS X 10.7 or
-	    # later (Darwin 11 and later) is required to have a 64-bit
-	    # processor. This is not true of the ARM version of Darwin
-	    # that Apple uses in portable devices.
-	    UNAME_PROCESSOR=x86_64
-	fi
-	echo ${UNAME_PROCESSOR}-apple-darwin${UNAME_RELEASE}
-	exit ;;
-    *:procnto*:*:* | *:QNX:[0123456789]*:*)
-	UNAME_PROCESSOR=`uname -p`
-	if test "$UNAME_PROCESSOR" = "x86"; then
-		UNAME_PROCESSOR=i386
-		UNAME_MACHINE=pc
-	fi
-	echo ${UNAME_PROCESSOR}-${UNAME_MACHINE}-nto-qnx${UNAME_RELEASE}
-	exit ;;
-    *:QNX:*:4*)
-	echo i386-pc-qnx
-	exit ;;
-    NEO-?:NONSTOP_KERNEL:*:*)
-	echo neo-tandem-nsk${UNAME_RELEASE}
-	exit ;;
-    NSE-*:NONSTOP_KERNEL:*:*)
-	echo nse-tandem-nsk${UNAME_RELEASE}
-	exit ;;
-    NSR-?:NONSTOP_KERNEL:*:*)
-	echo nsr-tandem-nsk${UNAME_RELEASE}
-	exit ;;
-    *:NonStop-UX:*:*)
-	echo mips-compaq-nonstopux
-	exit ;;
-    BS2000:POSIX*:*:*)
-	echo bs2000-siemens-sysv
-	exit ;;
-    DS/*:UNIX_System_V:*:*)
-	echo ${UNAME_MACHINE}-${UNAME_SYSTEM}-${UNAME_RELEASE}
-	exit ;;
-    *:Plan9:*:*)
-	# "uname -m" is not consistent, so use $cputype instead. 386
-	# is converted to i386 for consistency with other x86
-	# operating systems.
-	if test "$cputype" = "386"; then
-	    UNAME_MACHINE=i386
-	else
-	    UNAME_MACHINE="$cputype"
-	fi
-	echo ${UNAME_MACHINE}-unknown-plan9
-	exit ;;
-    *:TOPS-10:*:*)
-	echo pdp10-unknown-tops10
-	exit ;;
-    *:TENEX:*:*)
-	echo pdp10-unknown-tenex
-	exit ;;
-    KS10:TOPS-20:*:* | KL10:TOPS-20:*:* | TYPE4:TOPS-20:*:*)
-	echo pdp10-dec-tops20
-	exit ;;
-    XKL-1:TOPS-20:*:* | TYPE5:TOPS-20:*:*)
-	echo pdp10-xkl-tops20
-	exit ;;
-    *:TOPS-20:*:*)
-	echo pdp10-unknown-tops20
-	exit ;;
-    *:ITS:*:*)
-	echo pdp10-unknown-its
-	exit ;;
-    SEI:*:*:SEIUX)
-	echo mips-sei-seiux${UNAME_RELEASE}
-	exit ;;
-    *:DragonFly:*:*)
-	echo ${UNAME_MACHINE}-unknown-dragonfly`echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'`
-	exit ;;
-    *:*VMS:*:*)
-	UNAME_MACHINE=`(uname -p) 2>/dev/null`
-	case "${UNAME_MACHINE}" in
-	    A*) echo alpha-dec-vms ; exit ;;
-	    I*) echo ia64-dec-vms ; exit ;;
-	    V*) echo vax-dec-vms ; exit ;;
-	esac ;;
-    *:XENIX:*:SysV)
-	echo i386-pc-xenix
-	exit ;;
-    i*86:skyos:*:*)
-	echo ${UNAME_MACHINE}-pc-skyos`echo ${UNAME_RELEASE}` | sed -e 's/ .*$//'
-	exit ;;
-    i*86:rdos:*:*)
-	echo ${UNAME_MACHINE}-pc-rdos
-	exit ;;
-    i*86:AROS:*:*)
-	echo ${UNAME_MACHINE}-pc-aros
-	exit ;;
-    x86_64:VMkernel:*:*)
-	echo ${UNAME_MACHINE}-unknown-esx
-	exit ;;
-esac
-
-cat >&2 <<EOF
-$0: unable to guess system type
-
-This script, last modified $timestamp, has failed to recognize
-the operating system you are using. It is advised that you
-download the most up to date version of the config scripts from
-
-  http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.guess;hb=HEAD
-and
-  http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.sub;hb=HEAD
-
-If the version you run ($0) is already up to date, please
-send the following data and any information you think might be
-pertinent to <config-patches@gnu.org> in order to provide the needed
-information to handle your system.
-
-config.guess timestamp = $timestamp
-
-uname -m = `(uname -m) 2>/dev/null || echo unknown`
-uname -r = `(uname -r) 2>/dev/null || echo unknown`
-uname -s = `(uname -s) 2>/dev/null || echo unknown`
-uname -v = `(uname -v) 2>/dev/null || echo unknown`
-
-/usr/bin/uname -p = `(/usr/bin/uname -p) 2>/dev/null`
-/bin/uname -X     = `(/bin/uname -X) 2>/dev/null`
-
-hostinfo               = `(hostinfo) 2>/dev/null`
-/bin/universe          = `(/bin/universe) 2>/dev/null`
-/usr/bin/arch -k       = `(/usr/bin/arch -k) 2>/dev/null`
-/bin/arch              = `(/bin/arch) 2>/dev/null`
-/usr/bin/oslevel       = `(/usr/bin/oslevel) 2>/dev/null`
-/usr/convex/getsysinfo = `(/usr/convex/getsysinfo) 2>/dev/null`
-
-UNAME_MACHINE = ${UNAME_MACHINE}
-UNAME_RELEASE = ${UNAME_RELEASE}
-UNAME_SYSTEM  = ${UNAME_SYSTEM}
-UNAME_VERSION = ${UNAME_VERSION}
-EOF
-
-exit 1
-
-# Local variables:
-# eval: (add-hook 'write-file-hooks 'time-stamp)
-# time-stamp-start: "timestamp='"
-# time-stamp-format: "%:y-%02m-%02d"
-# time-stamp-end: "'"
-# End:
diff --git a/deps/jemalloc/config.sub b/deps/jemalloc/config.sub
deleted file mode 100755
index 0ccff7706..000000000
--- a/deps/jemalloc/config.sub
+++ /dev/null
@@ -1,1797 +0,0 @@
-#! /bin/sh
-# Configuration validation subroutine script.
-#   Copyright 1992-2014 Free Software Foundation, Inc.
-
-timestamp='2014-05-01'
-
-# This file is free software; you can redistribute it and/or modify it
-# under the terms of the GNU General Public License as published by
-# the Free Software Foundation; either version 3 of the License, or
-# (at your option) any later version.
-#
-# This program is distributed in the hope that it will be useful, but
-# WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
-# General Public License for more details.
-#
-# You should have received a copy of the GNU General Public License
-# along with this program; if not, see <http://www.gnu.org/licenses/>.
-#
-# As a special exception to the GNU General Public License, if you
-# distribute this file as part of a program that contains a
-# configuration script generated by Autoconf, you may include it under
-# the same distribution terms that you use for the rest of that
-# program.  This Exception is an additional permission under section 7
-# of the GNU General Public License, version 3 ("GPLv3").
-
-
-# Please send patches with a ChangeLog entry to config-patches@gnu.org.
-#
-# Configuration subroutine to validate and canonicalize a configuration type.
-# Supply the specified configuration type as an argument.
-# If it is invalid, we print an error message on stderr and exit with code 1.
-# Otherwise, we print the canonical config type on stdout and succeed.
-
-# You can get the latest version of this script from:
-# http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.sub;hb=HEAD
-
-# This file is supposed to be the same for all GNU packages
-# and recognize all the CPU types, system types and aliases
-# that are meaningful with *any* GNU software.
-# Each package is responsible for reporting which valid configurations
-# it does not support.  The user should be able to distinguish
-# a failure to support a valid configuration from a meaningless
-# configuration.
-
-# The goal of this file is to map all the various variations of a given
-# machine specification into a single specification in the form:
-#	CPU_TYPE-MANUFACTURER-OPERATING_SYSTEM
-# or in some cases, the newer four-part form:
-#	CPU_TYPE-MANUFACTURER-KERNEL-OPERATING_SYSTEM
-# It is wrong to echo any other type of specification.
-
-me=`echo "$0" | sed -e 's,.*/,,'`
-
-usage="\
-Usage: $0 [OPTION] CPU-MFR-OPSYS
-       $0 [OPTION] ALIAS
-
-Canonicalize a configuration name.
-
-Operation modes:
-  -h, --help         print this help, then exit
-  -t, --time-stamp   print date of last modification, then exit
-  -v, --version      print version number, then exit
-
-Report bugs and patches to <config-patches@gnu.org>."
-
-version="\
-GNU config.sub ($timestamp)
-
-Copyright 1992-2014 Free Software Foundation, Inc.
-
-This is free software; see the source for copying conditions.  There is NO
-warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE."
-
-help="
-Try \`$me --help' for more information."
-
-# Parse command line
-while test $# -gt 0 ; do
-  case $1 in
-    --time-stamp | --time* | -t )
-       echo "$timestamp" ; exit ;;
-    --version | -v )
-       echo "$version" ; exit ;;
-    --help | --h* | -h )
-       echo "$usage"; exit ;;
-    -- )     # Stop option processing
-       shift; break ;;
-    - )	# Use stdin as input.
-       break ;;
-    -* )
-       echo "$me: invalid option $1$help"
-       exit 1 ;;
-
-    *local*)
-       # First pass through any local machine types.
-       echo $1
-       exit ;;
-
-    * )
-       break ;;
-  esac
-done
-
-case $# in
- 0) echo "$me: missing argument$help" >&2
-    exit 1;;
- 1) ;;
- *) echo "$me: too many arguments$help" >&2
-    exit 1;;
-esac
-
-# Separate what the user gave into CPU-COMPANY and OS or KERNEL-OS (if any).
-# Here we must recognize all the valid KERNEL-OS combinations.
-maybe_os=`echo $1 | sed 's/^\(.*\)-\([^-]*-[^-]*\)$/\2/'`
-case $maybe_os in
-  nto-qnx* | linux-gnu* | linux-android* | linux-dietlibc | linux-newlib* | \
-  linux-musl* | linux-uclibc* | uclinux-uclibc* | uclinux-gnu* | kfreebsd*-gnu* | \
-  knetbsd*-gnu* | netbsd*-gnu* | \
-  kopensolaris*-gnu* | \
-  storm-chaos* | os2-emx* | rtmk-nova*)
-    os=-$maybe_os
-    basic_machine=`echo $1 | sed 's/^\(.*\)-\([^-]*-[^-]*\)$/\1/'`
-    ;;
-  android-linux)
-    os=-linux-android
-    basic_machine=`echo $1 | sed 's/^\(.*\)-\([^-]*-[^-]*\)$/\1/'`-unknown
-    ;;
-  *)
-    basic_machine=`echo $1 | sed 's/-[^-]*$//'`
-    if [ $basic_machine != $1 ]
-    then os=`echo $1 | sed 's/.*-/-/'`
-    else os=; fi
-    ;;
-esac
-
-### Let's recognize common machines as not being operating systems so
-### that things like config.sub decstation-3100 work.  We also
-### recognize some manufacturers as not being operating systems, so we
-### can provide default operating systems below.
-case $os in
-	-sun*os*)
-		# Prevent following clause from handling this invalid input.
-		;;
-	-dec* | -mips* | -sequent* | -encore* | -pc532* | -sgi* | -sony* | \
-	-att* | -7300* | -3300* | -delta* | -motorola* | -sun[234]* | \
-	-unicom* | -ibm* | -next | -hp | -isi* | -apollo | -altos* | \
-	-convergent* | -ncr* | -news | -32* | -3600* | -3100* | -hitachi* |\
-	-c[123]* | -convex* | -sun | -crds | -omron* | -dg | -ultra | -tti* | \
-	-harris | -dolphin | -highlevel | -gould | -cbm | -ns | -masscomp | \
-	-apple | -axis | -knuth | -cray | -microblaze*)
-		os=
-		basic_machine=$1
-		;;
-	-bluegene*)
-		os=-cnk
-		;;
-	-sim | -cisco | -oki | -wec | -winbond)
-		os=
-		basic_machine=$1
-		;;
-	-scout)
-		;;
-	-wrs)
-		os=-vxworks
-		basic_machine=$1
-		;;
-	-chorusos*)
-		os=-chorusos
-		basic_machine=$1
-		;;
-	-chorusrdb)
-		os=-chorusrdb
-		basic_machine=$1
-		;;
-	-hiux*)
-		os=-hiuxwe2
-		;;
-	-sco6)
-		os=-sco5v6
-		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
-		;;
-	-sco5)
-		os=-sco3.2v5
-		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
-		;;
-	-sco4)
-		os=-sco3.2v4
-		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
-		;;
-	-sco3.2.[4-9]*)
-		os=`echo $os | sed -e 's/sco3.2./sco3.2v/'`
-		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
-		;;
-	-sco3.2v[4-9]*)
-		# Don't forget version if it is 3.2v4 or newer.
-		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
-		;;
-	-sco5v6*)
-		# Don't forget version if it is 3.2v4 or newer.
-		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
-		;;
-	-sco*)
-		os=-sco3.2v2
-		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
-		;;
-	-udk*)
-		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
-		;;
-	-isc)
-		os=-isc2.2
-		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
-		;;
-	-clix*)
-		basic_machine=clipper-intergraph
-		;;
-	-isc*)
-		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
-		;;
-	-lynx*178)
-		os=-lynxos178
-		;;
-	-lynx*5)
-		os=-lynxos5
-		;;
-	-lynx*)
-		os=-lynxos
-		;;
-	-ptx*)
-		basic_machine=`echo $1 | sed -e 's/86-.*/86-sequent/'`
-		;;
-	-windowsnt*)
-		os=`echo $os | sed -e 's/windowsnt/winnt/'`
-		;;
-	-psos*)
-		os=-psos
-		;;
-	-mint | -mint[0-9]*)
-		basic_machine=m68k-atari
-		os=-mint
-		;;
-esac
-
-# Decode aliases for certain CPU-COMPANY combinations.
-case $basic_machine in
-	# Recognize the basic CPU types without company name.
-	# Some are omitted here because they have special meanings below.
-	1750a | 580 \
-	| a29k \
-	| aarch64 | aarch64_be \
-	| alpha | alphaev[4-8] | alphaev56 | alphaev6[78] | alphapca5[67] \
-	| alpha64 | alpha64ev[4-8] | alpha64ev56 | alpha64ev6[78] | alpha64pca5[67] \
-	| am33_2.0 \
-	| arc | arceb \
-	| arm | arm[bl]e | arme[lb] | armv[2-8] | armv[3-8][lb] | armv7[arm] \
-	| avr | avr32 \
-	| be32 | be64 \
-	| bfin \
-	| c4x | c8051 | clipper \
-	| d10v | d30v | dlx | dsp16xx \
-	| epiphany \
-	| fido | fr30 | frv \
-	| h8300 | h8500 | hppa | hppa1.[01] | hppa2.0 | hppa2.0[nw] | hppa64 \
-	| hexagon \
-	| i370 | i860 | i960 | ia64 \
-	| ip2k | iq2000 \
-	| k1om \
-	| le32 | le64 \
-	| lm32 \
-	| m32c | m32r | m32rle | m68000 | m68k | m88k \
-	| maxq | mb | microblaze | microblazeel | mcore | mep | metag \
-	| mips | mipsbe | mipseb | mipsel | mipsle \
-	| mips16 \
-	| mips64 | mips64el \
-	| mips64octeon | mips64octeonel \
-	| mips64orion | mips64orionel \
-	| mips64r5900 | mips64r5900el \
-	| mips64vr | mips64vrel \
-	| mips64vr4100 | mips64vr4100el \
-	| mips64vr4300 | mips64vr4300el \
-	| mips64vr5000 | mips64vr5000el \
-	| mips64vr5900 | mips64vr5900el \
-	| mipsisa32 | mipsisa32el \
-	| mipsisa32r2 | mipsisa32r2el \
-	| mipsisa32r6 | mipsisa32r6el \
-	| mipsisa64 | mipsisa64el \
-	| mipsisa64r2 | mipsisa64r2el \
-	| mipsisa64r6 | mipsisa64r6el \
-	| mipsisa64sb1 | mipsisa64sb1el \
-	| mipsisa64sr71k | mipsisa64sr71kel \
-	| mipsr5900 | mipsr5900el \
-	| mipstx39 | mipstx39el \
-	| mn10200 | mn10300 \
-	| moxie \
-	| mt \
-	| msp430 \
-	| nds32 | nds32le | nds32be \
-	| nios | nios2 | nios2eb | nios2el \
-	| ns16k | ns32k \
-	| open8 | or1k | or1knd | or32 \
-	| pdp10 | pdp11 | pj | pjl \
-	| powerpc | powerpc64 | powerpc64le | powerpcle \
-	| pyramid \
-	| rl78 | rx \
-	| score \
-	| sh | sh[1234] | sh[24]a | sh[24]aeb | sh[23]e | sh[34]eb | sheb | shbe | shle | sh[1234]le | sh3ele \
-	| sh64 | sh64le \
-	| sparc | sparc64 | sparc64b | sparc64v | sparc86x | sparclet | sparclite \
-	| sparcv8 | sparcv9 | sparcv9b | sparcv9v \
-	| spu \
-	| tahoe | tic4x | tic54x | tic55x | tic6x | tic80 | tron \
-	| ubicom32 \
-	| v850 | v850e | v850e1 | v850e2 | v850es | v850e2v3 \
-	| we32k \
-	| x86 | xc16x | xstormy16 | xtensa \
-	| z8k | z80)
-		basic_machine=$basic_machine-unknown
-		;;
-	c54x)
-		basic_machine=tic54x-unknown
-		;;
-	c55x)
-		basic_machine=tic55x-unknown
-		;;
-	c6x)
-		basic_machine=tic6x-unknown
-		;;
-	m6811 | m68hc11 | m6812 | m68hc12 | m68hcs12x | nvptx | picochip)
-		basic_machine=$basic_machine-unknown
-		os=-none
-		;;
-	m88110 | m680[12346]0 | m683?2 | m68360 | m5200 | v70 | w65 | z8k)
-		;;
-	ms1)
-		basic_machine=mt-unknown
-		;;
-
-	strongarm | thumb | xscale)
-		basic_machine=arm-unknown
-		;;
-	xgate)
-		basic_machine=$basic_machine-unknown
-		os=-none
-		;;
-	xscaleeb)
-		basic_machine=armeb-unknown
-		;;
-
-	xscaleel)
-		basic_machine=armel-unknown
-		;;
-
-	# We use `pc' rather than `unknown'
-	# because (1) that's what they normally are, and
-	# (2) the word "unknown" tends to confuse beginning users.
-	i*86 | x86_64)
-	  basic_machine=$basic_machine-pc
-	  ;;
-	# Object if more than one company name word.
-	*-*-*)
-		echo Invalid configuration \`$1\': machine \`$basic_machine\' not recognized 1>&2
-		exit 1
-		;;
-	# Recognize the basic CPU types with company name.
-	580-* \
-	| a29k-* \
-	| aarch64-* | aarch64_be-* \
-	| alpha-* | alphaev[4-8]-* | alphaev56-* | alphaev6[78]-* \
-	| alpha64-* | alpha64ev[4-8]-* | alpha64ev56-* | alpha64ev6[78]-* \
-	| alphapca5[67]-* | alpha64pca5[67]-* | arc-* | arceb-* \
-	| arm-*  | armbe-* | armle-* | armeb-* | armv*-* \
-	| avr-* | avr32-* \
-	| be32-* | be64-* \
-	| bfin-* | bs2000-* \
-	| c[123]* | c30-* | [cjt]90-* | c4x-* \
-	| c8051-* | clipper-* | craynv-* | cydra-* \
-	| d10v-* | d30v-* | dlx-* \
-	| elxsi-* \
-	| f30[01]-* | f700-* | fido-* | fr30-* | frv-* | fx80-* \
-	| h8300-* | h8500-* \
-	| hppa-* | hppa1.[01]-* | hppa2.0-* | hppa2.0[nw]-* | hppa64-* \
-	| hexagon-* \
-	| i*86-* | i860-* | i960-* | ia64-* \
-	| ip2k-* | iq2000-* \
-	| k1om-* \
-	| le32-* | le64-* \
-	| lm32-* \
-	| m32c-* | m32r-* | m32rle-* \
-	| m68000-* | m680[012346]0-* | m68360-* | m683?2-* | m68k-* \
-	| m88110-* | m88k-* | maxq-* | mcore-* | metag-* \
-	| microblaze-* | microblazeel-* \
-	| mips-* | mipsbe-* | mipseb-* | mipsel-* | mipsle-* \
-	| mips16-* \
-	| mips64-* | mips64el-* \
-	| mips64octeon-* | mips64octeonel-* \
-	| mips64orion-* | mips64orionel-* \
-	| mips64r5900-* | mips64r5900el-* \
-	| mips64vr-* | mips64vrel-* \
-	| mips64vr4100-* | mips64vr4100el-* \
-	| mips64vr4300-* | mips64vr4300el-* \
-	| mips64vr5000-* | mips64vr5000el-* \
-	| mips64vr5900-* | mips64vr5900el-* \
-	| mipsisa32-* | mipsisa32el-* \
-	| mipsisa32r2-* | mipsisa32r2el-* \
-	| mipsisa32r6-* | mipsisa32r6el-* \
-	| mipsisa64-* | mipsisa64el-* \
-	| mipsisa64r2-* | mipsisa64r2el-* \
-	| mipsisa64r6-* | mipsisa64r6el-* \
-	| mipsisa64sb1-* | mipsisa64sb1el-* \
-	| mipsisa64sr71k-* | mipsisa64sr71kel-* \
-	| mipsr5900-* | mipsr5900el-* \
-	| mipstx39-* | mipstx39el-* \
-	| mmix-* \
-	| mt-* \
-	| msp430-* \
-	| nds32-* | nds32le-* | nds32be-* \
-	| nios-* | nios2-* | nios2eb-* | nios2el-* \
-	| none-* | np1-* | ns16k-* | ns32k-* \
-	| open8-* \
-	| or1k*-* \
-	| orion-* \
-	| pdp10-* | pdp11-* | pj-* | pjl-* | pn-* | power-* \
-	| powerpc-* | powerpc64-* | powerpc64le-* | powerpcle-* \
-	| pyramid-* \
-	| rl78-* | romp-* | rs6000-* | rx-* \
-	| sh-* | sh[1234]-* | sh[24]a-* | sh[24]aeb-* | sh[23]e-* | sh[34]eb-* | sheb-* | shbe-* \
-	| shle-* | sh[1234]le-* | sh3ele-* | sh64-* | sh64le-* \
-	| sparc-* | sparc64-* | sparc64b-* | sparc64v-* | sparc86x-* | sparclet-* \
-	| sparclite-* \
-	| sparcv8-* | sparcv9-* | sparcv9b-* | sparcv9v-* | sv1-* | sx?-* \
-	| tahoe-* \
-	| tic30-* | tic4x-* | tic54x-* | tic55x-* | tic6x-* | tic80-* \
-	| tile*-* \
-	| tron-* \
-	| ubicom32-* \
-	| v850-* | v850e-* | v850e1-* | v850es-* | v850e2-* | v850e2v3-* \
-	| vax-* \
-	| we32k-* \
-	| x86-* | x86_64-* | xc16x-* | xps100-* \
-	| xstormy16-* | xtensa*-* \
-	| ymp-* \
-	| z8k-* | z80-*)
-		;;
-	# Recognize the basic CPU types without company name, with glob match.
-	xtensa*)
-		basic_machine=$basic_machine-unknown
-		;;
-	# Recognize the various machine names and aliases which stand
-	# for a CPU type and a company and sometimes even an OS.
-	386bsd)
-		basic_machine=i386-unknown
-		os=-bsd
-		;;
-	3b1 | 7300 | 7300-att | att-7300 | pc7300 | safari | unixpc)
-		basic_machine=m68000-att
-		;;
-	3b*)
-		basic_machine=we32k-att
-		;;
-	a29khif)
-		basic_machine=a29k-amd
-		os=-udi
-		;;
-	abacus)
-		basic_machine=abacus-unknown
-		;;
-	adobe68k)
-		basic_machine=m68010-adobe
-		os=-scout
-		;;
-	alliant | fx80)
-		basic_machine=fx80-alliant
-		;;
-	altos | altos3068)
-		basic_machine=m68k-altos
-		;;
-	am29k)
-		basic_machine=a29k-none
-		os=-bsd
-		;;
-	amd64)
-		basic_machine=x86_64-pc
-		;;
-	amd64-*)
-		basic_machine=x86_64-`echo $basic_machine | sed 's/^[^-]*-//'`
-		;;
-	amdahl)
-		basic_machine=580-amdahl
-		os=-sysv
-		;;
-	amiga | amiga-*)
-		basic_machine=m68k-unknown
-		;;
-	amigaos | amigados)
-		basic_machine=m68k-unknown
-		os=-amigaos
-		;;
-	amigaunix | amix)
-		basic_machine=m68k-unknown
-		os=-sysv4
-		;;
-	apollo68)
-		basic_machine=m68k-apollo
-		os=-sysv
-		;;
-	apollo68bsd)
-		basic_machine=m68k-apollo
-		os=-bsd
-		;;
-	aros)
-		basic_machine=i386-pc
-		os=-aros
-		;;
-	aux)
-		basic_machine=m68k-apple
-		os=-aux
-		;;
-	balance)
-		basic_machine=ns32k-sequent
-		os=-dynix
-		;;
-	blackfin)
-		basic_machine=bfin-unknown
-		os=-linux
-		;;
-	blackfin-*)
-		basic_machine=bfin-`echo $basic_machine | sed 's/^[^-]*-//'`
-		os=-linux
-		;;
-	bluegene*)
-		basic_machine=powerpc-ibm
-		os=-cnk
-		;;
-	c54x-*)
-		basic_machine=tic54x-`echo $basic_machine | sed 's/^[^-]*-//'`
-		;;
-	c55x-*)
-		basic_machine=tic55x-`echo $basic_machine | sed 's/^[^-]*-//'`
-		;;
-	c6x-*)
-		basic_machine=tic6x-`echo $basic_machine | sed 's/^[^-]*-//'`
-		;;
-	c90)
-		basic_machine=c90-cray
-		os=-unicos
-		;;
-	cegcc)
-		basic_machine=arm-unknown
-		os=-cegcc
-		;;
-	convex-c1)
-		basic_machine=c1-convex
-		os=-bsd
-		;;
-	convex-c2)
-		basic_machine=c2-convex
-		os=-bsd
-		;;
-	convex-c32)
-		basic_machine=c32-convex
-		os=-bsd
-		;;
-	convex-c34)
-		basic_machine=c34-convex
-		os=-bsd
-		;;
-	convex-c38)
-		basic_machine=c38-convex
-		os=-bsd
-		;;
-	cray | j90)
-		basic_machine=j90-cray
-		os=-unicos
-		;;
-	craynv)
-		basic_machine=craynv-cray
-		os=-unicosmp
-		;;
-	cr16 | cr16-*)
-		basic_machine=cr16-unknown
-		os=-elf
-		;;
-	crds | unos)
-		basic_machine=m68k-crds
-		;;
-	crisv32 | crisv32-* | etraxfs*)
-		basic_machine=crisv32-axis
-		;;
-	cris | cris-* | etrax*)
-		basic_machine=cris-axis
-		;;
-	crx)
-		basic_machine=crx-unknown
-		os=-elf
-		;;
-	da30 | da30-*)
-		basic_machine=m68k-da30
-		;;
-	decstation | decstation-3100 | pmax | pmax-* | pmin | dec3100 | decstatn)
-		basic_machine=mips-dec
-		;;
-	decsystem10* | dec10*)
-		basic_machine=pdp10-dec
-		os=-tops10
-		;;
-	decsystem20* | dec20*)
-		basic_machine=pdp10-dec
-		os=-tops20
-		;;
-	delta | 3300 | motorola-3300 | motorola-delta \
-	      | 3300-motorola | delta-motorola)
-		basic_machine=m68k-motorola
-		;;
-	delta88)
-		basic_machine=m88k-motorola
-		os=-sysv3
-		;;
-	dicos)
-		basic_machine=i686-pc
-		os=-dicos
-		;;
-	djgpp)
-		basic_machine=i586-pc
-		os=-msdosdjgpp
-		;;
-	dpx20 | dpx20-*)
-		basic_machine=rs6000-bull
-		os=-bosx
-		;;
-	dpx2* | dpx2*-bull)
-		basic_machine=m68k-bull
-		os=-sysv3
-		;;
-	ebmon29k)
-		basic_machine=a29k-amd
-		os=-ebmon
-		;;
-	elxsi)
-		basic_machine=elxsi-elxsi
-		os=-bsd
-		;;
-	encore | umax | mmax)
-		basic_machine=ns32k-encore
-		;;
-	es1800 | OSE68k | ose68k | ose | OSE)
-		basic_machine=m68k-ericsson
-		os=-ose
-		;;
-	fx2800)
-		basic_machine=i860-alliant
-		;;
-	genix)
-		basic_machine=ns32k-ns
-		;;
-	gmicro)
-		basic_machine=tron-gmicro
-		os=-sysv
-		;;
-	go32)
-		basic_machine=i386-pc
-		os=-go32
-		;;
-	h3050r* | hiux*)
-		basic_machine=hppa1.1-hitachi
-		os=-hiuxwe2
-		;;
-	h8300hms)
-		basic_machine=h8300-hitachi
-		os=-hms
-		;;
-	h8300xray)
-		basic_machine=h8300-hitachi
-		os=-xray
-		;;
-	h8500hms)
-		basic_machine=h8500-hitachi
-		os=-hms
-		;;
-	harris)
-		basic_machine=m88k-harris
-		os=-sysv3
-		;;
-	hp300-*)
-		basic_machine=m68k-hp
-		;;
-	hp300bsd)
-		basic_machine=m68k-hp
-		os=-bsd
-		;;
-	hp300hpux)
-		basic_machine=m68k-hp
-		os=-hpux
-		;;
-	hp3k9[0-9][0-9] | hp9[0-9][0-9])
-		basic_machine=hppa1.0-hp
-		;;
-	hp9k2[0-9][0-9] | hp9k31[0-9])
-		basic_machine=m68000-hp
-		;;
-	hp9k3[2-9][0-9])
-		basic_machine=m68k-hp
-		;;
-	hp9k6[0-9][0-9] | hp6[0-9][0-9])
-		basic_machine=hppa1.0-hp
-		;;
-	hp9k7[0-79][0-9] | hp7[0-79][0-9])
-		basic_machine=hppa1.1-hp
-		;;
-	hp9k78[0-9] | hp78[0-9])
-		# FIXME: really hppa2.0-hp
-		basic_machine=hppa1.1-hp
-		;;
-	hp9k8[67]1 | hp8[67]1 | hp9k80[24] | hp80[24] | hp9k8[78]9 | hp8[78]9 | hp9k893 | hp893)
-		# FIXME: really hppa2.0-hp
-		basic_machine=hppa1.1-hp
-		;;
-	hp9k8[0-9][13679] | hp8[0-9][13679])
-		basic_machine=hppa1.1-hp
-		;;
-	hp9k8[0-9][0-9] | hp8[0-9][0-9])
-		basic_machine=hppa1.0-hp
-		;;
-	hppa-next)
-		os=-nextstep3
-		;;
-	hppaosf)
-		basic_machine=hppa1.1-hp
-		os=-osf
-		;;
-	hppro)
-		basic_machine=hppa1.1-hp
-		os=-proelf
-		;;
-	i370-ibm* | ibm*)
-		basic_machine=i370-ibm
-		;;
-	i*86v32)
-		basic_machine=`echo $1 | sed -e 's/86.*/86-pc/'`
-		os=-sysv32
-		;;
-	i*86v4*)
-		basic_machine=`echo $1 | sed -e 's/86.*/86-pc/'`
-		os=-sysv4
-		;;
-	i*86v)
-		basic_machine=`echo $1 | sed -e 's/86.*/86-pc/'`
-		os=-sysv
-		;;
-	i*86sol2)
-		basic_machine=`echo $1 | sed -e 's/86.*/86-pc/'`
-		os=-solaris2
-		;;
-	i386mach)
-		basic_machine=i386-mach
-		os=-mach
-		;;
-	i386-vsta | vsta)
-		basic_machine=i386-unknown
-		os=-vsta
-		;;
-	iris | iris4d)
-		basic_machine=mips-sgi
-		case $os in
-		    -irix*)
-			;;
-		    *)
-			os=-irix4
-			;;
-		esac
-		;;
-	isi68 | isi)
-		basic_machine=m68k-isi
-		os=-sysv
-		;;
-	m68knommu)
-		basic_machine=m68k-unknown
-		os=-linux
-		;;
-	m68knommu-*)
-		basic_machine=m68k-`echo $basic_machine | sed 's/^[^-]*-//'`
-		os=-linux
-		;;
-	m88k-omron*)
-		basic_machine=m88k-omron
-		;;
-	magnum | m3230)
-		basic_machine=mips-mips
-		os=-sysv
-		;;
-	merlin)
-		basic_machine=ns32k-utek
-		os=-sysv
-		;;
-	microblaze*)
-		basic_machine=microblaze-xilinx
-		;;
-	mingw64)
-		basic_machine=x86_64-pc
-		os=-mingw64
-		;;
-	mingw32)
-		basic_machine=i686-pc
-		os=-mingw32
-		;;
-	mingw32ce)
-		basic_machine=arm-unknown
-		os=-mingw32ce
-		;;
-	miniframe)
-		basic_machine=m68000-convergent
-		;;
-	*mint | -mint[0-9]* | *MiNT | *MiNT[0-9]*)
-		basic_machine=m68k-atari
-		os=-mint
-		;;
-	mips3*-*)
-		basic_machine=`echo $basic_machine | sed -e 's/mips3/mips64/'`
-		;;
-	mips3*)
-		basic_machine=`echo $basic_machine | sed -e 's/mips3/mips64/'`-unknown
-		;;
-	monitor)
-		basic_machine=m68k-rom68k
-		os=-coff
-		;;
-	morphos)
-		basic_machine=powerpc-unknown
-		os=-morphos
-		;;
-	msdos)
-		basic_machine=i386-pc
-		os=-msdos
-		;;
-	ms1-*)
-		basic_machine=`echo $basic_machine | sed -e 's/ms1-/mt-/'`
-		;;
-	msys)
-		basic_machine=i686-pc
-		os=-msys
-		;;
-	mvs)
-		basic_machine=i370-ibm
-		os=-mvs
-		;;
-	nacl)
-		basic_machine=le32-unknown
-		os=-nacl
-		;;
-	ncr3000)
-		basic_machine=i486-ncr
-		os=-sysv4
-		;;
-	netbsd386)
-		basic_machine=i386-unknown
-		os=-netbsd
-		;;
-	netwinder)
-		basic_machine=armv4l-rebel
-		os=-linux
-		;;
-	news | news700 | news800 | news900)
-		basic_machine=m68k-sony
-		os=-newsos
-		;;
-	news1000)
-		basic_machine=m68030-sony
-		os=-newsos
-		;;
-	news-3600 | risc-news)
-		basic_machine=mips-sony
-		os=-newsos
-		;;
-	necv70)
-		basic_machine=v70-nec
-		os=-sysv
-		;;
-	next | m*-next )
-		basic_machine=m68k-next
-		case $os in
-		    -nextstep* )
-			;;
-		    -ns2*)
-		      os=-nextstep2
-			;;
-		    *)
-		      os=-nextstep3
-			;;
-		esac
-		;;
-	nh3000)
-		basic_machine=m68k-harris
-		os=-cxux
-		;;
-	nh[45]000)
-		basic_machine=m88k-harris
-		os=-cxux
-		;;
-	nindy960)
-		basic_machine=i960-intel
-		os=-nindy
-		;;
-	mon960)
-		basic_machine=i960-intel
-		os=-mon960
-		;;
-	nonstopux)
-		basic_machine=mips-compaq
-		os=-nonstopux
-		;;
-	np1)
-		basic_machine=np1-gould
-		;;
-	neo-tandem)
-		basic_machine=neo-tandem
-		;;
-	nse-tandem)
-		basic_machine=nse-tandem
-		;;
-	nsr-tandem)
-		basic_machine=nsr-tandem
-		;;
-	op50n-* | op60c-*)
-		basic_machine=hppa1.1-oki
-		os=-proelf
-		;;
-	openrisc | openrisc-*)
-		basic_machine=or32-unknown
-		;;
-	os400)
-		basic_machine=powerpc-ibm
-		os=-os400
-		;;
-	OSE68000 | ose68000)
-		basic_machine=m68000-ericsson
-		os=-ose
-		;;
-	os68k)
-		basic_machine=m68k-none
-		os=-os68k
-		;;
-	pa-hitachi)
-		basic_machine=hppa1.1-hitachi
-		os=-hiuxwe2
-		;;
-	paragon)
-		basic_machine=i860-intel
-		os=-osf
-		;;
-	parisc)
-		basic_machine=hppa-unknown
-		os=-linux
-		;;
-	parisc-*)
-		basic_machine=hppa-`echo $basic_machine | sed 's/^[^-]*-//'`
-		os=-linux
-		;;
-	pbd)
-		basic_machine=sparc-tti
-		;;
-	pbb)
-		basic_machine=m68k-tti
-		;;
-	pc532 | pc532-*)
-		basic_machine=ns32k-pc532
-		;;
-	pc98)
-		basic_machine=i386-pc
-		;;
-	pc98-*)
-		basic_machine=i386-`echo $basic_machine | sed 's/^[^-]*-//'`
-		;;
-	pentium | p5 | k5 | k6 | nexgen | viac3)
-		basic_machine=i586-pc
-		;;
-	pentiumpro | p6 | 6x86 | athlon | athlon_*)
-		basic_machine=i686-pc
-		;;
-	pentiumii | pentium2 | pentiumiii | pentium3)
-		basic_machine=i686-pc
-		;;
-	pentium4)
-		basic_machine=i786-pc
-		;;
-	pentium-* | p5-* | k5-* | k6-* | nexgen-* | viac3-*)
-		basic_machine=i586-`echo $basic_machine | sed 's/^[^-]*-//'`
-		;;
-	pentiumpro-* | p6-* | 6x86-* | athlon-*)
-		basic_machine=i686-`echo $basic_machine | sed 's/^[^-]*-//'`
-		;;
-	pentiumii-* | pentium2-* | pentiumiii-* | pentium3-*)
-		basic_machine=i686-`echo $basic_machine | sed 's/^[^-]*-//'`
-		;;
-	pentium4-*)
-		basic_machine=i786-`echo $basic_machine | sed 's/^[^-]*-//'`
-		;;
-	pn)
-		basic_machine=pn-gould
-		;;
-	power)	basic_machine=power-ibm
-		;;
-	ppc | ppcbe)	basic_machine=powerpc-unknown
-		;;
-	ppc-* | ppcbe-*)
-		basic_machine=powerpc-`echo $basic_machine | sed 's/^[^-]*-//'`
-		;;
-	ppcle | powerpclittle | ppc-le | powerpc-little)
-		basic_machine=powerpcle-unknown
-		;;
-	ppcle-* | powerpclittle-*)
-		basic_machine=powerpcle-`echo $basic_machine | sed 's/^[^-]*-//'`
-		;;
-	ppc64)	basic_machine=powerpc64-unknown
-		;;
-	ppc64-*) basic_machine=powerpc64-`echo $basic_machine | sed 's/^[^-]*-//'`
-		;;
-	ppc64le | powerpc64little | ppc64-le | powerpc64-little)
-		basic_machine=powerpc64le-unknown
-		;;
-	ppc64le-* | powerpc64little-*)
-		basic_machine=powerpc64le-`echo $basic_machine | sed 's/^[^-]*-//'`
-		;;
-	ps2)
-		basic_machine=i386-ibm
-		;;
-	pw32)
-		basic_machine=i586-unknown
-		os=-pw32
-		;;
-	rdos | rdos64)
-		basic_machine=x86_64-pc
-		os=-rdos
-		;;
-	rdos32)
-		basic_machine=i386-pc
-		os=-rdos
-		;;
-	rom68k)
-		basic_machine=m68k-rom68k
-		os=-coff
-		;;
-	rm[46]00)
-		basic_machine=mips-siemens
-		;;
-	rtpc | rtpc-*)
-		basic_machine=romp-ibm
-		;;
-	s390 | s390-*)
-		basic_machine=s390-ibm
-		;;
-	s390x | s390x-*)
-		basic_machine=s390x-ibm
-		;;
-	sa29200)
-		basic_machine=a29k-amd
-		os=-udi
-		;;
-	sb1)
-		basic_machine=mipsisa64sb1-unknown
-		;;
-	sb1el)
-		basic_machine=mipsisa64sb1el-unknown
-		;;
-	sde)
-		basic_machine=mipsisa32-sde
-		os=-elf
-		;;
-	sei)
-		basic_machine=mips-sei
-		os=-seiux
-		;;
-	sequent)
-		basic_machine=i386-sequent
-		;;
-	sh)
-		basic_machine=sh-hitachi
-		os=-hms
-		;;
-	sh5el)
-		basic_machine=sh5le-unknown
-		;;
-	sh64)
-		basic_machine=sh64-unknown
-		;;
-	sparclite-wrs | simso-wrs)
-		basic_machine=sparclite-wrs
-		os=-vxworks
-		;;
-	sps7)
-		basic_machine=m68k-bull
-		os=-sysv2
-		;;
-	spur)
-		basic_machine=spur-unknown
-		;;
-	st2000)
-		basic_machine=m68k-tandem
-		;;
-	stratus)
-		basic_machine=i860-stratus
-		os=-sysv4
-		;;
-	strongarm-* | thumb-*)
-		basic_machine=arm-`echo $basic_machine | sed 's/^[^-]*-//'`
-		;;
-	sun2)
-		basic_machine=m68000-sun
-		;;
-	sun2os3)
-		basic_machine=m68000-sun
-		os=-sunos3
-		;;
-	sun2os4)
-		basic_machine=m68000-sun
-		os=-sunos4
-		;;
-	sun3os3)
-		basic_machine=m68k-sun
-		os=-sunos3
-		;;
-	sun3os4)
-		basic_machine=m68k-sun
-		os=-sunos4
-		;;
-	sun4os3)
-		basic_machine=sparc-sun
-		os=-sunos3
-		;;
-	sun4os4)
-		basic_machine=sparc-sun
-		os=-sunos4
-		;;
-	sun4sol2)
-		basic_machine=sparc-sun
-		os=-solaris2
-		;;
-	sun3 | sun3-*)
-		basic_machine=m68k-sun
-		;;
-	sun4)
-		basic_machine=sparc-sun
-		;;
-	sun386 | sun386i | roadrunner)
-		basic_machine=i386-sun
-		;;
-	sv1)
-		basic_machine=sv1-cray
-		os=-unicos
-		;;
-	symmetry)
-		basic_machine=i386-sequent
-		os=-dynix
-		;;
-	t3e)
-		basic_machine=alphaev5-cray
-		os=-unicos
-		;;
-	t90)
-		basic_machine=t90-cray
-		os=-unicos
-		;;
-	tile*)
-		basic_machine=$basic_machine-unknown
-		os=-linux-gnu
-		;;
-	tx39)
-		basic_machine=mipstx39-unknown
-		;;
-	tx39el)
-		basic_machine=mipstx39el-unknown
-		;;
-	toad1)
-		basic_machine=pdp10-xkl
-		os=-tops20
-		;;
-	tower | tower-32)
-		basic_machine=m68k-ncr
-		;;
-	tpf)
-		basic_machine=s390x-ibm
-		os=-tpf
-		;;
-	udi29k)
-		basic_machine=a29k-amd
-		os=-udi
-		;;
-	ultra3)
-		basic_machine=a29k-nyu
-		os=-sym1
-		;;
-	v810 | necv810)
-		basic_machine=v810-nec
-		os=-none
-		;;
-	vaxv)
-		basic_machine=vax-dec
-		os=-sysv
-		;;
-	vms)
-		basic_machine=vax-dec
-		os=-vms
-		;;
-	vpp*|vx|vx-*)
-		basic_machine=f301-fujitsu
-		;;
-	vxworks960)
-		basic_machine=i960-wrs
-		os=-vxworks
-		;;
-	vxworks68)
-		basic_machine=m68k-wrs
-		os=-vxworks
-		;;
-	vxworks29k)
-		basic_machine=a29k-wrs
-		os=-vxworks
-		;;
-	w65*)
-		basic_machine=w65-wdc
-		os=-none
-		;;
-	w89k-*)
-		basic_machine=hppa1.1-winbond
-		os=-proelf
-		;;
-	xbox)
-		basic_machine=i686-pc
-		os=-mingw32
-		;;
-	xps | xps100)
-		basic_machine=xps100-honeywell
-		;;
-	xscale-* | xscalee[bl]-*)
-		basic_machine=`echo $basic_machine | sed 's/^xscale/arm/'`
-		;;
-	ymp)
-		basic_machine=ymp-cray
-		os=-unicos
-		;;
-	z8k-*-coff)
-		basic_machine=z8k-unknown
-		os=-sim
-		;;
-	z80-*-coff)
-		basic_machine=z80-unknown
-		os=-sim
-		;;
-	none)
-		basic_machine=none-none
-		os=-none
-		;;
-
-# Here we handle the default manufacturer of certain CPU types.  It is in
-# some cases the only manufacturer, in others, it is the most popular.
-	w89k)
-		basic_machine=hppa1.1-winbond
-		;;
-	op50n)
-		basic_machine=hppa1.1-oki
-		;;
-	op60c)
-		basic_machine=hppa1.1-oki
-		;;
-	romp)
-		basic_machine=romp-ibm
-		;;
-	mmix)
-		basic_machine=mmix-knuth
-		;;
-	rs6000)
-		basic_machine=rs6000-ibm
-		;;
-	vax)
-		basic_machine=vax-dec
-		;;
-	pdp10)
-		# there are many clones, so DEC is not a safe bet
-		basic_machine=pdp10-unknown
-		;;
-	pdp11)
-		basic_machine=pdp11-dec
-		;;
-	we32k)
-		basic_machine=we32k-att
-		;;
-	sh[1234] | sh[24]a | sh[24]aeb | sh[34]eb | sh[1234]le | sh[23]ele)
-		basic_machine=sh-unknown
-		;;
-	sparc | sparcv8 | sparcv9 | sparcv9b | sparcv9v)
-		basic_machine=sparc-sun
-		;;
-	cydra)
-		basic_machine=cydra-cydrome
-		;;
-	orion)
-		basic_machine=orion-highlevel
-		;;
-	orion105)
-		basic_machine=clipper-highlevel
-		;;
-	mac | mpw | mac-mpw)
-		basic_machine=m68k-apple
-		;;
-	pmac | pmac-mpw)
-		basic_machine=powerpc-apple
-		;;
-	*-unknown)
-		# Make sure to match an already-canonicalized machine name.
-		;;
-	*)
-		echo Invalid configuration \`$1\': machine \`$basic_machine\' not recognized 1>&2
-		exit 1
-		;;
-esac
-
-# Here we canonicalize certain aliases for manufacturers.
-case $basic_machine in
-	*-digital*)
-		basic_machine=`echo $basic_machine | sed 's/digital.*/dec/'`
-		;;
-	*-commodore*)
-		basic_machine=`echo $basic_machine | sed 's/commodore.*/cbm/'`
-		;;
-	*)
-		;;
-esac
-
-# Decode manufacturer-specific aliases for certain operating systems.
-
-if [ x"$os" != x"" ]
-then
-case $os in
-	# First match some system type aliases
-	# that might get confused with valid system types.
-	# -solaris* is a basic system type, with this one exception.
-	-auroraux)
-		os=-auroraux
-		;;
-	-solaris1 | -solaris1.*)
-		os=`echo $os | sed -e 's|solaris1|sunos4|'`
-		;;
-	-solaris)
-		os=-solaris2
-		;;
-	-svr4*)
-		os=-sysv4
-		;;
-	-unixware*)
-		os=-sysv4.2uw
-		;;
-	-gnu/linux*)
-		os=`echo $os | sed -e 's|gnu/linux|linux-gnu|'`
-		;;
-	# First accept the basic system types.
-	# The portable systems comes first.
-	# Each alternative MUST END IN A *, to match a version number.
-	# -sysv* is not here because it comes later, after sysvr4.
-	-gnu* | -bsd* | -mach* | -minix* | -genix* | -ultrix* | -irix* \
-	      | -*vms* | -sco* | -esix* | -isc* | -aix* | -cnk* | -sunos | -sunos[34]*\
-	      | -hpux* | -unos* | -osf* | -luna* | -dgux* | -auroraux* | -solaris* \
-	      | -sym* | -kopensolaris* | -plan9* \
-	      | -amigaos* | -amigados* | -msdos* | -newsos* | -unicos* | -aof* \
-	      | -aos* | -aros* \
-	      | -nindy* | -vxsim* | -vxworks* | -ebmon* | -hms* | -mvs* \
-	      | -clix* | -riscos* | -uniplus* | -iris* | -rtu* | -xenix* \
-	      | -hiux* | -386bsd* | -knetbsd* | -mirbsd* | -netbsd* \
-	      | -bitrig* | -openbsd* | -solidbsd* \
-	      | -ekkobsd* | -kfreebsd* | -freebsd* | -riscix* | -lynxos* \
-	      | -bosx* | -nextstep* | -cxux* | -aout* | -elf* | -oabi* \
-	      | -ptx* | -coff* | -ecoff* | -winnt* | -domain* | -vsta* \
-	      | -udi* | -eabi* | -lites* | -ieee* | -go32* | -aux* \
-	      | -chorusos* | -chorusrdb* | -cegcc* \
-	      | -cygwin* | -msys* | -pe* | -psos* | -moss* | -proelf* | -rtems* \
-	      | -mingw32* | -mingw64* | -linux-gnu* | -linux-android* \
-	      | -linux-newlib* | -linux-musl* | -linux-uclibc* \
-	      | -uxpv* | -beos* | -mpeix* | -udk* \
-	      | -interix* | -uwin* | -mks* | -rhapsody* | -darwin* | -opened* \
-	      | -openstep* | -oskit* | -conix* | -pw32* | -nonstopux* \
-	      | -storm-chaos* | -tops10* | -tenex* | -tops20* | -its* \
-	      | -os2* | -vos* | -palmos* | -uclinux* | -nucleus* \
-	      | -morphos* | -superux* | -rtmk* | -rtmk-nova* | -windiss* \
-	      | -powermax* | -dnix* | -nx6 | -nx7 | -sei* | -dragonfly* \
-	      | -skyos* | -haiku* | -rdos* | -toppers* | -drops* | -es* | -tirtos*)
-	# Remember, each alternative MUST END IN *, to match a version number.
-		;;
-	-qnx*)
-		case $basic_machine in
-		    x86-* | i*86-*)
-			;;
-		    *)
-			os=-nto$os
-			;;
-		esac
-		;;
-	-nto-qnx*)
-		;;
-	-nto*)
-		os=`echo $os | sed -e 's|nto|nto-qnx|'`
-		;;
-	-sim | -es1800* | -hms* | -xray | -os68k* | -none* | -v88r* \
-	      | -windows* | -osx | -abug | -netware* | -os9* | -beos* | -haiku* \
-	      | -macos* | -mpw* | -magic* | -mmixware* | -mon960* | -lnews*)
-		;;
-	-mac*)
-		os=`echo $os | sed -e 's|mac|macos|'`
-		;;
-	# Apple iOS
-	-ios*)
-		;;
-	-linux-dietlibc)
-		os=-linux-dietlibc
-		;;
-	-linux*)
-		os=`echo $os | sed -e 's|linux|linux-gnu|'`
-		;;
-	-sunos5*)
-		os=`echo $os | sed -e 's|sunos5|solaris2|'`
-		;;
-	-sunos6*)
-		os=`echo $os | sed -e 's|sunos6|solaris3|'`
-		;;
-	-opened*)
-		os=-openedition
-		;;
-	-os400*)
-		os=-os400
-		;;
-	-wince*)
-		os=-wince
-		;;
-	-osfrose*)
-		os=-osfrose
-		;;
-	-osf*)
-		os=-osf
-		;;
-	-utek*)
-		os=-bsd
-		;;
-	-dynix*)
-		os=-bsd
-		;;
-	-acis*)
-		os=-aos
-		;;
-	-atheos*)
-		os=-atheos
-		;;
-	-syllable*)
-		os=-syllable
-		;;
-	-386bsd)
-		os=-bsd
-		;;
-	-ctix* | -uts*)
-		os=-sysv
-		;;
-	-nova*)
-		os=-rtmk-nova
-		;;
-	-ns2 )
-		os=-nextstep2
-		;;
-	-nsk*)
-		os=-nsk
-		;;
-	# Preserve the version number of sinix5.
-	-sinix5.*)
-		os=`echo $os | sed -e 's|sinix|sysv|'`
-		;;
-	-sinix*)
-		os=-sysv4
-		;;
-	-tpf*)
-		os=-tpf
-		;;
-	-triton*)
-		os=-sysv3
-		;;
-	-oss*)
-		os=-sysv3
-		;;
-	-svr4)
-		os=-sysv4
-		;;
-	-svr3)
-		os=-sysv3
-		;;
-	-sysvr4)
-		os=-sysv4
-		;;
-	# This must come after -sysvr4.
-	-sysv*)
-		;;
-	-ose*)
-		os=-ose
-		;;
-	-es1800*)
-		os=-ose
-		;;
-	-xenix)
-		os=-xenix
-		;;
-	-*mint | -mint[0-9]* | -*MiNT | -MiNT[0-9]*)
-		os=-mint
-		;;
-	-aros*)
-		os=-aros
-		;;
-	-zvmoe)
-		os=-zvmoe
-		;;
-	-dicos*)
-		os=-dicos
-		;;
-	-nacl*)
-		;;
-	-none)
-		;;
-	*)
-		# Get rid of the `-' at the beginning of $os.
-		os=`echo $os | sed 's/[^-]*-//'`
-		echo Invalid configuration \`$1\': system \`$os\' not recognized 1>&2
-		exit 1
-		;;
-esac
-else
-
-# Here we handle the default operating systems that come with various machines.
-# The value should be what the vendor currently ships out the door with their
-# machine or put another way, the most popular os provided with the machine.
-
-# Note that if you're going to try to match "-MANUFACTURER" here (say,
-# "-sun"), then you have to tell the case statement up towards the top
-# that MANUFACTURER isn't an operating system.  Otherwise, code above
-# will signal an error saying that MANUFACTURER isn't an operating
-# system, and we'll never get to this point.
-
-case $basic_machine in
-	score-*)
-		os=-elf
-		;;
-	spu-*)
-		os=-elf
-		;;
-	*-acorn)
-		os=-riscix1.2
-		;;
-	arm*-rebel)
-		os=-linux
-		;;
-	arm*-semi)
-		os=-aout
-		;;
-	c4x-* | tic4x-*)
-		os=-coff
-		;;
-	c8051-*)
-		os=-elf
-		;;
-	hexagon-*)
-		os=-elf
-		;;
-	tic54x-*)
-		os=-coff
-		;;
-	tic55x-*)
-		os=-coff
-		;;
-	tic6x-*)
-		os=-coff
-		;;
-	# This must come before the *-dec entry.
-	pdp10-*)
-		os=-tops20
-		;;
-	pdp11-*)
-		os=-none
-		;;
-	*-dec | vax-*)
-		os=-ultrix4.2
-		;;
-	m68*-apollo)
-		os=-domain
-		;;
-	i386-sun)
-		os=-sunos4.0.2
-		;;
-	m68000-sun)
-		os=-sunos3
-		;;
-	m68*-cisco)
-		os=-aout
-		;;
-	mep-*)
-		os=-elf
-		;;
-	mips*-cisco)
-		os=-elf
-		;;
-	mips*-*)
-		os=-elf
-		;;
-	or32-*)
-		os=-coff
-		;;
-	*-tti)	# must be before sparc entry or we get the wrong os.
-		os=-sysv3
-		;;
-	sparc-* | *-sun)
-		os=-sunos4.1.1
-		;;
-	*-be)
-		os=-beos
-		;;
-	*-haiku)
-		os=-haiku
-		;;
-	*-ibm)
-		os=-aix
-		;;
-	*-knuth)
-		os=-mmixware
-		;;
-	*-wec)
-		os=-proelf
-		;;
-	*-winbond)
-		os=-proelf
-		;;
-	*-oki)
-		os=-proelf
-		;;
-	*-hp)
-		os=-hpux
-		;;
-	*-hitachi)
-		os=-hiux
-		;;
-	i860-* | *-att | *-ncr | *-altos | *-motorola | *-convergent)
-		os=-sysv
-		;;
-	*-cbm)
-		os=-amigaos
-		;;
-	*-dg)
-		os=-dgux
-		;;
-	*-dolphin)
-		os=-sysv3
-		;;
-	m68k-ccur)
-		os=-rtu
-		;;
-	m88k-omron*)
-		os=-luna
-		;;
-	*-next )
-		os=-nextstep
-		;;
-	*-sequent)
-		os=-ptx
-		;;
-	*-crds)
-		os=-unos
-		;;
-	*-ns)
-		os=-genix
-		;;
-	i370-*)
-		os=-mvs
-		;;
-	*-next)
-		os=-nextstep3
-		;;
-	*-gould)
-		os=-sysv
-		;;
-	*-highlevel)
-		os=-bsd
-		;;
-	*-encore)
-		os=-bsd
-		;;
-	*-sgi)
-		os=-irix
-		;;
-	*-siemens)
-		os=-sysv4
-		;;
-	*-masscomp)
-		os=-rtu
-		;;
-	f30[01]-fujitsu | f700-fujitsu)
-		os=-uxpv
-		;;
-	*-rom68k)
-		os=-coff
-		;;
-	*-*bug)
-		os=-coff
-		;;
-	*-apple)
-		os=-macos
-		;;
-	*-atari*)
-		os=-mint
-		;;
-	*)
-		os=-none
-		;;
-esac
-fi
-
-# Here we handle the case where we know the os, and the CPU type, but not the
-# manufacturer.  We pick the logical manufacturer.
-vendor=unknown
-case $basic_machine in
-	*-unknown)
-		case $os in
-			-riscix*)
-				vendor=acorn
-				;;
-			-sunos*)
-				vendor=sun
-				;;
-			-cnk*|-aix*)
-				vendor=ibm
-				;;
-			-beos*)
-				vendor=be
-				;;
-			-hpux*)
-				vendor=hp
-				;;
-			-mpeix*)
-				vendor=hp
-				;;
-			-hiux*)
-				vendor=hitachi
-				;;
-			-unos*)
-				vendor=crds
-				;;
-			-dgux*)
-				vendor=dg
-				;;
-			-luna*)
-				vendor=omron
-				;;
-			-genix*)
-				vendor=ns
-				;;
-			-mvs* | -opened*)
-				vendor=ibm
-				;;
-			-os400*)
-				vendor=ibm
-				;;
-			-ptx*)
-				vendor=sequent
-				;;
-			-tpf*)
-				vendor=ibm
-				;;
-			-vxsim* | -vxworks* | -windiss*)
-				vendor=wrs
-				;;
-			-aux*)
-				vendor=apple
-				;;
-			-hms*)
-				vendor=hitachi
-				;;
-			-mpw* | -macos*)
-				vendor=apple
-				;;
-			-*mint | -mint[0-9]* | -*MiNT | -MiNT[0-9]*)
-				vendor=atari
-				;;
-			-vos*)
-				vendor=stratus
-				;;
-		esac
-		basic_machine=`echo $basic_machine | sed "s/unknown/$vendor/"`
-		;;
-esac
-
-echo $basic_machine$os
-exit
-
-# Local variables:
-# eval: (add-hook 'write-file-hooks 'time-stamp)
-# time-stamp-start: "timestamp='"
-# time-stamp-format: "%:y-%02m-%02d"
-# time-stamp-end: "'"
-# End:
diff --git a/deps/jemalloc/configure b/deps/jemalloc/configure
index 8c56c92a1..d7391524d 100755
--- a/deps/jemalloc/configure
+++ b/deps/jemalloc/configure
@@ -628,7 +628,6 @@ cfghdrs_in
 enable_zone_allocator
 enable_tls
 enable_lazy_lock
-TESTLIBS
 jemalloc_version_gid
 jemalloc_version_nrev
 jemalloc_version_bugfix
@@ -658,16 +657,19 @@ INSTALL_SCRIPT
 INSTALL_PROGRAM
 enable_autogen
 RPATH_EXTRA
+LM
 CC_MM
 AROUT
 ARFLAGS
 MKLIB
+TEST_LD_MODE
 LDTARGET
 CTARGET
 PIC_CFLAGS
 SOREV
 EXTRA_LDFLAGS
 DSO_LDFLAGS
+link_whole_archive
 libprefix
 exe
 a
@@ -689,6 +691,7 @@ build
 EGREP
 GREP
 CPP
+EXTRA_CFLAGS
 OBJEXT
 EXEEXT
 ac_ct_CC
@@ -729,6 +732,7 @@ infodir
 docdir
 oldincludedir
 includedir
+runstatedir
 localstatedir
 sharedstatedir
 sysconfdir
@@ -760,6 +764,7 @@ with_jemalloc_prefix
 with_export
 with_private_namespace
 with_install_suffix
+with_malloc_conf
 enable_cc_silence
 enable_debug
 enable_ivsalloc
@@ -781,6 +786,8 @@ with_lg_quantum
 with_lg_page
 with_lg_page_sizes
 with_lg_size_class_group
+with_version
+enable_syscall
 enable_lazy_lock
 enable_tls
 enable_zone_allocator
@@ -832,6 +839,7 @@ datadir='${datarootdir}'
 sysconfdir='${prefix}/etc'
 sharedstatedir='${prefix}/com'
 localstatedir='${prefix}/var'
+runstatedir='${localstatedir}/run'
 includedir='${prefix}/include'
 oldincludedir='/usr/include'
 docdir='${datarootdir}/doc/${PACKAGE}'
@@ -1084,6 +1092,15 @@ do
   | -silent | --silent | --silen | --sile | --sil)
     silent=yes ;;
 
+  -runstatedir | --runstatedir | --runstatedi | --runstated \
+  | --runstate | --runstat | --runsta | --runst | --runs \
+  | --run | --ru | --r)
+    ac_prev=runstatedir ;;
+  -runstatedir=* | --runstatedir=* | --runstatedi=* | --runstated=* \
+  | --runstate=* | --runstat=* | --runsta=* | --runst=* | --runs=* \
+  | --run=* | --ru=* | --r=*)
+    runstatedir=$ac_optarg ;;
+
   -sbindir | --sbindir | --sbindi | --sbind | --sbin | --sbi | --sb)
     ac_prev=sbindir ;;
   -sbindir=* | --sbindir=* | --sbindi=* | --sbind=* | --sbin=* \
@@ -1221,7 +1238,7 @@ fi
 for ac_var in	exec_prefix prefix bindir sbindir libexecdir datarootdir \
 		datadir sysconfdir sharedstatedir localstatedir includedir \
 		oldincludedir docdir infodir htmldir dvidir pdfdir psdir \
-		libdir localedir mandir
+		libdir localedir mandir runstatedir
 do
   eval ac_val=\$$ac_var
   # Remove trailing slashes.
@@ -1374,6 +1391,7 @@ Fine tuning of the installation directories:
   --sysconfdir=DIR        read-only single-machine data [PREFIX/etc]
   --sharedstatedir=DIR    modifiable architecture-independent data [PREFIX/com]
   --localstatedir=DIR     modifiable single-machine data [PREFIX/var]
+  --runstatedir=DIR       modifiable per-process data [LOCALSTATEDIR/run]
   --libdir=DIR            object code libraries [EPREFIX/lib]
   --includedir=DIR        C header files [PREFIX/include]
   --oldincludedir=DIR     C header files for non-gcc [/usr/include]
@@ -1425,6 +1443,7 @@ Optional Features:
   --disable-cache-oblivious
                           Disable support for cache-oblivious allocation
                           alignment
+  --disable-syscall       Disable use of syscall(2)
   --enable-lazy-lock      Enable lazy locking (only lock when multi-threaded)
   --disable-tls           Disable thread-local storage (__thread keyword)
   --disable-zone-allocator
@@ -1443,6 +1462,8 @@ Optional Packages:
                           Prefix to prepend to all library-private APIs
   --with-install-suffix=<suffix>
                           Suffix to append to all installed files
+  --with-malloc-conf=<malloc_conf>
+                          config.malloc_conf options string
   --with-static-libunwind=<libunwind.a>
                           Path to static libunwind library; use rather than
                           dynamically linking
@@ -1456,6 +1477,8 @@ Optional Packages:
                           Base 2 logs of system page sizes to support
   --with-lg-size-class-group=<lg-size-class-group>
                           Base 2 log of size classes per doubling
+  --with-version=<major>.<minor>.<bugfix>-<nrev>-g<gid>
+                          Version string
 
 Some influential environment variables:
   CC          C compiler command
@@ -2484,6 +2507,36 @@ ac_compiler_gnu=$ac_cv_c_compiler_gnu
 
 
 
+ac_aux_dir=
+for ac_dir in build-aux "$srcdir"/build-aux; do
+  if test -f "$ac_dir/install-sh"; then
+    ac_aux_dir=$ac_dir
+    ac_install_sh="$ac_aux_dir/install-sh -c"
+    break
+  elif test -f "$ac_dir/install.sh"; then
+    ac_aux_dir=$ac_dir
+    ac_install_sh="$ac_aux_dir/install.sh -c"
+    break
+  elif test -f "$ac_dir/shtool"; then
+    ac_aux_dir=$ac_dir
+    ac_install_sh="$ac_aux_dir/shtool install -c"
+    break
+  fi
+done
+if test -z "$ac_aux_dir"; then
+  as_fn_error $? "cannot find install-sh, install.sh, or shtool in build-aux \"$srcdir\"/build-aux" "$LINENO" 5
+fi
+
+# These three variables are undocumented and unsupported,
+# and are intended to be withdrawn in a future Autoconf release.
+# They can cause serious problems if a builder's source tree is in a directory
+# whose full name contains unusual characters.
+ac_config_guess="$SHELL $ac_aux_dir/config.guess"  # Please don't use this var.
+ac_config_sub="$SHELL $ac_aux_dir/config.sub"  # Please don't use this var.
+ac_configure="$SHELL $ac_aux_dir/configure"  # Please don't use this var.
+
+
+
 
 
 
@@ -3390,6 +3443,7 @@ ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5'
 ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5'
 ac_compiler_gnu=$ac_cv_c_compiler_gnu
 
+
 if test "x$GCC" != "xyes" ; then
 
 { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler is MSVC" >&5
@@ -3423,10 +3477,125 @@ fi
 $as_echo "$je_cv_msvc" >&6; }
 fi
 
+je_cv_cray_prgenv_wrapper=""
+if test "x${PE_ENV}" != "x" ; then
+  case "${CC}" in
+    CC|cc)
+	je_cv_cray_prgenv_wrapper="yes"
+	;;
+    *)
+       ;;
+  esac
+fi
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler is cray" >&5
+$as_echo_n "checking whether compiler is cray... " >&6; }
+if ${je_cv_cray+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+int
+main ()
+{
+
+#ifndef _CRAYC
+  int fail-1;
+#endif
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+  je_cv_cray=yes
+else
+  je_cv_cray=no
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $je_cv_cray" >&5
+$as_echo "$je_cv_cray" >&6; }
+
+if test "x${je_cv_cray}" = "xyes" ; then
+  { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether cray compiler version is 8.4" >&5
+$as_echo_n "checking whether cray compiler version is 8.4... " >&6; }
+if ${je_cv_cray_84+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+int
+main ()
+{
+
+#if !(_RELEASE_MAJOR == 8 && _RELEASE_MINOR == 4)
+  int fail-1;
+#endif
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+  je_cv_cray_84=yes
+else
+  je_cv_cray_84=no
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $je_cv_cray_84" >&5
+$as_echo "$je_cv_cray_84" >&6; }
+fi
+
 if test "x$CFLAGS" = "x" ; then
   no_CFLAGS="yes"
   if test "x$GCC" = "xyes" ; then
 
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler supports -std=gnu11" >&5
+$as_echo_n "checking whether compiler supports -std=gnu11... " >&6; }
+TCFLAGS="${CFLAGS}"
+if test "x${CFLAGS}" = "x" ; then
+  CFLAGS="-std=gnu11"
+else
+  CFLAGS="${CFLAGS} -std=gnu11"
+fi
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+
+int
+main ()
+{
+
+    return 0;
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+  je_cv_cflags_appended=-std=gnu11
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
+$as_echo "yes" >&6; }
+else
+  je_cv_cflags_appended=
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
+$as_echo "no" >&6; }
+              CFLAGS="${TCFLAGS}"
+
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+
+    if test "x$je_cv_cflags_appended" = "x-std=gnu11" ; then
+      cat >>confdefs.h <<_ACEOF
+#define JEMALLOC_HAS_RESTRICT 1
+_ACEOF
+
+    else
+
 { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler supports -std=gnu99" >&5
 $as_echo_n "checking whether compiler supports -std=gnu99... " >&6; }
 TCFLAGS="${CFLAGS}"
@@ -3462,11 +3631,12 @@ $as_echo "no" >&6; }
 fi
 rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
 
-    if test "x$je_cv_cflags_appended" = "x-std=gnu99" ; then
-      cat >>confdefs.h <<_ACEOF
+      if test "x$je_cv_cflags_appended" = "x-std=gnu99" ; then
+        cat >>confdefs.h <<_ACEOF
 #define JEMALLOC_HAS_RESTRICT 1
 _ACEOF
 
+      fi
     fi
 
 { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler supports -Wall" >&5
@@ -3541,6 +3711,78 @@ fi
 rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
 
 
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler supports -Wshorten-64-to-32" >&5
+$as_echo_n "checking whether compiler supports -Wshorten-64-to-32... " >&6; }
+TCFLAGS="${CFLAGS}"
+if test "x${CFLAGS}" = "x" ; then
+  CFLAGS="-Wshorten-64-to-32"
+else
+  CFLAGS="${CFLAGS} -Wshorten-64-to-32"
+fi
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+
+int
+main ()
+{
+
+    return 0;
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+  je_cv_cflags_appended=-Wshorten-64-to-32
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
+$as_echo "yes" >&6; }
+else
+  je_cv_cflags_appended=
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
+$as_echo "no" >&6; }
+              CFLAGS="${TCFLAGS}"
+
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler supports -Wsign-compare" >&5
+$as_echo_n "checking whether compiler supports -Wsign-compare... " >&6; }
+TCFLAGS="${CFLAGS}"
+if test "x${CFLAGS}" = "x" ; then
+  CFLAGS="-Wsign-compare"
+else
+  CFLAGS="${CFLAGS} -Wsign-compare"
+fi
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+
+int
+main ()
+{
+
+    return 0;
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+  je_cv_cflags_appended=-Wsign-compare
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
+$as_echo "yes" >&6; }
+else
+  je_cv_cflags_appended=
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
+$as_echo "no" >&6; }
+              CFLAGS="${TCFLAGS}"
+
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+
+
 { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler supports -pipe" >&5
 $as_echo_n "checking whether compiler supports -pipe... " >&6; }
 TCFLAGS="${CFLAGS}"
@@ -3760,16 +4002,16 @@ rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
 
     CPPFLAGS="$CPPFLAGS -I${srcdir}/include/msvc_compat"
   fi
-fi
-if test "x$EXTRA_CFLAGS" != "x" ; then
+  if test "x$je_cv_cray" = "xyes" ; then
+        if test "x$je_cv_cray_84" = "xyes" ; then
 
-{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler supports $EXTRA_CFLAGS" >&5
-$as_echo_n "checking whether compiler supports $EXTRA_CFLAGS... " >&6; }
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler supports -hipa2" >&5
+$as_echo_n "checking whether compiler supports -hipa2... " >&6; }
 TCFLAGS="${CFLAGS}"
 if test "x${CFLAGS}" = "x" ; then
-  CFLAGS="$EXTRA_CFLAGS"
+  CFLAGS="-hipa2"
 else
-  CFLAGS="${CFLAGS} $EXTRA_CFLAGS"
+  CFLAGS="${CFLAGS} -hipa2"
 fi
 cat confdefs.h - <<_ACEOF >conftest.$ac_ext
 /* end confdefs.h.  */
@@ -3786,7 +4028,7 @@ main ()
 }
 _ACEOF
 if ac_fn_c_try_compile "$LINENO"; then :
-  je_cv_cflags_appended=$EXTRA_CFLAGS
+  je_cv_cflags_appended=-hipa2
               { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
 $as_echo "yes" >&6; }
 else
@@ -3798,82 +4040,195 @@ $as_echo "no" >&6; }
 fi
 rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
 
-fi
-ac_ext=c
-ac_cpp='$CPP $CPPFLAGS'
-ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5'
-ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5'
-ac_compiler_gnu=$ac_cv_c_compiler_gnu
-{ $as_echo "$as_me:${as_lineno-$LINENO}: checking how to run the C preprocessor" >&5
-$as_echo_n "checking how to run the C preprocessor... " >&6; }
-# On Suns, sometimes $CPP names a directory.
-if test -n "$CPP" && test -d "$CPP"; then
-  CPP=
-fi
-if test -z "$CPP"; then
-  if ${ac_cv_prog_CPP+:} false; then :
-  $as_echo_n "(cached) " >&6
-else
-      # Double quotes because CPP needs to be expanded
-    for CPP in "$CC -E" "$CC -E -traditional-cpp" "/lib/cpp"
-    do
-      ac_preproc_ok=false
-for ac_c_preproc_warn_flag in '' yes
-do
-  # Use a header file that comes with gcc, so configuring glibc
-  # with a fresh cross-compiler works.
-  # Prefer <limits.h> to <assert.h> if __STDC__ is defined, since
-  # <limits.h> exists even on freestanding compilers.
-  # On the NeXT, cc -E runs the code through the compiler's parser,
-  # not just through cpp. "Syntax error" is here to catch this case.
-  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
-/* end confdefs.h.  */
-#ifdef __STDC__
-# include <limits.h>
-#else
-# include <assert.h>
-#endif
-		     Syntax error
-_ACEOF
-if ac_fn_c_try_cpp "$LINENO"; then :
 
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler supports -hnognu" >&5
+$as_echo_n "checking whether compiler supports -hnognu... " >&6; }
+TCFLAGS="${CFLAGS}"
+if test "x${CFLAGS}" = "x" ; then
+  CFLAGS="-hnognu"
 else
-  # Broken: fails on valid input.
-continue
+  CFLAGS="${CFLAGS} -hnognu"
 fi
-rm -f conftest.err conftest.i conftest.$ac_ext
-
-  # OK, works on sane cases.  Now check whether nonexistent headers
-  # can be detected and how.
-  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
 /* end confdefs.h.  */
-#include <ac_nonexistent.h>
-_ACEOF
-if ac_fn_c_try_cpp "$LINENO"; then :
-  # Broken: success on invalid input.
-continue
-else
-  # Passes both tests.
-ac_preproc_ok=:
-break
-fi
-rm -f conftest.err conftest.i conftest.$ac_ext
 
-done
-# Because of `break', _AC_PREPROC_IFELSE's cleaning code was skipped.
-rm -f conftest.i conftest.err conftest.$ac_ext
-if $ac_preproc_ok; then :
-  break
-fi
 
-    done
-    ac_cv_prog_CPP=$CPP
+int
+main ()
+{
 
-fi
-  CPP=$ac_cv_prog_CPP
-else
-  ac_cv_prog_CPP=$CPP
-fi
+    return 0;
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+  je_cv_cflags_appended=-hnognu
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
+$as_echo "yes" >&6; }
+else
+  je_cv_cflags_appended=
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
+$as_echo "no" >&6; }
+              CFLAGS="${TCFLAGS}"
+
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+
+    fi
+    if test "x$enable_cc_silence" != "xno" ; then
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler supports -hnomessage=128" >&5
+$as_echo_n "checking whether compiler supports -hnomessage=128... " >&6; }
+TCFLAGS="${CFLAGS}"
+if test "x${CFLAGS}" = "x" ; then
+  CFLAGS="-hnomessage=128"
+else
+  CFLAGS="${CFLAGS} -hnomessage=128"
+fi
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+
+int
+main ()
+{
+
+    return 0;
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+  je_cv_cflags_appended=-hnomessage=128
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
+$as_echo "yes" >&6; }
+else
+  je_cv_cflags_appended=
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
+$as_echo "no" >&6; }
+              CFLAGS="${TCFLAGS}"
+
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler supports -hnomessage=1357" >&5
+$as_echo_n "checking whether compiler supports -hnomessage=1357... " >&6; }
+TCFLAGS="${CFLAGS}"
+if test "x${CFLAGS}" = "x" ; then
+  CFLAGS="-hnomessage=1357"
+else
+  CFLAGS="${CFLAGS} -hnomessage=1357"
+fi
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+
+int
+main ()
+{
+
+    return 0;
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+  je_cv_cflags_appended=-hnomessage=1357
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
+$as_echo "yes" >&6; }
+else
+  je_cv_cflags_appended=
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
+$as_echo "no" >&6; }
+              CFLAGS="${TCFLAGS}"
+
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+
+    fi
+  fi
+fi
+
+ac_ext=c
+ac_cpp='$CPP $CPPFLAGS'
+ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5'
+ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5'
+ac_compiler_gnu=$ac_cv_c_compiler_gnu
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking how to run the C preprocessor" >&5
+$as_echo_n "checking how to run the C preprocessor... " >&6; }
+# On Suns, sometimes $CPP names a directory.
+if test -n "$CPP" && test -d "$CPP"; then
+  CPP=
+fi
+if test -z "$CPP"; then
+  if ${ac_cv_prog_CPP+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+      # Double quotes because CPP needs to be expanded
+    for CPP in "$CC -E" "$CC -E -traditional-cpp" "/lib/cpp"
+    do
+      ac_preproc_ok=false
+for ac_c_preproc_warn_flag in '' yes
+do
+  # Use a header file that comes with gcc, so configuring glibc
+  # with a fresh cross-compiler works.
+  # Prefer <limits.h> to <assert.h> if __STDC__ is defined, since
+  # <limits.h> exists even on freestanding compilers.
+  # On the NeXT, cc -E runs the code through the compiler's parser,
+  # not just through cpp. "Syntax error" is here to catch this case.
+  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+#ifdef __STDC__
+# include <limits.h>
+#else
+# include <assert.h>
+#endif
+		     Syntax error
+_ACEOF
+if ac_fn_c_try_cpp "$LINENO"; then :
+
+else
+  # Broken: fails on valid input.
+continue
+fi
+rm -f conftest.err conftest.i conftest.$ac_ext
+
+  # OK, works on sane cases.  Now check whether nonexistent headers
+  # can be detected and how.
+  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+#include <ac_nonexistent.h>
+_ACEOF
+if ac_fn_c_try_cpp "$LINENO"; then :
+  # Broken: success on invalid input.
+continue
+else
+  # Passes both tests.
+ac_preproc_ok=:
+break
+fi
+rm -f conftest.err conftest.i conftest.$ac_ext
+
+done
+# Because of `break', _AC_PREPROC_IFELSE's cleaning code was skipped.
+rm -f conftest.i conftest.err conftest.$ac_ext
+if $ac_preproc_ok; then :
+  break
+fi
+
+    done
+    ac_cv_prog_CPP=$CPP
+
+fi
+  CPP=$ac_cv_prog_CPP
+else
+  ac_cv_prog_CPP=$CPP
+fi
 { $as_echo "$as_me:${as_lineno-$LINENO}: result: $CPP" >&5
 $as_echo "$CPP" >&6; }
 ac_preproc_ok=false
@@ -4431,7 +4786,12 @@ if test "x${je_cv_msvc}" = "xyes" -a "x${ac_cv_header_inttypes_h}" = "xno"; then
   CPPFLAGS="$CPPFLAGS -I${srcdir}/include/msvc_compat/C99"
 fi
 
-# The cast to long int works around a bug in the HP C Compiler
+if test "x${je_cv_msvc}" = "xyes" ; then
+  LG_SIZEOF_PTR=LG_SIZEOF_PTR_WIN
+  { $as_echo "$as_me:${as_lineno-$LINENO}: result: Using a predefined value for sizeof(void *): 4 for 32-bit, 8 for 64-bit" >&5
+$as_echo "Using a predefined value for sizeof(void *): 4 for 32-bit, 8 for 64-bit" >&6; }
+else
+  # The cast to long int works around a bug in the HP C Compiler
 # version HP92453-01 B.11.11.23709.GP, which incorrectly rejects
 # declarations like `int a3[[(sizeof (unsigned char)) >= 0]];'.
 # This bug is HP SR number 8606223364.
@@ -4464,12 +4824,13 @@ cat >>confdefs.h <<_ACEOF
 _ACEOF
 
 
-if test "x${ac_cv_sizeof_void_p}" = "x8" ; then
-  LG_SIZEOF_PTR=3
-elif test "x${ac_cv_sizeof_void_p}" = "x4" ; then
-  LG_SIZEOF_PTR=2
-else
-  as_fn_error $? "Unsupported pointer size: ${ac_cv_sizeof_void_p}" "$LINENO" 5
+  if test "x${ac_cv_sizeof_void_p}" = "x8" ; then
+    LG_SIZEOF_PTR=3
+  elif test "x${ac_cv_sizeof_void_p}" = "x4" ; then
+    LG_SIZEOF_PTR=2
+  else
+    as_fn_error $? "Unsupported pointer size: ${ac_cv_sizeof_void_p}" "$LINENO" 5
+  fi
 fi
 cat >>confdefs.h <<_ACEOF
 #define LG_SIZEOF_PTR $LG_SIZEOF_PTR
@@ -4566,6 +4927,51 @@ cat >>confdefs.h <<_ACEOF
 _ACEOF
 
 
+# The cast to long int works around a bug in the HP C Compiler
+# version HP92453-01 B.11.11.23709.GP, which incorrectly rejects
+# declarations like `int a3[[(sizeof (unsigned char)) >= 0]];'.
+# This bug is HP SR number 8606223364.
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking size of long long" >&5
+$as_echo_n "checking size of long long... " >&6; }
+if ${ac_cv_sizeof_long_long+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+  if ac_fn_c_compute_int "$LINENO" "(long int) (sizeof (long long))" "ac_cv_sizeof_long_long"        "$ac_includes_default"; then :
+
+else
+  if test "$ac_cv_type_long_long" = yes; then
+     { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5
+$as_echo "$as_me: error: in \`$ac_pwd':" >&2;}
+as_fn_error 77 "cannot compute sizeof (long long)
+See \`config.log' for more details" "$LINENO" 5; }
+   else
+     ac_cv_sizeof_long_long=0
+   fi
+fi
+
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_sizeof_long_long" >&5
+$as_echo "$ac_cv_sizeof_long_long" >&6; }
+
+
+
+cat >>confdefs.h <<_ACEOF
+#define SIZEOF_LONG_LONG $ac_cv_sizeof_long_long
+_ACEOF
+
+
+if test "x${ac_cv_sizeof_long_long}" = "x8" ; then
+  LG_SIZEOF_LONG_LONG=3
+elif test "x${ac_cv_sizeof_long_long}" = "x4" ; then
+  LG_SIZEOF_LONG_LONG=2
+else
+  as_fn_error $? "Unsupported long long size: ${ac_cv_sizeof_long_long}" "$LINENO" 5
+fi
+cat >>confdefs.h <<_ACEOF
+#define LG_SIZEOF_LONG_LONG $LG_SIZEOF_LONG_LONG
+_ACEOF
+
+
 # The cast to long int works around a bug in the HP C Compiler
 # version HP92453-01 B.11.11.23709.GP, which incorrectly rejects
 # declarations like `int a3[[(sizeof (unsigned char)) >= 0]];'.
@@ -4613,35 +5019,6 @@ cat >>confdefs.h <<_ACEOF
 _ACEOF
 
 
-ac_aux_dir=
-for ac_dir in "$srcdir" "$srcdir/.." "$srcdir/../.."; do
-  if test -f "$ac_dir/install-sh"; then
-    ac_aux_dir=$ac_dir
-    ac_install_sh="$ac_aux_dir/install-sh -c"
-    break
-  elif test -f "$ac_dir/install.sh"; then
-    ac_aux_dir=$ac_dir
-    ac_install_sh="$ac_aux_dir/install.sh -c"
-    break
-  elif test -f "$ac_dir/shtool"; then
-    ac_aux_dir=$ac_dir
-    ac_install_sh="$ac_aux_dir/shtool install -c"
-    break
-  fi
-done
-if test -z "$ac_aux_dir"; then
-  as_fn_error $? "cannot find install-sh, install.sh, or shtool in \"$srcdir\" \"$srcdir/..\" \"$srcdir/../..\"" "$LINENO" 5
-fi
-
-# These three variables are undocumented and unsupported,
-# and are intended to be withdrawn in a future Autoconf release.
-# They can cause serious problems if a builder's source tree is in a directory
-# whose full name contains unusual characters.
-ac_config_guess="$SHELL $ac_aux_dir/config.guess"  # Please don't use this var.
-ac_config_sub="$SHELL $ac_aux_dir/config.sub"  # Please don't use this var.
-ac_configure="$SHELL $ac_aux_dir/configure"  # Please don't use this var.
-
-
 # Make sure we can run config.sub.
 $SHELL "$ac_aux_dir/config.sub" sun4 >/dev/null 2>&1 ||
   as_fn_error $? "cannot run $SHELL $ac_aux_dir/config.sub" "$LINENO" 5
@@ -4716,7 +5093,45 @@ case $host_os in *\ *) host_os=`echo "$host_os" | sed 's/ /-/g'`;; esac
 CPU_SPINWAIT=""
 case "${host_cpu}" in
   i686|x86_64)
-	if ${je_cv_pause+:} false; then :
+	if test "x${je_cv_msvc}" = "xyes" ; then
+	    if ${je_cv_pause_msvc+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether pause instruction MSVC is compilable" >&5
+$as_echo_n "checking whether pause instruction MSVC is compilable... " >&6; }
+if ${je_cv_pause_msvc+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+int
+main ()
+{
+_mm_pause(); return 0;
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_link "$LINENO"; then :
+  je_cv_pause_msvc=yes
+else
+  je_cv_pause_msvc=no
+fi
+rm -f core conftest.err conftest.$ac_objext \
+    conftest$ac_exeext conftest.$ac_ext
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $je_cv_pause_msvc" >&5
+$as_echo "$je_cv_pause_msvc" >&6; }
+
+fi
+
+	    if test "x${je_cv_pause_msvc}" = "xyes" ; then
+		CPU_SPINWAIT='_mm_pause()'
+	    fi
+	else
+	    if ${je_cv_pause+:} false; then :
   $as_echo_n "(cached) " >&6
 else
 
@@ -4749,8 +5164,9 @@ $as_echo "$je_cv_pause" >&6; }
 
 fi
 
-	if test "x${je_cv_pause}" = "xyes" ; then
-	    CPU_SPINWAIT='__asm__ volatile("pause")'
+	    if test "x${je_cv_pause}" = "xyes" ; then
+		CPU_SPINWAIT='__asm__ volatile("pause")'
+	    fi
 	fi
 	;;
   powerpc)
@@ -4774,17 +5190,27 @@ o="$ac_objext"
 a="a"
 exe="$ac_exeext"
 libprefix="lib"
+link_whole_archive="0"
 DSO_LDFLAGS='-shared -Wl,-soname,$(@F)'
 RPATH='-Wl,-rpath,$(1)'
 SOREV="${so}.${rev}"
 PIC_CFLAGS='-fPIC -DPIC'
 CTARGET='-o $@'
 LDTARGET='-o $@'
+TEST_LD_MODE=
 EXTRA_LDFLAGS=
 ARFLAGS='crus'
 AROUT=' $@'
 CC_MM=1
 
+if test "x$je_cv_cray_prgenv_wrapper" = "xyes" ; then
+  TEST_LD_MODE='-dynamic'
+fi
+
+if test "x${je_cv_cray}" = "xyes" ; then
+  CC_MM=
+fi
+
 
 
 
@@ -4881,14 +5307,12 @@ else
 fi
 
 
+CFLAGS="$CFLAGS"
 default_munmap="1"
 maps_coalesce="1"
 case "${host}" in
   *-*-darwin* | *-*-ios*)
-	CFLAGS="$CFLAGS"
 	abi="macho"
-	$as_echo "#define JEMALLOC_PURGE_MADVISE_FREE  " >>confdefs.h
-
 	RPATH=""
 	LD_PRELOAD_VAR="DYLD_INSERT_LIBRARIES"
 	so="dylib"
@@ -4899,38 +5323,41 @@ case "${host}" in
 	sbrk_deprecated="1"
 	;;
   *-*-freebsd*)
-	CFLAGS="$CFLAGS"
 	abi="elf"
-	$as_echo "#define JEMALLOC_PURGE_MADVISE_FREE  " >>confdefs.h
+	$as_echo "#define JEMALLOC_SYSCTL_VM_OVERCOMMIT  " >>confdefs.h
 
 	force_lazy_lock="1"
 	;;
   *-*-dragonfly*)
-	CFLAGS="$CFLAGS"
 	abi="elf"
-	$as_echo "#define JEMALLOC_PURGE_MADVISE_FREE  " >>confdefs.h
-
 	;;
   *-*-openbsd*)
-	CFLAGS="$CFLAGS"
 	abi="elf"
-	$as_echo "#define JEMALLOC_PURGE_MADVISE_FREE  " >>confdefs.h
-
 	force_tls="0"
 	;;
   *-*-bitrig*)
-	CFLAGS="$CFLAGS"
 	abi="elf"
-	$as_echo "#define JEMALLOC_PURGE_MADVISE_FREE  " >>confdefs.h
-
 	;;
-  *-*-linux*)
-	CFLAGS="$CFLAGS"
-	CPPFLAGS="$CPPFLAGS -D_GNU_SOURCE"
+  *-*-linux-android)
+		CPPFLAGS="$CPPFLAGS -D_GNU_SOURCE"
 	abi="elf"
 	$as_echo "#define JEMALLOC_HAS_ALLOCA_H 1" >>confdefs.h
 
-	$as_echo "#define JEMALLOC_PURGE_MADVISE_DONTNEED  " >>confdefs.h
+	$as_echo "#define JEMALLOC_PROC_SYS_VM_OVERCOMMIT_MEMORY  " >>confdefs.h
+
+	$as_echo "#define JEMALLOC_THREADED_INIT  " >>confdefs.h
+
+	$as_echo "#define JEMALLOC_C11ATOMICS 1" >>confdefs.h
+
+	force_tls="0"
+	default_munmap="0"
+	;;
+  *-*-linux* | *-*-kfreebsd*)
+		CPPFLAGS="$CPPFLAGS -D_GNU_SOURCE"
+	abi="elf"
+	$as_echo "#define JEMALLOC_HAS_ALLOCA_H 1" >>confdefs.h
+
+	$as_echo "#define JEMALLOC_PROC_SYS_VM_OVERCOMMIT_MEMORY  " >>confdefs.h
 
 	$as_echo "#define JEMALLOC_THREADED_INIT  " >>confdefs.h
 
@@ -4958,21 +5385,16 @@ main ()
 }
 _ACEOF
 if ac_fn_c_try_compile "$LINENO"; then :
-  CFLAGS="$CFLAGS"; abi="elf"
+  abi="elf"
 else
   abi="aout"
 fi
 rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
 	{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $abi" >&5
 $as_echo "$abi" >&6; }
-	$as_echo "#define JEMALLOC_PURGE_MADVISE_FREE  " >>confdefs.h
-
 	;;
   *-*-solaris2*)
-	CFLAGS="$CFLAGS"
 	abi="elf"
-	$as_echo "#define JEMALLOC_PURGE_MADVISE_FREE  " >>confdefs.h
-
 	RPATH='-Wl,-R,$(1)'
 		CPPFLAGS="$CPPFLAGS -D_POSIX_PTHREAD_SEMANTICS"
 	LIBS="$LIBS -lposix4 -lsocket -lnsl"
@@ -4988,7 +5410,6 @@ $as_echo "$abi" >&6; }
   *-*-mingw* | *-*-cygwin*)
 	abi="pecoff"
 	force_tls="0"
-	force_lazy_lock="1"
 	maps_coalesce="0"
 	RPATH=""
 	so="dll"
@@ -5005,6 +5426,7 @@ $as_echo "$abi" >&6; }
         else
 	  importlib="${so}"
 	  DSO_LDFLAGS="-shared"
+	  link_whole_archive="1"
 	fi
 	a="lib"
 	libprefix=""
@@ -5084,6 +5506,73 @@ _ACEOF
 
 
 
+
+
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for library containing log" >&5
+$as_echo_n "checking for library containing log... " >&6; }
+if ${ac_cv_search_log+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+  ac_func_search_save_LIBS=$LIBS
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+/* Override any GCC internal prototype to avoid an error.
+   Use char because int might match the return type of a GCC
+   builtin and then its argument prototype would still apply.  */
+#ifdef __cplusplus
+extern "C"
+#endif
+char log ();
+int
+main ()
+{
+return log ();
+  ;
+  return 0;
+}
+_ACEOF
+for ac_lib in '' m; do
+  if test -z "$ac_lib"; then
+    ac_res="none required"
+  else
+    ac_res=-l$ac_lib
+    LIBS="-l$ac_lib  $ac_func_search_save_LIBS"
+  fi
+  if ac_fn_c_try_link "$LINENO"; then :
+  ac_cv_search_log=$ac_res
+fi
+rm -f core conftest.err conftest.$ac_objext \
+    conftest$ac_exeext
+  if ${ac_cv_search_log+:} false; then :
+  break
+fi
+done
+if ${ac_cv_search_log+:} false; then :
+
+else
+  ac_cv_search_log=no
+fi
+rm conftest.$ac_ext
+LIBS=$ac_func_search_save_LIBS
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_search_log" >&5
+$as_echo "$ac_cv_search_log" >&6; }
+ac_res=$ac_cv_search_log
+if test "$ac_res" != no; then :
+  test "$ac_res" = "none required" || LIBS="$ac_res $LIBS"
+
+else
+  as_fn_error $? "Missing math functions" "$LINENO" 5
+fi
+
+if test "x$ac_cv_search_log" != "xnone required" ; then
+  LM="$ac_cv_search_log"
+else
+  LM=
+fi
+
 
 
 { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether __attribute__ syntax is compilable" >&5
@@ -5193,6 +5682,42 @@ fi
 rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
 
 
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler supports -herror_on_warning" >&5
+$as_echo_n "checking whether compiler supports -herror_on_warning... " >&6; }
+TCFLAGS="${CFLAGS}"
+if test "x${CFLAGS}" = "x" ; then
+  CFLAGS="-herror_on_warning"
+else
+  CFLAGS="${CFLAGS} -herror_on_warning"
+fi
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+
+int
+main ()
+{
+
+    return 0;
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+  je_cv_cflags_appended=-herror_on_warning
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
+$as_echo "yes" >&6; }
+else
+  je_cv_cflags_appended=
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
+$as_echo "no" >&6; }
+              CFLAGS="${TCFLAGS}"
+
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+
+
 { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether tls_model attribute is compilable" >&5
 $as_echo_n "checking whether tls_model attribute is compilable... " >&6; }
 if ${je_cv_tls_model+:} false; then :
@@ -5268,6 +5793,42 @@ fi
 rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
 
 
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler supports -herror_on_warning" >&5
+$as_echo_n "checking whether compiler supports -herror_on_warning... " >&6; }
+TCFLAGS="${CFLAGS}"
+if test "x${CFLAGS}" = "x" ; then
+  CFLAGS="-herror_on_warning"
+else
+  CFLAGS="${CFLAGS} -herror_on_warning"
+fi
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+
+int
+main ()
+{
+
+    return 0;
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+  je_cv_cflags_appended=-herror_on_warning
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
+$as_echo "yes" >&6; }
+else
+  je_cv_cflags_appended=
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
+$as_echo "no" >&6; }
+              CFLAGS="${TCFLAGS}"
+
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+
+
 { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether alloc_size attribute is compilable" >&5
 $as_echo_n "checking whether alloc_size attribute is compilable... " >&6; }
 if ${je_cv_alloc_size+:} false; then :
@@ -5338,6 +5899,42 @@ fi
 rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
 
 
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler supports -herror_on_warning" >&5
+$as_echo_n "checking whether compiler supports -herror_on_warning... " >&6; }
+TCFLAGS="${CFLAGS}"
+if test "x${CFLAGS}" = "x" ; then
+  CFLAGS="-herror_on_warning"
+else
+  CFLAGS="${CFLAGS} -herror_on_warning"
+fi
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+
+int
+main ()
+{
+
+    return 0;
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+  je_cv_cflags_appended=-herror_on_warning
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
+$as_echo "yes" >&6; }
+else
+  je_cv_cflags_appended=
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
+$as_echo "no" >&6; }
+              CFLAGS="${TCFLAGS}"
+
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+
+
 { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether format(gnu_printf, ...) attribute is compilable" >&5
 $as_echo_n "checking whether format(gnu_printf, ...) attribute is compilable... " >&6; }
 if ${je_cv_format_gnu_printf+:} false; then :
@@ -5408,6 +6005,42 @@ fi
 rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
 
 
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler supports -herror_on_warning" >&5
+$as_echo_n "checking whether compiler supports -herror_on_warning... " >&6; }
+TCFLAGS="${CFLAGS}"
+if test "x${CFLAGS}" = "x" ; then
+  CFLAGS="-herror_on_warning"
+else
+  CFLAGS="${CFLAGS} -herror_on_warning"
+fi
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+
+int
+main ()
+{
+
+    return 0;
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+  je_cv_cflags_appended=-herror_on_warning
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
+$as_echo "yes" >&6; }
+else
+  je_cv_cflags_appended=
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
+$as_echo "no" >&6; }
+              CFLAGS="${TCFLAGS}"
+
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+
+
 { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether format(printf, ...) attribute is compilable" >&5
 $as_echo_n "checking whether format(printf, ...) attribute is compilable... " >&6; }
 if ${je_cv_format_printf+:} false; then :
@@ -5927,6 +6560,21 @@ fi
 install_suffix="$INSTALL_SUFFIX"
 
 
+
+# Check whether --with-malloc_conf was given.
+if test "${with_malloc_conf+set}" = set; then :
+  withval=$with_malloc_conf; JEMALLOC_CONFIG_MALLOC_CONF="$with_malloc_conf"
+else
+  JEMALLOC_CONFIG_MALLOC_CONF=""
+
+fi
+
+config_malloc_conf="$JEMALLOC_CONFIG_MALLOC_CONF"
+cat >>confdefs.h <<_ACEOF
+#define JEMALLOC_CONFIG_MALLOC_CONF "$config_malloc_conf"
+_ACEOF
+
+
 je_="je_"
 
 
@@ -6495,8 +7143,8 @@ $as_echo_n "checking configured backtracing method... " >&6; }
 { $as_echo "$as_me:${as_lineno-$LINENO}: result: $backtrace_method" >&5
 $as_echo "$backtrace_method" >&6; }
 if test "x$enable_prof" = "x1" ; then
-  if test "x$abi" != "xpecoff"; then
-        LIBS="$LIBS -lm"
+    if test "x$LM" != "x" ; then
+    LIBS="$LIBS $LM"
   fi
 
   $as_echo "#define JEMALLOC_PROF  " >>confdefs.h
@@ -6745,6 +7393,52 @@ fi
 
 
 
+
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether a program using __builtin_unreachable is compilable" >&5
+$as_echo_n "checking whether a program using __builtin_unreachable is compilable... " >&6; }
+if ${je_cv_gcc_builtin_unreachable+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+void foo (void) {
+  __builtin_unreachable();
+}
+
+int
+main ()
+{
+
+	{
+		foo();
+	}
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_link "$LINENO"; then :
+  je_cv_gcc_builtin_unreachable=yes
+else
+  je_cv_gcc_builtin_unreachable=no
+fi
+rm -f core conftest.err conftest.$ac_objext \
+    conftest$ac_exeext conftest.$ac_ext
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $je_cv_gcc_builtin_unreachable" >&5
+$as_echo "$je_cv_gcc_builtin_unreachable" >&6; }
+
+if test "x${je_cv_gcc_builtin_unreachable}" = "xyes" ; then
+  $as_echo "#define JEMALLOC_INTERNAL_UNREACHABLE __builtin_unreachable" >>confdefs.h
+
+else
+  $as_echo "#define JEMALLOC_INTERNAL_UNREACHABLE abort" >>confdefs.h
+
+fi
+
+
 { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether a program using __builtin_ffsl is compilable" >&5
 $as_echo_n "checking whether a program using __builtin_ffsl is compilable... " >&6; }
 if ${je_cv_gcc_builtin_ffsl+:} false; then :
@@ -6782,6 +7476,8 @@ fi
 $as_echo "$je_cv_gcc_builtin_ffsl" >&6; }
 
 if test "x${je_cv_gcc_builtin_ffsl}" = "xyes" ; then
+  $as_echo "#define JEMALLOC_INTERNAL_FFSLL __builtin_ffsll" >>confdefs.h
+
   $as_echo "#define JEMALLOC_INTERNAL_FFSL __builtin_ffsl" >>confdefs.h
 
   $as_echo "#define JEMALLOC_INTERNAL_FFS __builtin_ffs" >>confdefs.h
@@ -6825,6 +7521,8 @@ fi
 $as_echo "$je_cv_function_ffsl" >&6; }
 
   if test "x${je_cv_function_ffsl}" = "xyes" ; then
+    $as_echo "#define JEMALLOC_INTERNAL_FFSLL ffsll" >>confdefs.h
+
     $as_echo "#define JEMALLOC_INTERNAL_FFSL ffsl" >>confdefs.h
 
     $as_echo "#define JEMALLOC_INTERNAL_FFS ffs" >>confdefs.h
@@ -6913,7 +7611,7 @@ main ()
     if (f == NULL) {
 	return 1;
     }
-    fprintf(f, "%d\n", result);
+    fprintf(f, "%d", result);
     fclose(f);
 
     return 0;
@@ -6964,7 +7662,6 @@ else
   LG_SIZE_CLASS_GROUP="2"
 fi
 
-
 if test ! -e "${objroot}VERSION" ; then
   if test ! -e "${srcroot}VERSION" ; then
     { $as_echo "$as_me:${as_lineno-$LINENO}: result: Missing VERSION file, and unable to generate it; creating bogus VERSION" >&5
@@ -7102,39 +7799,73 @@ fi
 
 fi
 
-fi
-
-CPPFLAGS="$CPPFLAGS -D_REENTRANT"
 
-SAVED_LIBS="${LIBS}"
-LIBS=
-{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for library containing clock_gettime" >&5
-$as_echo_n "checking for library containing clock_gettime... " >&6; }
-if ${ac_cv_search_clock_gettime+:} false; then :
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether pthread_atfork(3) is compilable" >&5
+$as_echo_n "checking whether pthread_atfork(3) is compilable... " >&6; }
+if ${je_cv_pthread_atfork+:} false; then :
   $as_echo_n "(cached) " >&6
 else
-  ac_func_search_save_LIBS=$LIBS
-cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
 /* end confdefs.h.  */
 
-/* Override any GCC internal prototype to avoid an error.
-   Use char because int might match the return type of a GCC
-   builtin and then its argument prototype would still apply.  */
-#ifdef __cplusplus
-extern "C"
-#endif
-char clock_gettime ();
+#include <pthread.h>
+
 int
 main ()
 {
-return clock_gettime ();
+
+  pthread_atfork((void *)0, (void *)0, (void *)0);
+
   ;
   return 0;
 }
 _ACEOF
-for ac_lib in '' rt; do
-  if test -z "$ac_lib"; then
-    ac_res="none required"
+if ac_fn_c_try_link "$LINENO"; then :
+  je_cv_pthread_atfork=yes
+else
+  je_cv_pthread_atfork=no
+fi
+rm -f core conftest.err conftest.$ac_objext \
+    conftest$ac_exeext conftest.$ac_ext
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $je_cv_pthread_atfork" >&5
+$as_echo "$je_cv_pthread_atfork" >&6; }
+
+  if test "x${je_cv_pthread_atfork}" = "xyes" ; then
+    $as_echo "#define JEMALLOC_HAVE_PTHREAD_ATFORK  " >>confdefs.h
+
+  fi
+fi
+
+CPPFLAGS="$CPPFLAGS -D_REENTRANT"
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for library containing clock_gettime" >&5
+$as_echo_n "checking for library containing clock_gettime... " >&6; }
+if ${ac_cv_search_clock_gettime+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+  ac_func_search_save_LIBS=$LIBS
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+/* Override any GCC internal prototype to avoid an error.
+   Use char because int might match the return type of a GCC
+   builtin and then its argument prototype would still apply.  */
+#ifdef __cplusplus
+extern "C"
+#endif
+char clock_gettime ();
+int
+main ()
+{
+return clock_gettime ();
+  ;
+  return 0;
+}
+_ACEOF
+for ac_lib in '' rt; do
+  if test -z "$ac_lib"; then
+    ac_res="none required"
   else
     ac_res=-l$ac_lib
     LIBS="-l$ac_lib  $ac_func_search_save_LIBS"
@@ -7161,11 +7892,321 @@ $as_echo "$ac_cv_search_clock_gettime" >&6; }
 ac_res=$ac_cv_search_clock_gettime
 if test "$ac_res" != no; then :
   test "$ac_res" = "none required" || LIBS="$ac_res $LIBS"
-  TESTLIBS="${LIBS}"
+
 fi
 
 
-LIBS="${SAVED_LIBS}"
+if test "x$je_cv_cray_prgenv_wrapper" = "xyes" ; then
+  if test "$ac_cv_search_clock_gettime" != "-lrt"; then
+    SAVED_CFLAGS="${CFLAGS}"
+
+    unset ac_cv_search_clock_gettime
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler supports -dynamic" >&5
+$as_echo_n "checking whether compiler supports -dynamic... " >&6; }
+TCFLAGS="${CFLAGS}"
+if test "x${CFLAGS}" = "x" ; then
+  CFLAGS="-dynamic"
+else
+  CFLAGS="${CFLAGS} -dynamic"
+fi
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+
+int
+main ()
+{
+
+    return 0;
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+  je_cv_cflags_appended=-dynamic
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
+$as_echo "yes" >&6; }
+else
+  je_cv_cflags_appended=
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
+$as_echo "no" >&6; }
+              CFLAGS="${TCFLAGS}"
+
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+
+    { $as_echo "$as_me:${as_lineno-$LINENO}: checking for library containing clock_gettime" >&5
+$as_echo_n "checking for library containing clock_gettime... " >&6; }
+if ${ac_cv_search_clock_gettime+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+  ac_func_search_save_LIBS=$LIBS
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+/* Override any GCC internal prototype to avoid an error.
+   Use char because int might match the return type of a GCC
+   builtin and then its argument prototype would still apply.  */
+#ifdef __cplusplus
+extern "C"
+#endif
+char clock_gettime ();
+int
+main ()
+{
+return clock_gettime ();
+  ;
+  return 0;
+}
+_ACEOF
+for ac_lib in '' rt; do
+  if test -z "$ac_lib"; then
+    ac_res="none required"
+  else
+    ac_res=-l$ac_lib
+    LIBS="-l$ac_lib  $ac_func_search_save_LIBS"
+  fi
+  if ac_fn_c_try_link "$LINENO"; then :
+  ac_cv_search_clock_gettime=$ac_res
+fi
+rm -f core conftest.err conftest.$ac_objext \
+    conftest$ac_exeext
+  if ${ac_cv_search_clock_gettime+:} false; then :
+  break
+fi
+done
+if ${ac_cv_search_clock_gettime+:} false; then :
+
+else
+  ac_cv_search_clock_gettime=no
+fi
+rm conftest.$ac_ext
+LIBS=$ac_func_search_save_LIBS
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_search_clock_gettime" >&5
+$as_echo "$ac_cv_search_clock_gettime" >&6; }
+ac_res=$ac_cv_search_clock_gettime
+if test "$ac_res" != no; then :
+  test "$ac_res" = "none required" || LIBS="$ac_res $LIBS"
+
+fi
+
+
+    CFLAGS="${SAVED_CFLAGS}"
+  fi
+fi
+
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether clock_gettime(CLOCK_MONOTONIC_COARSE, ...) is compilable" >&5
+$as_echo_n "checking whether clock_gettime(CLOCK_MONOTONIC_COARSE, ...) is compilable... " >&6; }
+if ${je_cv_clock_monotonic_coarse+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+#include <time.h>
+
+int
+main ()
+{
+
+	struct timespec ts;
+
+	clock_gettime(CLOCK_MONOTONIC_COARSE, &ts);
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_link "$LINENO"; then :
+  je_cv_clock_monotonic_coarse=yes
+else
+  je_cv_clock_monotonic_coarse=no
+fi
+rm -f core conftest.err conftest.$ac_objext \
+    conftest$ac_exeext conftest.$ac_ext
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $je_cv_clock_monotonic_coarse" >&5
+$as_echo "$je_cv_clock_monotonic_coarse" >&6; }
+
+if test "x${je_cv_clock_monotonic_coarse}" = "xyes" ; then
+  $as_echo "#define JEMALLOC_HAVE_CLOCK_MONOTONIC_COARSE 1" >>confdefs.h
+
+fi
+
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether clock_gettime(CLOCK_MONOTONIC, ...) is compilable" >&5
+$as_echo_n "checking whether clock_gettime(CLOCK_MONOTONIC, ...) is compilable... " >&6; }
+if ${je_cv_clock_monotonic+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+#include <unistd.h>
+#include <time.h>
+
+int
+main ()
+{
+
+	struct timespec ts;
+
+	clock_gettime(CLOCK_MONOTONIC, &ts);
+#if !defined(_POSIX_MONOTONIC_CLOCK) || _POSIX_MONOTONIC_CLOCK < 0
+#  error _POSIX_MONOTONIC_CLOCK missing/invalid
+#endif
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_link "$LINENO"; then :
+  je_cv_clock_monotonic=yes
+else
+  je_cv_clock_monotonic=no
+fi
+rm -f core conftest.err conftest.$ac_objext \
+    conftest$ac_exeext conftest.$ac_ext
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $je_cv_clock_monotonic" >&5
+$as_echo "$je_cv_clock_monotonic" >&6; }
+
+if test "x${je_cv_clock_monotonic}" = "xyes" ; then
+  $as_echo "#define JEMALLOC_HAVE_CLOCK_MONOTONIC 1" >>confdefs.h
+
+fi
+
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether mach_absolute_time() is compilable" >&5
+$as_echo_n "checking whether mach_absolute_time() is compilable... " >&6; }
+if ${je_cv_mach_absolute_time+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+#include <mach/mach_time.h>
+
+int
+main ()
+{
+
+	mach_absolute_time();
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_link "$LINENO"; then :
+  je_cv_mach_absolute_time=yes
+else
+  je_cv_mach_absolute_time=no
+fi
+rm -f core conftest.err conftest.$ac_objext \
+    conftest$ac_exeext conftest.$ac_ext
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $je_cv_mach_absolute_time" >&5
+$as_echo "$je_cv_mach_absolute_time" >&6; }
+
+if test "x${je_cv_mach_absolute_time}" = "xyes" ; then
+  $as_echo "#define JEMALLOC_HAVE_MACH_ABSOLUTE_TIME 1" >>confdefs.h
+
+fi
+
+# Check whether --enable-syscall was given.
+if test "${enable_syscall+set}" = set; then :
+  enableval=$enable_syscall; if test "x$enable_syscall" = "xno" ; then
+  enable_syscall="0"
+else
+  enable_syscall="1"
+fi
+
+else
+  enable_syscall="1"
+
+fi
+
+if test "x$enable_syscall" = "x1" ; then
+      SAVED_CFLAGS="${CFLAGS}"
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether compiler supports -Werror" >&5
+$as_echo_n "checking whether compiler supports -Werror... " >&6; }
+TCFLAGS="${CFLAGS}"
+if test "x${CFLAGS}" = "x" ; then
+  CFLAGS="-Werror"
+else
+  CFLAGS="${CFLAGS} -Werror"
+fi
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+
+int
+main ()
+{
+
+    return 0;
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+  je_cv_cflags_appended=-Werror
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
+$as_echo "yes" >&6; }
+else
+  je_cv_cflags_appended=
+              { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
+$as_echo "no" >&6; }
+              CFLAGS="${TCFLAGS}"
+
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether syscall(2) is compilable" >&5
+$as_echo_n "checking whether syscall(2) is compilable... " >&6; }
+if ${je_cv_syscall+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+#include <sys/syscall.h>
+#include <unistd.h>
+
+int
+main ()
+{
+
+	syscall(SYS_write, 2, "hello", 5);
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_link "$LINENO"; then :
+  je_cv_syscall=yes
+else
+  je_cv_syscall=no
+fi
+rm -f core conftest.err conftest.$ac_objext \
+    conftest$ac_exeext conftest.$ac_ext
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $je_cv_syscall" >&5
+$as_echo "$je_cv_syscall" >&6; }
+
+  CFLAGS="${SAVED_CFLAGS}"
+  if test "x$je_cv_syscall" = "xyes" ; then
+    $as_echo "#define JEMALLOC_USE_SYSCALL  " >>confdefs.h
+
+  fi
+fi
 
 ac_fn_c_check_func "$LINENO" "secure_getenv" "ac_cv_func_secure_getenv"
 if test "x$ac_cv_func_secure_getenv" = xyes; then :
@@ -7233,10 +8274,19 @@ else
 
 fi
 
-if test "x$enable_lazy_lock" = "x" -a "x${force_lazy_lock}" = "x1" ; then
-  { $as_echo "$as_me:${as_lineno-$LINENO}: result: Forcing lazy-lock to avoid allocator/threading bootstrap issues" >&5
+if test "x${enable_lazy_lock}" = "x" ; then
+  if test "x${force_lazy_lock}" = "x1" ; then
+    { $as_echo "$as_me:${as_lineno-$LINENO}: result: Forcing lazy-lock to avoid allocator/threading bootstrap issues" >&5
 $as_echo "Forcing lazy-lock to avoid allocator/threading bootstrap issues" >&6; }
-  enable_lazy_lock="1"
+    enable_lazy_lock="1"
+  else
+    enable_lazy_lock="0"
+  fi
+fi
+if test "x${enable_lazy_lock}" = "x1" -a "x${abi}" = "xpecoff" ; then
+  { $as_echo "$as_me:${as_lineno-$LINENO}: result: Forcing no lazy-lock because thread creation monitoring is unimplemented" >&5
+$as_echo "Forcing no lazy-lock because thread creation monitoring is unimplemented" >&6; }
+  enable_lazy_lock="0"
 fi
 if test "x$enable_lazy_lock" = "x1" ; then
   if test "x$abi" != "xpecoff" ; then
@@ -7306,8 +8356,6 @@ fi
   fi
   $as_echo "#define JEMALLOC_LAZY_LOCK  " >>confdefs.h
 
-else
-  enable_lazy_lock="0"
 fi
 
 
@@ -7544,9 +8592,7 @@ int
 main ()
 {
 
-	{
-		madvise((void *)0, 0, 0);
-	}
+	madvise((void *)0, 0, 0);
 
   ;
   return 0;
@@ -7566,6 +8612,118 @@ $as_echo "$je_cv_madvise" >&6; }
 if test "x${je_cv_madvise}" = "xyes" ; then
   $as_echo "#define JEMALLOC_HAVE_MADVISE  " >>confdefs.h
 
+
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether madvise(..., MADV_FREE) is compilable" >&5
+$as_echo_n "checking whether madvise(..., MADV_FREE) is compilable... " >&6; }
+if ${je_cv_madv_free+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+#include <sys/mman.h>
+
+int
+main ()
+{
+
+	madvise((void *)0, 0, MADV_FREE);
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_link "$LINENO"; then :
+  je_cv_madv_free=yes
+else
+  je_cv_madv_free=no
+fi
+rm -f core conftest.err conftest.$ac_objext \
+    conftest$ac_exeext conftest.$ac_ext
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $je_cv_madv_free" >&5
+$as_echo "$je_cv_madv_free" >&6; }
+
+  if test "x${je_cv_madv_free}" = "xyes" ; then
+    $as_echo "#define JEMALLOC_PURGE_MADVISE_FREE  " >>confdefs.h
+
+  fi
+
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether madvise(..., MADV_DONTNEED) is compilable" >&5
+$as_echo_n "checking whether madvise(..., MADV_DONTNEED) is compilable... " >&6; }
+if ${je_cv_madv_dontneed+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+#include <sys/mman.h>
+
+int
+main ()
+{
+
+	madvise((void *)0, 0, MADV_DONTNEED);
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_link "$LINENO"; then :
+  je_cv_madv_dontneed=yes
+else
+  je_cv_madv_dontneed=no
+fi
+rm -f core conftest.err conftest.$ac_objext \
+    conftest$ac_exeext conftest.$ac_ext
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $je_cv_madv_dontneed" >&5
+$as_echo "$je_cv_madv_dontneed" >&6; }
+
+  if test "x${je_cv_madv_dontneed}" = "xyes" ; then
+    $as_echo "#define JEMALLOC_PURGE_MADVISE_DONTNEED  " >>confdefs.h
+
+  fi
+
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether madvise(..., MADV_[NO]HUGEPAGE) is compilable" >&5
+$as_echo_n "checking whether madvise(..., MADV_[NO]HUGEPAGE) is compilable... " >&6; }
+if ${je_cv_thp+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+#include <sys/mman.h>
+
+int
+main ()
+{
+
+	madvise((void *)0, 0, MADV_HUGEPAGE);
+	madvise((void *)0, 0, MADV_NOHUGEPAGE);
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_link "$LINENO"; then :
+  je_cv_thp=yes
+else
+  je_cv_thp=no
+fi
+rm -f core conftest.err conftest.$ac_objext \
+    conftest$ac_exeext conftest.$ac_ext
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $je_cv_thp" >&5
+$as_echo "$je_cv_thp" >&6; }
+
+  if test "x${je_cv_thp}" = "xyes" ; then
+    $as_echo "#define JEMALLOC_THP  " >>confdefs.h
+
+  fi
 fi
 
 
@@ -7708,6 +8866,51 @@ fi
 
 
 
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking whether Darwin os_unfair_lock_*() is compilable" >&5
+$as_echo_n "checking whether Darwin os_unfair_lock_*() is compilable... " >&6; }
+if ${je_cv_os_unfair_lock+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+#include <os/lock.h>
+#include <AvailabilityMacros.h>
+
+int
+main ()
+{
+
+	#if MAC_OS_X_VERSION_MIN_REQUIRED < 101200
+	#error "os_unfair_lock is not supported"
+	#else
+	os_unfair_lock lock = OS_UNFAIR_LOCK_INIT;
+	os_unfair_lock_lock(&lock);
+	os_unfair_lock_unlock(&lock);
+	#endif
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_link "$LINENO"; then :
+  je_cv_os_unfair_lock=yes
+else
+  je_cv_os_unfair_lock=no
+fi
+rm -f core conftest.err conftest.$ac_objext \
+    conftest$ac_exeext conftest.$ac_ext
+fi
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $je_cv_os_unfair_lock" >&5
+$as_echo "$je_cv_os_unfair_lock" >&6; }
+
+if test "x${je_cv_os_unfair_lock}" = "xyes" ; then
+  $as_echo "#define JEMALLOC_OS_UNFAIR_LOCK  " >>confdefs.h
+
+fi
+
+
+
 { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether Darwin OSSpin*() is compilable" >&5
 $as_echo_n "checking whether Darwin OSSpin*() is compilable... " >&6; }
 if ${je_cv_osspin+:} false; then :
@@ -9610,6 +10813,8 @@ $as_echo "CONFIG             : ${CONFIG}" >&6; }
 $as_echo "CC                 : ${CC}" >&6; }
 { $as_echo "$as_me:${as_lineno-$LINENO}: result: CFLAGS             : ${CFLAGS}" >&5
 $as_echo "CFLAGS             : ${CFLAGS}" >&6; }
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: EXTRA_CFLAGS       : ${EXTRA_CFLAGS}" >&5
+$as_echo "EXTRA_CFLAGS       : ${EXTRA_CFLAGS}" >&6; }
 { $as_echo "$as_me:${as_lineno-$LINENO}: result: CPPFLAGS           : ${CPPFLAGS}" >&5
 $as_echo "CPPFLAGS           : ${CPPFLAGS}" >&6; }
 { $as_echo "$as_me:${as_lineno-$LINENO}: result: LDFLAGS            : ${LDFLAGS}" >&5
@@ -9618,8 +10823,6 @@ $as_echo "LDFLAGS            : ${LDFLAGS}" >&6; }
 $as_echo "EXTRA_LDFLAGS      : ${EXTRA_LDFLAGS}" >&6; }
 { $as_echo "$as_me:${as_lineno-$LINENO}: result: LIBS               : ${LIBS}" >&5
 $as_echo "LIBS               : ${LIBS}" >&6; }
-{ $as_echo "$as_me:${as_lineno-$LINENO}: result: TESTLIBS           : ${TESTLIBS}" >&5
-$as_echo "TESTLIBS           : ${TESTLIBS}" >&6; }
 { $as_echo "$as_me:${as_lineno-$LINENO}: result: RPATH_EXTRA        : ${RPATH_EXTRA}" >&5
 $as_echo "RPATH_EXTRA        : ${RPATH_EXTRA}" >&6; }
 { $as_echo "$as_me:${as_lineno-$LINENO}: result: " >&5
@@ -9662,6 +10865,8 @@ $as_echo "JEMALLOC_PRIVATE_NAMESPACE" >&6; }
 $as_echo "                   : ${JEMALLOC_PRIVATE_NAMESPACE}" >&6; }
 { $as_echo "$as_me:${as_lineno-$LINENO}: result: install_suffix     : ${install_suffix}" >&5
 $as_echo "install_suffix     : ${install_suffix}" >&6; }
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: malloc_conf        : ${config_malloc_conf}" >&5
+$as_echo "malloc_conf        : ${config_malloc_conf}" >&6; }
 { $as_echo "$as_me:${as_lineno-$LINENO}: result: autogen            : ${enable_autogen}" >&5
 $as_echo "autogen            : ${enable_autogen}" >&6; }
 { $as_echo "$as_me:${as_lineno-$LINENO}: result: cc-silence         : ${enable_cc_silence}" >&5
diff --git a/deps/jemalloc/configure.ac b/deps/jemalloc/configure.ac
index 7a1290e0d..9573c3020 100644
--- a/deps/jemalloc/configure.ac
+++ b/deps/jemalloc/configure.ac
@@ -1,6 +1,8 @@
 dnl Process this file with autoconf to produce a configure script.
 AC_INIT([Makefile.in])
 
+AC_CONFIG_AUX_DIR([build-aux])
+
 dnl ============================================================================
 dnl Custom macro definitions.
 
@@ -116,6 +118,7 @@ dnl If CFLAGS isn't defined, set CFLAGS to something reasonable.  Otherwise,
 dnl just prevent autoconf from molesting CFLAGS.
 CFLAGS=$CFLAGS
 AC_PROG_CC
+
 if test "x$GCC" != "xyes" ; then
   AC_CACHE_CHECK([whether compiler is MSVC],
                  [je_cv_msvc],
@@ -129,15 +132,58 @@ if test "x$GCC" != "xyes" ; then
                                [je_cv_msvc=no])])
 fi
 
+dnl check if a cray prgenv wrapper compiler is being used
+je_cv_cray_prgenv_wrapper=""
+if test "x${PE_ENV}" != "x" ; then
+  case "${CC}" in
+    CC|cc)
+	je_cv_cray_prgenv_wrapper="yes"
+	;;
+    *)
+       ;;
+  esac
+fi
+
+AC_CACHE_CHECK([whether compiler is cray],
+              [je_cv_cray],
+              [AC_COMPILE_IFELSE([AC_LANG_PROGRAM([],
+                                                  [
+#ifndef _CRAYC
+  int fail[-1];
+#endif
+])],
+                            [je_cv_cray=yes],
+                            [je_cv_cray=no])])
+
+if test "x${je_cv_cray}" = "xyes" ; then
+  AC_CACHE_CHECK([whether cray compiler version is 8.4],
+                [je_cv_cray_84],
+                [AC_COMPILE_IFELSE([AC_LANG_PROGRAM([],
+                                                      [
+#if !(_RELEASE_MAJOR == 8 && _RELEASE_MINOR == 4)
+  int fail[-1];
+#endif
+])],
+                              [je_cv_cray_84=yes],
+                              [je_cv_cray_84=no])])
+fi
+
 if test "x$CFLAGS" = "x" ; then
   no_CFLAGS="yes"
   if test "x$GCC" = "xyes" ; then
-    JE_CFLAGS_APPEND([-std=gnu99])
-    if test "x$je_cv_cflags_appended" = "x-std=gnu99" ; then
+    JE_CFLAGS_APPEND([-std=gnu11])
+    if test "x$je_cv_cflags_appended" = "x-std=gnu11" ; then
       AC_DEFINE_UNQUOTED([JEMALLOC_HAS_RESTRICT])
+    else
+      JE_CFLAGS_APPEND([-std=gnu99])
+      if test "x$je_cv_cflags_appended" = "x-std=gnu99" ; then
+        AC_DEFINE_UNQUOTED([JEMALLOC_HAS_RESTRICT])
+      fi
     fi
     JE_CFLAGS_APPEND([-Wall])
     JE_CFLAGS_APPEND([-Werror=declaration-after-statement])
+    JE_CFLAGS_APPEND([-Wshorten-64-to-32])
+    JE_CFLAGS_APPEND([-Wsign-compare])
     JE_CFLAGS_APPEND([-pipe])
     JE_CFLAGS_APPEND([-g3])
   elif test "x$je_cv_msvc" = "xyes" ; then
@@ -148,11 +194,21 @@ if test "x$CFLAGS" = "x" ; then
     JE_CFLAGS_APPEND([-FS])
     CPPFLAGS="$CPPFLAGS -I${srcdir}/include/msvc_compat"
   fi
+  if test "x$je_cv_cray" = "xyes" ; then
+    dnl cray compiler 8.4 has an inlining bug
+    if test "x$je_cv_cray_84" = "xyes" ; then
+      JE_CFLAGS_APPEND([-hipa2])
+      JE_CFLAGS_APPEND([-hnognu])
+    fi
+    if test "x$enable_cc_silence" != "xno" ; then
+      dnl ignore unreachable code warning
+      JE_CFLAGS_APPEND([-hnomessage=128])
+      dnl ignore redefinition of "malloc", "free", etc warning
+      JE_CFLAGS_APPEND([-hnomessage=1357])
+    fi
+  fi
 fi
-dnl Append EXTRA_CFLAGS to CFLAGS, if defined.
-if test "x$EXTRA_CFLAGS" != "x" ; then
-  JE_CFLAGS_APPEND([$EXTRA_CFLAGS])
-fi
+AC_SUBST([EXTRA_CFLAGS])
 AC_PROG_CPP
 
 AC_C_BIGENDIAN([ac_cv_big_endian=1], [ac_cv_big_endian=0])
@@ -164,13 +220,18 @@ if test "x${je_cv_msvc}" = "xyes" -a "x${ac_cv_header_inttypes_h}" = "xno"; then
   CPPFLAGS="$CPPFLAGS -I${srcdir}/include/msvc_compat/C99"
 fi
 
-AC_CHECK_SIZEOF([void *])
-if test "x${ac_cv_sizeof_void_p}" = "x8" ; then
-  LG_SIZEOF_PTR=3
-elif test "x${ac_cv_sizeof_void_p}" = "x4" ; then
-  LG_SIZEOF_PTR=2
+if test "x${je_cv_msvc}" = "xyes" ; then
+  LG_SIZEOF_PTR=LG_SIZEOF_PTR_WIN
+  AC_MSG_RESULT([Using a predefined value for sizeof(void *): 4 for 32-bit, 8 for 64-bit])
 else
-  AC_MSG_ERROR([Unsupported pointer size: ${ac_cv_sizeof_void_p}])
+  AC_CHECK_SIZEOF([void *])
+  if test "x${ac_cv_sizeof_void_p}" = "x8" ; then
+    LG_SIZEOF_PTR=3
+  elif test "x${ac_cv_sizeof_void_p}" = "x4" ; then
+    LG_SIZEOF_PTR=2
+  else
+    AC_MSG_ERROR([Unsupported pointer size: ${ac_cv_sizeof_void_p}])
+  fi
 fi
 AC_DEFINE_UNQUOTED([LG_SIZEOF_PTR], [$LG_SIZEOF_PTR])
 
@@ -194,6 +255,16 @@ else
 fi
 AC_DEFINE_UNQUOTED([LG_SIZEOF_LONG], [$LG_SIZEOF_LONG])
 
+AC_CHECK_SIZEOF([long long])
+if test "x${ac_cv_sizeof_long_long}" = "x8" ; then
+  LG_SIZEOF_LONG_LONG=3
+elif test "x${ac_cv_sizeof_long_long}" = "x4" ; then
+  LG_SIZEOF_LONG_LONG=2
+else
+  AC_MSG_ERROR([Unsupported long long size: ${ac_cv_sizeof_long_long}])
+fi
+AC_DEFINE_UNQUOTED([LG_SIZEOF_LONG_LONG], [$LG_SIZEOF_LONG_LONG])
+
 AC_CHECK_SIZEOF([intmax_t])
 if test "x${ac_cv_sizeof_intmax_t}" = "x16" ; then
   LG_SIZEOF_INTMAX_T=4
@@ -211,12 +282,22 @@ dnl CPU-specific settings.
 CPU_SPINWAIT=""
 case "${host_cpu}" in
   i686|x86_64)
-	AC_CACHE_VAL([je_cv_pause],
-	  [JE_COMPILABLE([pause instruction], [],
-	                [[__asm__ volatile("pause"); return 0;]],
-	                [je_cv_pause])])
-	if test "x${je_cv_pause}" = "xyes" ; then
-	    CPU_SPINWAIT='__asm__ volatile("pause")'
+	if test "x${je_cv_msvc}" = "xyes" ; then
+	    AC_CACHE_VAL([je_cv_pause_msvc],
+	      [JE_COMPILABLE([pause instruction MSVC], [],
+					[[_mm_pause(); return 0;]],
+					[je_cv_pause_msvc])])
+	    if test "x${je_cv_pause_msvc}" = "xyes" ; then
+		CPU_SPINWAIT='_mm_pause()'
+	    fi
+	else
+	    AC_CACHE_VAL([je_cv_pause],
+	      [JE_COMPILABLE([pause instruction], [],
+					[[__asm__ volatile("pause"); return 0;]],
+					[je_cv_pause])])
+	    if test "x${je_cv_pause}" = "xyes" ; then
+		CPU_SPINWAIT='__asm__ volatile("pause")'
+	    fi
 	fi
 	;;
   powerpc)
@@ -234,17 +315,27 @@ o="$ac_objext"
 a="a"
 exe="$ac_exeext"
 libprefix="lib"
+link_whole_archive="0"
 DSO_LDFLAGS='-shared -Wl,-soname,$(@F)'
 RPATH='-Wl,-rpath,$(1)'
 SOREV="${so}.${rev}"
 PIC_CFLAGS='-fPIC -DPIC'
 CTARGET='-o $@'
 LDTARGET='-o $@'
+TEST_LD_MODE=
 EXTRA_LDFLAGS=
 ARFLAGS='crus'
 AROUT=' $@'
 CC_MM=1
 
+if test "x$je_cv_cray_prgenv_wrapper" = "xyes" ; then
+  TEST_LD_MODE='-dynamic'
+fi
+
+if test "x${je_cv_cray}" = "xyes" ; then
+  CC_MM=
+fi
+
 AN_MAKEVAR([AR], [AC_PROG_AR])
 AN_PROGRAM([ar], [AC_PROG_AR])
 AC_DEFUN([AC_PROG_AR], [AC_CHECK_TOOL(AR, ar, :)])
@@ -257,13 +348,12 @@ dnl
 dnl Define cpp macros in CPPFLAGS, rather than doing AC_DEFINE(macro), since the
 dnl definitions need to be seen before any headers are included, which is a pain
 dnl to make happen otherwise.
+CFLAGS="$CFLAGS"
 default_munmap="1"
 maps_coalesce="1"
 case "${host}" in
   *-*-darwin* | *-*-ios*)
-	CFLAGS="$CFLAGS"
 	abi="macho"
-	AC_DEFINE([JEMALLOC_PURGE_MADVISE_FREE], [ ])
 	RPATH=""
 	LD_PRELOAD_VAR="DYLD_INSERT_LIBRARIES"
 	so="dylib"
@@ -274,33 +364,37 @@ case "${host}" in
 	sbrk_deprecated="1"
 	;;
   *-*-freebsd*)
-	CFLAGS="$CFLAGS"
 	abi="elf"
-	AC_DEFINE([JEMALLOC_PURGE_MADVISE_FREE], [ ])
+	AC_DEFINE([JEMALLOC_SYSCTL_VM_OVERCOMMIT], [ ])
 	force_lazy_lock="1"
 	;;
   *-*-dragonfly*)
-	CFLAGS="$CFLAGS"
 	abi="elf"
-	AC_DEFINE([JEMALLOC_PURGE_MADVISE_FREE], [ ])
 	;;
   *-*-openbsd*)
-	CFLAGS="$CFLAGS"
 	abi="elf"
-	AC_DEFINE([JEMALLOC_PURGE_MADVISE_FREE], [ ])
 	force_tls="0"
 	;;
   *-*-bitrig*)
-	CFLAGS="$CFLAGS"
 	abi="elf"
-	AC_DEFINE([JEMALLOC_PURGE_MADVISE_FREE], [ ])
 	;;
-  *-*-linux*)
-	CFLAGS="$CFLAGS"
+  *-*-linux-android)
+	dnl syscall(2) and secure_getenv(3) are exposed by _GNU_SOURCE.
 	CPPFLAGS="$CPPFLAGS -D_GNU_SOURCE"
 	abi="elf"
 	AC_DEFINE([JEMALLOC_HAS_ALLOCA_H])
-	AC_DEFINE([JEMALLOC_PURGE_MADVISE_DONTNEED], [ ])
+	AC_DEFINE([JEMALLOC_PROC_SYS_VM_OVERCOMMIT_MEMORY], [ ])
+	AC_DEFINE([JEMALLOC_THREADED_INIT], [ ])
+	AC_DEFINE([JEMALLOC_C11ATOMICS])
+	force_tls="0"
+	default_munmap="0"
+	;;
+  *-*-linux* | *-*-kfreebsd*)
+	dnl syscall(2) and secure_getenv(3) are exposed by _GNU_SOURCE.
+	CPPFLAGS="$CPPFLAGS -D_GNU_SOURCE"
+	abi="elf"
+	AC_DEFINE([JEMALLOC_HAS_ALLOCA_H])
+	AC_DEFINE([JEMALLOC_PROC_SYS_VM_OVERCOMMIT_MEMORY], [ ])
 	AC_DEFINE([JEMALLOC_THREADED_INIT], [ ])
 	AC_DEFINE([JEMALLOC_USE_CXX_THROW], [ ])
 	default_munmap="0"
@@ -314,15 +408,12 @@ case "${host}" in
 #error aout
 #endif
 ]])],
-                          [CFLAGS="$CFLAGS"; abi="elf"],
+                          [abi="elf"],
                           [abi="aout"])
 	AC_MSG_RESULT([$abi])
-	AC_DEFINE([JEMALLOC_PURGE_MADVISE_FREE], [ ])
 	;;
   *-*-solaris2*)
-	CFLAGS="$CFLAGS"
 	abi="elf"
-	AC_DEFINE([JEMALLOC_PURGE_MADVISE_FREE], [ ])
 	RPATH='-Wl,-R,$(1)'
 	dnl Solaris needs this for sigwait().
 	CPPFLAGS="$CPPFLAGS -D_POSIX_PTHREAD_SEMANTICS"
@@ -341,7 +432,6 @@ case "${host}" in
   *-*-mingw* | *-*-cygwin*)
 	abi="pecoff"
 	force_tls="0"
-	force_lazy_lock="1"
 	maps_coalesce="0"
 	RPATH=""
 	so="dll"
@@ -358,6 +448,7 @@ case "${host}" in
         else
 	  importlib="${so}"
 	  DSO_LDFLAGS="-shared"
+	  link_whole_archive="1"
 	fi
 	a="lib"
 	libprefix=""
@@ -395,17 +486,28 @@ AC_SUBST([o])
 AC_SUBST([a])
 AC_SUBST([exe])
 AC_SUBST([libprefix])
+AC_SUBST([link_whole_archive])
 AC_SUBST([DSO_LDFLAGS])
 AC_SUBST([EXTRA_LDFLAGS])
 AC_SUBST([SOREV])
 AC_SUBST([PIC_CFLAGS])
 AC_SUBST([CTARGET])
 AC_SUBST([LDTARGET])
+AC_SUBST([TEST_LD_MODE])
 AC_SUBST([MKLIB])
 AC_SUBST([ARFLAGS])
 AC_SUBST([AROUT])
 AC_SUBST([CC_MM])
 
+dnl Determine whether libm must be linked to use e.g. log(3).
+AC_SEARCH_LIBS([log], [m], , [AC_MSG_ERROR([Missing math functions])])
+if test "x$ac_cv_search_log" != "xnone required" ; then
+  LM="$ac_cv_search_log"
+else
+  LM=
+fi
+AC_SUBST(LM)
+
 JE_COMPILABLE([__attribute__ syntax],
               [static __attribute__((unused)) void foo(void){}],
               [],
@@ -419,6 +521,7 @@ fi
 dnl Check for tls_model attribute support (clang 3.0 still lacks support).
 SAVED_CFLAGS="${CFLAGS}"
 JE_CFLAGS_APPEND([-Werror])
+JE_CFLAGS_APPEND([-herror_on_warning])
 JE_COMPILABLE([tls_model attribute], [],
               [static __thread int
                __attribute__((tls_model("initial-exec"), unused)) foo;
@@ -434,6 +537,7 @@ fi
 dnl Check for alloc_size attribute support.
 SAVED_CFLAGS="${CFLAGS}"
 JE_CFLAGS_APPEND([-Werror])
+JE_CFLAGS_APPEND([-herror_on_warning])
 JE_COMPILABLE([alloc_size attribute], [#include <stdlib.h>],
               [void *foo(size_t size) __attribute__((alloc_size(1)));],
               [je_cv_alloc_size])
@@ -444,6 +548,7 @@ fi
 dnl Check for format(gnu_printf, ...) attribute support.
 SAVED_CFLAGS="${CFLAGS}"
 JE_CFLAGS_APPEND([-Werror])
+JE_CFLAGS_APPEND([-herror_on_warning])
 JE_COMPILABLE([format(gnu_printf, ...) attribute], [#include <stdlib.h>],
               [void *foo(const char *format, ...) __attribute__((format(gnu_printf, 1, 2)));],
               [je_cv_format_gnu_printf])
@@ -454,6 +559,7 @@ fi
 dnl Check for format(printf, ...) attribute support.
 SAVED_CFLAGS="${CFLAGS}"
 JE_CFLAGS_APPEND([-Werror])
+JE_CFLAGS_APPEND([-herror_on_warning])
 JE_COMPILABLE([format(printf, ...) attribute], [#include <stdlib.h>],
               [void *foo(const char *format, ...) __attribute__((format(printf, 1, 2)));],
               [je_cv_format_printf])
@@ -575,6 +681,15 @@ AC_ARG_WITH([install_suffix],
 install_suffix="$INSTALL_SUFFIX"
 AC_SUBST([install_suffix])
 
+dnl Specify default malloc_conf.
+AC_ARG_WITH([malloc_conf],
+  [AS_HELP_STRING([--with-malloc-conf=<malloc_conf>], [config.malloc_conf options string])],
+  [JEMALLOC_CONFIG_MALLOC_CONF="$with_malloc_conf"],
+  [JEMALLOC_CONFIG_MALLOC_CONF=""]
+)
+config_malloc_conf="$JEMALLOC_CONFIG_MALLOC_CONF"
+AC_DEFINE_UNQUOTED([JEMALLOC_CONFIG_MALLOC_CONF], ["$config_malloc_conf"])
+
 dnl Substitute @je_@ in jemalloc_protos.h.in, primarily to make generation of
 dnl jemalloc_protos_jet.h easy.
 je_="je_"
@@ -839,9 +954,9 @@ fi
 AC_MSG_CHECKING([configured backtracing method])
 AC_MSG_RESULT([$backtrace_method])
 if test "x$enable_prof" = "x1" ; then
-  if test "x$abi" != "xpecoff"; then
-    dnl Heap profiling uses the log(3) function.
-    LIBS="$LIBS -lm"
+  dnl Heap profiling uses the log(3) function.
+  if test "x$LM" != "x" ; then
+    LIBS="$LIBS $LM"
   fi
 
   AC_DEFINE([JEMALLOC_PROF], [ ])
@@ -1010,11 +1125,28 @@ if test "x$enable_cache_oblivious" = "x1" ; then
 fi
 AC_SUBST([enable_cache_oblivious])
 
+
+
+JE_COMPILABLE([a program using __builtin_unreachable], [
+void foo (void) {
+  __builtin_unreachable();
+}
+], [
+	{
+		foo();
+	}
+], [je_cv_gcc_builtin_unreachable])
+if test "x${je_cv_gcc_builtin_unreachable}" = "xyes" ; then
+  AC_DEFINE([JEMALLOC_INTERNAL_UNREACHABLE], [__builtin_unreachable])
+else
+  AC_DEFINE([JEMALLOC_INTERNAL_UNREACHABLE], [abort])
+fi
+
 dnl ============================================================================
 dnl Check for  __builtin_ffsl(), then ffsl(3), and fail if neither are found.
 dnl One of those two functions should (theoretically) exist on all platforms
 dnl that jemalloc currently has a chance of functioning on without modification.
-dnl We additionally assume ffs() or __builtin_ffs() are defined if
+dnl We additionally assume ffs[ll]() or __builtin_ffs[ll]() are defined if
 dnl ffsl() or __builtin_ffsl() are defined, respectively.
 JE_COMPILABLE([a program using __builtin_ffsl], [
 #include <stdio.h>
@@ -1027,6 +1159,7 @@ JE_COMPILABLE([a program using __builtin_ffsl], [
 	}
 ], [je_cv_gcc_builtin_ffsl])
 if test "x${je_cv_gcc_builtin_ffsl}" = "xyes" ; then
+  AC_DEFINE([JEMALLOC_INTERNAL_FFSLL], [__builtin_ffsll])
   AC_DEFINE([JEMALLOC_INTERNAL_FFSL], [__builtin_ffsl])
   AC_DEFINE([JEMALLOC_INTERNAL_FFS], [__builtin_ffs])
 else
@@ -1041,6 +1174,7 @@ else
 	}
   ], [je_cv_function_ffsl])
   if test "x${je_cv_function_ffsl}" = "xyes" ; then
+    AC_DEFINE([JEMALLOC_INTERNAL_FFSLL], [ffsll])
     AC_DEFINE([JEMALLOC_INTERNAL_FFSL], [ffsl])
     AC_DEFINE([JEMALLOC_INTERNAL_FFS], [ffs])
   else
@@ -1100,7 +1234,7 @@ if test "x$LG_PAGE" = "xdetect"; then
     if (f == NULL) {
 	return 1;
     }
-    fprintf(f, "%d\n", result);
+    fprintf(f, "%d", result);
     fclose(f);
 
     return 0;
@@ -1133,27 +1267,36 @@ dnl ============================================================================
 dnl jemalloc configuration.
 dnl 
 
-dnl Set VERSION if source directory is inside a git repository.
-if test "x`test ! \"${srcroot}\" && cd \"${srcroot}\"; git rev-parse --is-inside-work-tree 2>/dev/null`" = "xtrue" ; then
-  dnl Pattern globs aren't powerful enough to match both single- and
-  dnl double-digit version numbers, so iterate over patterns to support up to
-  dnl version 99.99.99 without any accidental matches.
-  rm -f "${objroot}VERSION"
-  for pattern in ['[0-9].[0-9].[0-9]' '[0-9].[0-9].[0-9][0-9]' \
-                 '[0-9].[0-9][0-9].[0-9]' '[0-9].[0-9][0-9].[0-9][0-9]' \
-                 '[0-9][0-9].[0-9].[0-9]' '[0-9][0-9].[0-9].[0-9][0-9]' \
-                 '[0-9][0-9].[0-9][0-9].[0-9]' \
-                 '[0-9][0-9].[0-9][0-9].[0-9][0-9]']; do
-    if test ! -e "${objroot}VERSION" ; then
-      (test ! "${srcroot}" && cd "${srcroot}"; git describe --long --abbrev=40 --match="${pattern}") > "${objroot}VERSION.tmp" 2>/dev/null
-      if test $? -eq 0 ; then
-        mv "${objroot}VERSION.tmp" "${objroot}VERSION"
-        break
-      fi
+AC_ARG_WITH([version],
+  [AS_HELP_STRING([--with-version=<major>.<minor>.<bugfix>-<nrev>-g<gid>],
+   [Version string])],
+  [
+    echo "${with_version}" | grep ['^[0-9]\+\.[0-9]\+\.[0-9]\+-[0-9]\+-g[0-9a-f]\+$'] 2>&1 1>/dev/null
+    if test $? -ne 0 ; then
+      AC_MSG_ERROR([${with_version} does not match <major>.<minor>.<bugfix>-<nrev>-g<gid>])
     fi
-  done
-fi
-rm -f "${objroot}VERSION.tmp"
+    echo "$with_version" > "${objroot}VERSION"
+  ], [
+    dnl Set VERSION if source directory is inside a git repository.
+    if test "x`test ! \"${srcroot}\" && cd \"${srcroot}\"; git rev-parse --is-inside-work-tree 2>/dev/null`" = "xtrue" ; then
+      dnl Pattern globs aren't powerful enough to match both single- and
+      dnl double-digit version numbers, so iterate over patterns to support up
+      dnl to version 99.99.99 without any accidental matches.
+      for pattern in ['[0-9].[0-9].[0-9]' '[0-9].[0-9].[0-9][0-9]' \
+                     '[0-9].[0-9][0-9].[0-9]' '[0-9].[0-9][0-9].[0-9][0-9]' \
+                     '[0-9][0-9].[0-9].[0-9]' '[0-9][0-9].[0-9].[0-9][0-9]' \
+                     '[0-9][0-9].[0-9][0-9].[0-9]' \
+                     '[0-9][0-9].[0-9][0-9].[0-9][0-9]']; do
+        (test ! "${srcroot}" && cd "${srcroot}"; git describe --long --abbrev=40 --match="${pattern}") > "${objroot}VERSION.tmp" 2>/dev/null
+        if test $? -eq 0 ; then
+          mv "${objroot}VERSION.tmp" "${objroot}VERSION"
+          break
+        fi
+      done
+    fi
+    rm -f "${objroot}VERSION.tmp"
+  ])
+
 if test ! -e "${objroot}VERSION" ; then
   if test ! -e "${srcroot}VERSION" ; then
     AC_MSG_RESULT(
@@ -1186,17 +1329,101 @@ if test "x$abi" != "xpecoff" ; then
   AC_CHECK_LIB([pthread], [pthread_create], [LIBS="$LIBS -lpthread"],
                [AC_SEARCH_LIBS([pthread_create], , ,
                                AC_MSG_ERROR([libpthread is missing]))])
+  JE_COMPILABLE([pthread_atfork(3)], [
+#include <pthread.h>
+], [
+  pthread_atfork((void *)0, (void *)0, (void *)0);
+], [je_cv_pthread_atfork])
+  if test "x${je_cv_pthread_atfork}" = "xyes" ; then
+    AC_DEFINE([JEMALLOC_HAVE_PTHREAD_ATFORK], [ ])
+  fi
 fi
 
 CPPFLAGS="$CPPFLAGS -D_REENTRANT"
 
-dnl Check whether clock_gettime(2) is in libc or librt.  This function is only
-dnl used in test code, so save the result to TESTLIBS to avoid poluting LIBS.
-SAVED_LIBS="${LIBS}"
-LIBS=
-AC_SEARCH_LIBS([clock_gettime], [rt], [TESTLIBS="${LIBS}"])
-AC_SUBST([TESTLIBS])
-LIBS="${SAVED_LIBS}"
+dnl Check whether clock_gettime(2) is in libc or librt.
+AC_SEARCH_LIBS([clock_gettime], [rt])
+
+dnl Cray wrapper compiler often adds `-lrt` when using `-static`. Check with
+dnl `-dynamic` as well in case a user tries to dynamically link in jemalloc
+if test "x$je_cv_cray_prgenv_wrapper" = "xyes" ; then
+  if test "$ac_cv_search_clock_gettime" != "-lrt"; then
+    SAVED_CFLAGS="${CFLAGS}"
+
+    unset ac_cv_search_clock_gettime
+    JE_CFLAGS_APPEND([-dynamic])
+    AC_SEARCH_LIBS([clock_gettime], [rt])
+
+    CFLAGS="${SAVED_CFLAGS}"
+  fi
+fi
+
+dnl check for CLOCK_MONOTONIC_COARSE (Linux-specific).
+JE_COMPILABLE([clock_gettime(CLOCK_MONOTONIC_COARSE, ...)], [
+#include <time.h>
+], [
+	struct timespec ts;
+
+	clock_gettime(CLOCK_MONOTONIC_COARSE, &ts);
+], [je_cv_clock_monotonic_coarse])
+if test "x${je_cv_clock_monotonic_coarse}" = "xyes" ; then
+  AC_DEFINE([JEMALLOC_HAVE_CLOCK_MONOTONIC_COARSE])
+fi
+
+dnl check for CLOCK_MONOTONIC.
+JE_COMPILABLE([clock_gettime(CLOCK_MONOTONIC, ...)], [
+#include <unistd.h>
+#include <time.h>
+], [
+	struct timespec ts;
+
+	clock_gettime(CLOCK_MONOTONIC, &ts);
+#if !defined(_POSIX_MONOTONIC_CLOCK) || _POSIX_MONOTONIC_CLOCK < 0
+#  error _POSIX_MONOTONIC_CLOCK missing/invalid
+#endif
+], [je_cv_clock_monotonic])
+if test "x${je_cv_clock_monotonic}" = "xyes" ; then
+  AC_DEFINE([JEMALLOC_HAVE_CLOCK_MONOTONIC])
+fi
+
+dnl Check for mach_absolute_time().
+JE_COMPILABLE([mach_absolute_time()], [
+#include <mach/mach_time.h>
+], [
+	mach_absolute_time();
+], [je_cv_mach_absolute_time])
+if test "x${je_cv_mach_absolute_time}" = "xyes" ; then
+  AC_DEFINE([JEMALLOC_HAVE_MACH_ABSOLUTE_TIME])
+fi
+
+dnl Use syscall(2) (if available) by default.
+AC_ARG_ENABLE([syscall],
+  [AS_HELP_STRING([--disable-syscall], [Disable use of syscall(2)])],
+[if test "x$enable_syscall" = "xno" ; then
+  enable_syscall="0"
+else
+  enable_syscall="1"
+fi
+],
+[enable_syscall="1"]
+)
+if test "x$enable_syscall" = "x1" ; then
+  dnl Check if syscall(2) is usable.  Treat warnings as errors, so that e.g. OS
+  dnl X 10.12's deprecation warning prevents use.
+  SAVED_CFLAGS="${CFLAGS}"
+  JE_CFLAGS_APPEND([-Werror])
+  JE_COMPILABLE([syscall(2)], [
+#include <sys/syscall.h>
+#include <unistd.h>
+], [
+	syscall(SYS_write, 2, "hello", 5);
+],
+                [je_cv_syscall])
+  CFLAGS="${SAVED_CFLAGS}"
+  if test "x$je_cv_syscall" = "xyes" ; then
+    AC_DEFINE([JEMALLOC_USE_SYSCALL], [ ])
+  fi
+fi
 
 dnl Check if the GNU-specific secure_getenv function exists.
 AC_CHECK_FUNC([secure_getenv],
@@ -1252,9 +1479,17 @@ fi
 ],
 [enable_lazy_lock=""]
 )
-if test "x$enable_lazy_lock" = "x" -a "x${force_lazy_lock}" = "x1" ; then
-  AC_MSG_RESULT([Forcing lazy-lock to avoid allocator/threading bootstrap issues])
-  enable_lazy_lock="1"
+if test "x${enable_lazy_lock}" = "x" ; then
+  if test "x${force_lazy_lock}" = "x1" ; then
+    AC_MSG_RESULT([Forcing lazy-lock to avoid allocator/threading bootstrap issues])
+    enable_lazy_lock="1"
+  else
+    enable_lazy_lock="0"
+  fi
+fi
+if test "x${enable_lazy_lock}" = "x1" -a "x${abi}" = "xpecoff" ; then
+  AC_MSG_RESULT([Forcing no lazy-lock because thread creation monitoring is unimplemented])
+  enable_lazy_lock="0"
 fi
 if test "x$enable_lazy_lock" = "x1" ; then
   if test "x$abi" != "xpecoff" ; then
@@ -1265,8 +1500,6 @@ if test "x$enable_lazy_lock" = "x1" ; then
       ])
   fi
   AC_DEFINE([JEMALLOC_LAZY_LOCK], [ ])
-else
-  enable_lazy_lock="0"
 fi
 AC_SUBST([enable_lazy_lock])
 
@@ -1389,12 +1622,41 @@ dnl Check for madvise(2).
 JE_COMPILABLE([madvise(2)], [
 #include <sys/mman.h>
 ], [
-	{
-		madvise((void *)0, 0, 0);
-	}
+	madvise((void *)0, 0, 0);
 ], [je_cv_madvise])
 if test "x${je_cv_madvise}" = "xyes" ; then
   AC_DEFINE([JEMALLOC_HAVE_MADVISE], [ ])
+
+  dnl Check for madvise(..., MADV_FREE).
+  JE_COMPILABLE([madvise(..., MADV_FREE)], [
+#include <sys/mman.h>
+], [
+	madvise((void *)0, 0, MADV_FREE);
+], [je_cv_madv_free])
+  if test "x${je_cv_madv_free}" = "xyes" ; then
+    AC_DEFINE([JEMALLOC_PURGE_MADVISE_FREE], [ ])
+  fi
+
+  dnl Check for madvise(..., MADV_DONTNEED).
+  JE_COMPILABLE([madvise(..., MADV_DONTNEED)], [
+#include <sys/mman.h>
+], [
+	madvise((void *)0, 0, MADV_DONTNEED);
+], [je_cv_madv_dontneed])
+  if test "x${je_cv_madv_dontneed}" = "xyes" ; then
+    AC_DEFINE([JEMALLOC_PURGE_MADVISE_DONTNEED], [ ])
+  fi
+
+  dnl Check for madvise(..., MADV_[NO]HUGEPAGE).
+  JE_COMPILABLE([madvise(..., MADV_[[NO]]HUGEPAGE)], [
+#include <sys/mman.h>
+], [
+	madvise((void *)0, 0, MADV_HUGEPAGE);
+	madvise((void *)0, 0, MADV_NOHUGEPAGE);
+], [je_cv_thp])
+  if test "x${je_cv_thp}" = "xyes" ; then
+    AC_DEFINE([JEMALLOC_THP], [ ])
+  fi
 fi
 
 dnl ============================================================================
@@ -1454,6 +1716,25 @@ if test "x${je_cv_builtin_clz}" = "xyes" ; then
   AC_DEFINE([JEMALLOC_HAVE_BUILTIN_CLZ], [ ])
 fi
 
+dnl ============================================================================
+dnl Check for os_unfair_lock operations as provided on Darwin.
+
+JE_COMPILABLE([Darwin os_unfair_lock_*()], [
+#include <os/lock.h>
+#include <AvailabilityMacros.h>
+], [
+	#if MAC_OS_X_VERSION_MIN_REQUIRED < 101200
+	#error "os_unfair_lock is not supported"
+	#else
+	os_unfair_lock lock = OS_UNFAIR_LOCK_INIT;
+	os_unfair_lock_lock(&lock);
+	os_unfair_lock_unlock(&lock);
+	#endif
+], [je_cv_os_unfair_lock])
+if test "x${je_cv_os_unfair_lock}" = "xyes" ; then
+  AC_DEFINE([JEMALLOC_OS_UNFAIR_LOCK], [ ])
+fi
+
 dnl ============================================================================
 dnl Check for spinlock(3) operations as provided on Darwin.
 
@@ -1698,11 +1979,11 @@ AC_MSG_RESULT([])
 AC_MSG_RESULT([CONFIG             : ${CONFIG}])
 AC_MSG_RESULT([CC                 : ${CC}])
 AC_MSG_RESULT([CFLAGS             : ${CFLAGS}])
+AC_MSG_RESULT([EXTRA_CFLAGS       : ${EXTRA_CFLAGS}])
 AC_MSG_RESULT([CPPFLAGS           : ${CPPFLAGS}])
 AC_MSG_RESULT([LDFLAGS            : ${LDFLAGS}])
 AC_MSG_RESULT([EXTRA_LDFLAGS      : ${EXTRA_LDFLAGS}])
 AC_MSG_RESULT([LIBS               : ${LIBS}])
-AC_MSG_RESULT([TESTLIBS           : ${TESTLIBS}])
 AC_MSG_RESULT([RPATH_EXTRA        : ${RPATH_EXTRA}])
 AC_MSG_RESULT([])
 AC_MSG_RESULT([XSLTPROC           : ${XSLTPROC}])
@@ -1724,6 +2005,7 @@ AC_MSG_RESULT([JEMALLOC_PREFIX    : ${JEMALLOC_PREFIX}])
 AC_MSG_RESULT([JEMALLOC_PRIVATE_NAMESPACE])
 AC_MSG_RESULT([                   : ${JEMALLOC_PRIVATE_NAMESPACE}])
 AC_MSG_RESULT([install_suffix     : ${install_suffix}])
+AC_MSG_RESULT([malloc_conf        : ${config_malloc_conf}])
 AC_MSG_RESULT([autogen            : ${enable_autogen}])
 AC_MSG_RESULT([cc-silence         : ${enable_cc_silence}])
 AC_MSG_RESULT([debug              : ${enable_debug}])
diff --git a/deps/jemalloc/doc/html.xsl.in b/deps/jemalloc/doc/html.xsl.in
index a91d9746f..ec4fa6552 100644
--- a/deps/jemalloc/doc/html.xsl.in
+++ b/deps/jemalloc/doc/html.xsl.in
@@ -1,4 +1,5 @@
 <xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version="1.0">
   <xsl:import href="@XSLROOT@/html/docbook.xsl"/>
   <xsl:import href="@abs_srcroot@doc/stylesheet.xsl"/>
+  <xsl:output method="xml" encoding="utf-8"/>
 </xsl:stylesheet>
diff --git a/deps/jemalloc/doc/jemalloc.3 b/deps/jemalloc/doc/jemalloc.3
index 2e6b2c0e8..3709f6692 100644
--- a/deps/jemalloc/doc/jemalloc.3
+++ b/deps/jemalloc/doc/jemalloc.3
@@ -1,13 +1,13 @@
 '\" t
 .\"     Title: JEMALLOC
 .\"    Author: Jason Evans
-.\" Generator: DocBook XSL Stylesheets v1.78.1 <http://docbook.sf.net/>
-.\"      Date: 09/24/2015
+.\" Generator: DocBook XSL Stylesheets v1.79.1 <http://docbook.sf.net/>
+.\"      Date: 12/03/2016
 .\"    Manual: User Manual
-.\"    Source: jemalloc 4.0.3-0-ge9192eacf8935e29fc62fddc2701f7942b1cc02c
+.\"    Source: jemalloc 4.4.0-0-gf1f76357313e7dcad7262f17a48ff0a2e005fcdc
 .\"  Language: English
 .\"
-.TH "JEMALLOC" "3" "09/24/2015" "jemalloc 4.0.3-0-ge9192eacf893" "User Manual"
+.TH "JEMALLOC" "3" "12/03/2016" "jemalloc 4.4.0-0-gf1f76357313e" "User Manual"
 .\" -----------------------------------------------------------------
 .\" * Define some portability stuff
 .\" -----------------------------------------------------------------
@@ -31,7 +31,7 @@
 jemalloc \- general purpose memory allocation functions
 .SH "LIBRARY"
 .PP
-This manual describes jemalloc 4\&.0\&.3\-0\-ge9192eacf8935e29fc62fddc2701f7942b1cc02c\&. More information can be found at the
+This manual describes jemalloc 4\&.4\&.0\-0\-gf1f76357313e7dcad7262f17a48ff0a2e005fcdc\&. More information can be found at the
 \m[blue]\fBjemalloc website\fR\m[]\&\s-2\u[1]\d\s+2\&.
 .SH "SYNOPSIS"
 .sp
@@ -86,26 +86,26 @@ const char *\fImalloc_conf\fR;
 .SS "Standard API"
 .PP
 The
-\fBmalloc\fR\fB\fR
+malloc()
 function allocates
 \fIsize\fR
 bytes of uninitialized memory\&. The allocated space is suitably aligned (after possible pointer coercion) for storage of any type of object\&.
 .PP
 The
-\fBcalloc\fR\fB\fR
+calloc()
 function allocates space for
 \fInumber\fR
 objects, each
 \fIsize\fR
 bytes in length\&. The result is identical to calling
-\fBmalloc\fR\fB\fR
+malloc()
 with an argument of
 \fInumber\fR
 *
 \fIsize\fR, with the exception that the allocated memory is explicitly initialized to zero bytes\&.
 .PP
 The
-\fBposix_memalign\fR\fB\fR
+posix_memalign()
 function allocates
 \fIsize\fR
 bytes of memory such that the allocation\*(Aqs base address is a multiple of
@@ -116,7 +116,7 @@ must be a power of 2 at least as large as
 sizeof(\fBvoid *\fR)\&.
 .PP
 The
-\fBaligned_alloc\fR\fB\fR
+aligned_alloc()
 function allocates
 \fIsize\fR
 bytes of memory such that the allocation\*(Aqs base address is a multiple of
@@ -128,7 +128,7 @@ is not an integral multiple of
 \fIalignment\fR\&.
 .PP
 The
-\fBrealloc\fR\fB\fR
+realloc()
 function changes the size of the previously allocated memory referenced by
 \fIptr\fR
 to
@@ -136,19 +136,19 @@ to
 bytes\&. The contents of the memory are unchanged up to the lesser of the new and old sizes\&. If the new size is larger, the contents of the newly allocated portion of the memory are undefined\&. Upon success, the memory referenced by
 \fIptr\fR
 is freed and a pointer to the newly allocated memory is returned\&. Note that
-\fBrealloc\fR\fB\fR
+realloc()
 may move the memory allocation, resulting in a different return value than
 \fIptr\fR\&. If
 \fIptr\fR
 is
 \fBNULL\fR, the
-\fBrealloc\fR\fB\fR
+realloc()
 function behaves identically to
-\fBmalloc\fR\fB\fR
+malloc()
 for the specified size\&.
 .PP
 The
-\fBfree\fR\fB\fR
+free()
 function causes the allocated memory referenced by
 \fIptr\fR
 to be made available for future allocations\&. If
@@ -158,13 +158,13 @@ is
 .SS "Non\-standard API"
 .PP
 The
-\fBmallocx\fR\fB\fR,
-\fBrallocx\fR\fB\fR,
-\fBxallocx\fR\fB\fR,
-\fBsallocx\fR\fB\fR,
-\fBdallocx\fR\fB\fR,
-\fBsdallocx\fR\fB\fR, and
-\fBnallocx\fR\fB\fR
+mallocx(),
+rallocx(),
+xallocx(),
+sallocx(),
+dallocx(),
+sdallocx(), and
+nallocx()
 functions all have a
 \fIflags\fR
 argument that can be used to specify options\&. The functions only check the options that are contextually relevant\&. Use bitwise or (|) operations to specify one or more of the following:
@@ -196,7 +196,7 @@ Initialize newly allocated memory to contain zero bytes\&. In the growing reallo
 .RS 4
 Use the thread\-specific cache (tcache) specified by the identifier
 \fItc\fR, which must have been acquired via the
-"tcache\&.create"
+tcache\&.create
 mallctl\&. This macro does not validate that
 \fItc\fR
 specifies a valid identifier\&.
@@ -223,16 +223,16 @@ specifies an arena index in the valid range\&.
 .RE
 .PP
 The
-\fBmallocx\fR\fB\fR
+mallocx()
 function allocates at least
 \fIsize\fR
 bytes of memory, and returns a pointer to the base address of the allocation\&. Behavior is undefined if
 \fIsize\fR
 is
-\fB0\fR, or if request size overflows due to size class and/or alignment constraints\&.
+\fB0\fR\&.
 .PP
 The
-\fBrallocx\fR\fB\fR
+rallocx()
 function resizes the allocation at
 \fIptr\fR
 to be at least
@@ -240,10 +240,10 @@ to be at least
 bytes, and returns a pointer to the base address of the resulting allocation, which may or may not have moved from its original location\&. Behavior is undefined if
 \fIsize\fR
 is
-\fB0\fR, or if request size overflows due to size class and/or alignment constraints\&.
+\fB0\fR\&.
 .PP
 The
-\fBxallocx\fR\fB\fR
+xallocx()
 function resizes the allocation at
 \fIptr\fR
 in place to be at least
@@ -259,40 +259,42 @@ is
 (\fIsize\fR + \fIextra\fR > \fBSIZE_T_MAX\fR)\&.
 .PP
 The
-\fBsallocx\fR\fB\fR
+sallocx()
 function returns the real size of the allocation at
 \fIptr\fR\&.
 .PP
 The
-\fBdallocx\fR\fB\fR
+dallocx()
 function causes the memory referenced by
 \fIptr\fR
 to be made available for future allocations\&.
 .PP
 The
-\fBsdallocx\fR\fB\fR
+sdallocx()
 function is an extension of
-\fBdallocx\fR\fB\fR
+dallocx()
 with a
 \fIsize\fR
 parameter to allow the caller to pass in the allocation size as an optimization\&. The minimum valid input size is the original requested size of the allocation, and the maximum valid input size is the corresponding value returned by
-\fBnallocx\fR\fB\fR
+nallocx()
 or
-\fBsallocx\fR\fB\fR\&.
+sallocx()\&.
 .PP
 The
-\fBnallocx\fR\fB\fR
+nallocx()
 function allocates no memory, but it performs the same size computation as the
-\fBmallocx\fR\fB\fR
+mallocx()
 function, and returns the real size of the allocation that would result from the equivalent
-\fBmallocx\fR\fB\fR
-function call\&. Behavior is undefined if
+mallocx()
+function call, or
+\fB0\fR
+if the inputs exceed the maximum supported size class and/or alignment\&. Behavior is undefined if
 \fIsize\fR
 is
-\fB0\fR, or if request size overflows due to size class and/or alignment constraints\&.
+\fB0\fR\&.
 .PP
 The
-\fBmallctl\fR\fB\fR
+mallctl()
 function provides a general interface for introspecting the memory allocator, as well as setting modifiable parameters and triggering actions\&. The period\-separated
 \fIname\fR
 argument specifies a location in a tree\-structured namespace; see the
@@ -311,10 +313,12 @@ and
 \fB0\fR\&.
 .PP
 The
-\fBmallctlnametomib\fR\fB\fR
-function provides a way to avoid repeated name lookups for applications that repeatedly query the same portion of the namespace, by translating a name to a \(lqManagement Information Base\(rq (MIB) that can be passed repeatedly to
-\fBmallctlbymib\fR\fB\fR\&. Upon successful return from
-\fBmallctlnametomib\fR\fB\fR,
+mallctlnametomib()
+function provides a way to avoid repeated name lookups for applications that repeatedly query the same portion of the namespace, by translating a name to a
+\(lqManagement Information Base\(rq
+(MIB) that can be passed repeatedly to
+mallctlbymib()\&. Upon successful return from
+mallctlnametomib(),
 \fImibp\fR
 contains an array of
 \fI*miblenp\fR
@@ -326,7 +330,7 @@ and the input value of
 \fI*miblenp\fR\&. Thus it is possible to pass a
 \fI*miblenp\fR
 that is smaller than the number of period\-separated name components, which results in a partial MIB that can be used as the basis for constructing a complete MIB\&. For name components that are integers (e\&.g\&. the 2 in
-"arenas\&.bin\&.2\&.size"), the corresponding MIB component will always be that integer\&. Therefore, it is legitimate to construct code like the following:
+arenas\&.bin\&.2\&.size), the corresponding MIB component will always be that integer\&. Therefore, it is legitimate to construct code like the following:
 .sp
 .if n \{\
 .RS 4
@@ -346,7 +350,7 @@ for (i = 0; i < nbins; i++) {
 
 	mib[2] = i;
 	len = sizeof(bin_size);
-	mallctlbymib(mib, miblen, &bin_size, &len, NULL, 0);
+	mallctlbymib(mib, miblen, (void *)&bin_size, &len, NULL, 0);
 	/* Do something with bin_size\&.\&.\&. */
 }
 .fi
@@ -355,67 +359,87 @@ for (i = 0; i < nbins; i++) {
 .\}
 .PP
 The
-\fBmalloc_stats_print\fR\fB\fR
-function writes human\-readable summary statistics via the
+malloc_stats_print()
+function writes summary statistics via the
 \fIwrite_cb\fR
 callback function pointer and
 \fIcbopaque\fR
 data passed to
 \fIwrite_cb\fR, or
-\fBmalloc_message\fR\fB\fR
+malloc_message()
 if
 \fIwrite_cb\fR
 is
-\fBNULL\fR\&. This function can be called repeatedly\&. General information that never changes during execution can be omitted by specifying "g" as a character within the
+\fBNULL\fR\&. The statistics are presented in human\-readable form unless
+\(lqJ\(rq
+is specified as a character within the
+\fIopts\fR
+string, in which case the statistics are presented in
+\m[blue]\fBJSON format\fR\m[]\&\s-2\u[2]\d\s+2\&. This function can be called repeatedly\&. General information that never changes during execution can be omitted by specifying
+\(lqg\(rq
+as a character within the
 \fIopts\fR
 string\&. Note that
-\fBmalloc_message\fR\fB\fR
+malloc_message()
 uses the
-\fBmallctl*\fR\fB\fR
+mallctl*()
 functions internally, so inconsistent statistics can be reported if multiple threads use these functions simultaneously\&. If
 \fB\-\-enable\-stats\fR
-is specified during configuration, \(lqm\(rq and \(lqa\(rq can be specified to omit merged arena and per arena statistics, respectively; \(lqb\(rq, \(lql\(rq, and \(lqh\(rq can be specified to omit per size class statistics for bins, large objects, and huge objects, respectively\&. Unrecognized characters are silently ignored\&. Note that thread caching may prevent some statistics from being completely up to date, since extra locking would be required to merge counters that track thread cache operations\&.
+is specified during configuration,
+\(lqm\(rq
+and
+\(lqa\(rq
+can be specified to omit merged arena and per arena statistics, respectively;
+\(lqb\(rq,
+\(lql\(rq, and
+\(lqh\(rq
+can be specified to omit per size class statistics for bins, large objects, and huge objects, respectively\&. Unrecognized characters are silently ignored\&. Note that thread caching may prevent some statistics from being completely up to date, since extra locking would be required to merge counters that track thread cache operations\&.
 .PP
 The
-\fBmalloc_usable_size\fR\fB\fR
+malloc_usable_size()
 function returns the usable size of the allocation pointed to by
 \fIptr\fR\&. The return value may be larger than the size that was requested during allocation\&. The
-\fBmalloc_usable_size\fR\fB\fR
+malloc_usable_size()
 function is not a mechanism for in\-place
-\fBrealloc\fR\fB\fR; rather it is provided solely as a tool for introspection purposes\&. Any discrepancy between the requested allocation size and the size reported by
-\fBmalloc_usable_size\fR\fB\fR
+realloc(); rather it is provided solely as a tool for introspection purposes\&. Any discrepancy between the requested allocation size and the size reported by
+malloc_usable_size()
 should not be depended on, since such behavior is entirely implementation\-dependent\&.
 .SH "TUNING"
 .PP
 Once, when the first call is made to one of the memory allocation routines, the allocator initializes its internals based in part on various options that can be specified at compile\- or run\-time\&.
 .PP
-The string pointed to by the global variable
-\fImalloc_conf\fR, the \(lqname\(rq of the file referenced by the symbolic link named
+The string specified via
+\fB\-\-with\-malloc\-conf\fR, the string pointed to by the global variable
+\fImalloc_conf\fR, the
+\(lqname\(rq
+of the file referenced by the symbolic link named
 /etc/malloc\&.conf, and the value of the environment variable
 \fBMALLOC_CONF\fR, will be interpreted, in that order, from left to right as options\&. Note that
 \fImalloc_conf\fR
 may be read before
-\fBmain\fR\fB\fR
+main()
 is entered, so the declaration of
 \fImalloc_conf\fR
 should specify an initializer that contains the final value to be read by jemalloc\&.
+\fB\-\-with\-malloc\-conf\fR
+and
 \fImalloc_conf\fR
-is a compile\-time setting, whereas
+are compile\-time mechanisms, whereas
 /etc/malloc\&.conf
 and
 \fBMALLOC_CONF\fR
 can be safely set any time prior to program invocation\&.
 .PP
 An options string is a comma\-separated list of option:value pairs\&. There is one key corresponding to each
-"opt\&.*"
+opt\&.*
 mallctl (see the
 MALLCTL NAMESPACE
 section for options documentation)\&. For example,
 abort:true,narenas:1
 sets the
-"opt\&.abort"
+opt\&.abort
 and
-"opt\&.narenas"
+opt\&.narenas
 options\&. Some options have boolean values (true/false), others have integer values (base 8, 10, or 16, depending on prefix), and yet others have raw string values\&.
 .SH "IMPLEMENTATION NOTES"
 .PP
@@ -436,29 +460,26 @@ In addition to multiple arenas, unless
 \fB\-\-disable\-tcache\fR
 is specified during configuration, this allocator supports thread\-specific caching for small and large objects, in order to make it possible to completely avoid synchronization for most allocation requests\&. Such caching allows very fast allocation in the common case, but it increases memory usage and fragmentation, since a bounded number of objects can remain allocated in each thread cache\&.
 .PP
-Memory is conceptually broken into equal\-sized chunks, where the chunk size is a power of two that is greater than the page size\&. Chunks are always aligned to multiples of the chunk size\&. This alignment makes it possible to find metadata for user objects very quickly\&.
-.PP
-User objects are broken into three categories according to size: small, large, and huge\&. Small and large objects are managed entirely by arenas; huge objects are additionally aggregated in a single data structure that is shared by all threads\&. Huge objects are typically used by applications infrequently enough that this single data structure is not a scalability issue\&.
-.PP
-Each chunk that is managed by an arena tracks its contents as runs of contiguous pages (unused, backing a set of small objects, or backing one large object)\&. The combination of chunk alignment and chunk page maps makes it possible to determine all metadata regarding small and large allocations in constant time\&.
+Memory is conceptually broken into equal\-sized chunks, where the chunk size is a power of two that is greater than the page size\&. Chunks are always aligned to multiples of the chunk size\&. This alignment makes it possible to find metadata for user objects very quickly\&. User objects are broken into three categories according to size: small, large, and huge\&. Multiple small and large objects can reside within a single chunk, whereas huge objects each have one or more chunks backing them\&. Each chunk that contains small and/or large objects tracks its contents as runs of contiguous pages (unused, backing a set of small objects, or backing one large object)\&. The combination of chunk alignment and chunk page maps makes it possible to determine all metadata regarding small and large allocations in constant time\&.
 .PP
 Small objects are managed in groups by page runs\&. Each run maintains a bitmap to track which regions are in use\&. Allocation requests that are no more than half the quantum (8 or 16, depending on architecture) are rounded up to the nearest power of two that is at least
 sizeof(\fBdouble\fR)\&. All other object size classes are multiples of the quantum, spaced such that there are four size classes for each doubling in size, which limits internal fragmentation to approximately 20% for all but the smallest size classes\&. Small size classes are smaller than four times the page size, large size classes are smaller than the chunk size (see the
-"opt\&.lg_chunk"
-option), and huge size classes extend from the chunk size up to one size class less than the full address space size\&.
+opt\&.lg_chunk
+option), and huge size classes extend from the chunk size up to the largest size class that does not exceed
+\fBPTRDIFF_MAX\fR\&.
 .PP
 Allocations are packed tightly together, which can be an issue for multi\-threaded applications\&. If you need to assure that allocations do not suffer from cacheline sharing, round your allocation requests up to the nearest multiple of the cacheline size, or specify cacheline alignment when allocating\&.
 .PP
 The
-\fBrealloc\fR\fB\fR,
-\fBrallocx\fR\fB\fR, and
-\fBxallocx\fR\fB\fR
+realloc(),
+rallocx(), and
+xallocx()
 functions may resize allocations without moving them under limited circumstances\&. Unlike the
-\fB*allocx\fR\fB\fR
+*allocx()
 API, the standard API does not officially round up the usable size of an allocation to the nearest size class, so technically it is necessary to call
-\fBrealloc\fR\fB\fR
+realloc()
 to grow e\&.g\&. a 9\-byte allocation to 16 bytes, or shrink a 16\-byte allocation to 9 bytes\&. Growth and shrinkage trivially succeeds in place as long as the pre\-size and post\-size both round up to the same size class\&. No other API guarantees are made regarding in\-place resizing, but the current implementation also tries to resize large and huge allocations in place, as long as the pre\-size and post\-size are both large or both huge\&. In such cases shrinkage always succeeds for large size classes, but for huge size classes the chunk allocator must support splitting (see
-"arena\&.<i>\&.chunk_hooks")\&. Growth only succeeds if the trailing memory is currently available, and additionally for huge size classes the chunk allocator must support merging\&.
+arena\&.<i>\&.chunk_hooks)\&. Growth only succeeds if the trailing memory is currently available, and additionally for huge size classes the chunk allocator must support merging\&.
 .PP
 Assuming 2 MiB chunks, 4 KiB pages, and a 16\-byte quantum on a 64\-bit system, the size classes in each category are as shown in
 Table 1\&.
@@ -502,6 +523,8 @@ l r l
 ^ r l
 ^ r l
 ^ r l
+^ r l
+^ r l
 ^ r l.
 T{
 Small
@@ -629,12 +652,22 @@ T}
 T}:T{
 \&.\&.\&.
 T}
+:T{
+512 PiB
+T}:T{
+[2560 PiB, 3 EiB, 3584 PiB, 4 EiB]
+T}
+:T{
+1 EiB
+T}:T{
+[5 EiB, 6 EiB, 7 EiB]
+T}
 .TE
 .sp 1
 .SH "MALLCTL NAMESPACE"
 .PP
 The following names are defined in the namespace accessible via the
-\fBmallctl*\fR\fB\fR
+mallctl*()
 functions\&. Value types are specified in parentheses, their readable/writable statuses are encoded as
 rw,
 r\-,
@@ -644,111 +677,118 @@ r\-,
 or
 <j>
 indicates an integer component, where the integer varies from 0 to some upper value that must be determined via introspection\&. In the case of
-"stats\&.arenas\&.<i>\&.*",
+stats\&.arenas\&.<i>\&.*,
 <i>
 equal to
-"arenas\&.narenas"
+arenas\&.narenas
 can be used to access the summation of statistics from all arenas\&. Take special note of the
-"epoch"
+epoch
 mallctl, which controls refreshing of cached dynamic statistics\&.
 .PP
-"version" (\fBconst char *\fR) r\-
+version (\fBconst char *\fR) r\-
 .RS 4
 Return the jemalloc version string\&.
 .RE
 .PP
-"epoch" (\fBuint64_t\fR) rw
+epoch (\fBuint64_t\fR) rw
 .RS 4
 If a value is passed in, refresh the data from which the
-\fBmallctl*\fR\fB\fR
+mallctl*()
 functions report values, and increment the epoch\&. Return the current epoch\&. This is useful for detecting whether another thread caused a refresh\&.
 .RE
 .PP
-"config\&.cache_oblivious" (\fBbool\fR) r\-
+config\&.cache_oblivious (\fBbool\fR) r\-
 .RS 4
 \fB\-\-enable\-cache\-oblivious\fR
 was specified during build configuration\&.
 .RE
 .PP
-"config\&.debug" (\fBbool\fR) r\-
+config\&.debug (\fBbool\fR) r\-
 .RS 4
 \fB\-\-enable\-debug\fR
 was specified during build configuration\&.
 .RE
 .PP
-"config\&.fill" (\fBbool\fR) r\-
+config\&.fill (\fBbool\fR) r\-
 .RS 4
 \fB\-\-enable\-fill\fR
 was specified during build configuration\&.
 .RE
 .PP
-"config\&.lazy_lock" (\fBbool\fR) r\-
+config\&.lazy_lock (\fBbool\fR) r\-
 .RS 4
 \fB\-\-enable\-lazy\-lock\fR
 was specified during build configuration\&.
 .RE
 .PP
-"config\&.munmap" (\fBbool\fR) r\-
+config\&.malloc_conf (\fBconst char *\fR) r\-
+.RS 4
+Embedded configure\-time\-specified run\-time options string, empty unless
+\fB\-\-with\-malloc\-conf\fR
+was specified during build configuration\&.
+.RE
+.PP
+config\&.munmap (\fBbool\fR) r\-
 .RS 4
 \fB\-\-enable\-munmap\fR
 was specified during build configuration\&.
 .RE
 .PP
-"config\&.prof" (\fBbool\fR) r\-
+config\&.prof (\fBbool\fR) r\-
 .RS 4
 \fB\-\-enable\-prof\fR
 was specified during build configuration\&.
 .RE
 .PP
-"config\&.prof_libgcc" (\fBbool\fR) r\-
+config\&.prof_libgcc (\fBbool\fR) r\-
 .RS 4
 \fB\-\-disable\-prof\-libgcc\fR
 was not specified during build configuration\&.
 .RE
 .PP
-"config\&.prof_libunwind" (\fBbool\fR) r\-
+config\&.prof_libunwind (\fBbool\fR) r\-
 .RS 4
 \fB\-\-enable\-prof\-libunwind\fR
 was specified during build configuration\&.
 .RE
 .PP
-"config\&.stats" (\fBbool\fR) r\-
+config\&.stats (\fBbool\fR) r\-
 .RS 4
 \fB\-\-enable\-stats\fR
 was specified during build configuration\&.
 .RE
 .PP
-"config\&.tcache" (\fBbool\fR) r\-
+config\&.tcache (\fBbool\fR) r\-
 .RS 4
 \fB\-\-disable\-tcache\fR
 was not specified during build configuration\&.
 .RE
 .PP
-"config\&.tls" (\fBbool\fR) r\-
+config\&.tls (\fBbool\fR) r\-
 .RS 4
 \fB\-\-disable\-tls\fR
 was not specified during build configuration\&.
 .RE
 .PP
-"config\&.utrace" (\fBbool\fR) r\-
+config\&.utrace (\fBbool\fR) r\-
 .RS 4
 \fB\-\-enable\-utrace\fR
 was specified during build configuration\&.
 .RE
 .PP
-"config\&.valgrind" (\fBbool\fR) r\-
+config\&.valgrind (\fBbool\fR) r\-
 .RS 4
 \fB\-\-enable\-valgrind\fR
 was specified during build configuration\&.
 .RE
 .PP
-"config\&.xmalloc" (\fBbool\fR) r\-
+config\&.xmalloc (\fBbool\fR) r\-
 .RS 4
 \fB\-\-enable\-xmalloc\fR
 was specified during build configuration\&.
 .RE
 .PP
-"opt\&.abort" (\fBbool\fR) r\-
+opt\&.abort (\fBbool\fR) r\-
 .RS 4
 Abort\-on\-warning enabled/disabled\&. If true, most warnings are fatal\&. The process will call
 \fBabort\fR(3)
@@ -757,97 +797,132 @@ in these cases\&. This option is disabled by default unless
 is specified during configuration, in which case it is enabled by default\&.
 .RE
 .PP
-"opt\&.dss" (\fBconst char *\fR) r\-
+opt\&.dss (\fBconst char *\fR) r\-
 .RS 4
 dss (\fBsbrk\fR(2)) allocation precedence as related to
 \fBmmap\fR(2)
 allocation\&. The following settings are supported if
 \fBsbrk\fR(2)
-is supported by the operating system: \(lqdisabled\(rq, \(lqprimary\(rq, and \(lqsecondary\(rq; otherwise only \(lqdisabled\(rq is supported\&. The default is \(lqsecondary\(rq if
+is supported by the operating system:
+\(lqdisabled\(rq,
+\(lqprimary\(rq, and
+\(lqsecondary\(rq; otherwise only
+\(lqdisabled\(rq
+is supported\&. The default is
+\(lqsecondary\(rq
+if
 \fBsbrk\fR(2)
-is supported by the operating system; \(lqdisabled\(rq otherwise\&.
+is supported by the operating system;
+\(lqdisabled\(rq
+otherwise\&.
 .RE
 .PP
-"opt\&.lg_chunk" (\fBsize_t\fR) r\-
+opt\&.lg_chunk (\fBsize_t\fR) r\-
 .RS 4
 Virtual memory chunk size (log base 2)\&. If a chunk size outside the supported size range is specified, the size is silently clipped to the minimum/maximum supported size\&. The default chunk size is 2 MiB (2^21)\&.
 .RE
 .PP
-"opt\&.narenas" (\fBsize_t\fR) r\-
+opt\&.narenas (\fBunsigned\fR) r\-
 .RS 4
 Maximum number of arenas to use for automatic multiplexing of threads and arenas\&. The default is four times the number of CPUs, or one if there is a single CPU\&.
 .RE
 .PP
-"opt\&.lg_dirty_mult" (\fBssize_t\fR) r\-
+opt\&.purge (\fBconst char *\fR) r\-
+.RS 4
+Purge mode is \(lqratio\(rq (default) or \(lqdecay\(rq\&. See
+opt\&.lg_dirty_mult
+for details of the ratio mode\&. See
+opt\&.decay_time
+for details of the decay mode\&.
+.RE
+.PP
+opt\&.lg_dirty_mult (\fBssize_t\fR) r\-
 .RS 4
 Per\-arena minimum ratio (log base 2) of active to dirty pages\&. Some dirty unused pages may be allowed to accumulate, within the limit set by the ratio (or one chunk worth of dirty pages, whichever is greater), before informing the kernel about some of those pages via
 \fBmadvise\fR(2)
 or a similar system call\&. This provides the kernel with sufficient information to recycle dirty pages if physical memory becomes scarce and the pages remain unused\&. The default minimum ratio is 8:1 (2^3:1); an option value of \-1 will disable dirty page purging\&. See
-"arenas\&.lg_dirty_mult"
+arenas\&.lg_dirty_mult
 and
-"arena\&.<i>\&.lg_dirty_mult"
+arena\&.<i>\&.lg_dirty_mult
 for related dynamic control options\&.
 .RE
 .PP
-"opt\&.stats_print" (\fBbool\fR) r\-
+opt\&.decay_time (\fBssize_t\fR) r\-
+.RS 4
+Approximate time in seconds from the creation of a set of unused dirty pages until an equivalent set of unused dirty pages is purged and/or reused\&. The pages are incrementally purged according to a sigmoidal decay curve that starts and ends with zero purge rate\&. A decay time of 0 causes all unused dirty pages to be purged immediately upon creation\&. A decay time of \-1 disables purging\&. The default decay time is 10 seconds\&. See
+arenas\&.decay_time
+and
+arena\&.<i>\&.decay_time
+for related dynamic control options\&.
+.RE
+.PP
+opt\&.stats_print (\fBbool\fR) r\-
 .RS 4
 Enable/disable statistics printing at exit\&. If enabled, the
-\fBmalloc_stats_print\fR\fB\fR
+malloc_stats_print()
 function is called at program exit via an
 \fBatexit\fR(3)
 function\&. If
 \fB\-\-enable\-stats\fR
 is specified during configuration, this has the potential to cause deadlock for a multi\-threaded process that exits while one or more threads are executing in the memory allocation functions\&. Furthermore,
-\fBatexit\fR\fB\fR
+atexit()
 may allocate memory during application initialization and then deadlock internally when jemalloc in turn calls
-\fBatexit\fR\fB\fR, so this option is not univerally usable (though the application can register its own
-\fBatexit\fR\fB\fR
+atexit(), so this option is not universally usable (though the application can register its own
+atexit()
 function with equivalent functionality)\&. Therefore, this option should only be used with care; it is primarily intended as a performance tuning aid during application development\&. This option is disabled by default\&.
 .RE
 .PP
-"opt\&.junk" (\fBconst char *\fR) r\- [\fB\-\-enable\-fill\fR]
+opt\&.junk (\fBconst char *\fR) r\- [\fB\-\-enable\-fill\fR]
 .RS 4
-Junk filling\&. If set to "alloc", each byte of uninitialized allocated memory will be initialized to
-0xa5\&. If set to "free", all deallocated memory will be initialized to
-0x5a\&. If set to "true", both allocated and deallocated memory will be initialized, and if set to "false", junk filling be disabled entirely\&. This is intended for debugging and will impact performance negatively\&. This option is "false" by default unless
+Junk filling\&. If set to
+\(lqalloc\(rq, each byte of uninitialized allocated memory will be initialized to
+0xa5\&. If set to
+\(lqfree\(rq, all deallocated memory will be initialized to
+0x5a\&. If set to
+\(lqtrue\(rq, both allocated and deallocated memory will be initialized, and if set to
+\(lqfalse\(rq, junk filling be disabled entirely\&. This is intended for debugging and will impact performance negatively\&. This option is
+\(lqfalse\(rq
+by default unless
 \fB\-\-enable\-debug\fR
-is specified during configuration, in which case it is "true" by default unless running inside
-\m[blue]\fBValgrind\fR\m[]\&\s-2\u[2]\d\s+2\&.
+is specified during configuration, in which case it is
+\(lqtrue\(rq
+by default unless running inside
+\m[blue]\fBValgrind\fR\m[]\&\s-2\u[3]\d\s+2\&.
 .RE
 .PP
-"opt\&.quarantine" (\fBsize_t\fR) r\- [\fB\-\-enable\-fill\fR]
+opt\&.quarantine (\fBsize_t\fR) r\- [\fB\-\-enable\-fill\fR]
 .RS 4
 Per thread quarantine size in bytes\&. If non\-zero, each thread maintains a FIFO object quarantine that stores up to the specified number of bytes of memory\&. The quarantined memory is not freed until it is released from quarantine, though it is immediately junk\-filled if the
-"opt\&.junk"
+opt\&.junk
 option is enabled\&. This feature is of particular use in combination with
-\m[blue]\fBValgrind\fR\m[]\&\s-2\u[2]\d\s+2, which can detect attempts to access quarantined objects\&. This is intended for debugging and will impact performance negatively\&. The default quarantine size is 0 unless running inside Valgrind, in which case the default is 16 MiB\&.
+\m[blue]\fBValgrind\fR\m[]\&\s-2\u[3]\d\s+2, which can detect attempts to access quarantined objects\&. This is intended for debugging and will impact performance negatively\&. The default quarantine size is 0 unless running inside Valgrind, in which case the default is 16 MiB\&.
 .RE
 .PP
-"opt\&.redzone" (\fBbool\fR) r\- [\fB\-\-enable\-fill\fR]
+opt\&.redzone (\fBbool\fR) r\- [\fB\-\-enable\-fill\fR]
 .RS 4
 Redzones enabled/disabled\&. If enabled, small allocations have redzones before and after them\&. Furthermore, if the
-"opt\&.junk"
+opt\&.junk
 option is enabled, the redzones are checked for corruption during deallocation\&. However, the primary intended purpose of this feature is to be used in combination with
-\m[blue]\fBValgrind\fR\m[]\&\s-2\u[2]\d\s+2, which needs redzones in order to do effective buffer overflow/underflow detection\&. This option is intended for debugging and will impact performance negatively\&. This option is disabled by default unless running inside Valgrind\&.
+\m[blue]\fBValgrind\fR\m[]\&\s-2\u[3]\d\s+2, which needs redzones in order to do effective buffer overflow/underflow detection\&. This option is intended for debugging and will impact performance negatively\&. This option is disabled by default unless running inside Valgrind\&.
 .RE
 .PP
-"opt\&.zero" (\fBbool\fR) r\- [\fB\-\-enable\-fill\fR]
+opt\&.zero (\fBbool\fR) r\- [\fB\-\-enable\-fill\fR]
 .RS 4
 Zero filling enabled/disabled\&. If enabled, each byte of uninitialized allocated memory will be initialized to 0\&. Note that this initialization only happens once for each byte, so
-\fBrealloc\fR\fB\fR
+realloc()
 and
-\fBrallocx\fR\fB\fR
+rallocx()
 calls do not zero memory that was previously allocated\&. This is intended for debugging and will impact performance negatively\&. This option is disabled by default\&.
 .RE
 .PP
-"opt\&.utrace" (\fBbool\fR) r\- [\fB\-\-enable\-utrace\fR]
+opt\&.utrace (\fBbool\fR) r\- [\fB\-\-enable\-utrace\fR]
 .RS 4
 Allocation tracing based on
 \fButrace\fR(2)
 enabled/disabled\&. This option is disabled by default\&.
 .RE
 .PP
-"opt\&.xmalloc" (\fBbool\fR) r\- [\fB\-\-enable\-xmalloc\fR]
+opt\&.xmalloc (\fBbool\fR) r\- [\fB\-\-enable\-xmalloc\fR]
 .RS 4
 Abort\-on\-out\-of\-memory enabled/disabled\&. If enabled, rather than returning failure for any allocation function, display a diagnostic message on
 \fBSTDERR_FILENO\fR
@@ -867,92 +942,94 @@ malloc_conf = "xmalloc:true";
 This option is disabled by default\&.
 .RE
 .PP
-"opt\&.tcache" (\fBbool\fR) r\- [\fB\-\-enable\-tcache\fR]
+opt\&.tcache (\fBbool\fR) r\- [\fB\-\-enable\-tcache\fR]
 .RS 4
 Thread\-specific caching (tcache) enabled/disabled\&. When there are multiple threads, each thread uses a tcache for objects up to a certain size\&. Thread\-specific caching allows many allocations to be satisfied without performing any thread synchronization, at the cost of increased memory use\&. See the
-"opt\&.lg_tcache_max"
+opt\&.lg_tcache_max
 option for related tuning information\&. This option is enabled by default unless running inside
-\m[blue]\fBValgrind\fR\m[]\&\s-2\u[2]\d\s+2, in which case it is forcefully disabled\&.
+\m[blue]\fBValgrind\fR\m[]\&\s-2\u[3]\d\s+2, in which case it is forcefully disabled\&.
 .RE
 .PP
-"opt\&.lg_tcache_max" (\fBsize_t\fR) r\- [\fB\-\-enable\-tcache\fR]
+opt\&.lg_tcache_max (\fBsize_t\fR) r\- [\fB\-\-enable\-tcache\fR]
 .RS 4
 Maximum size class (log base 2) to cache in the thread\-specific cache (tcache)\&. At a minimum, all small size classes are cached, and at a maximum all large size classes are cached\&. The default maximum is 32 KiB (2^15)\&.
 .RE
 .PP
-"opt\&.prof" (\fBbool\fR) r\- [\fB\-\-enable\-prof\fR]
+opt\&.prof (\fBbool\fR) r\- [\fB\-\-enable\-prof\fR]
 .RS 4
 Memory profiling enabled/disabled\&. If enabled, profile memory allocation activity\&. See the
-"opt\&.prof_active"
+opt\&.prof_active
 option for on\-the\-fly activation/deactivation\&. See the
-"opt\&.lg_prof_sample"
+opt\&.lg_prof_sample
 option for probabilistic sampling control\&. See the
-"opt\&.prof_accum"
+opt\&.prof_accum
 option for control of cumulative sample reporting\&. See the
-"opt\&.lg_prof_interval"
+opt\&.lg_prof_interval
 option for information on interval\-triggered profile dumping, the
-"opt\&.prof_gdump"
+opt\&.prof_gdump
 option for information on high\-water\-triggered profile dumping, and the
-"opt\&.prof_final"
+opt\&.prof_final
 option for final profile dumping\&. Profile output is compatible with the
 \fBjeprof\fR
 command, which is based on the
 \fBpprof\fR
 that is developed as part of the
-\m[blue]\fBgperftools package\fR\m[]\&\s-2\u[3]\d\s+2\&.
+\m[blue]\fBgperftools package\fR\m[]\&\s-2\u[4]\d\s+2\&. See
+HEAP PROFILE FORMAT
+for heap profile format documentation\&.
 .RE
 .PP
-"opt\&.prof_prefix" (\fBconst char *\fR) r\- [\fB\-\-enable\-prof\fR]
+opt\&.prof_prefix (\fBconst char *\fR) r\- [\fB\-\-enable\-prof\fR]
 .RS 4
 Filename prefix for profile dumps\&. If the prefix is set to the empty string, no automatic dumps will occur; this is primarily useful for disabling the automatic final heap dump (which also disables leak reporting, if enabled)\&. The default prefix is
 jeprof\&.
 .RE
 .PP
-"opt\&.prof_active" (\fBbool\fR) r\- [\fB\-\-enable\-prof\fR]
+opt\&.prof_active (\fBbool\fR) r\- [\fB\-\-enable\-prof\fR]
 .RS 4
 Profiling activated/deactivated\&. This is a secondary control mechanism that makes it possible to start the application with profiling enabled (see the
-"opt\&.prof"
+opt\&.prof
 option) but inactive, then toggle profiling at any time during program execution with the
-"prof\&.active"
+prof\&.active
 mallctl\&. This option is enabled by default\&.
 .RE
 .PP
-"opt\&.prof_thread_active_init" (\fBbool\fR) r\- [\fB\-\-enable\-prof\fR]
+opt\&.prof_thread_active_init (\fBbool\fR) r\- [\fB\-\-enable\-prof\fR]
 .RS 4
 Initial setting for
-"thread\&.prof\&.active"
+thread\&.prof\&.active
 in newly created threads\&. The initial setting for newly created threads can also be changed during execution via the
-"prof\&.thread_active_init"
+prof\&.thread_active_init
 mallctl\&. This option is enabled by default\&.
 .RE
 .PP
-"opt\&.lg_prof_sample" (\fBsize_t\fR) r\- [\fB\-\-enable\-prof\fR]
+opt\&.lg_prof_sample (\fBsize_t\fR) r\- [\fB\-\-enable\-prof\fR]
 .RS 4
 Average interval (log base 2) between allocation samples, as measured in bytes of allocation activity\&. Increasing the sampling interval decreases profile fidelity, but also decreases the computational overhead\&. The default sample interval is 512 KiB (2^19 B)\&.
 .RE
 .PP
-"opt\&.prof_accum" (\fBbool\fR) r\- [\fB\-\-enable\-prof\fR]
+opt\&.prof_accum (\fBbool\fR) r\- [\fB\-\-enable\-prof\fR]
 .RS 4
 Reporting of cumulative object/byte counts in profile dumps enabled/disabled\&. If this option is enabled, every unique backtrace must be stored for the duration of execution\&. Depending on the application, this can impose a large memory overhead, and the cumulative counts are not always of interest\&. This option is disabled by default\&.
 .RE
 .PP
-"opt\&.lg_prof_interval" (\fBssize_t\fR) r\- [\fB\-\-enable\-prof\fR]
+opt\&.lg_prof_interval (\fBssize_t\fR) r\- [\fB\-\-enable\-prof\fR]
 .RS 4
 Average interval (log base 2) between memory profile dumps, as measured in bytes of allocation activity\&. The actual interval between dumps may be sporadic because decentralized allocation counters are used to avoid synchronization bottlenecks\&. Profiles are dumped to files named according to the pattern
 <prefix>\&.<pid>\&.<seq>\&.i<iseq>\&.heap, where
 <prefix>
 is controlled by the
-"opt\&.prof_prefix"
+opt\&.prof_prefix
 option\&. By default, interval\-triggered profile dumping is disabled (encoded as \-1)\&.
 .RE
 .PP
-"opt\&.prof_gdump" (\fBbool\fR) r\- [\fB\-\-enable\-prof\fR]
+opt\&.prof_gdump (\fBbool\fR) r\- [\fB\-\-enable\-prof\fR]
 .RS 4
 Set the initial state of
-"prof\&.gdump", which when enabled triggers a memory profile dump every time the total virtual memory exceeds the previous maximum\&. This option is disabled by default\&.
+prof\&.gdump, which when enabled triggers a memory profile dump every time the total virtual memory exceeds the previous maximum\&. This option is disabled by default\&.
 .RE
 .PP
-"opt\&.prof_final" (\fBbool\fR) r\- [\fB\-\-enable\-prof\fR]
+opt\&.prof_final (\fBbool\fR) r\- [\fB\-\-enable\-prof\fR]
 .RS 4
 Use an
 \fBatexit\fR(3)
@@ -960,127 +1037,150 @@ function to dump final memory usage to a file named according to the pattern
 <prefix>\&.<pid>\&.<seq>\&.f\&.heap, where
 <prefix>
 is controlled by the
-"opt\&.prof_prefix"
+opt\&.prof_prefix
 option\&. Note that
-\fBatexit\fR\fB\fR
+atexit()
 may allocate memory during application initialization and then deadlock internally when jemalloc in turn calls
-\fBatexit\fR\fB\fR, so this option is not univerally usable (though the application can register its own
-\fBatexit\fR\fB\fR
+atexit(), so this option is not universally usable (though the application can register its own
+atexit()
 function with equivalent functionality)\&. This option is disabled by default\&.
 .RE
 .PP
-"opt\&.prof_leak" (\fBbool\fR) r\- [\fB\-\-enable\-prof\fR]
+opt\&.prof_leak (\fBbool\fR) r\- [\fB\-\-enable\-prof\fR]
 .RS 4
 Leak reporting enabled/disabled\&. If enabled, use an
 \fBatexit\fR(3)
 function to report memory leaks detected by allocation sampling\&. See the
-"opt\&.prof"
+opt\&.prof
 option for information on analyzing heap profile output\&. This option is disabled by default\&.
 .RE
 .PP
-"thread\&.arena" (\fBunsigned\fR) rw
+thread\&.arena (\fBunsigned\fR) rw
 .RS 4
 Get or set the arena associated with the calling thread\&. If the specified arena was not initialized beforehand (see the
-"arenas\&.initialized"
+arenas\&.initialized
 mallctl), it will be automatically initialized as a side effect of calling this interface\&.
 .RE
 .PP
-"thread\&.allocated" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+thread\&.allocated (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Get the total number of bytes ever allocated by the calling thread\&. This counter has the potential to wrap around; it is up to the application to appropriately interpret the counter in such cases\&.
 .RE
 .PP
-"thread\&.allocatedp" (\fBuint64_t *\fR) r\- [\fB\-\-enable\-stats\fR]
+thread\&.allocatedp (\fBuint64_t *\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Get a pointer to the the value that is returned by the
-"thread\&.allocated"
+thread\&.allocated
 mallctl\&. This is useful for avoiding the overhead of repeated
-\fBmallctl*\fR\fB\fR
+mallctl*()
 calls\&.
 .RE
 .PP
-"thread\&.deallocated" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+thread\&.deallocated (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Get the total number of bytes ever deallocated by the calling thread\&. This counter has the potential to wrap around; it is up to the application to appropriately interpret the counter in such cases\&.
 .RE
 .PP
-"thread\&.deallocatedp" (\fBuint64_t *\fR) r\- [\fB\-\-enable\-stats\fR]
+thread\&.deallocatedp (\fBuint64_t *\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Get a pointer to the the value that is returned by the
-"thread\&.deallocated"
+thread\&.deallocated
 mallctl\&. This is useful for avoiding the overhead of repeated
-\fBmallctl*\fR\fB\fR
+mallctl*()
 calls\&.
 .RE
 .PP
-"thread\&.tcache\&.enabled" (\fBbool\fR) rw [\fB\-\-enable\-tcache\fR]
+thread\&.tcache\&.enabled (\fBbool\fR) rw [\fB\-\-enable\-tcache\fR]
 .RS 4
 Enable/disable calling thread\*(Aqs tcache\&. The tcache is implicitly flushed as a side effect of becoming disabled (see
-"thread\&.tcache\&.flush")\&.
+thread\&.tcache\&.flush)\&.
 .RE
 .PP
-"thread\&.tcache\&.flush" (\fBvoid\fR) \-\- [\fB\-\-enable\-tcache\fR]
+thread\&.tcache\&.flush (\fBvoid\fR) \-\- [\fB\-\-enable\-tcache\fR]
 .RS 4
 Flush calling thread\*(Aqs thread\-specific cache (tcache)\&. This interface releases all cached objects and internal data structures associated with the calling thread\*(Aqs tcache\&. Ordinarily, this interface need not be called, since automatic periodic incremental garbage collection occurs, and the thread cache is automatically discarded when a thread exits\&. However, garbage collection is triggered by allocation activity, so it is possible for a thread that stops allocating/deallocating to retain its cache indefinitely, in which case the developer may find manual flushing useful\&.
 .RE
 .PP
-"thread\&.prof\&.name" (\fBconst char *\fR) r\- or \-w [\fB\-\-enable\-prof\fR]
+thread\&.prof\&.name (\fBconst char *\fR) r\- or \-w [\fB\-\-enable\-prof\fR]
 .RS 4
-Get/set the descriptive name associated with the calling thread in memory profile dumps\&. An internal copy of the name string is created, so the input string need not be maintained after this interface completes execution\&. The output string of this interface should be copied for non\-ephemeral uses, because multiple implementation details can cause asynchronous string deallocation\&. Furthermore, each invocation of this interface can only read or write; simultaneous read/write is not supported due to string lifetime limitations\&. The name string must nil\-terminated and comprised only of characters in the sets recognized by
+Get/set the descriptive name associated with the calling thread in memory profile dumps\&. An internal copy of the name string is created, so the input string need not be maintained after this interface completes execution\&. The output string of this interface should be copied for non\-ephemeral uses, because multiple implementation details can cause asynchronous string deallocation\&. Furthermore, each invocation of this interface can only read or write; simultaneous read/write is not supported due to string lifetime limitations\&. The name string must be nil\-terminated and comprised only of characters in the sets recognized by
 \fBisgraph\fR(3)
 and
 \fBisblank\fR(3)\&.
 .RE
 .PP
-"thread\&.prof\&.active" (\fBbool\fR) rw [\fB\-\-enable\-prof\fR]
+thread\&.prof\&.active (\fBbool\fR) rw [\fB\-\-enable\-prof\fR]
 .RS 4
 Control whether sampling is currently active for the calling thread\&. This is an activation mechanism in addition to
-"prof\&.active"; both must be active for the calling thread to sample\&. This flag is enabled by default\&.
+prof\&.active; both must be active for the calling thread to sample\&. This flag is enabled by default\&.
 .RE
 .PP
-"tcache\&.create" (\fBunsigned\fR) r\- [\fB\-\-enable\-tcache\fR]
+tcache\&.create (\fBunsigned\fR) r\- [\fB\-\-enable\-tcache\fR]
 .RS 4
 Create an explicit thread\-specific cache (tcache) and return an identifier that can be passed to the
 \fBMALLOCX_TCACHE(\fR\fB\fItc\fR\fR\fB)\fR
 macro to explicitly use the specified cache rather than the automatically managed one that is used by default\&. Each explicit cache can be used by only one thread at a time; the application must assure that this constraint holds\&.
 .RE
 .PP
-"tcache\&.flush" (\fBunsigned\fR) \-w [\fB\-\-enable\-tcache\fR]
+tcache\&.flush (\fBunsigned\fR) \-w [\fB\-\-enable\-tcache\fR]
 .RS 4
 Flush the specified thread\-specific cache (tcache)\&. The same considerations apply to this interface as to
-"thread\&.tcache\&.flush", except that the tcache will never be automatically be discarded\&.
+thread\&.tcache\&.flush, except that the tcache will never be automatically discarded\&.
 .RE
 .PP
-"tcache\&.destroy" (\fBunsigned\fR) \-w [\fB\-\-enable\-tcache\fR]
+tcache\&.destroy (\fBunsigned\fR) \-w [\fB\-\-enable\-tcache\fR]
 .RS 4
 Flush the specified thread\-specific cache (tcache) and make the identifier available for use during a future tcache creation\&.
 .RE
 .PP
-"arena\&.<i>\&.purge" (\fBvoid\fR) \-\-
+arena\&.<i>\&.purge (\fBvoid\fR) \-\-
 .RS 4
-Purge unused dirty pages for arena <i>, or for all arenas if <i> equals
-"arenas\&.narenas"\&.
+Purge all unused dirty pages for arena <i>, or for all arenas if <i> equals
+arenas\&.narenas\&.
+.RE
+.PP
+arena\&.<i>\&.decay (\fBvoid\fR) \-\-
+.RS 4
+Trigger decay\-based purging of unused dirty pages for arena <i>, or for all arenas if <i> equals
+arenas\&.narenas\&. The proportion of unused dirty pages to be purged depends on the current time; see
+opt\&.decay_time
+for details\&.
 .RE
 .PP
-"arena\&.<i>\&.dss" (\fBconst char *\fR) rw
+arena\&.<i>\&.reset (\fBvoid\fR) \-\-
+.RS 4
+Discard all of the arena\*(Aqs extant allocations\&. This interface can only be used with arenas created via
+arenas\&.extend\&. None of the arena\*(Aqs discarded/cached allocations may accessed afterward\&. As part of this requirement, all thread caches which were used to allocate/deallocate in conjunction with the arena must be flushed beforehand\&. This interface cannot be used if running inside Valgrind, nor if the
+quarantine
+size is non\-zero\&.
+.RE
+.PP
+arena\&.<i>\&.dss (\fBconst char *\fR) rw
 .RS 4
 Set the precedence of dss allocation as related to mmap allocation for arena <i>, or for all arenas if <i> equals
-"arenas\&.narenas"\&. See
-"opt\&.dss"
+arenas\&.narenas\&. See
+opt\&.dss
 for supported settings\&.
 .RE
 .PP
-"arena\&.<i>\&.lg_dirty_mult" (\fBssize_t\fR) rw
+arena\&.<i>\&.lg_dirty_mult (\fBssize_t\fR) rw
 .RS 4
 Current per\-arena minimum ratio (log base 2) of active to dirty pages for arena <i>\&. Each time this interface is set and the ratio is increased, pages are synchronously purged as necessary to impose the new ratio\&. See
-"opt\&.lg_dirty_mult"
+opt\&.lg_dirty_mult
+for additional information\&.
+.RE
+.PP
+arena\&.<i>\&.decay_time (\fBssize_t\fR) rw
+.RS 4
+Current per\-arena approximate time in seconds from the creation of a set of unused dirty pages until an equivalent set of unused dirty pages is purged and/or reused\&. Each time this interface is set, all currently unused dirty pages are considered to have fully decayed, which causes immediate purging of all unused dirty pages unless the decay time is set to \-1 (i\&.e\&. purging disabled)\&. See
+opt\&.decay_time
 for additional information\&.
 .RE
 .PP
-"arena\&.<i>\&.chunk_hooks" (\fBchunk_hooks_t\fR) rw
+arena\&.<i>\&.chunk_hooks (\fBchunk_hooks_t\fR) rw
 .RS 4
 Get or set the chunk management hook functions for arena <i>\&. The functions must be capable of operating on all extant chunks associated with arena <i>, usually by passing unknown chunks to the replaced functions\&. In practice, it is feasible to control allocation for arenas created via
-"arenas\&.extend"
+arenas\&.extend
 such that all chunks originate from an application\-supplied chunk allocator (by setting custom chunk hook functions just after arena creation), but the automatically created arenas may have already created chunks prior to the application having an opportunity to take over chunk allocation\&.
 .sp
 .if n \{\
@@ -1149,7 +1249,7 @@ is not
 on success or
 \fBNULL\fR
 on error\&. Committed memory may be committed in absolute terms as on a system that does not overcommit, or in implicit terms as on a system that overcommits and satisfies physical memory needs on demand via soft page faults\&. Note that replacing the default chunk allocation function makes the arena\*(Aqs
-"arena\&.<i>\&.dss"
+arena\&.<i>\&.dss
 setting irrelevant\&.
 .HP \w'typedef\ bool\ (chunk_dalloc_t)('u
 .BI "typedef bool (chunk_dalloc_t)(void\ *" "chunk" ", size_t\ " "size" ", bool\ " "committed" ", unsigned\ " "arena_ind" ");"
@@ -1296,407 +1396,504 @@ into one contiguous chunk, operating on
 \fIarena_ind\fR, returning false upon success\&. If the function returns true, this indicates that the chunks remain distinct mappings and therefore should continue to be operated on independently\&.
 .RE
 .PP
-"arenas\&.narenas" (\fBunsigned\fR) r\-
+arenas\&.narenas (\fBunsigned\fR) r\-
 .RS 4
 Current limit on number of arenas\&.
 .RE
 .PP
-"arenas\&.initialized" (\fBbool *\fR) r\-
+arenas\&.initialized (\fBbool *\fR) r\-
 .RS 4
 An array of
-"arenas\&.narenas"
+arenas\&.narenas
 booleans\&. Each boolean indicates whether the corresponding arena is initialized\&.
 .RE
 .PP
-"arenas\&.lg_dirty_mult" (\fBssize_t\fR) rw
+arenas\&.lg_dirty_mult (\fBssize_t\fR) rw
 .RS 4
 Current default per\-arena minimum ratio (log base 2) of active to dirty pages, used to initialize
-"arena\&.<i>\&.lg_dirty_mult"
+arena\&.<i>\&.lg_dirty_mult
 during arena creation\&. See
-"opt\&.lg_dirty_mult"
+opt\&.lg_dirty_mult
 for additional information\&.
 .RE
 .PP
-"arenas\&.quantum" (\fBsize_t\fR) r\-
+arenas\&.decay_time (\fBssize_t\fR) rw
+.RS 4
+Current default per\-arena approximate time in seconds from the creation of a set of unused dirty pages until an equivalent set of unused dirty pages is purged and/or reused, used to initialize
+arena\&.<i>\&.decay_time
+during arena creation\&. See
+opt\&.decay_time
+for additional information\&.
+.RE
+.PP
+arenas\&.quantum (\fBsize_t\fR) r\-
 .RS 4
 Quantum size\&.
 .RE
 .PP
-"arenas\&.page" (\fBsize_t\fR) r\-
+arenas\&.page (\fBsize_t\fR) r\-
 .RS 4
 Page size\&.
 .RE
 .PP
-"arenas\&.tcache_max" (\fBsize_t\fR) r\- [\fB\-\-enable\-tcache\fR]
+arenas\&.tcache_max (\fBsize_t\fR) r\- [\fB\-\-enable\-tcache\fR]
 .RS 4
 Maximum thread\-cached size class\&.
 .RE
 .PP
-"arenas\&.nbins" (\fBunsigned\fR) r\-
+arenas\&.nbins (\fBunsigned\fR) r\-
 .RS 4
 Number of bin size classes\&.
 .RE
 .PP
-"arenas\&.nhbins" (\fBunsigned\fR) r\- [\fB\-\-enable\-tcache\fR]
+arenas\&.nhbins (\fBunsigned\fR) r\- [\fB\-\-enable\-tcache\fR]
 .RS 4
 Total number of thread cache bin size classes\&.
 .RE
 .PP
-"arenas\&.bin\&.<i>\&.size" (\fBsize_t\fR) r\-
+arenas\&.bin\&.<i>\&.size (\fBsize_t\fR) r\-
 .RS 4
 Maximum size supported by size class\&.
 .RE
 .PP
-"arenas\&.bin\&.<i>\&.nregs" (\fBuint32_t\fR) r\-
+arenas\&.bin\&.<i>\&.nregs (\fBuint32_t\fR) r\-
 .RS 4
 Number of regions per page run\&.
 .RE
 .PP
-"arenas\&.bin\&.<i>\&.run_size" (\fBsize_t\fR) r\-
+arenas\&.bin\&.<i>\&.run_size (\fBsize_t\fR) r\-
 .RS 4
 Number of bytes per page run\&.
 .RE
 .PP
-"arenas\&.nlruns" (\fBunsigned\fR) r\-
+arenas\&.nlruns (\fBunsigned\fR) r\-
 .RS 4
 Total number of large size classes\&.
 .RE
 .PP
-"arenas\&.lrun\&.<i>\&.size" (\fBsize_t\fR) r\-
+arenas\&.lrun\&.<i>\&.size (\fBsize_t\fR) r\-
 .RS 4
 Maximum size supported by this large size class\&.
 .RE
 .PP
-"arenas\&.nhchunks" (\fBunsigned\fR) r\-
+arenas\&.nhchunks (\fBunsigned\fR) r\-
 .RS 4
 Total number of huge size classes\&.
 .RE
 .PP
-"arenas\&.hchunk\&.<i>\&.size" (\fBsize_t\fR) r\-
+arenas\&.hchunk\&.<i>\&.size (\fBsize_t\fR) r\-
 .RS 4
 Maximum size supported by this huge size class\&.
 .RE
 .PP
-"arenas\&.extend" (\fBunsigned\fR) r\-
+arenas\&.extend (\fBunsigned\fR) r\-
 .RS 4
 Extend the array of arenas by appending a new arena, and returning the new arena index\&.
 .RE
 .PP
-"prof\&.thread_active_init" (\fBbool\fR) rw [\fB\-\-enable\-prof\fR]
+prof\&.thread_active_init (\fBbool\fR) rw [\fB\-\-enable\-prof\fR]
 .RS 4
 Control the initial setting for
-"thread\&.prof\&.active"
+thread\&.prof\&.active
 in newly created threads\&. See the
-"opt\&.prof_thread_active_init"
+opt\&.prof_thread_active_init
 option for additional information\&.
 .RE
 .PP
-"prof\&.active" (\fBbool\fR) rw [\fB\-\-enable\-prof\fR]
+prof\&.active (\fBbool\fR) rw [\fB\-\-enable\-prof\fR]
 .RS 4
 Control whether sampling is currently active\&. See the
-"opt\&.prof_active"
+opt\&.prof_active
 option for additional information, as well as the interrelated
-"thread\&.prof\&.active"
+thread\&.prof\&.active
 mallctl\&.
 .RE
 .PP
-"prof\&.dump" (\fBconst char *\fR) \-w [\fB\-\-enable\-prof\fR]
+prof\&.dump (\fBconst char *\fR) \-w [\fB\-\-enable\-prof\fR]
 .RS 4
 Dump a memory profile to the specified file, or if NULL is specified, to a file according to the pattern
 <prefix>\&.<pid>\&.<seq>\&.m<mseq>\&.heap, where
 <prefix>
 is controlled by the
-"opt\&.prof_prefix"
+opt\&.prof_prefix
 option\&.
 .RE
 .PP
-"prof\&.gdump" (\fBbool\fR) rw [\fB\-\-enable\-prof\fR]
+prof\&.gdump (\fBbool\fR) rw [\fB\-\-enable\-prof\fR]
 .RS 4
 When enabled, trigger a memory profile dump every time the total virtual memory exceeds the previous maximum\&. Profiles are dumped to files named according to the pattern
 <prefix>\&.<pid>\&.<seq>\&.u<useq>\&.heap, where
 <prefix>
 is controlled by the
-"opt\&.prof_prefix"
+opt\&.prof_prefix
 option\&.
 .RE
 .PP
-"prof\&.reset" (\fBsize_t\fR) \-w [\fB\-\-enable\-prof\fR]
+prof\&.reset (\fBsize_t\fR) \-w [\fB\-\-enable\-prof\fR]
 .RS 4
 Reset all memory profile statistics, and optionally update the sample rate (see
-"opt\&.lg_prof_sample"
+opt\&.lg_prof_sample
 and
-"prof\&.lg_sample")\&.
+prof\&.lg_sample)\&.
 .RE
 .PP
-"prof\&.lg_sample" (\fBsize_t\fR) r\- [\fB\-\-enable\-prof\fR]
+prof\&.lg_sample (\fBsize_t\fR) r\- [\fB\-\-enable\-prof\fR]
 .RS 4
 Get the current sample rate (see
-"opt\&.lg_prof_sample")\&.
+opt\&.lg_prof_sample)\&.
 .RE
 .PP
-"prof\&.interval" (\fBuint64_t\fR) r\- [\fB\-\-enable\-prof\fR]
+prof\&.interval (\fBuint64_t\fR) r\- [\fB\-\-enable\-prof\fR]
 .RS 4
-Average number of bytes allocated between inverval\-based profile dumps\&. See the
-"opt\&.lg_prof_interval"
+Average number of bytes allocated between interval\-based profile dumps\&. See the
+opt\&.lg_prof_interval
 option for additional information\&.
 .RE
 .PP
-"stats\&.cactive" (\fBsize_t *\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.cactive (\fBsize_t *\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Pointer to a counter that contains an approximate count of the current number of bytes in active pages\&. The estimate may be high, but never low, because each arena rounds up when computing its contribution to the counter\&. Note that the
-"epoch"
+epoch
 mallctl has no bearing on this counter\&. Furthermore, counter consistency is maintained via atomic operations, so it is necessary to use an atomic operation in order to guarantee a consistent read when dereferencing the pointer\&.
 .RE
 .PP
-"stats\&.allocated" (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.allocated (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Total number of bytes allocated by the application\&.
 .RE
 .PP
-"stats\&.active" (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.active (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Total number of bytes in active pages allocated by the application\&. This is a multiple of the page size, and greater than or equal to
-"stats\&.allocated"\&. This does not include
-"stats\&.arenas\&.<i>\&.pdirty", nor pages entirely devoted to allocator metadata\&.
+stats\&.allocated\&. This does not include
+stats\&.arenas\&.<i>\&.pdirty, nor pages entirely devoted to allocator metadata\&.
 .RE
 .PP
-"stats\&.metadata" (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.metadata (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Total number of bytes dedicated to metadata, which comprise base allocations used for bootstrap\-sensitive internal allocator data structures, arena chunk headers (see
-"stats\&.arenas\&.<i>\&.metadata\&.mapped"), and internal allocations (see
-"stats\&.arenas\&.<i>\&.metadata\&.allocated")\&.
+stats\&.arenas\&.<i>\&.metadata\&.mapped), and internal allocations (see
+stats\&.arenas\&.<i>\&.metadata\&.allocated)\&.
 .RE
 .PP
-"stats\&.resident" (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.resident (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Maximum number of bytes in physically resident data pages mapped by the allocator, comprising all pages dedicated to allocator metadata, pages backing active allocations, and unused dirty pages\&. This is a maximum rather than precise because pages may not actually be physically resident if they correspond to demand\-zeroed virtual memory that has not yet been touched\&. This is a multiple of the page size, and is larger than
-"stats\&.active"\&.
+stats\&.active\&.
 .RE
 .PP
-"stats\&.mapped" (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.mapped (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Total number of bytes in active chunks mapped by the allocator\&. This is a multiple of the chunk size, and is larger than
-"stats\&.active"\&. This does not include inactive chunks, even those that contain unused dirty pages, which means that there is no strict ordering between this and
-"stats\&.resident"\&.
+stats\&.active\&. This does not include inactive chunks, even those that contain unused dirty pages, which means that there is no strict ordering between this and
+stats\&.resident\&.
+.RE
+.PP
+stats\&.retained (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
+.RS 4
+Total number of bytes in virtual memory mappings that were retained rather than being returned to the operating system via e\&.g\&.
+\fBmunmap\fR(2)\&. Retained virtual memory is typically untouched, decommitted, or purged, so it has no strongly associated physical memory (see
+chunk hooks
+for details)\&. Retained memory is excluded from mapped memory statistics, e\&.g\&.
+stats\&.mapped\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.dss" (\fBconst char *\fR) r\-
+stats\&.arenas\&.<i>\&.dss (\fBconst char *\fR) r\-
 .RS 4
 dss (\fBsbrk\fR(2)) allocation precedence as related to
 \fBmmap\fR(2)
 allocation\&. See
-"opt\&.dss"
+opt\&.dss
 for details\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.lg_dirty_mult" (\fBssize_t\fR) r\-
+stats\&.arenas\&.<i>\&.lg_dirty_mult (\fBssize_t\fR) r\-
 .RS 4
 Minimum ratio (log base 2) of active to dirty pages\&. See
-"opt\&.lg_dirty_mult"
+opt\&.lg_dirty_mult
 for details\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.nthreads" (\fBunsigned\fR) r\-
+stats\&.arenas\&.<i>\&.decay_time (\fBssize_t\fR) r\-
+.RS 4
+Approximate time in seconds from the creation of a set of unused dirty pages until an equivalent set of unused dirty pages is purged and/or reused\&. See
+opt\&.decay_time
+for details\&.
+.RE
+.PP
+stats\&.arenas\&.<i>\&.nthreads (\fBunsigned\fR) r\-
 .RS 4
 Number of threads currently assigned to arena\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.pactive" (\fBsize_t\fR) r\-
+stats\&.arenas\&.<i>\&.pactive (\fBsize_t\fR) r\-
 .RS 4
 Number of pages in active runs\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.pdirty" (\fBsize_t\fR) r\-
+stats\&.arenas\&.<i>\&.pdirty (\fBsize_t\fR) r\-
 .RS 4
 Number of pages within unused runs that are potentially dirty, and for which
-\fBmadvise\fR\fB\fI\&.\&.\&.\fR\fR\fB \fR\fB\fI\fBMADV_DONTNEED\fR\fR\fR
+madvise\fI\&.\&.\&.\fR \fI\fBMADV_DONTNEED\fR\fR
 or similar has not been called\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.mapped" (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.mapped (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Number of mapped bytes\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.metadata\&.mapped" (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.retained (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
+.RS 4
+Number of retained bytes\&. See
+stats\&.retained
+for details\&.
+.RE
+.PP
+stats\&.arenas\&.<i>\&.metadata\&.mapped (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Number of mapped bytes in arena chunk headers, which track the states of the non\-metadata pages\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.metadata\&.allocated" (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.metadata\&.allocated (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Number of bytes dedicated to internal allocations\&. Internal allocations differ from application\-originated allocations in that they are for internal use, and that they are omitted from heap profiles\&. This statistic is reported separately from
-"stats\&.metadata"
+stats\&.metadata
 and
-"stats\&.arenas\&.<i>\&.metadata\&.mapped"
+stats\&.arenas\&.<i>\&.metadata\&.mapped
 because it overlaps with e\&.g\&. the
-"stats\&.allocated"
+stats\&.allocated
 and
-"stats\&.active"
+stats\&.active
 statistics, whereas the other metadata statistics do not\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.npurge" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.npurge (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Number of dirty page purge sweeps performed\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.nmadvise" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.nmadvise (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Number of
-\fBmadvise\fR\fB\fI\&.\&.\&.\fR\fR\fB \fR\fB\fI\fBMADV_DONTNEED\fR\fR\fR
+madvise\fI\&.\&.\&.\fR \fI\fBMADV_DONTNEED\fR\fR
 or similar calls made to purge dirty pages\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.purged" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.purged (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Number of pages purged\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.small\&.allocated" (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.small\&.allocated (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Number of bytes currently allocated by small objects\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.small\&.nmalloc" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.small\&.nmalloc (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of allocation requests served by small bins\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.small\&.ndalloc" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.small\&.ndalloc (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of small objects returned to bins\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.small\&.nrequests" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.small\&.nrequests (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of small allocation requests\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.large\&.allocated" (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.large\&.allocated (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Number of bytes currently allocated by large objects\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.large\&.nmalloc" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.large\&.nmalloc (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of large allocation requests served directly by the arena\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.large\&.ndalloc" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.large\&.ndalloc (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of large deallocation requests served directly by the arena\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.large\&.nrequests" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.large\&.nrequests (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of large allocation requests\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.huge\&.allocated" (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.huge\&.allocated (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Number of bytes currently allocated by huge objects\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.huge\&.nmalloc" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.huge\&.nmalloc (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of huge allocation requests served directly by the arena\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.huge\&.ndalloc" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.huge\&.ndalloc (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of huge deallocation requests served directly by the arena\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.huge\&.nrequests" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.huge\&.nrequests (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of huge allocation requests\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.bins\&.<j>\&.nmalloc" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.bins\&.<j>\&.nmalloc (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of allocations served by bin\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.bins\&.<j>\&.ndalloc" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.bins\&.<j>\&.ndalloc (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of allocations returned to bin\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.bins\&.<j>\&.nrequests" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.bins\&.<j>\&.nrequests (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of allocation requests\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.bins\&.<j>\&.curregs" (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.bins\&.<j>\&.curregs (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Current number of regions for this size class\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.bins\&.<j>\&.nfills" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR \fB\-\-enable\-tcache\fR]
+stats\&.arenas\&.<i>\&.bins\&.<j>\&.nfills (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR \fB\-\-enable\-tcache\fR]
 .RS 4
 Cumulative number of tcache fills\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.bins\&.<j>\&.nflushes" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR \fB\-\-enable\-tcache\fR]
+stats\&.arenas\&.<i>\&.bins\&.<j>\&.nflushes (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR \fB\-\-enable\-tcache\fR]
 .RS 4
 Cumulative number of tcache flushes\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.bins\&.<j>\&.nruns" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.bins\&.<j>\&.nruns (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of runs created\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.bins\&.<j>\&.nreruns" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.bins\&.<j>\&.nreruns (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of times the current run from which to allocate changed\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.bins\&.<j>\&.curruns" (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.bins\&.<j>\&.curruns (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Current number of runs\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.lruns\&.<j>\&.nmalloc" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.lruns\&.<j>\&.nmalloc (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of allocation requests for this size class served directly by the arena\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.lruns\&.<j>\&.ndalloc" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.lruns\&.<j>\&.ndalloc (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of deallocation requests for this size class served directly by the arena\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.lruns\&.<j>\&.nrequests" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.lruns\&.<j>\&.nrequests (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of allocation requests for this size class\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.lruns\&.<j>\&.curruns" (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.lruns\&.<j>\&.curruns (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Current number of runs for this size class\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.hchunks\&.<j>\&.nmalloc" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.hchunks\&.<j>\&.nmalloc (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of allocation requests for this size class served directly by the arena\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.hchunks\&.<j>\&.ndalloc" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.hchunks\&.<j>\&.ndalloc (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of deallocation requests for this size class served directly by the arena\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.hchunks\&.<j>\&.nrequests" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.hchunks\&.<j>\&.nrequests (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Cumulative number of allocation requests for this size class\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.hchunks\&.<j>\&.curhchunks" (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
+stats\&.arenas\&.<i>\&.hchunks\&.<j>\&.curhchunks (\fBsize_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Current number of huge allocations for this size class\&.
 .RE
+.SH "HEAP PROFILE FORMAT"
+.PP
+Although the heap profiling functionality was originally designed to be compatible with the
+\fBpprof\fR
+command that is developed as part of the
+\m[blue]\fBgperftools package\fR\m[]\&\s-2\u[4]\d\s+2, the addition of per thread heap profiling functionality required a different heap profile format\&. The
+\fBjeprof\fR
+command is derived from
+\fBpprof\fR, with enhancements to support the heap profile format described here\&.
+.PP
+In the following hypothetical heap profile,
+\fB[\&.\&.\&.]\fR
+indicates elision for the sake of compactness\&.
+.sp
+.if n \{\
+.RS 4
+.\}
+.nf
+heap_v2/524288
+  t*: 28106: 56637512 [0: 0]
+  [\&.\&.\&.]
+  t3: 352: 16777344 [0: 0]
+  [\&.\&.\&.]
+  t99: 17754: 29341640 [0: 0]
+  [\&.\&.\&.]
+@ 0x5f86da8 0x5f5a1dc [\&.\&.\&.] 0x29e4d4e 0xa200316 0xabb2988 [\&.\&.\&.]
+  t*: 13: 6688 [0: 0]
+  t3: 12: 6496 [0: ]
+  t99: 1: 192 [0: 0]
+[\&.\&.\&.]
+
+MAPPED_LIBRARIES:
+[\&.\&.\&.]
+.fi
+.if n \{\
+.RE
+.\}
+.sp
+The following matches the above heap profile, but most tokens are replaced with
+\fB<description>\fR
+to indicate descriptions of the corresponding fields\&.
+.sp
+.if n \{\
+.RS 4
+.\}
+.nf
+<heap_profile_format_version>/<mean_sample_interval>
+  <aggregate>: <curobjs>: <curbytes> [<cumobjs>: <cumbytes>]
+  [\&.\&.\&.]
+  <thread_3_aggregate>: <curobjs>: <curbytes>[<cumobjs>: <cumbytes>]
+  [\&.\&.\&.]
+  <thread_99_aggregate>: <curobjs>: <curbytes>[<cumobjs>: <cumbytes>]
+  [\&.\&.\&.]
+@ <top_frame> <frame> [\&.\&.\&.] <frame> <frame> <frame> [\&.\&.\&.]
+  <backtrace_aggregate>: <curobjs>: <curbytes> [<cumobjs>: <cumbytes>]
+  <backtrace_thread_3>: <curobjs>: <curbytes> [<cumobjs>: <cumbytes>]
+  <backtrace_thread_99>: <curobjs>: <curbytes> [<cumobjs>: <cumbytes>]
+[\&.\&.\&.]
+
+MAPPED_LIBRARIES:
+</proc/<pid>/maps>
+.fi
+.if n \{\
+.RE
+.\}
 .SH "DEBUGGING MALLOC PROBLEMS"
 .PP
 When debugging, it is a good idea to configure/build jemalloc with the
@@ -1705,14 +1902,16 @@ and
 \fB\-\-enable\-fill\fR
 options, and recompile the program with suitable options and symbols for debugger support\&. When so configured, jemalloc incorporates a wide variety of run\-time assertions that catch application errors such as double\-free, write\-after\-free, etc\&.
 .PP
-Programs often accidentally depend on \(lquninitialized\(rq memory actually being filled with zero bytes\&. Junk filling (see the
-"opt\&.junk"
+Programs often accidentally depend on
+\(lquninitialized\(rq
+memory actually being filled with zero bytes\&. Junk filling (see the
+opt\&.junk
 option) tends to expose such bugs in the form of obviously incorrect results and/or coredumps\&. Conversely, zero filling (see the
-"opt\&.zero"
+opt\&.zero
 option) eliminates the symptoms of such bugs\&. Between these two options, it is usually possible to quickly detect, diagnose, and eliminate such bugs\&.
 .PP
 This implementation does not provide much detail about the problems it detects, because the performance impact for storing such information would be prohibitive\&. However, jemalloc does integrate with the most excellent
-\m[blue]\fBValgrind\fR\m[]\&\s-2\u[2]\d\s+2
+\m[blue]\fBValgrind\fR\m[]\&\s-2\u[3]\d\s+2
 tool if the
 \fB\-\-enable\-valgrind\fR
 configuration option is enabled\&.
@@ -1720,7 +1919,7 @@ configuration option is enabled\&.
 .PP
 If any of the memory allocation/deallocation functions detect an error or warning condition, a message will be printed to file descriptor
 \fBSTDERR_FILENO\fR\&. Errors will result in the process dumping core\&. If the
-"opt\&.abort"
+opt\&.abort
 option is set, most warnings are treated as errors\&.
 .PP
 The
@@ -1728,22 +1927,23 @@ The
 variable allows the programmer to override the function which emits the text strings forming the errors and warnings if for some reason the
 \fBSTDERR_FILENO\fR
 file descriptor is not suitable for this\&.
-\fBmalloc_message\fR\fB\fR
+malloc_message()
 takes the
 \fIcbopaque\fR
 pointer argument that is
 \fBNULL\fR
 unless overridden by the arguments in a call to
-\fBmalloc_stats_print\fR\fB\fR, followed by a string pointer\&. Please note that doing anything which tries to allocate memory in this function is likely to result in a crash or deadlock\&.
+malloc_stats_print(), followed by a string pointer\&. Please note that doing anything which tries to allocate memory in this function is likely to result in a crash or deadlock\&.
 .PP
-All messages are prefixed by \(lq<jemalloc>:\(rq\&.
+All messages are prefixed by
+\(lq<jemalloc>: \(rq\&.
 .SH "RETURN VALUES"
 .SS "Standard API"
 .PP
 The
-\fBmalloc\fR\fB\fR
+malloc()
 and
-\fBcalloc\fR\fB\fR
+calloc()
 functions return a pointer to the allocated memory if successful; otherwise a
 \fBNULL\fR
 pointer is returned and
@@ -1752,9 +1952,9 @@ is set to
 ENOMEM\&.
 .PP
 The
-\fBposix_memalign\fR\fB\fR
+posix_memalign()
 function returns the value 0 if successful; otherwise it returns an error value\&. The
-\fBposix_memalign\fR\fB\fR
+posix_memalign()
 function will fail if:
 .PP
 EINVAL
@@ -1771,13 +1971,13 @@ Memory allocation error\&.
 .RE
 .PP
 The
-\fBaligned_alloc\fR\fB\fR
+aligned_alloc()
 function returns a pointer to the allocated memory if successful; otherwise a
 \fBNULL\fR
 pointer is returned and
 \fIerrno\fR
 is set\&. The
-\fBaligned_alloc\fR\fB\fR
+aligned_alloc()
 function will fail if:
 .PP
 EINVAL
@@ -1793,7 +1993,7 @@ Memory allocation error\&.
 .RE
 .PP
 The
-\fBrealloc\fR\fB\fR
+realloc()
 function returns a pointer, possibly identical to
 \fIptr\fR, to the allocated memory if successful; otherwise a
 \fBNULL\fR
@@ -1802,44 +2002,44 @@ pointer is returned, and
 is set to
 ENOMEM
 if the error was the result of an allocation failure\&. The
-\fBrealloc\fR\fB\fR
+realloc()
 function always leaves the original buffer intact when an error occurs\&.
 .PP
 The
-\fBfree\fR\fB\fR
+free()
 function returns no value\&.
 .SS "Non\-standard API"
 .PP
 The
-\fBmallocx\fR\fB\fR
+mallocx()
 and
-\fBrallocx\fR\fB\fR
+rallocx()
 functions return a pointer to the allocated memory if successful; otherwise a
 \fBNULL\fR
 pointer is returned to indicate insufficient contiguous memory was available to service the allocation request\&.
 .PP
 The
-\fBxallocx\fR\fB\fR
+xallocx()
 function returns the real size of the resulting resized allocation pointed to by
 \fIptr\fR, which is a value less than
 \fIsize\fR
 if the allocation could not be adequately grown in place\&.
 .PP
 The
-\fBsallocx\fR\fB\fR
+sallocx()
 function returns the real size of the allocation pointed to by
 \fIptr\fR\&.
 .PP
 The
-\fBnallocx\fR\fB\fR
+nallocx()
 returns the real size that would result from a successful equivalent
-\fBmallocx\fR\fB\fR
+mallocx()
 function call, or zero if insufficient memory is available to perform the size computation\&.
 .PP
 The
-\fBmallctl\fR\fB\fR,
-\fBmallctlnametomib\fR\fB\fR, and
-\fBmallctlbymib\fR\fB\fR
+mallctl(),
+mallctlnametomib(), and
+mallctlbymib()
 functions return 0 on success; otherwise they return an error value\&. The functions will fail if:
 .PP
 EINVAL
@@ -1874,12 +2074,12 @@ A memory allocation failure occurred\&.
 EFAULT
 .RS 4
 An interface with side effects failed in some way not directly related to
-\fBmallctl*\fR\fB\fR
+mallctl*()
 read/write processing\&.
 .RE
 .PP
 The
-\fBmalloc_usable_size\fR\fB\fR
+malloc_usable_size()
 function returns the usable size of the allocation pointed to by
 \fIptr\fR\&.
 .SH "ENVIRONMENT"
@@ -1929,14 +2129,14 @@ malloc_conf = "lg_chunk:24";
 .SH "STANDARDS"
 .PP
 The
-\fBmalloc\fR\fB\fR,
-\fBcalloc\fR\fB\fR,
-\fBrealloc\fR\fB\fR, and
-\fBfree\fR\fB\fR
+malloc(),
+calloc(),
+realloc(), and
+free()
 functions conform to ISO/IEC 9899:1990 (\(lqISO C90\(rq)\&.
 .PP
 The
-\fBposix_memalign\fR\fB\fR
+posix_memalign()
 function conforms to IEEE Std 1003\&.1\-2001 (\(lqPOSIX\&.1\(rq)\&.
 .SH "AUTHOR"
 .PP
@@ -1947,14 +2147,19 @@ function conforms to IEEE Std 1003\&.1\-2001 (\(lqPOSIX\&.1\(rq)\&.
 .IP " 1." 4
 jemalloc website
 .RS 4
-\%http://www.canonware.com/jemalloc/
+\%http://jemalloc.net/
 .RE
 .IP " 2." 4
+JSON format
+.RS 4
+\%http://www.json.org/
+.RE
+.IP " 3." 4
 Valgrind
 .RS 4
 \%http://valgrind.org/
 .RE
-.IP " 3." 4
+.IP " 4." 4
 gperftools package
 .RS 4
 \%http://code.google.com/p/gperftools/
diff --git a/deps/jemalloc/doc/jemalloc.html b/deps/jemalloc/doc/jemalloc.html
index 7b8e2be8c..db2504f6e 100644
--- a/deps/jemalloc/doc/jemalloc.html
+++ b/deps/jemalloc/doc/jemalloc.html
@@ -1,28 +1,29 @@
-<html><head><meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1"><title>JEMALLOC</title><meta name="generator" content="DocBook XSL Stylesheets V1.78.1"></head><body bgcolor="white" text="black" link="#0000FF" vlink="#840084" alink="#0000FF"><div class="refentry"><a name="idp45223136"></a><div class="titlepage"></div><div class="refnamediv"><h2>Name</h2><p>jemalloc &#8212; general purpose memory allocation functions</p></div><div class="refsect1"><a name="library"></a><h2>LIBRARY</h2><p>This manual describes jemalloc 4.0.3-0-ge9192eacf8935e29fc62fddc2701f7942b1cc02c.  More information
-    can be found at the <a class="ulink" href="http://www.canonware.com/jemalloc/" target="_top">jemalloc website</a>.</p></div><div class="refsynopsisdiv"><h2>SYNOPSIS</h2><div class="funcsynopsis"><pre class="funcsynopsisinfo">#include &lt;<code class="filename">jemalloc/jemalloc.h</code>&gt;</pre><div class="refsect2"><a name="idp44244480"></a><h3>Standard API</h3><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void *<b class="fsfunc">malloc</b>(</code></td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void *<b class="fsfunc">calloc</b>(</code></td><td>size_t <var class="pdparam">number</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">int <b class="fsfunc">posix_memalign</b>(</code></td><td>void **<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">alignment</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void *<b class="fsfunc">aligned_alloc</b>(</code></td><td>size_t <var class="pdparam">alignment</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void *<b class="fsfunc">realloc</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void <b class="fsfunc">free</b>(</code></td><td>void *<var class="pdparam">ptr</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="refsect2"><a name="idp46062768"></a><h3>Non-standard API</h3><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void *<b class="fsfunc">mallocx</b>(</code></td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void *<b class="fsfunc">rallocx</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">size_t <b class="fsfunc">xallocx</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">extra</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">size_t <b class="fsfunc">sallocx</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void <b class="fsfunc">dallocx</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void <b class="fsfunc">sdallocx</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">size_t <b class="fsfunc">nallocx</b>(</code></td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">int <b class="fsfunc">mallctl</b>(</code></td><td>const char *<var class="pdparam">name</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">oldp</var>, </td></tr><tr><td> </td><td>size_t *<var class="pdparam">oldlenp</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">newp</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">newlen</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">int <b class="fsfunc">mallctlnametomib</b>(</code></td><td>const char *<var class="pdparam">name</var>, </td></tr><tr><td> </td><td>size_t *<var class="pdparam">mibp</var>, </td></tr><tr><td> </td><td>size_t *<var class="pdparam">miblenp</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">int <b class="fsfunc">mallctlbymib</b>(</code></td><td>const size_t *<var class="pdparam">mib</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">miblen</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">oldp</var>, </td></tr><tr><td> </td><td>size_t *<var class="pdparam">oldlenp</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">newp</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">newlen</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void <b class="fsfunc">malloc_stats_print</b>(</code></td><td>void <var class="pdparam">(*write_cb)</var>
+<?xml version="1.0" encoding="utf-8"?>
+<html><head><title>JEMALLOC</title><meta name="generator" content="DocBook XSL Stylesheets V1.79.1"/></head><body bgcolor="white" text="black" link="#0000FF" vlink="#840084" alink="#0000FF"><div class="refentry"><a name="idm45291902356496"/><div class="titlepage"/><div class="refnamediv"><h2>Name</h2><p>jemalloc — general purpose memory allocation functions</p></div><div class="refsect1"><a name="library"/><h2>LIBRARY</h2><p>This manual describes jemalloc 4.4.0-0-gf1f76357313e7dcad7262f17a48ff0a2e005fcdc.  More information
+    can be found at the <a class="ulink" href="http://jemalloc.net/" target="_top">jemalloc website</a>.</p></div><div class="refsynopsisdiv"><h2>SYNOPSIS</h2><div class="funcsynopsis"><pre class="funcsynopsisinfo">#include &lt;<code class="filename">jemalloc/jemalloc.h</code>&gt;</pre><div class="refsect2"><a name="idm45291899537440"/><h3>Standard API</h3><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void *<b class="fsfunc">malloc</b>(</code></td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void *<b class="fsfunc">calloc</b>(</code></td><td>size_t <var class="pdparam">number</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">int <b class="fsfunc">posix_memalign</b>(</code></td><td>void **<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">alignment</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void *<b class="fsfunc">aligned_alloc</b>(</code></td><td>size_t <var class="pdparam">alignment</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void *<b class="fsfunc">realloc</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void <b class="fsfunc">free</b>(</code></td><td>void *<var class="pdparam">ptr</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="refsect2"><a name="idm45291903783248"/><h3>Non-standard API</h3><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void *<b class="fsfunc">mallocx</b>(</code></td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void *<b class="fsfunc">rallocx</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">size_t <b class="fsfunc">xallocx</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">extra</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">size_t <b class="fsfunc">sallocx</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void <b class="fsfunc">dallocx</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void <b class="fsfunc">sdallocx</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">size_t <b class="fsfunc">nallocx</b>(</code></td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">int <b class="fsfunc">mallctl</b>(</code></td><td>const char *<var class="pdparam">name</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">oldp</var>, </td></tr><tr><td> </td><td>size_t *<var class="pdparam">oldlenp</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">newp</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">newlen</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">int <b class="fsfunc">mallctlnametomib</b>(</code></td><td>const char *<var class="pdparam">name</var>, </td></tr><tr><td> </td><td>size_t *<var class="pdparam">mibp</var>, </td></tr><tr><td> </td><td>size_t *<var class="pdparam">miblenp</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">int <b class="fsfunc">mallctlbymib</b>(</code></td><td>const size_t *<var class="pdparam">mib</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">miblen</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">oldp</var>, </td></tr><tr><td> </td><td>size_t *<var class="pdparam">oldlenp</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">newp</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">newlen</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void <b class="fsfunc">malloc_stats_print</b>(</code></td><td>void <var class="pdparam">(*write_cb)</var>
             <code>(</code>void *, const char *<code>)</code>
-          , </td></tr><tr><td> </td><td>void *<var class="pdparam">cbopaque</var>, </td></tr><tr><td> </td><td>const char *<var class="pdparam">opts</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">size_t <b class="fsfunc">malloc_usable_size</b>(</code></td><td>const void *<var class="pdparam">ptr</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void <b class="fsfunc">(*malloc_message)</b>(</code></td><td>void *<var class="pdparam">cbopaque</var>, </td></tr><tr><td> </td><td>const char *<var class="pdparam">s</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><p><span class="type">const char *</span><code class="varname">malloc_conf</code>;</p></div></div></div><div class="refsect1"><a name="description"></a><h2>DESCRIPTION</h2><div class="refsect2"><a name="idp46115952"></a><h3>Standard API</h3><p>The <code class="function">malloc</code>(<em class="parameter"><code></code></em>) function allocates
+          , </td></tr><tr><td> </td><td>void *<var class="pdparam">cbopaque</var>, </td></tr><tr><td> </td><td>const char *<var class="pdparam">opts</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">size_t <b class="fsfunc">malloc_usable_size</b>(</code></td><td>const void *<var class="pdparam">ptr</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void <b class="fsfunc">(*malloc_message)</b>(</code></td><td>void *<var class="pdparam">cbopaque</var>, </td></tr><tr><td> </td><td>const char *<var class="pdparam">s</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><p><span class="type">const char *</span><code class="varname">malloc_conf</code>;</p></div></div></div><div class="refsect1"><a name="description"/><h2>DESCRIPTION</h2><div class="refsect2"><a name="idm45291898401136"/><h3>Standard API</h3><p>The <code class="function">malloc()</code> function allocates
       <em class="parameter"><code>size</code></em> bytes of uninitialized memory.  The allocated
       space is suitably aligned (after possible pointer coercion) for storage
-      of any type of object.</p><p>The <code class="function">calloc</code>(<em class="parameter"><code></code></em>) function allocates
+      of any type of object.</p><p>The <code class="function">calloc()</code> function allocates
       space for <em class="parameter"><code>number</code></em> objects, each
       <em class="parameter"><code>size</code></em> bytes in length.  The result is identical to
-      calling <code class="function">malloc</code>(<em class="parameter"><code></code></em>) with an argument of
+      calling <code class="function">malloc()</code> with an argument of
       <em class="parameter"><code>number</code></em> * <em class="parameter"><code>size</code></em>, with the
       exception that the allocated memory is explicitly initialized to zero
-      bytes.</p><p>The <code class="function">posix_memalign</code>(<em class="parameter"><code></code></em>) function
+      bytes.</p><p>The <code class="function">posix_memalign()</code> function
       allocates <em class="parameter"><code>size</code></em> bytes of memory such that the
       allocation's base address is a multiple of
       <em class="parameter"><code>alignment</code></em>, and returns the allocation in the value
       pointed to by <em class="parameter"><code>ptr</code></em>.  The requested
       <em class="parameter"><code>alignment</code></em> must be a power of 2 at least as large as
-      <code class="code">sizeof(<span class="type">void *</span>)</code>.</p><p>The <code class="function">aligned_alloc</code>(<em class="parameter"><code></code></em>) function
+      <code class="code">sizeof(<span class="type">void *</span>)</code>.</p><p>The <code class="function">aligned_alloc()</code> function
       allocates <em class="parameter"><code>size</code></em> bytes of memory such that the
       allocation's base address is a multiple of
       <em class="parameter"><code>alignment</code></em>.  The requested
       <em class="parameter"><code>alignment</code></em> must be a power of 2.  Behavior is
       undefined if <em class="parameter"><code>size</code></em> is not an integral multiple of
-      <em class="parameter"><code>alignment</code></em>.</p><p>The <code class="function">realloc</code>(<em class="parameter"><code></code></em>) function changes the
+      <em class="parameter"><code>alignment</code></em>.</p><p>The <code class="function">realloc()</code> function changes the
       size of the previously allocated memory referenced by
       <em class="parameter"><code>ptr</code></em> to <em class="parameter"><code>size</code></em> bytes.  The
       contents of the memory are unchanged up to the lesser of the new and old
@@ -30,69 +31,65 @@
       portion of the memory are undefined.  Upon success, the memory referenced
       by <em class="parameter"><code>ptr</code></em> is freed and a pointer to the newly
       allocated memory is returned.  Note that
-      <code class="function">realloc</code>(<em class="parameter"><code></code></em>) may move the memory allocation,
+      <code class="function">realloc()</code> may move the memory allocation,
       resulting in a different return value than <em class="parameter"><code>ptr</code></em>.
       If <em class="parameter"><code>ptr</code></em> is <code class="constant">NULL</code>, the
-      <code class="function">realloc</code>(<em class="parameter"><code></code></em>) function behaves identically to
-      <code class="function">malloc</code>(<em class="parameter"><code></code></em>) for the specified size.</p><p>The <code class="function">free</code>(<em class="parameter"><code></code></em>) function causes the
+      <code class="function">realloc()</code> function behaves identically to
+      <code class="function">malloc()</code> for the specified size.</p><p>The <code class="function">free()</code> function causes the
       allocated memory referenced by <em class="parameter"><code>ptr</code></em> to be made
       available for future allocations.  If <em class="parameter"><code>ptr</code></em> is
-      <code class="constant">NULL</code>, no action occurs.</p></div><div class="refsect2"><a name="idp46144704"></a><h3>Non-standard API</h3><p>The <code class="function">mallocx</code>(<em class="parameter"><code></code></em>),
-      <code class="function">rallocx</code>(<em class="parameter"><code></code></em>),
-      <code class="function">xallocx</code>(<em class="parameter"><code></code></em>),
-      <code class="function">sallocx</code>(<em class="parameter"><code></code></em>),
-      <code class="function">dallocx</code>(<em class="parameter"><code></code></em>),
-      <code class="function">sdallocx</code>(<em class="parameter"><code></code></em>), and
-      <code class="function">nallocx</code>(<em class="parameter"><code></code></em>) functions all have a
+      <code class="constant">NULL</code>, no action occurs.</p></div><div class="refsect2"><a name="idm45291898376160"/><h3>Non-standard API</h3><p>The <code class="function">mallocx()</code>,
+      <code class="function">rallocx()</code>,
+      <code class="function">xallocx()</code>,
+      <code class="function">sallocx()</code>,
+      <code class="function">dallocx()</code>,
+      <code class="function">sdallocx()</code>, and
+      <code class="function">nallocx()</code> functions all have a
       <em class="parameter"><code>flags</code></em> argument that can be used to specify
       options.  The functions only check the options that are contextually
       relevant.  Use bitwise or (<code class="code">|</code>) operations to
       specify one or more of the following:
-        </p><div class="variablelist"><dl class="variablelist"><dt><a name="MALLOCX_LG_ALIGN"></a><span class="term"><code class="constant">MALLOCX_LG_ALIGN(<em class="parameter"><code>la</code></em>)
+        </p><div class="variablelist"><dl class="variablelist"><dt><a name="MALLOCX_LG_ALIGN"/><span class="term"><code class="constant">MALLOCX_LG_ALIGN(<em class="parameter"><code>la</code></em>)
             </code></span></dt><dd><p>Align the memory allocation to start at an address
             that is a multiple of <code class="code">(1 &lt;&lt;
             <em class="parameter"><code>la</code></em>)</code>.  This macro does not validate
             that <em class="parameter"><code>la</code></em> is within the valid
-            range.</p></dd><dt><a name="MALLOCX_ALIGN"></a><span class="term"><code class="constant">MALLOCX_ALIGN(<em class="parameter"><code>a</code></em>)
+            range.</p></dd><dt><a name="MALLOCX_ALIGN"/><span class="term"><code class="constant">MALLOCX_ALIGN(<em class="parameter"><code>a</code></em>)
             </code></span></dt><dd><p>Align the memory allocation to start at an address
             that is a multiple of <em class="parameter"><code>a</code></em>, where
             <em class="parameter"><code>a</code></em> is a power of two.  This macro does not
             validate that <em class="parameter"><code>a</code></em> is a power of 2.
-            </p></dd><dt><a name="MALLOCX_ZERO"></a><span class="term"><code class="constant">MALLOCX_ZERO</code></span></dt><dd><p>Initialize newly allocated memory to contain zero
+            </p></dd><dt><a name="MALLOCX_ZERO"/><span class="term"><code class="constant">MALLOCX_ZERO</code></span></dt><dd><p>Initialize newly allocated memory to contain zero
             bytes.  In the growing reallocation case, the real size prior to
             reallocation defines the boundary between untouched bytes and those
             that are initialized to contain zero bytes.  If this macro is
-            absent, newly allocated memory is uninitialized.</p></dd><dt><a name="MALLOCX_TCACHE"></a><span class="term"><code class="constant">MALLOCX_TCACHE(<em class="parameter"><code>tc</code></em>)
+            absent, newly allocated memory is uninitialized.</p></dd><dt><a name="MALLOCX_TCACHE"/><span class="term"><code class="constant">MALLOCX_TCACHE(<em class="parameter"><code>tc</code></em>)
             </code></span></dt><dd><p>Use the thread-specific cache (tcache) specified by
             the identifier <em class="parameter"><code>tc</code></em>, which must have been
-            acquired via the <a class="link" href="#tcache.create">
-    "<code class="mallctl">tcache.create</code>"
-  </a>
+            acquired via the <a class="link" href="#tcache.create"><quote><code class="mallctl">tcache.create</code></quote></a>
             mallctl.  This macro does not validate that
             <em class="parameter"><code>tc</code></em> specifies a valid
-            identifier.</p></dd><dt><a name="MALLOC_TCACHE_NONE"></a><span class="term"><code class="constant">MALLOCX_TCACHE_NONE</code></span></dt><dd><p>Do not use a thread-specific cache (tcache).  Unless
+            identifier.</p></dd><dt><a name="MALLOC_TCACHE_NONE"/><span class="term"><code class="constant">MALLOCX_TCACHE_NONE</code></span></dt><dd><p>Do not use a thread-specific cache (tcache).  Unless
             <code class="constant">MALLOCX_TCACHE(<em class="parameter"><code>tc</code></em>)</code> or
             <code class="constant">MALLOCX_TCACHE_NONE</code> is specified, an
             automatically managed tcache will be used under many circumstances.
             This macro cannot be used in the same <em class="parameter"><code>flags</code></em>
             argument as
-            <code class="constant">MALLOCX_TCACHE(<em class="parameter"><code>tc</code></em>)</code>.</p></dd><dt><a name="MALLOCX_ARENA"></a><span class="term"><code class="constant">MALLOCX_ARENA(<em class="parameter"><code>a</code></em>)
+            <code class="constant">MALLOCX_TCACHE(<em class="parameter"><code>tc</code></em>)</code>.</p></dd><dt><a name="MALLOCX_ARENA"/><span class="term"><code class="constant">MALLOCX_ARENA(<em class="parameter"><code>a</code></em>)
             </code></span></dt><dd><p>Use the arena specified by the index
             <em class="parameter"><code>a</code></em>.  This macro has no effect for regions that
             were allocated via an arena other than the one specified.  This
             macro does not validate that <em class="parameter"><code>a</code></em> specifies an
             arena index in the valid range.</p></dd></dl></div><p>
-      </p><p>The <code class="function">mallocx</code>(<em class="parameter"><code></code></em>) function allocates at
+      </p><p>The <code class="function">mallocx()</code> function allocates at
       least <em class="parameter"><code>size</code></em> bytes of memory, and returns a pointer
       to the base address of the allocation.  Behavior is undefined if
-      <em class="parameter"><code>size</code></em> is <code class="constant">0</code>, or if request size
-      overflows due to size class and/or alignment constraints.</p><p>The <code class="function">rallocx</code>(<em class="parameter"><code></code></em>) function resizes the
+      <em class="parameter"><code>size</code></em> is <code class="constant">0</code>.</p><p>The <code class="function">rallocx()</code> function resizes the
       allocation at <em class="parameter"><code>ptr</code></em> to be at least
       <em class="parameter"><code>size</code></em> bytes, and returns a pointer to the base
       address of the resulting allocation, which may or may not have moved from
       its original location.  Behavior is undefined if
-      <em class="parameter"><code>size</code></em> is <code class="constant">0</code>, or if request size
-      overflows due to size class and/or alignment constraints.</p><p>The <code class="function">xallocx</code>(<em class="parameter"><code></code></em>) function resizes the
+      <em class="parameter"><code>size</code></em> is <code class="constant">0</code>.</p><p>The <code class="function">xallocx()</code> function resizes the
       allocation at <em class="parameter"><code>ptr</code></em> in place to be at least
       <em class="parameter"><code>size</code></em> bytes, and returns the real size of the
       allocation.  If <em class="parameter"><code>extra</code></em> is non-zero, an attempt is
@@ -101,24 +98,24 @@
       the extra byte(s) will not by itself result in failure to resize.
       Behavior is undefined if <em class="parameter"><code>size</code></em> is
       <code class="constant">0</code>, or if <code class="code">(<em class="parameter"><code>size</code></em> + <em class="parameter"><code>extra</code></em>
-      &gt; <code class="constant">SIZE_T_MAX</code>)</code>.</p><p>The <code class="function">sallocx</code>(<em class="parameter"><code></code></em>) function returns the
-      real size of the allocation at <em class="parameter"><code>ptr</code></em>.</p><p>The <code class="function">dallocx</code>(<em class="parameter"><code></code></em>) function causes the
+      &gt; <code class="constant">SIZE_T_MAX</code>)</code>.</p><p>The <code class="function">sallocx()</code> function returns the
+      real size of the allocation at <em class="parameter"><code>ptr</code></em>.</p><p>The <code class="function">dallocx()</code> function causes the
       memory referenced by <em class="parameter"><code>ptr</code></em> to be made available for
-      future allocations.</p><p>The <code class="function">sdallocx</code>(<em class="parameter"><code></code></em>) function is an
-      extension of <code class="function">dallocx</code>(<em class="parameter"><code></code></em>) with a
+      future allocations.</p><p>The <code class="function">sdallocx()</code> function is an
+      extension of <code class="function">dallocx()</code> with a
       <em class="parameter"><code>size</code></em> parameter to allow the caller to pass in the
       allocation size as an optimization.  The minimum valid input size is the
       original requested size of the allocation, and the maximum valid input
       size is the corresponding value returned by
-      <code class="function">nallocx</code>(<em class="parameter"><code></code></em>) or
-      <code class="function">sallocx</code>(<em class="parameter"><code></code></em>).</p><p>The <code class="function">nallocx</code>(<em class="parameter"><code></code></em>) function allocates no
+      <code class="function">nallocx()</code> or
+      <code class="function">sallocx()</code>.</p><p>The <code class="function">nallocx()</code> function allocates no
       memory, but it performs the same size computation as the
-      <code class="function">mallocx</code>(<em class="parameter"><code></code></em>) function, and returns the real
+      <code class="function">mallocx()</code> function, and returns the real
       size of the allocation that would result from the equivalent
-      <code class="function">mallocx</code>(<em class="parameter"><code></code></em>) function call.  Behavior is
-      undefined if <em class="parameter"><code>size</code></em> is <code class="constant">0</code>, or if
-      request size overflows due to size class and/or alignment
-      constraints.</p><p>The <code class="function">mallctl</code>(<em class="parameter"><code></code></em>) function provides a
+      <code class="function">mallocx()</code> function call, or
+      <code class="constant">0</code> if the inputs exceed the maximum supported size
+      class and/or alignment.  Behavior is undefined if
+      <em class="parameter"><code>size</code></em> is <code class="constant">0</code>.</p><p>The <code class="function">mallctl()</code> function provides a
       general interface for introspecting the memory allocator, as well as
       setting modifiable parameters and triggering actions.  The
       period-separated <em class="parameter"><code>name</code></em> argument specifies a
@@ -130,12 +127,12 @@
       write a value, pass a pointer to the value via
       <em class="parameter"><code>newp</code></em>, and its length via
       <em class="parameter"><code>newlen</code></em>; otherwise pass <code class="constant">NULL</code>
-      and <code class="constant">0</code>.</p><p>The <code class="function">mallctlnametomib</code>(<em class="parameter"><code></code></em>) function
+      and <code class="constant">0</code>.</p><p>The <code class="function">mallctlnametomib()</code> function
       provides a way to avoid repeated name lookups for applications that
       repeatedly query the same portion of the namespace, by translating a name
-      to a &#8220;Management Information Base&#8221; (MIB) that can be passed
-      repeatedly to <code class="function">mallctlbymib</code>(<em class="parameter"><code></code></em>).  Upon
-      successful return from <code class="function">mallctlnametomib</code>(<em class="parameter"><code></code></em>),
+      to a <span class="quote">“<span class="quote">Management Information Base</span>�</span> (MIB) that can be passed
+      repeatedly to <code class="function">mallctlbymib()</code>.  Upon
+      successful return from <code class="function">mallctlnametomib()</code>,
       <em class="parameter"><code>mibp</code></em> contains an array of
       <em class="parameter"><code>*miblenp</code></em> integers, where
       <em class="parameter"><code>*miblenp</code></em> is the lesser of the number of components
@@ -145,9 +142,7 @@
       period-separated name components, which results in a partial MIB that can
       be used as the basis for constructing a complete MIB.  For name
       components that are integers (e.g. the 2 in
-      <a class="link" href="#arenas.bin.i.size">
-    "<code class="mallctl">arenas.bin.2.size</code>"
-  </a>),
+      <a class="link" href="#arenas.bin.i.size"><quote><code class="mallctl">arenas.bin.2.size</code></quote></a>),
       the corresponding MIB component will always be that integer.  Therefore,
       it is legitimate to construct code like the following: </p><pre class="programlisting">
 unsigned nbins, i;
@@ -164,65 +159,62 @@ for (i = 0; i &lt; nbins; i++) {
 
 	mib[2] = i;
 	len = sizeof(bin_size);
-	mallctlbymib(mib, miblen, &amp;bin_size, &amp;len, NULL, 0);
+	mallctlbymib(mib, miblen, (void *)&amp;bin_size, &amp;len, NULL, 0);
 	/* Do something with bin_size... */
-}</pre><p>The <code class="function">malloc_stats_print</code>(<em class="parameter"><code></code></em>) function
-      writes human-readable summary statistics via the
-      <em class="parameter"><code>write_cb</code></em> callback function pointer and
-      <em class="parameter"><code>cbopaque</code></em> data passed to
-      <em class="parameter"><code>write_cb</code></em>, or
-      <code class="function">malloc_message</code>(<em class="parameter"><code></code></em>) if
-      <em class="parameter"><code>write_cb</code></em> is <code class="constant">NULL</code>.  This
-      function can be called repeatedly.  General information that never
-      changes during execution can be omitted by specifying "g" as a character
+}</pre><p>The <code class="function">malloc_stats_print()</code> function writes
+      summary statistics via the <em class="parameter"><code>write_cb</code></em> callback
+      function pointer and <em class="parameter"><code>cbopaque</code></em> data passed to
+      <em class="parameter"><code>write_cb</code></em>, or <code class="function">malloc_message()</code>
+      if <em class="parameter"><code>write_cb</code></em> is <code class="constant">NULL</code>.  The
+      statistics are presented in human-readable form unless <span class="quote">“<span class="quote">J</span>�</span> is
+      specified as a character within the <em class="parameter"><code>opts</code></em> string, in
+      which case the statistics are presented in <a class="ulink" href="http://www.json.org/" target="_top">JSON format</a>.  This function can be
+      called repeatedly.  General information that never changes during
+      execution can be omitted by specifying <span class="quote">“<span class="quote">g</span>�</span> as a character
       within the <em class="parameter"><code>opts</code></em> string.  Note that
-      <code class="function">malloc_message</code>(<em class="parameter"><code></code></em>) uses the
-      <code class="function">mallctl*</code>(<em class="parameter"><code></code></em>) functions internally, so
-      inconsistent statistics can be reported if multiple threads use these
-      functions simultaneously.  If <code class="option">--enable-stats</code> is
-      specified during configuration, &#8220;m&#8221; and &#8220;a&#8221; can
-      be specified to omit merged arena and per arena statistics, respectively;
-      &#8220;b&#8221;, &#8220;l&#8221;, and &#8220;h&#8221; can be specified to
-      omit per size class statistics for bins, large objects, and huge objects,
-      respectively.  Unrecognized characters are silently ignored.  Note that
-      thread caching may prevent some statistics from being completely up to
-      date, since extra locking would be required to merge counters that track
-      thread cache operations.
-      </p><p>The <code class="function">malloc_usable_size</code>(<em class="parameter"><code></code></em>) function
+      <code class="function">malloc_message()</code> uses the
+      <code class="function">mallctl*()</code> functions internally, so inconsistent
+      statistics can be reported if multiple threads use these functions
+      simultaneously.  If <code class="option">--enable-stats</code> is specified during
+      configuration, <span class="quote">“<span class="quote">m</span>�</span> and <span class="quote">“<span class="quote">a</span>�</span> can be specified to
+      omit merged arena and per arena statistics, respectively;
+      <span class="quote">“<span class="quote">b</span>�</span>, <span class="quote">“<span class="quote">l</span>�</span>, and <span class="quote">“<span class="quote">h</span>�</span> can be specified
+      to omit per size class statistics for bins, large objects, and huge
+      objects, respectively.  Unrecognized characters are silently ignored.
+      Note that thread caching may prevent some statistics from being completely
+      up to date, since extra locking would be required to merge counters that
+      track thread cache operations.</p><p>The <code class="function">malloc_usable_size()</code> function
       returns the usable size of the allocation pointed to by
       <em class="parameter"><code>ptr</code></em>.  The return value may be larger than the size
       that was requested during allocation.  The
-      <code class="function">malloc_usable_size</code>(<em class="parameter"><code></code></em>) function is not a
-      mechanism for in-place <code class="function">realloc</code>(<em class="parameter"><code></code></em>); rather
+      <code class="function">malloc_usable_size()</code> function is not a
+      mechanism for in-place <code class="function">realloc()</code>; rather
       it is provided solely as a tool for introspection purposes.  Any
       discrepancy between the requested allocation size and the size reported
-      by <code class="function">malloc_usable_size</code>(<em class="parameter"><code></code></em>) should not be
+      by <code class="function">malloc_usable_size()</code> should not be
       depended on, since such behavior is entirely implementation-dependent.
-      </p></div></div><div class="refsect1"><a name="tuning"></a><h2>TUNING</h2><p>Once, when the first call is made to one of the memory allocation
+      </p></div></div><div class="refsect1"><a name="tuning"/><h2>TUNING</h2><p>Once, when the first call is made to one of the memory allocation
     routines, the allocator initializes its internals based in part on various
-    options that can be specified at compile- or run-time.</p><p>The string pointed to by the global variable
-    <code class="varname">malloc_conf</code>, the &#8220;name&#8221; of the file
-    referenced by the symbolic link named <code class="filename">/etc/malloc.conf</code>, and the value of the
+    options that can be specified at compile- or run-time.</p><p>The string specified via <code class="option">--with-malloc-conf</code>, the
+    string pointed to by the global variable <code class="varname">malloc_conf</code>, the
+    <span class="quote">“<span class="quote">name</span>�</span> of the file referenced by the symbolic link named
+    <code class="filename">/etc/malloc.conf</code>, and the value of the
     environment variable <code class="envar">MALLOC_CONF</code>, will be interpreted, in
     that order, from left to right as options.  Note that
     <code class="varname">malloc_conf</code> may be read before
-    <code class="function">main</code>(<em class="parameter"><code></code></em>) is entered, so the declaration of
+    <code class="function">main()</code> is entered, so the declaration of
     <code class="varname">malloc_conf</code> should specify an initializer that contains
-    the final value to be read by jemalloc.  <code class="varname">malloc_conf</code> is
-    a compile-time setting, whereas <code class="filename">/etc/malloc.conf</code> and <code class="envar">MALLOC_CONF</code>
-    can be safely set any time prior to program invocation.</p><p>An options string is a comma-separated list of option:value pairs.
-    There is one key corresponding to each <a class="link" href="#opt.abort">
-    "<code class="mallctl">opt.*</code>"
-  </a> mallctl (see the <a class="xref" href="#mallctl_namespace" title="MALLCTL NAMESPACE">MALLCTL NAMESPACE</a> section for options
+    the final value to be read by jemalloc.  <code class="option">--with-malloc-conf</code>
+    and <code class="varname">malloc_conf</code> are compile-time mechanisms, whereas
+    <code class="filename">/etc/malloc.conf</code> and
+    <code class="envar">MALLOC_CONF</code> can be safely set any time prior to program
+    invocation.</p><p>An options string is a comma-separated list of option:value pairs.
+    There is one key corresponding to each <a class="link" href="#opt.abort"><quote><code class="mallctl">opt.*</code></quote></a> mallctl (see the <a class="xref" href="#mallctl_namespace" title="MALLCTL NAMESPACE">MALLCTL NAMESPACE</a> section for options
     documentation).  For example, <code class="literal">abort:true,narenas:1</code> sets
-    the <a class="link" href="#opt.abort">
-    "<code class="mallctl">opt.abort</code>"
-  </a> and <a class="link" href="#opt.narenas">
-    "<code class="mallctl">opt.narenas</code>"
-  </a> options.  Some
+    the <a class="link" href="#opt.abort"><quote><code class="mallctl">opt.abort</code></quote></a> and <a class="link" href="#opt.narenas"><quote><code class="mallctl">opt.narenas</code></quote></a> options.  Some
     options have boolean values (true/false), others have integer values (base
     8, 10, or 16, depending on prefix), and yet others have raw string
-    values.</p></div><div class="refsect1"><a name="implementation_notes"></a><h2>IMPLEMENTATION NOTES</h2><p>Traditionally, allocators have used
+    values.</p></div><div class="refsect1"><a name="implementation_notes"/><h2>IMPLEMENTATION NOTES</h2><p>Traditionally, allocators have used
     <span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span> to obtain memory, which is
     suboptimal for several reasons, including race conditions, increased
     fragmentation, and artificial limitations on maximum usable memory.  If
@@ -246,19 +238,18 @@ for (i = 0; i &lt; nbins; i++) {
     order to make it possible to completely avoid synchronization for most
     allocation requests.  Such caching allows very fast allocation in the
     common case, but it increases memory usage and fragmentation, since a
-    bounded number of objects can remain allocated in each thread cache.</p><p>Memory is conceptually broken into equal-sized chunks, where the
-    chunk size is a power of two that is greater than the page size.  Chunks
-    are always aligned to multiples of the chunk size.  This alignment makes it
-    possible to find metadata for user objects very quickly.</p><p>User objects are broken into three categories according to size:
-    small, large, and huge.  Small and large objects are managed entirely by
-    arenas; huge objects are additionally aggregated in a single data structure
-    that is shared by all threads.  Huge objects are typically used by
-    applications infrequently enough that this single data structure is not a
-    scalability issue.</p><p>Each chunk that is managed by an arena tracks its contents as runs of
+    bounded number of objects can remain allocated in each thread cache.</p><p>Memory is conceptually broken into equal-sized chunks, where the chunk
+    size is a power of two that is greater than the page size.  Chunks are
+    always aligned to multiples of the chunk size.  This alignment makes it
+    possible to find metadata for user objects very quickly.  User objects are
+    broken into three categories according to size: small, large, and huge.
+    Multiple small and large objects can reside within a single chunk, whereas
+    huge objects each have one or more chunks backing them.  Each chunk that
+    contains small and/or large objects tracks its contents as runs of
     contiguous pages (unused, backing a set of small objects, or backing one
-    large object).  The combination of chunk alignment and chunk page maps
-    makes it possible to determine all metadata regarding small and large
-    allocations in constant time.</p><p>Small objects are managed in groups by page runs.  Each run maintains
+    large object).  The combination of chunk alignment and chunk page maps makes
+    it possible to determine all metadata regarding small and large allocations
+    in constant time.</p><p>Small objects are managed in groups by page runs.  Each run maintains
     a bitmap to track which regions are in use.  Allocation requests that are no
     more than half the quantum (8 or 16, depending on architecture) are rounded
     up to the nearest power of two that is at least <code class="code">sizeof(<span class="type">double</span>)</code>.  All other object size
@@ -266,22 +257,20 @@ for (i = 0; i &lt; nbins; i++) {
     classes for each doubling in size, which limits internal fragmentation to
     approximately 20% for all but the smallest size classes.  Small size classes
     are smaller than four times the page size, large size classes are smaller
-    than the chunk size (see the <a class="link" href="#opt.lg_chunk">
-    "<code class="mallctl">opt.lg_chunk</code>"
-  </a> option), and
-    huge size classes extend from the chunk size up to one size class less than
-    the full address space size.</p><p>Allocations are packed tightly together, which can be an issue for
+    than the chunk size (see the <a class="link" href="#opt.lg_chunk"><quote><code class="mallctl">opt.lg_chunk</code></quote></a> option), and
+    huge size classes extend from the chunk size up to the largest size class
+    that does not exceed <code class="constant">PTRDIFF_MAX</code>.</p><p>Allocations are packed tightly together, which can be an issue for
     multi-threaded applications.  If you need to assure that allocations do not
     suffer from cacheline sharing, round your allocation requests up to the
     nearest multiple of the cacheline size, or specify cacheline alignment when
-    allocating.</p><p>The <code class="function">realloc</code>(<em class="parameter"><code></code></em>),
-    <code class="function">rallocx</code>(<em class="parameter"><code></code></em>), and
-    <code class="function">xallocx</code>(<em class="parameter"><code></code></em>) functions may resize allocations
+    allocating.</p><p>The <code class="function">realloc()</code>,
+    <code class="function">rallocx()</code>, and
+    <code class="function">xallocx()</code> functions may resize allocations
     without moving them under limited circumstances.  Unlike the
-    <code class="function">*allocx</code>(<em class="parameter"><code></code></em>) API, the standard API does not
+    <code class="function">*allocx()</code> API, the standard API does not
     officially round up the usable size of an allocation to the nearest size
     class, so technically it is necessary to call
-    <code class="function">realloc</code>(<em class="parameter"><code></code></em>) to grow e.g. a 9-byte allocation to
+    <code class="function">realloc()</code> to grow e.g. a 9-byte allocation to
     16 bytes, or shrink a 16-byte allocation to 9 bytes.  Growth and shrinkage
     trivially succeeds in place as long as the pre-size and post-size both round
     up to the same size class.  No other API guarantees are made regarding
@@ -289,147 +278,111 @@ for (i = 0; i &lt; nbins; i++) {
     and huge allocations in place, as long as the pre-size and post-size are
     both large or both huge.  In such cases shrinkage always succeeds for large
     size classes, but for huge size classes the chunk allocator must support
-    splitting (see <a class="link" href="#arena.i.chunk_hooks">
-    "<code class="mallctl">arena.&lt;i&gt;.chunk_hooks</code>"
-  </a>).
+    splitting (see <a class="link" href="#arena.i.chunk_hooks"><quote><code class="mallctl">arena.&lt;i&gt;.chunk_hooks</code></quote></a>).
     Growth only succeeds if the trailing memory is currently available, and
     additionally for huge size classes the chunk allocator must support
     merging.</p><p>Assuming 2 MiB chunks, 4 KiB pages, and a 16-byte quantum on a
-    64-bit system, the size classes in each category are as shown in <a class="xref" href="#size_classes" title="Table 1. Size classes">Table 1</a>.</p><div class="table"><a name="size_classes"></a><p class="title"><b>Table 1. Size classes</b></p><div class="table-contents"><table summary="Size classes" border="1"><colgroup><col align="left" class="c1"><col align="right" class="c2"><col align="left" class="c3"></colgroup><thead><tr><th align="left">Category</th><th align="right">Spacing</th><th align="left">Size</th></tr></thead><tbody><tr><td rowspan="9" align="left">Small</td><td align="right">lg</td><td align="left">[8]</td></tr><tr><td align="right">16</td><td align="left">[16, 32, 48, 64, 80, 96, 112, 128]</td></tr><tr><td align="right">32</td><td align="left">[160, 192, 224, 256]</td></tr><tr><td align="right">64</td><td align="left">[320, 384, 448, 512]</td></tr><tr><td align="right">128</td><td align="left">[640, 768, 896, 1024]</td></tr><tr><td align="right">256</td><td align="left">[1280, 1536, 1792, 2048]</td></tr><tr><td align="right">512</td><td align="left">[2560, 3072, 3584, 4096]</td></tr><tr><td align="right">1 KiB</td><td align="left">[5 KiB, 6 KiB, 7 KiB, 8 KiB]</td></tr><tr><td align="right">2 KiB</td><td align="left">[10 KiB, 12 KiB, 14 KiB]</td></tr><tr><td rowspan="8" align="left">Large</td><td align="right">2 KiB</td><td align="left">[16 KiB]</td></tr><tr><td align="right">4 KiB</td><td align="left">[20 KiB, 24 KiB, 28 KiB, 32 KiB]</td></tr><tr><td align="right">8 KiB</td><td align="left">[40 KiB, 48 KiB, 54 KiB, 64 KiB]</td></tr><tr><td align="right">16 KiB</td><td align="left">[80 KiB, 96 KiB, 112 KiB, 128 KiB]</td></tr><tr><td align="right">32 KiB</td><td align="left">[160 KiB, 192 KiB, 224 KiB, 256 KiB]</td></tr><tr><td align="right">64 KiB</td><td align="left">[320 KiB, 384 KiB, 448 KiB, 512 KiB]</td></tr><tr><td align="right">128 KiB</td><td align="left">[640 KiB, 768 KiB, 896 KiB, 1 MiB]</td></tr><tr><td align="right">256 KiB</td><td align="left">[1280 KiB, 1536 KiB, 1792 KiB]</td></tr><tr><td rowspan="7" align="left">Huge</td><td align="right">256 KiB</td><td align="left">[2 MiB]</td></tr><tr><td align="right">512 KiB</td><td align="left">[2560 KiB, 3 MiB, 3584 KiB, 4 MiB]</td></tr><tr><td align="right">1 MiB</td><td align="left">[5 MiB, 6 MiB, 7 MiB, 8 MiB]</td></tr><tr><td align="right">2 MiB</td><td align="left">[10 MiB, 12 MiB, 14 MiB, 16 MiB]</td></tr><tr><td align="right">4 MiB</td><td align="left">[20 MiB, 24 MiB, 28 MiB, 32 MiB]</td></tr><tr><td align="right">8 MiB</td><td align="left">[40 MiB, 48 MiB, 56 MiB, 64 MiB]</td></tr><tr><td align="right">...</td><td align="left">...</td></tr></tbody></table></div></div><br class="table-break"></div><div class="refsect1"><a name="mallctl_namespace"></a><h2>MALLCTL NAMESPACE</h2><p>The following names are defined in the namespace accessible via the
-    <code class="function">mallctl*</code>(<em class="parameter"><code></code></em>) functions.  Value types are
+    64-bit system, the size classes in each category are as shown in <a class="xref" href="#size_classes" title="Table 1. Size classes">Table 1</a>.</p><div class="table"><a name="size_classes"/><p class="title"><b>Table 1. Size classes</b></p><div class="table-contents"><table class="table" summary="Size classes" border="1"><colgroup><col align="left" class="c1"/><col align="right" class="c2"/><col align="left" class="c3"/></colgroup><thead><tr><th align="left">Category</th><th align="right">Spacing</th><th align="left">Size</th></tr></thead><tbody><tr><td rowspan="9" align="left">Small</td><td align="right">lg</td><td align="left">[8]</td></tr><tr><td align="right">16</td><td align="left">[16, 32, 48, 64, 80, 96, 112, 128]</td></tr><tr><td align="right">32</td><td align="left">[160, 192, 224, 256]</td></tr><tr><td align="right">64</td><td align="left">[320, 384, 448, 512]</td></tr><tr><td align="right">128</td><td align="left">[640, 768, 896, 1024]</td></tr><tr><td align="right">256</td><td align="left">[1280, 1536, 1792, 2048]</td></tr><tr><td align="right">512</td><td align="left">[2560, 3072, 3584, 4096]</td></tr><tr><td align="right">1 KiB</td><td align="left">[5 KiB, 6 KiB, 7 KiB, 8 KiB]</td></tr><tr><td align="right">2 KiB</td><td align="left">[10 KiB, 12 KiB, 14 KiB]</td></tr><tr><td rowspan="8" align="left">Large</td><td align="right">2 KiB</td><td align="left">[16 KiB]</td></tr><tr><td align="right">4 KiB</td><td align="left">[20 KiB, 24 KiB, 28 KiB, 32 KiB]</td></tr><tr><td align="right">8 KiB</td><td align="left">[40 KiB, 48 KiB, 54 KiB, 64 KiB]</td></tr><tr><td align="right">16 KiB</td><td align="left">[80 KiB, 96 KiB, 112 KiB, 128 KiB]</td></tr><tr><td align="right">32 KiB</td><td align="left">[160 KiB, 192 KiB, 224 KiB, 256 KiB]</td></tr><tr><td align="right">64 KiB</td><td align="left">[320 KiB, 384 KiB, 448 KiB, 512 KiB]</td></tr><tr><td align="right">128 KiB</td><td align="left">[640 KiB, 768 KiB, 896 KiB, 1 MiB]</td></tr><tr><td align="right">256 KiB</td><td align="left">[1280 KiB, 1536 KiB, 1792 KiB]</td></tr><tr><td rowspan="9" align="left">Huge</td><td align="right">256 KiB</td><td align="left">[2 MiB]</td></tr><tr><td align="right">512 KiB</td><td align="left">[2560 KiB, 3 MiB, 3584 KiB, 4 MiB]</td></tr><tr><td align="right">1 MiB</td><td align="left">[5 MiB, 6 MiB, 7 MiB, 8 MiB]</td></tr><tr><td align="right">2 MiB</td><td align="left">[10 MiB, 12 MiB, 14 MiB, 16 MiB]</td></tr><tr><td align="right">4 MiB</td><td align="left">[20 MiB, 24 MiB, 28 MiB, 32 MiB]</td></tr><tr><td align="right">8 MiB</td><td align="left">[40 MiB, 48 MiB, 56 MiB, 64 MiB]</td></tr><tr><td align="right">...</td><td align="left">...</td></tr><tr><td align="right">512 PiB</td><td align="left">[2560 PiB, 3 EiB, 3584 PiB, 4 EiB]</td></tr><tr><td align="right">1 EiB</td><td align="left">[5 EiB, 6 EiB, 7 EiB]</td></tr></tbody></table></div></div><br class="table-break"/></div><div class="refsect1"><a name="mallctl_namespace"/><h2>MALLCTL NAMESPACE</h2><p>The following names are defined in the namespace accessible via the
+    <code class="function">mallctl*()</code> functions.  Value types are
     specified in parentheses, their readable/writable statuses are encoded as
     <code class="literal">rw</code>, <code class="literal">r-</code>, <code class="literal">-w</code>, or
     <code class="literal">--</code>, and required build configuration flags follow, if
     any.  A name element encoded as <code class="literal">&lt;i&gt;</code> or
     <code class="literal">&lt;j&gt;</code> indicates an integer component, where the
     integer varies from 0 to some upper value that must be determined via
-    introspection.  In the case of 
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.*</code>"
-  ,
-    <code class="literal">&lt;i&gt;</code> equal to <a class="link" href="#arenas.narenas">
-    "<code class="mallctl">arenas.narenas</code>"
-  </a> can be
+    introspection.  In the case of <quote><code class="mallctl">stats.arenas.&lt;i&gt;.*</code></quote>,
+    <code class="literal">&lt;i&gt;</code> equal to <a class="link" href="#arenas.narenas"><quote><code class="mallctl">arenas.narenas</code></quote></a> can be
     used to access the summation of statistics from all arenas.  Take special
-    note of the <a class="link" href="#epoch">
-    "<code class="mallctl">epoch</code>"
-  </a> mallctl,
-    which controls refreshing of cached dynamic statistics.</p><div class="variablelist"><dl class="variablelist"><dt><a name="version"></a><span class="term">
-          
-    "<code class="mallctl">version</code>"
-  
+    note of the <a class="link" href="#epoch"><quote><code class="mallctl">epoch</code></quote></a> mallctl,
+    which controls refreshing of cached dynamic statistics.</p><div class="variablelist"><dl class="variablelist"><dt><a name="version"/><span class="term">
+          <quote><code class="mallctl">version</code></quote>
           (<span class="type">const char *</span>)
           <code class="literal">r-</code>
-        </span></dt><dd><p>Return the jemalloc version string.</p></dd><dt><a name="epoch"></a><span class="term">
-          
-    "<code class="mallctl">epoch</code>"
-  
+        </span></dt><dd><p>Return the jemalloc version string.</p></dd><dt><a name="epoch"/><span class="term">
+          <quote><code class="mallctl">epoch</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">rw</code>
         </span></dt><dd><p>If a value is passed in, refresh the data from which
-        the <code class="function">mallctl*</code>(<em class="parameter"><code></code></em>) functions report values,
+        the <code class="function">mallctl*()</code> functions report values,
         and increment the epoch.  Return the current epoch.  This is useful for
-        detecting whether another thread caused a refresh.</p></dd><dt><a name="config.cache_oblivious"></a><span class="term">
-          
-    "<code class="mallctl">config.cache_oblivious</code>"
-  
+        detecting whether another thread caused a refresh.</p></dd><dt><a name="config.cache_oblivious"/><span class="term">
+          <quote><code class="mallctl">config.cache_oblivious</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p><code class="option">--enable-cache-oblivious</code> was specified
-        during build configuration.</p></dd><dt><a name="config.debug"></a><span class="term">
-          
-    "<code class="mallctl">config.debug</code>"
-  
+        during build configuration.</p></dd><dt><a name="config.debug"/><span class="term">
+          <quote><code class="mallctl">config.debug</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p><code class="option">--enable-debug</code> was specified during
-        build configuration.</p></dd><dt><a name="config.fill"></a><span class="term">
-          
-    "<code class="mallctl">config.fill</code>"
-  
+        build configuration.</p></dd><dt><a name="config.fill"/><span class="term">
+          <quote><code class="mallctl">config.fill</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p><code class="option">--enable-fill</code> was specified during
-        build configuration.</p></dd><dt><a name="config.lazy_lock"></a><span class="term">
-          
-    "<code class="mallctl">config.lazy_lock</code>"
-  
+        build configuration.</p></dd><dt><a name="config.lazy_lock"/><span class="term">
+          <quote><code class="mallctl">config.lazy_lock</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p><code class="option">--enable-lazy-lock</code> was specified
-        during build configuration.</p></dd><dt><a name="config.munmap"></a><span class="term">
-          
-    "<code class="mallctl">config.munmap</code>"
-  
+        during build configuration.</p></dd><dt><a name="config.malloc_conf"/><span class="term">
+          <quote><code class="mallctl">config.malloc_conf</code></quote>
+          (<span class="type">const char *</span>)
+          <code class="literal">r-</code>
+        </span></dt><dd><p>Embedded configure-time-specified run-time options
+        string, empty unless <code class="option">--with-malloc-conf</code> was specified
+        during build configuration.</p></dd><dt><a name="config.munmap"/><span class="term">
+          <quote><code class="mallctl">config.munmap</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p><code class="option">--enable-munmap</code> was specified during
-        build configuration.</p></dd><dt><a name="config.prof"></a><span class="term">
-          
-    "<code class="mallctl">config.prof</code>"
-  
+        build configuration.</p></dd><dt><a name="config.prof"/><span class="term">
+          <quote><code class="mallctl">config.prof</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p><code class="option">--enable-prof</code> was specified during
-        build configuration.</p></dd><dt><a name="config.prof_libgcc"></a><span class="term">
-          
-    "<code class="mallctl">config.prof_libgcc</code>"
-  
+        build configuration.</p></dd><dt><a name="config.prof_libgcc"/><span class="term">
+          <quote><code class="mallctl">config.prof_libgcc</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p><code class="option">--disable-prof-libgcc</code> was not
-        specified during build configuration.</p></dd><dt><a name="config.prof_libunwind"></a><span class="term">
-          
-    "<code class="mallctl">config.prof_libunwind</code>"
-  
+        specified during build configuration.</p></dd><dt><a name="config.prof_libunwind"/><span class="term">
+          <quote><code class="mallctl">config.prof_libunwind</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p><code class="option">--enable-prof-libunwind</code> was specified
-        during build configuration.</p></dd><dt><a name="config.stats"></a><span class="term">
-          
-    "<code class="mallctl">config.stats</code>"
-  
+        during build configuration.</p></dd><dt><a name="config.stats"/><span class="term">
+          <quote><code class="mallctl">config.stats</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p><code class="option">--enable-stats</code> was specified during
-        build configuration.</p></dd><dt><a name="config.tcache"></a><span class="term">
-          
-    "<code class="mallctl">config.tcache</code>"
-  
+        build configuration.</p></dd><dt><a name="config.tcache"/><span class="term">
+          <quote><code class="mallctl">config.tcache</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p><code class="option">--disable-tcache</code> was not specified
-        during build configuration.</p></dd><dt><a name="config.tls"></a><span class="term">
-          
-    "<code class="mallctl">config.tls</code>"
-  
+        during build configuration.</p></dd><dt><a name="config.tls"/><span class="term">
+          <quote><code class="mallctl">config.tls</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p><code class="option">--disable-tls</code> was not specified during
-        build configuration.</p></dd><dt><a name="config.utrace"></a><span class="term">
-          
-    "<code class="mallctl">config.utrace</code>"
-  
+        build configuration.</p></dd><dt><a name="config.utrace"/><span class="term">
+          <quote><code class="mallctl">config.utrace</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p><code class="option">--enable-utrace</code> was specified during
-        build configuration.</p></dd><dt><a name="config.valgrind"></a><span class="term">
-          
-    "<code class="mallctl">config.valgrind</code>"
-  
+        build configuration.</p></dd><dt><a name="config.valgrind"/><span class="term">
+          <quote><code class="mallctl">config.valgrind</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p><code class="option">--enable-valgrind</code> was specified during
-        build configuration.</p></dd><dt><a name="config.xmalloc"></a><span class="term">
-          
-    "<code class="mallctl">config.xmalloc</code>"
-  
+        build configuration.</p></dd><dt><a name="config.xmalloc"/><span class="term">
+          <quote><code class="mallctl">config.xmalloc</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p><code class="option">--enable-xmalloc</code> was specified during
-        build configuration.</p></dd><dt><a name="opt.abort"></a><span class="term">
-          
-    "<code class="mallctl">opt.abort</code>"
-  
+        build configuration.</p></dd><dt><a name="opt.abort"/><span class="term">
+          <quote><code class="mallctl">opt.abort</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p>Abort-on-warning enabled/disabled.  If true, most
@@ -437,43 +390,42 @@ for (i = 0; i &lt; nbins; i++) {
         <span class="citerefentry"><span class="refentrytitle">abort</span>(3)</span> in these cases.  This option is
         disabled by default unless <code class="option">--enable-debug</code> is
         specified during configuration, in which case it is enabled by default.
-        </p></dd><dt><a name="opt.dss"></a><span class="term">
-          
-    "<code class="mallctl">opt.dss</code>"
-  
+        </p></dd><dt><a name="opt.dss"/><span class="term">
+          <quote><code class="mallctl">opt.dss</code></quote>
           (<span class="type">const char *</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p>dss (<span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span>) allocation precedence as
         related to <span class="citerefentry"><span class="refentrytitle">mmap</span>(2)</span> allocation.  The following
         settings are supported if
         <span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span> is supported by the operating
-        system: &#8220;disabled&#8221;, &#8220;primary&#8221;, and
-        &#8220;secondary&#8221;; otherwise only &#8220;disabled&#8221; is
-        supported.  The default is &#8220;secondary&#8221; if
+        system: <span class="quote">“<span class="quote">disabled</span>�</span>, <span class="quote">“<span class="quote">primary</span>�</span>, and
+        <span class="quote">“<span class="quote">secondary</span>�</span>; otherwise only <span class="quote">“<span class="quote">disabled</span>�</span> is
+        supported.  The default is <span class="quote">“<span class="quote">secondary</span>�</span> if
         <span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span> is supported by the operating
-        system; &#8220;disabled&#8221; otherwise.
-        </p></dd><dt><a name="opt.lg_chunk"></a><span class="term">
-          
-    "<code class="mallctl">opt.lg_chunk</code>"
-  
+        system; <span class="quote">“<span class="quote">disabled</span>�</span> otherwise.
+        </p></dd><dt><a name="opt.lg_chunk"/><span class="term">
+          <quote><code class="mallctl">opt.lg_chunk</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p>Virtual memory chunk size (log base 2).  If a chunk
         size outside the supported size range is specified, the size is
         silently clipped to the minimum/maximum supported size.  The default
         chunk size is 2 MiB (2^21).
-        </p></dd><dt><a name="opt.narenas"></a><span class="term">
-          
-    "<code class="mallctl">opt.narenas</code>"
-  
-          (<span class="type">size_t</span>)
+        </p></dd><dt><a name="opt.narenas"/><span class="term">
+          <quote><code class="mallctl">opt.narenas</code></quote>
+          (<span class="type">unsigned</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p>Maximum number of arenas to use for automatic
         multiplexing of threads and arenas.  The default is four times the
-        number of CPUs, or one if there is a single CPU.</p></dd><dt><a name="opt.lg_dirty_mult"></a><span class="term">
-          
-    "<code class="mallctl">opt.lg_dirty_mult</code>"
-  
+        number of CPUs, or one if there is a single CPU.</p></dd><dt><a name="opt.purge"/><span class="term">
+          <quote><code class="mallctl">opt.purge</code></quote>
+          (<span class="type">const char *</span>)
+          <code class="literal">r-</code>
+        </span></dt><dd><p>Purge mode is “ratio� (default) or
+        “decay�.  See <a class="link" href="#opt.lg_dirty_mult"><quote><code class="mallctl">opt.lg_dirty_mult</code></quote></a>
+        for details of the ratio mode.  See <a class="link" href="#opt.decay_time"><quote><code class="mallctl">opt.decay_time</code></quote></a> for
+        details of the decay mode.</p></dd><dt><a name="opt.lg_dirty_mult"/><span class="term">
+          <quote><code class="mallctl">opt.lg_dirty_mult</code></quote>
           (<span class="type">ssize_t</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p>Per-arena minimum ratio (log base 2) of active to dirty
@@ -484,53 +436,57 @@ for (i = 0; i &lt; nbins; i++) {
         provides the kernel with sufficient information to recycle dirty pages
         if physical memory becomes scarce and the pages remain unused.  The
         default minimum ratio is 8:1 (2^3:1); an option value of -1 will
-        disable dirty page purging.  See <a class="link" href="#arenas.lg_dirty_mult">
-    "<code class="mallctl">arenas.lg_dirty_mult</code>"
-  </a>
-        and <a class="link" href="#arena.i.lg_dirty_mult">
-    "<code class="mallctl">arena.&lt;i&gt;.lg_dirty_mult</code>"
-  </a>
-        for related dynamic control options.</p></dd><dt><a name="opt.stats_print"></a><span class="term">
-          
-    "<code class="mallctl">opt.stats_print</code>"
-  
+        disable dirty page purging.  See <a class="link" href="#arenas.lg_dirty_mult"><quote><code class="mallctl">arenas.lg_dirty_mult</code></quote></a>
+        and <a class="link" href="#arena.i.lg_dirty_mult"><quote><code class="mallctl">arena.&lt;i&gt;.lg_dirty_mult</code></quote></a>
+        for related dynamic control options.</p></dd><dt><a name="opt.decay_time"/><span class="term">
+          <quote><code class="mallctl">opt.decay_time</code></quote>
+          (<span class="type">ssize_t</span>)
+          <code class="literal">r-</code>
+        </span></dt><dd><p>Approximate time in seconds from the creation of a set
+        of unused dirty pages until an equivalent set of unused dirty pages is
+        purged and/or reused.  The pages are incrementally purged according to a
+        sigmoidal decay curve that starts and ends with zero purge rate.  A
+        decay time of 0 causes all unused dirty pages to be purged immediately
+        upon creation.  A decay time of -1 disables purging.  The default decay
+        time is 10 seconds.  See <a class="link" href="#arenas.decay_time"><quote><code class="mallctl">arenas.decay_time</code></quote></a>
+        and <a class="link" href="#arena.i.decay_time"><quote><code class="mallctl">arena.&lt;i&gt;.decay_time</code></quote></a>
+        for related dynamic control options.
+        </p></dd><dt><a name="opt.stats_print"/><span class="term">
+          <quote><code class="mallctl">opt.stats_print</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p>Enable/disable statistics printing at exit.  If
-        enabled, the <code class="function">malloc_stats_print</code>(<em class="parameter"><code></code></em>)
+        enabled, the <code class="function">malloc_stats_print()</code>
         function is called at program exit via an
         <span class="citerefentry"><span class="refentrytitle">atexit</span>(3)</span> function.  If
         <code class="option">--enable-stats</code> is specified during configuration, this
         has the potential to cause deadlock for a multi-threaded process that
         exits while one or more threads are executing in the memory allocation
-        functions.  Furthermore, <code class="function">atexit</code>(<em class="parameter"><code></code></em>) may
+        functions.  Furthermore, <code class="function">atexit()</code> may
         allocate memory during application initialization and then deadlock
         internally when jemalloc in turn calls
-        <code class="function">atexit</code>(<em class="parameter"><code></code></em>), so this option is not
-        univerally usable (though the application can register its own
-        <code class="function">atexit</code>(<em class="parameter"><code></code></em>) function with equivalent
+        <code class="function">atexit()</code>, so this option is not
+        universally usable (though the application can register its own
+        <code class="function">atexit()</code> function with equivalent
         functionality).  Therefore, this option should only be used with care;
         it is primarily intended as a performance tuning aid during application
-        development.  This option is disabled by default.</p></dd><dt><a name="opt.junk"></a><span class="term">
-          
-    "<code class="mallctl">opt.junk</code>"
-  
+        development.  This option is disabled by default.</p></dd><dt><a name="opt.junk"/><span class="term">
+          <quote><code class="mallctl">opt.junk</code></quote>
           (<span class="type">const char *</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-fill</code>]
-        </span></dt><dd><p>Junk filling.  If set to "alloc", each byte of
-        uninitialized allocated memory will be initialized to
-        <code class="literal">0xa5</code>.  If set to "free", all deallocated memory will
-        be initialized to <code class="literal">0x5a</code>.  If set to "true", both
-        allocated and deallocated memory will be initialized, and if set to
-        "false", junk filling be disabled entirely.  This is intended for
-        debugging and will impact performance negatively.  This option is
-        "false" by default unless <code class="option">--enable-debug</code> is specified
-        during configuration, in which case it is "true" by default unless
-        running inside <a class="ulink" href="http://valgrind.org/" target="_top">Valgrind</a>.</p></dd><dt><a name="opt.quarantine"></a><span class="term">
-          
-    "<code class="mallctl">opt.quarantine</code>"
-  
+        </span></dt><dd><p>Junk filling.  If set to <span class="quote">“<span class="quote">alloc</span>�</span>, each byte
+        of uninitialized allocated memory will be initialized to
+        <code class="literal">0xa5</code>.  If set to <span class="quote">“<span class="quote">free</span>�</span>, all deallocated
+        memory will be initialized to <code class="literal">0x5a</code>.  If set to
+        <span class="quote">“<span class="quote">true</span>�</span>, both allocated and deallocated memory will be
+        initialized, and if set to <span class="quote">“<span class="quote">false</span>�</span>, junk filling be
+        disabled entirely.  This is intended for debugging and will impact
+        performance negatively.  This option is <span class="quote">“<span class="quote">false</span>�</span> by default
+        unless <code class="option">--enable-debug</code> is specified during
+        configuration, in which case it is <span class="quote">“<span class="quote">true</span>�</span> by default unless
+        running inside <a class="ulink" href="http://valgrind.org/" target="_top">Valgrind</a>.</p></dd><dt><a name="opt.quarantine"/><span class="term">
+          <quote><code class="mallctl">opt.quarantine</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-fill</code>]
@@ -538,58 +494,46 @@ for (i = 0; i &lt; nbins; i++) {
         thread maintains a FIFO object quarantine that stores up to the
         specified number of bytes of memory.  The quarantined memory is not
         freed until it is released from quarantine, though it is immediately
-        junk-filled if the <a class="link" href="#opt.junk">
-    "<code class="mallctl">opt.junk</code>"
-  </a> option is
+        junk-filled if the <a class="link" href="#opt.junk"><quote><code class="mallctl">opt.junk</code></quote></a> option is
         enabled.  This feature is of particular use in combination with <a class="ulink" href="http://valgrind.org/" target="_top">Valgrind</a>, which can detect attempts
         to access quarantined objects.  This is intended for debugging and will
         impact performance negatively.  The default quarantine size is 0 unless
         running inside Valgrind, in which case the default is 16
-        MiB.</p></dd><dt><a name="opt.redzone"></a><span class="term">
-          
-    "<code class="mallctl">opt.redzone</code>"
-  
+        MiB.</p></dd><dt><a name="opt.redzone"/><span class="term">
+          <quote><code class="mallctl">opt.redzone</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-fill</code>]
         </span></dt><dd><p>Redzones enabled/disabled.  If enabled, small
         allocations have redzones before and after them.  Furthermore, if the
-        <a class="link" href="#opt.junk">
-    "<code class="mallctl">opt.junk</code>"
-  </a> option is
+        <a class="link" href="#opt.junk"><quote><code class="mallctl">opt.junk</code></quote></a> option is
         enabled, the redzones are checked for corruption during deallocation.
         However, the primary intended purpose of this feature is to be used in
         combination with <a class="ulink" href="http://valgrind.org/" target="_top">Valgrind</a>,
         which needs redzones in order to do effective buffer overflow/underflow
         detection.  This option is intended for debugging and will impact
         performance negatively.  This option is disabled by
-        default unless running inside Valgrind.</p></dd><dt><a name="opt.zero"></a><span class="term">
-          
-    "<code class="mallctl">opt.zero</code>"
-  
+        default unless running inside Valgrind.</p></dd><dt><a name="opt.zero"/><span class="term">
+          <quote><code class="mallctl">opt.zero</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-fill</code>]
         </span></dt><dd><p>Zero filling enabled/disabled.  If enabled, each byte
         of uninitialized allocated memory will be initialized to 0.  Note that
         this initialization only happens once for each byte, so
-        <code class="function">realloc</code>(<em class="parameter"><code></code></em>) and
-        <code class="function">rallocx</code>(<em class="parameter"><code></code></em>) calls do not zero memory that
+        <code class="function">realloc()</code> and
+        <code class="function">rallocx()</code> calls do not zero memory that
         was previously allocated.  This is intended for debugging and will
         impact performance negatively.  This option is disabled by default.
-        </p></dd><dt><a name="opt.utrace"></a><span class="term">
-          
-    "<code class="mallctl">opt.utrace</code>"
-  
+        </p></dd><dt><a name="opt.utrace"/><span class="term">
+          <quote><code class="mallctl">opt.utrace</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-utrace</code>]
         </span></dt><dd><p>Allocation tracing based on
         <span class="citerefentry"><span class="refentrytitle">utrace</span>(2)</span> enabled/disabled.  This option
-        is disabled by default.</p></dd><dt><a name="opt.xmalloc"></a><span class="term">
-          
-    "<code class="mallctl">opt.xmalloc</code>"
-  
+        is disabled by default.</p></dd><dt><a name="opt.xmalloc"/><span class="term">
+          <quote><code class="mallctl">opt.xmalloc</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-xmalloc</code>]
@@ -602,10 +546,8 @@ for (i = 0; i &lt; nbins; i++) {
         including the following in the source code:
         </p><pre class="programlisting">
 malloc_conf = "xmalloc:true";</pre><p>
-        This option is disabled by default.</p></dd><dt><a name="opt.tcache"></a><span class="term">
-          
-    "<code class="mallctl">opt.tcache</code>"
-  
+        This option is disabled by default.</p></dd><dt><a name="opt.tcache"/><span class="term">
+          <quote><code class="mallctl">opt.tcache</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-tcache</code>]
@@ -613,55 +555,36 @@ malloc_conf = "xmalloc:true";</pre><p>
         there are multiple threads, each thread uses a tcache for objects up to
         a certain size.  Thread-specific caching allows many allocations to be
         satisfied without performing any thread synchronization, at the cost of
-        increased memory use.  See the <a class="link" href="#opt.lg_tcache_max">
-    "<code class="mallctl">opt.lg_tcache_max</code>"
-  </a>
+        increased memory use.  See the <a class="link" href="#opt.lg_tcache_max"><quote><code class="mallctl">opt.lg_tcache_max</code></quote></a>
         option for related tuning information.  This option is enabled by
         default unless running inside <a class="ulink" href="http://valgrind.org/" target="_top">Valgrind</a>, in which case it is
-        forcefully disabled.</p></dd><dt><a name="opt.lg_tcache_max"></a><span class="term">
-          
-    "<code class="mallctl">opt.lg_tcache_max</code>"
-  
+        forcefully disabled.</p></dd><dt><a name="opt.lg_tcache_max"/><span class="term">
+          <quote><code class="mallctl">opt.lg_tcache_max</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-tcache</code>]
         </span></dt><dd><p>Maximum size class (log base 2) to cache in the
         thread-specific cache (tcache).  At a minimum, all small size classes
         are cached, and at a maximum all large size classes are cached.  The
-        default maximum is 32 KiB (2^15).</p></dd><dt><a name="opt.prof"></a><span class="term">
-          
-    "<code class="mallctl">opt.prof</code>"
-  
+        default maximum is 32 KiB (2^15).</p></dd><dt><a name="opt.prof"/><span class="term">
+          <quote><code class="mallctl">opt.prof</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-prof</code>]
         </span></dt><dd><p>Memory profiling enabled/disabled.  If enabled, profile
-        memory allocation activity.  See the <a class="link" href="#opt.prof_active">
-    "<code class="mallctl">opt.prof_active</code>"
-  </a>
-        option for on-the-fly activation/deactivation.  See the <a class="link" href="#opt.lg_prof_sample">
-    "<code class="mallctl">opt.lg_prof_sample</code>"
-  </a>
-        option for probabilistic sampling control.  See the <a class="link" href="#opt.prof_accum">
-    "<code class="mallctl">opt.prof_accum</code>"
-  </a>
-        option for control of cumulative sample reporting.  See the <a class="link" href="#opt.lg_prof_interval">
-    "<code class="mallctl">opt.lg_prof_interval</code>"
-  </a>
-        option for information on interval-triggered profile dumping, the <a class="link" href="#opt.prof_gdump">
-    "<code class="mallctl">opt.prof_gdump</code>"
-  </a>
+        memory allocation activity.  See the <a class="link" href="#opt.prof_active"><quote><code class="mallctl">opt.prof_active</code></quote></a>
+        option for on-the-fly activation/deactivation.  See the <a class="link" href="#opt.lg_prof_sample"><quote><code class="mallctl">opt.lg_prof_sample</code></quote></a>
+        option for probabilistic sampling control.  See the <a class="link" href="#opt.prof_accum"><quote><code class="mallctl">opt.prof_accum</code></quote></a>
+        option for control of cumulative sample reporting.  See the <a class="link" href="#opt.lg_prof_interval"><quote><code class="mallctl">opt.lg_prof_interval</code></quote></a>
+        option for information on interval-triggered profile dumping, the <a class="link" href="#opt.prof_gdump"><quote><code class="mallctl">opt.prof_gdump</code></quote></a>
         option for information on high-water-triggered profile dumping, and the
-        <a class="link" href="#opt.prof_final">
-    "<code class="mallctl">opt.prof_final</code>"
-  </a>
+        <a class="link" href="#opt.prof_final"><quote><code class="mallctl">opt.prof_final</code></quote></a>
         option for final profile dumping.  Profile output is compatible with
         the <span class="command"><strong>jeprof</strong></span> command, which is based on the
         <span class="command"><strong>pprof</strong></span> that is developed as part of the <a class="ulink" href="http://code.google.com/p/gperftools/" target="_top">gperftools
-        package</a>.</p></dd><dt><a name="opt.prof_prefix"></a><span class="term">
-          
-    "<code class="mallctl">opt.prof_prefix</code>"
-  
+        package</a>.  See <a class="link" href="#heap_profile_format" title="HEAP PROFILE FORMAT">HEAP PROFILE
+        FORMAT</a> for heap profile format documentation.</p></dd><dt><a name="opt.prof_prefix"/><span class="term">
+          <quote><code class="mallctl">opt.prof_prefix</code></quote>
           (<span class="type">const char *</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-prof</code>]
@@ -669,40 +592,26 @@ malloc_conf = "xmalloc:true";</pre><p>
         set to the empty string, no automatic dumps will occur; this is
         primarily useful for disabling the automatic final heap dump (which
         also disables leak reporting, if enabled).  The default prefix is
-        <code class="filename">jeprof</code>.</p></dd><dt><a name="opt.prof_active"></a><span class="term">
-          
-    "<code class="mallctl">opt.prof_active</code>"
-  
+        <code class="filename">jeprof</code>.</p></dd><dt><a name="opt.prof_active"/><span class="term">
+          <quote><code class="mallctl">opt.prof_active</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-prof</code>]
         </span></dt><dd><p>Profiling activated/deactivated.  This is a secondary
         control mechanism that makes it possible to start the application with
-        profiling enabled (see the <a class="link" href="#opt.prof">
-    "<code class="mallctl">opt.prof</code>"
-  </a> option) but
+        profiling enabled (see the <a class="link" href="#opt.prof"><quote><code class="mallctl">opt.prof</code></quote></a> option) but
         inactive, then toggle profiling at any time during program execution
-        with the <a class="link" href="#prof.active">
-    "<code class="mallctl">prof.active</code>"
-  </a> mallctl.
-        This option is enabled by default.</p></dd><dt><a name="opt.prof_thread_active_init"></a><span class="term">
-          
-    "<code class="mallctl">opt.prof_thread_active_init</code>"
-  
+        with the <a class="link" href="#prof.active"><quote><code class="mallctl">prof.active</code></quote></a> mallctl.
+        This option is enabled by default.</p></dd><dt><a name="opt.prof_thread_active_init"/><span class="term">
+          <quote><code class="mallctl">opt.prof_thread_active_init</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-prof</code>]
-        </span></dt><dd><p>Initial setting for <a class="link" href="#thread.prof.active">
-    "<code class="mallctl">thread.prof.active</code>"
-  </a>
+        </span></dt><dd><p>Initial setting for <a class="link" href="#thread.prof.active"><quote><code class="mallctl">thread.prof.active</code></quote></a>
         in newly created threads.  The initial setting for newly created threads
-        can also be changed during execution via the <a class="link" href="#prof.thread_active_init">
-    "<code class="mallctl">prof.thread_active_init</code>"
-  </a>
-        mallctl.  This option is enabled by default.</p></dd><dt><a name="opt.lg_prof_sample"></a><span class="term">
-          
-    "<code class="mallctl">opt.lg_prof_sample</code>"
-  
+        can also be changed during execution via the <a class="link" href="#prof.thread_active_init"><quote><code class="mallctl">prof.thread_active_init</code></quote></a>
+        mallctl.  This option is enabled by default.</p></dd><dt><a name="opt.lg_prof_sample"/><span class="term">
+          <quote><code class="mallctl">opt.lg_prof_sample</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-prof</code>]
@@ -710,10 +619,8 @@ malloc_conf = "xmalloc:true";</pre><p>
         samples, as measured in bytes of allocation activity.  Increasing the
         sampling interval decreases profile fidelity, but also decreases the
         computational overhead.  The default sample interval is 512 KiB (2^19
-        B).</p></dd><dt><a name="opt.prof_accum"></a><span class="term">
-          
-    "<code class="mallctl">opt.prof_accum</code>"
-  
+        B).</p></dd><dt><a name="opt.prof_accum"/><span class="term">
+          <quote><code class="mallctl">opt.prof_accum</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-prof</code>]
@@ -722,10 +629,8 @@ malloc_conf = "xmalloc:true";</pre><p>
         backtrace must be stored for the duration of execution.  Depending on
         the application, this can impose a large memory overhead, and the
         cumulative counts are not always of interest.  This option is disabled
-        by default.</p></dd><dt><a name="opt.lg_prof_interval"></a><span class="term">
-          
-    "<code class="mallctl">opt.lg_prof_interval</code>"
-  
+        by default.</p></dd><dt><a name="opt.lg_prof_interval"/><span class="term">
+          <quote><code class="mallctl">opt.lg_prof_interval</code></quote>
           (<span class="type">ssize_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-prof</code>]
@@ -736,27 +641,19 @@ malloc_conf = "xmalloc:true";</pre><p>
         dumped to files named according to the pattern
         <code class="filename">&lt;prefix&gt;.&lt;pid&gt;.&lt;seq&gt;.i&lt;iseq&gt;.heap</code>,
         where <code class="literal">&lt;prefix&gt;</code> is controlled by the
-        <a class="link" href="#opt.prof_prefix">
-    "<code class="mallctl">opt.prof_prefix</code>"
-  </a>
+        <a class="link" href="#opt.prof_prefix"><quote><code class="mallctl">opt.prof_prefix</code></quote></a>
         option.  By default, interval-triggered profile dumping is disabled
         (encoded as -1).
-        </p></dd><dt><a name="opt.prof_gdump"></a><span class="term">
-          
-    "<code class="mallctl">opt.prof_gdump</code>"
-  
+        </p></dd><dt><a name="opt.prof_gdump"/><span class="term">
+          <quote><code class="mallctl">opt.prof_gdump</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-prof</code>]
-        </span></dt><dd><p>Set the initial state of <a class="link" href="#prof.gdump">
-    "<code class="mallctl">prof.gdump</code>"
-  </a>, which when
+        </span></dt><dd><p>Set the initial state of <a class="link" href="#prof.gdump"><quote><code class="mallctl">prof.gdump</code></quote></a>, which when
         enabled triggers a memory profile dump every time the total virtual
         memory exceeds the previous maximum.  This option is disabled by
-        default.</p></dd><dt><a name="opt.prof_final"></a><span class="term">
-          
-    "<code class="mallctl">opt.prof_final</code>"
-  
+        default.</p></dd><dt><a name="opt.prof_final"/><span class="term">
+          <quote><code class="mallctl">opt.prof_final</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-prof</code>]
@@ -764,101 +661,73 @@ malloc_conf = "xmalloc:true";</pre><p>
         <span class="citerefentry"><span class="refentrytitle">atexit</span>(3)</span> function to dump final memory
         usage to a file named according to the pattern
         <code class="filename">&lt;prefix&gt;.&lt;pid&gt;.&lt;seq&gt;.f.heap</code>,
-        where <code class="literal">&lt;prefix&gt;</code> is controlled by the <a class="link" href="#opt.prof_prefix">
-    "<code class="mallctl">opt.prof_prefix</code>"
-  </a>
-        option.  Note that <code class="function">atexit</code>(<em class="parameter"><code></code></em>) may allocate
+        where <code class="literal">&lt;prefix&gt;</code> is controlled by the <a class="link" href="#opt.prof_prefix"><quote><code class="mallctl">opt.prof_prefix</code></quote></a>
+        option.  Note that <code class="function">atexit()</code> may allocate
         memory during application initialization and then deadlock internally
-        when jemalloc in turn calls <code class="function">atexit</code>(<em class="parameter"><code></code></em>), so
-        this option is not univerally usable (though the application can
-        register its own <code class="function">atexit</code>(<em class="parameter"><code></code></em>) function with
+        when jemalloc in turn calls <code class="function">atexit()</code>, so
+        this option is not universally usable (though the application can
+        register its own <code class="function">atexit()</code> function with
         equivalent functionality).  This option is disabled by
-        default.</p></dd><dt><a name="opt.prof_leak"></a><span class="term">
-          
-    "<code class="mallctl">opt.prof_leak</code>"
-  
+        default.</p></dd><dt><a name="opt.prof_leak"/><span class="term">
+          <quote><code class="mallctl">opt.prof_leak</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-prof</code>]
         </span></dt><dd><p>Leak reporting enabled/disabled.  If enabled, use an
         <span class="citerefentry"><span class="refentrytitle">atexit</span>(3)</span> function to report memory leaks
         detected by allocation sampling.  See the
-        <a class="link" href="#opt.prof">
-    "<code class="mallctl">opt.prof</code>"
-  </a> option for
+        <a class="link" href="#opt.prof"><quote><code class="mallctl">opt.prof</code></quote></a> option for
         information on analyzing heap profile output.  This option is disabled
-        by default.</p></dd><dt><a name="thread.arena"></a><span class="term">
-          
-    "<code class="mallctl">thread.arena</code>"
-  
+        by default.</p></dd><dt><a name="thread.arena"/><span class="term">
+          <quote><code class="mallctl">thread.arena</code></quote>
           (<span class="type">unsigned</span>)
           <code class="literal">rw</code>
         </span></dt><dd><p>Get or set the arena associated with the calling
         thread.  If the specified arena was not initialized beforehand (see the
-        <a class="link" href="#arenas.initialized">
-    "<code class="mallctl">arenas.initialized</code>"
-  </a>
+        <a class="link" href="#arenas.initialized"><quote><code class="mallctl">arenas.initialized</code></quote></a>
         mallctl), it will be automatically initialized as a side effect of
-        calling this interface.</p></dd><dt><a name="thread.allocated"></a><span class="term">
-          
-    "<code class="mallctl">thread.allocated</code>"
-  
+        calling this interface.</p></dd><dt><a name="thread.allocated"/><span class="term">
+          <quote><code class="mallctl">thread.allocated</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Get the total number of bytes ever allocated by the
         calling thread.  This counter has the potential to wrap around; it is
         up to the application to appropriately interpret the counter in such
-        cases.</p></dd><dt><a name="thread.allocatedp"></a><span class="term">
-          
-    "<code class="mallctl">thread.allocatedp</code>"
-  
+        cases.</p></dd><dt><a name="thread.allocatedp"/><span class="term">
+          <quote><code class="mallctl">thread.allocatedp</code></quote>
           (<span class="type">uint64_t *</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Get a pointer to the the value that is returned by the
-        <a class="link" href="#thread.allocated">
-    "<code class="mallctl">thread.allocated</code>"
-  </a>
+        <a class="link" href="#thread.allocated"><quote><code class="mallctl">thread.allocated</code></quote></a>
         mallctl.  This is useful for avoiding the overhead of repeated
-        <code class="function">mallctl*</code>(<em class="parameter"><code></code></em>) calls.</p></dd><dt><a name="thread.deallocated"></a><span class="term">
-          
-    "<code class="mallctl">thread.deallocated</code>"
-  
+        <code class="function">mallctl*()</code> calls.</p></dd><dt><a name="thread.deallocated"/><span class="term">
+          <quote><code class="mallctl">thread.deallocated</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Get the total number of bytes ever deallocated by the
         calling thread.  This counter has the potential to wrap around; it is
         up to the application to appropriately interpret the counter in such
-        cases.</p></dd><dt><a name="thread.deallocatedp"></a><span class="term">
-          
-    "<code class="mallctl">thread.deallocatedp</code>"
-  
+        cases.</p></dd><dt><a name="thread.deallocatedp"/><span class="term">
+          <quote><code class="mallctl">thread.deallocatedp</code></quote>
           (<span class="type">uint64_t *</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Get a pointer to the the value that is returned by the
-        <a class="link" href="#thread.deallocated">
-    "<code class="mallctl">thread.deallocated</code>"
-  </a>
+        <a class="link" href="#thread.deallocated"><quote><code class="mallctl">thread.deallocated</code></quote></a>
         mallctl.  This is useful for avoiding the overhead of repeated
-        <code class="function">mallctl*</code>(<em class="parameter"><code></code></em>) calls.</p></dd><dt><a name="thread.tcache.enabled"></a><span class="term">
-          
-    "<code class="mallctl">thread.tcache.enabled</code>"
-  
+        <code class="function">mallctl*()</code> calls.</p></dd><dt><a name="thread.tcache.enabled"/><span class="term">
+          <quote><code class="mallctl">thread.tcache.enabled</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">rw</code>
           [<code class="option">--enable-tcache</code>]
         </span></dt><dd><p>Enable/disable calling thread's tcache.  The tcache is
         implicitly flushed as a side effect of becoming
-        disabled (see <a class="link" href="#thread.tcache.flush">
-    "<code class="mallctl">thread.tcache.flush</code>"
-  </a>).
-        </p></dd><dt><a name="thread.tcache.flush"></a><span class="term">
-          
-    "<code class="mallctl">thread.tcache.flush</code>"
-  
+        disabled (see <a class="link" href="#thread.tcache.flush"><quote><code class="mallctl">thread.tcache.flush</code></quote></a>).
+        </p></dd><dt><a name="thread.tcache.flush"/><span class="term">
+          <quote><code class="mallctl">thread.tcache.flush</code></quote>
           (<span class="type">void</span>)
           <code class="literal">--</code>
           [<code class="option">--enable-tcache</code>]
@@ -870,10 +739,8 @@ malloc_conf = "xmalloc:true";</pre><p>
         a thread exits.  However, garbage collection is triggered by allocation
         activity, so it is possible for a thread that stops
         allocating/deallocating to retain its cache indefinitely, in which case
-        the developer may find manual flushing useful.</p></dd><dt><a name="thread.prof.name"></a><span class="term">
-          
-    "<code class="mallctl">thread.prof.name</code>"
-  
+        the developer may find manual flushing useful.</p></dd><dt><a name="thread.prof.name"/><span class="term">
+          <quote><code class="mallctl">thread.prof.name</code></quote>
           (<span class="type">const char *</span>)
           <code class="literal">r-</code> or
           <code class="literal">-w</code>
@@ -886,25 +753,19 @@ malloc_conf = "xmalloc:true";</pre><p>
         can cause asynchronous string deallocation.  Furthermore, each
         invocation of this interface can only read or write; simultaneous
         read/write is not supported due to string lifetime limitations.  The
-        name string must nil-terminated and comprised only of characters in the
-        sets recognized
+        name string must be nil-terminated and comprised only of characters in
+        the sets recognized
         by <span class="citerefentry"><span class="refentrytitle">isgraph</span>(3)</span> and
-        <span class="citerefentry"><span class="refentrytitle">isblank</span>(3)</span>.</p></dd><dt><a name="thread.prof.active"></a><span class="term">
-          
-    "<code class="mallctl">thread.prof.active</code>"
-  
+        <span class="citerefentry"><span class="refentrytitle">isblank</span>(3)</span>.</p></dd><dt><a name="thread.prof.active"/><span class="term">
+          <quote><code class="mallctl">thread.prof.active</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">rw</code>
           [<code class="option">--enable-prof</code>]
         </span></dt><dd><p>Control whether sampling is currently active for the
-        calling thread.  This is an activation mechanism in addition to <a class="link" href="#prof.active">
-    "<code class="mallctl">prof.active</code>"
-  </a>; both must
+        calling thread.  This is an activation mechanism in addition to <a class="link" href="#prof.active"><quote><code class="mallctl">prof.active</code></quote></a>; both must
         be active for the calling thread to sample.  This flag is enabled by
-        default.</p></dd><dt><a name="tcache.create"></a><span class="term">
-          
-    "<code class="mallctl">tcache.create</code>"
-  
+        default.</p></dd><dt><a name="tcache.create"/><span class="term">
+          <quote><code class="mallctl">tcache.create</code></quote>
           (<span class="type">unsigned</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-tcache</code>]
@@ -914,76 +775,81 @@ malloc_conf = "xmalloc:true";</pre><p>
         automatically managed one that is used by default.  Each explicit cache
         can be used by only one thread at a time; the application must assure
         that this constraint holds.
-        </p></dd><dt><a name="tcache.flush"></a><span class="term">
-          
-    "<code class="mallctl">tcache.flush</code>"
-  
+        </p></dd><dt><a name="tcache.flush"/><span class="term">
+          <quote><code class="mallctl">tcache.flush</code></quote>
           (<span class="type">unsigned</span>)
           <code class="literal">-w</code>
           [<code class="option">--enable-tcache</code>]
         </span></dt><dd><p>Flush the specified thread-specific cache (tcache).  The
-        same considerations apply to this interface as to <a class="link" href="#thread.tcache.flush">
-    "<code class="mallctl">thread.tcache.flush</code>"
-  </a>,
-        except that the tcache will never be automatically be discarded.
-        </p></dd><dt><a name="tcache.destroy"></a><span class="term">
-          
-    "<code class="mallctl">tcache.destroy</code>"
-  
+        same considerations apply to this interface as to <a class="link" href="#thread.tcache.flush"><quote><code class="mallctl">thread.tcache.flush</code></quote></a>,
+        except that the tcache will never be automatically discarded.
+        </p></dd><dt><a name="tcache.destroy"/><span class="term">
+          <quote><code class="mallctl">tcache.destroy</code></quote>
           (<span class="type">unsigned</span>)
           <code class="literal">-w</code>
           [<code class="option">--enable-tcache</code>]
         </span></dt><dd><p>Flush the specified thread-specific cache (tcache) and
         make the identifier available for use during a future tcache creation.
-        </p></dd><dt><a name="arena.i.purge"></a><span class="term">
-          
-    "<code class="mallctl">arena.&lt;i&gt;.purge</code>"
-  
+        </p></dd><dt><a name="arena.i.purge"/><span class="term">
+          <quote><code class="mallctl">arena.&lt;i&gt;.purge</code></quote>
+          (<span class="type">void</span>)
+          <code class="literal">--</code>
+        </span></dt><dd><p>Purge all unused dirty pages for arena &lt;i&gt;, or for
+        all arenas if &lt;i&gt; equals <a class="link" href="#arenas.narenas"><quote><code class="mallctl">arenas.narenas</code></quote></a>.
+        </p></dd><dt><a name="arena.i.decay"/><span class="term">
+          <quote><code class="mallctl">arena.&lt;i&gt;.decay</code></quote>
           (<span class="type">void</span>)
           <code class="literal">--</code>
-        </span></dt><dd><p>Purge unused dirty pages for arena &lt;i&gt;, or for
-        all arenas if &lt;i&gt; equals <a class="link" href="#arenas.narenas">
-    "<code class="mallctl">arenas.narenas</code>"
-  </a>.
-        </p></dd><dt><a name="arena.i.dss"></a><span class="term">
-          
-    "<code class="mallctl">arena.&lt;i&gt;.dss</code>"
-  
+        </span></dt><dd><p>Trigger decay-based purging of unused dirty pages for
+        arena &lt;i&gt;, or for all arenas if &lt;i&gt; equals <a class="link" href="#arenas.narenas"><quote><code class="mallctl">arenas.narenas</code></quote></a>.
+        The proportion of unused dirty pages to be purged depends on the current
+        time; see <a class="link" href="#opt.decay_time"><quote><code class="mallctl">opt.decay_time</code></quote></a> for
+        details.</p></dd><dt><a name="arena.i.reset"/><span class="term">
+          <quote><code class="mallctl">arena.&lt;i&gt;.reset</code></quote>
+          (<span class="type">void</span>)
+          <code class="literal">--</code>
+        </span></dt><dd><p>Discard all of the arena's extant allocations.  This
+        interface can only be used with arenas created via <a class="link" href="#arenas.extend"><quote><code class="mallctl">arenas.extend</code></quote></a>.  None
+        of the arena's discarded/cached allocations may accessed afterward.  As
+        part of this requirement, all thread caches which were used to
+        allocate/deallocate in conjunction with the arena must be flushed
+        beforehand.  This interface cannot be used if running inside Valgrind,
+        nor if the <a class="link" href="#opt.quarantine">quarantine</a> size is
+        non-zero.</p></dd><dt><a name="arena.i.dss"/><span class="term">
+          <quote><code class="mallctl">arena.&lt;i&gt;.dss</code></quote>
           (<span class="type">const char *</span>)
           <code class="literal">rw</code>
         </span></dt><dd><p>Set the precedence of dss allocation as related to mmap
         allocation for arena &lt;i&gt;, or for all arenas if &lt;i&gt; equals
-        <a class="link" href="#arenas.narenas">
-    "<code class="mallctl">arenas.narenas</code>"
-  </a>.  See
-        <a class="link" href="#opt.dss">
-    "<code class="mallctl">opt.dss</code>"
-  </a> for supported
-        settings.</p></dd><dt><a name="arena.i.lg_dirty_mult"></a><span class="term">
-          
-    "<code class="mallctl">arena.&lt;i&gt;.lg_dirty_mult</code>"
-  
+        <a class="link" href="#arenas.narenas"><quote><code class="mallctl">arenas.narenas</code></quote></a>.  See
+        <a class="link" href="#opt.dss"><quote><code class="mallctl">opt.dss</code></quote></a> for supported
+        settings.</p></dd><dt><a name="arena.i.lg_dirty_mult"/><span class="term">
+          <quote><code class="mallctl">arena.&lt;i&gt;.lg_dirty_mult</code></quote>
           (<span class="type">ssize_t</span>)
           <code class="literal">rw</code>
         </span></dt><dd><p>Current per-arena minimum ratio (log base 2) of active
         to dirty pages for arena &lt;i&gt;.  Each time this interface is set and
         the ratio is increased, pages are synchronously purged as necessary to
-        impose the new ratio.  See <a class="link" href="#opt.lg_dirty_mult">
-    "<code class="mallctl">opt.lg_dirty_mult</code>"
-  </a>
-        for additional information.</p></dd><dt><a name="arena.i.chunk_hooks"></a><span class="term">
-          
-    "<code class="mallctl">arena.&lt;i&gt;.chunk_hooks</code>"
-  
+        impose the new ratio.  See <a class="link" href="#opt.lg_dirty_mult"><quote><code class="mallctl">opt.lg_dirty_mult</code></quote></a>
+        for additional information.</p></dd><dt><a name="arena.i.decay_time"/><span class="term">
+          <quote><code class="mallctl">arena.&lt;i&gt;.decay_time</code></quote>
+          (<span class="type">ssize_t</span>)
+          <code class="literal">rw</code>
+        </span></dt><dd><p>Current per-arena approximate time in seconds from the
+        creation of a set of unused dirty pages until an equivalent set of
+        unused dirty pages is purged and/or reused.  Each time this interface is
+        set, all currently unused dirty pages are considered to have fully
+        decayed, which causes immediate purging of all unused dirty pages unless
+        the decay time is set to -1 (i.e. purging disabled).  See <a class="link" href="#opt.decay_time"><quote><code class="mallctl">opt.decay_time</code></quote></a> for
+        additional information.</p></dd><dt><a name="arena.i.chunk_hooks"/><span class="term">
+          <quote><code class="mallctl">arena.&lt;i&gt;.chunk_hooks</code></quote>
           (<span class="type">chunk_hooks_t</span>)
           <code class="literal">rw</code>
         </span></dt><dd><p>Get or set the chunk management hook functions for arena
         &lt;i&gt;.  The functions must be capable of operating on all extant
         chunks associated with arena &lt;i&gt;, usually by passing unknown
         chunks to the replaced functions.  In practice, it is feasible to
-        control allocation for arenas created via <a class="link" href="#arenas.extend">
-    "<code class="mallctl">arenas.extend</code>"
-  </a> such
+        control allocation for arenas created via <a class="link" href="#arenas.extend"><quote><code class="mallctl">arenas.extend</code></quote></a> such
         that all chunks originate from an application-supplied chunk allocator
         (by setting custom chunk hook functions just after arena creation), but
         the automatically created arenas may have already created chunks prior
@@ -1009,7 +875,7 @@ typedef struct {
         operations can also be opted out of, but this is mainly intended to
         support platforms on which virtual memory mappings provided by the
         operating system kernel do not automatically coalesce and split, e.g.
-        Windows.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef void *<b class="fsfunc">(chunk_alloc_t)</b>(</code></td><td>void *<var class="pdparam">chunk</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">alignment</var>, </td></tr><tr><td> </td><td>bool *<var class="pdparam">zero</var>, </td></tr><tr><td> </td><td>bool *<var class="pdparam">commit</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p></p></div><p>A chunk allocation function conforms to the
+        Windows.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef void *<b class="fsfunc">(chunk_alloc_t)</b>(</code></td><td>void *<var class="pdparam">chunk</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">alignment</var>, </td></tr><tr><td> </td><td>bool *<var class="pdparam">zero</var>, </td></tr><tr><td> </td><td>bool *<var class="pdparam">commit</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p/></div><p>A chunk allocation function conforms to the
         <span class="type">chunk_alloc_t</span> type and upon success returns a pointer to
         <em class="parameter"><code>size</code></em> bytes of mapped memory on behalf of arena
         <em class="parameter"><code>arena_ind</code></em> such that the chunk's base address is a
@@ -1030,10 +896,8 @@ typedef struct {
         in absolute terms as on a system that does not overcommit, or in
         implicit terms as on a system that overcommits and satisfies physical
         memory needs on demand via soft page faults.  Note that replacing the
-        default chunk allocation function makes the arena's <a class="link" href="#arena.i.dss">
-    "<code class="mallctl">arena.&lt;i&gt;.dss</code>"
-  </a>
-        setting irrelevant.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef bool <b class="fsfunc">(chunk_dalloc_t)</b>(</code></td><td>void *<var class="pdparam">chunk</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>bool <var class="pdparam">committed</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p></p></div><p>
+        default chunk allocation function makes the arena's <a class="link" href="#arena.i.dss"><quote><code class="mallctl">arena.&lt;i&gt;.dss</code></quote></a>
+        setting irrelevant.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef bool <b class="fsfunc">(chunk_dalloc_t)</b>(</code></td><td>void *<var class="pdparam">chunk</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>bool <var class="pdparam">committed</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p/></div><p>
         A chunk deallocation function conforms to the
         <span class="type">chunk_dalloc_t</span> type and deallocates a
         <em class="parameter"><code>chunk</code></em> of given <em class="parameter"><code>size</code></em> with
@@ -1042,7 +906,7 @@ typedef struct {
         success.  If the function returns true, this indicates opt-out from
         deallocation; the virtual memory mapping associated with the chunk
         remains mapped, in the same commit state, and available for future use,
-        in which case it will be automatically retained for later reuse.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef bool <b class="fsfunc">(chunk_commit_t)</b>(</code></td><td>void *<var class="pdparam">chunk</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">offset</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">length</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p></p></div><p>A chunk commit function conforms to the
+        in which case it will be automatically retained for later reuse.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef bool <b class="fsfunc">(chunk_commit_t)</b>(</code></td><td>void *<var class="pdparam">chunk</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">offset</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">length</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p/></div><p>A chunk commit function conforms to the
         <span class="type">chunk_commit_t</span> type and commits zeroed physical memory to
         back pages within a <em class="parameter"><code>chunk</code></em> of given
         <em class="parameter"><code>size</code></em> at <em class="parameter"><code>offset</code></em> bytes,
@@ -1052,7 +916,7 @@ typedef struct {
         does not overcommit, or in implicit terms as on a system that
         overcommits and satisfies physical memory needs on demand via soft page
         faults. If the function returns true, this indicates insufficient
-        physical memory to satisfy the request.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef bool <b class="fsfunc">(chunk_decommit_t)</b>(</code></td><td>void *<var class="pdparam">chunk</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">offset</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">length</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p></p></div><p>A chunk decommit function conforms to the
+        physical memory to satisfy the request.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef bool <b class="fsfunc">(chunk_decommit_t)</b>(</code></td><td>void *<var class="pdparam">chunk</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">offset</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">length</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p/></div><p>A chunk decommit function conforms to the
         <span class="type">chunk_decommit_t</span> type and decommits any physical memory
         that is backing pages within a <em class="parameter"><code>chunk</code></em> of given
         <em class="parameter"><code>size</code></em> at <em class="parameter"><code>offset</code></em> bytes,
@@ -1061,14 +925,14 @@ typedef struct {
         case the pages will be committed via the chunk commit function before
         being reused.  If the function returns true, this indicates opt-out from
         decommit; the memory remains committed and available for future use, in
-        which case it will be automatically retained for later reuse.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef bool <b class="fsfunc">(chunk_purge_t)</b>(</code></td><td>void *<var class="pdparam">chunk</var>, </td></tr><tr><td> </td><td>size_t<var class="pdparam">size</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">offset</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">length</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p></p></div><p>A chunk purge function conforms to the <span class="type">chunk_purge_t</span>
+        which case it will be automatically retained for later reuse.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef bool <b class="fsfunc">(chunk_purge_t)</b>(</code></td><td>void *<var class="pdparam">chunk</var>, </td></tr><tr><td> </td><td>size_t<var class="pdparam">size</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">offset</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">length</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p/></div><p>A chunk purge function conforms to the <span class="type">chunk_purge_t</span>
         type and optionally discards physical pages within the virtual memory
         mapping associated with <em class="parameter"><code>chunk</code></em> of given
         <em class="parameter"><code>size</code></em> at <em class="parameter"><code>offset</code></em> bytes,
         extending for <em class="parameter"><code>length</code></em> on behalf of arena
         <em class="parameter"><code>arena_ind</code></em>, returning false if pages within the
         purged virtual memory range will be zero-filled the next time they are
-        accessed.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef bool <b class="fsfunc">(chunk_split_t)</b>(</code></td><td>void *<var class="pdparam">chunk</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size_a</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size_b</var>, </td></tr><tr><td> </td><td>bool <var class="pdparam">committed</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p></p></div><p>A chunk split function conforms to the <span class="type">chunk_split_t</span>
+        accessed.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef bool <b class="fsfunc">(chunk_split_t)</b>(</code></td><td>void *<var class="pdparam">chunk</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size_a</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size_b</var>, </td></tr><tr><td> </td><td>bool <var class="pdparam">committed</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p/></div><p>A chunk split function conforms to the <span class="type">chunk_split_t</span>
         type and optionally splits <em class="parameter"><code>chunk</code></em> of given
         <em class="parameter"><code>size</code></em> into two adjacent chunks, the first of
         <em class="parameter"><code>size_a</code></em> bytes, and the second of
@@ -1077,7 +941,7 @@ typedef struct {
         behalf of arena <em class="parameter"><code>arena_ind</code></em>, returning false upon
         success.  If the function returns true, this indicates that the chunk
         remains unsplit and therefore should continue to be operated on as a
-        whole.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef bool <b class="fsfunc">(chunk_merge_t)</b>(</code></td><td>void *<var class="pdparam">chunk_a</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size_a</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">chunk_b</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size_b</var>, </td></tr><tr><td> </td><td>bool <var class="pdparam">committed</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p></p></div><p>A chunk merge function conforms to the <span class="type">chunk_merge_t</span>
+        whole.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef bool <b class="fsfunc">(chunk_merge_t)</b>(</code></td><td>void *<var class="pdparam">chunk_a</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size_a</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">chunk_b</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size_b</var>, </td></tr><tr><td> </td><td>bool <var class="pdparam">committed</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p/></div><p>A chunk merge function conforms to the <span class="type">chunk_merge_t</span>
         type and optionally merges adjacent chunks,
         <em class="parameter"><code>chunk_a</code></em> of given <em class="parameter"><code>size_a</code></em>
         and <em class="parameter"><code>chunk_b</code></em> of given
@@ -1086,150 +950,106 @@ typedef struct {
         behalf of arena <em class="parameter"><code>arena_ind</code></em>, returning false upon
         success.  If the function returns true, this indicates that the chunks
         remain distinct mappings and therefore should continue to be operated on
-        independently.</p></dd><dt><a name="arenas.narenas"></a><span class="term">
-          
-    "<code class="mallctl">arenas.narenas</code>"
-  
+        independently.</p></dd><dt><a name="arenas.narenas"/><span class="term">
+          <quote><code class="mallctl">arenas.narenas</code></quote>
           (<span class="type">unsigned</span>)
           <code class="literal">r-</code>
-        </span></dt><dd><p>Current limit on number of arenas.</p></dd><dt><a name="arenas.initialized"></a><span class="term">
-          
-    "<code class="mallctl">arenas.initialized</code>"
-  
+        </span></dt><dd><p>Current limit on number of arenas.</p></dd><dt><a name="arenas.initialized"/><span class="term">
+          <quote><code class="mallctl">arenas.initialized</code></quote>
           (<span class="type">bool *</span>)
           <code class="literal">r-</code>
-        </span></dt><dd><p>An array of <a class="link" href="#arenas.narenas">
-    "<code class="mallctl">arenas.narenas</code>"
-  </a>
+        </span></dt><dd><p>An array of <a class="link" href="#arenas.narenas"><quote><code class="mallctl">arenas.narenas</code></quote></a>
         booleans.  Each boolean indicates whether the corresponding arena is
-        initialized.</p></dd><dt><a name="arenas.lg_dirty_mult"></a><span class="term">
-          
-    "<code class="mallctl">arenas.lg_dirty_mult</code>"
-  
+        initialized.</p></dd><dt><a name="arenas.lg_dirty_mult"/><span class="term">
+          <quote><code class="mallctl">arenas.lg_dirty_mult</code></quote>
           (<span class="type">ssize_t</span>)
           <code class="literal">rw</code>
         </span></dt><dd><p>Current default per-arena minimum ratio (log base 2) of
-        active to dirty pages, used to initialize <a class="link" href="#arena.i.lg_dirty_mult">
-    "<code class="mallctl">arena.&lt;i&gt;.lg_dirty_mult</code>"
-  </a>
-        during arena creation.  See <a class="link" href="#opt.lg_dirty_mult">
-    "<code class="mallctl">opt.lg_dirty_mult</code>"
-  </a>
-        for additional information.</p></dd><dt><a name="arenas.quantum"></a><span class="term">
-          
-    "<code class="mallctl">arenas.quantum</code>"
-  
+        active to dirty pages, used to initialize <a class="link" href="#arena.i.lg_dirty_mult"><quote><code class="mallctl">arena.&lt;i&gt;.lg_dirty_mult</code></quote></a>
+        during arena creation.  See <a class="link" href="#opt.lg_dirty_mult"><quote><code class="mallctl">opt.lg_dirty_mult</code></quote></a>
+        for additional information.</p></dd><dt><a name="arenas.decay_time"/><span class="term">
+          <quote><code class="mallctl">arenas.decay_time</code></quote>
+          (<span class="type">ssize_t</span>)
+          <code class="literal">rw</code>
+        </span></dt><dd><p>Current default per-arena approximate time in seconds
+        from the creation of a set of unused dirty pages until an equivalent set
+        of unused dirty pages is purged and/or reused, used to initialize <a class="link" href="#arena.i.decay_time"><quote><code class="mallctl">arena.&lt;i&gt;.decay_time</code></quote></a>
+        during arena creation.  See <a class="link" href="#opt.decay_time"><quote><code class="mallctl">opt.decay_time</code></quote></a> for
+        additional information.</p></dd><dt><a name="arenas.quantum"/><span class="term">
+          <quote><code class="mallctl">arenas.quantum</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
-        </span></dt><dd><p>Quantum size.</p></dd><dt><a name="arenas.page"></a><span class="term">
-          
-    "<code class="mallctl">arenas.page</code>"
-  
+        </span></dt><dd><p>Quantum size.</p></dd><dt><a name="arenas.page"/><span class="term">
+          <quote><code class="mallctl">arenas.page</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
-        </span></dt><dd><p>Page size.</p></dd><dt><a name="arenas.tcache_max"></a><span class="term">
-          
-    "<code class="mallctl">arenas.tcache_max</code>"
-  
+        </span></dt><dd><p>Page size.</p></dd><dt><a name="arenas.tcache_max"/><span class="term">
+          <quote><code class="mallctl">arenas.tcache_max</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-tcache</code>]
-        </span></dt><dd><p>Maximum thread-cached size class.</p></dd><dt><a name="arenas.nbins"></a><span class="term">
-          
-    "<code class="mallctl">arenas.nbins</code>"
-  
+        </span></dt><dd><p>Maximum thread-cached size class.</p></dd><dt><a name="arenas.nbins"/><span class="term">
+          <quote><code class="mallctl">arenas.nbins</code></quote>
           (<span class="type">unsigned</span>)
           <code class="literal">r-</code>
-        </span></dt><dd><p>Number of bin size classes.</p></dd><dt><a name="arenas.nhbins"></a><span class="term">
-          
-    "<code class="mallctl">arenas.nhbins</code>"
-  
+        </span></dt><dd><p>Number of bin size classes.</p></dd><dt><a name="arenas.nhbins"/><span class="term">
+          <quote><code class="mallctl">arenas.nhbins</code></quote>
           (<span class="type">unsigned</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-tcache</code>]
         </span></dt><dd><p>Total number of thread cache bin size
-        classes.</p></dd><dt><a name="arenas.bin.i.size"></a><span class="term">
-          
-    "<code class="mallctl">arenas.bin.&lt;i&gt;.size</code>"
-  
+        classes.</p></dd><dt><a name="arenas.bin.i.size"/><span class="term">
+          <quote><code class="mallctl">arenas.bin.&lt;i&gt;.size</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
-        </span></dt><dd><p>Maximum size supported by size class.</p></dd><dt><a name="arenas.bin.i.nregs"></a><span class="term">
-          
-    "<code class="mallctl">arenas.bin.&lt;i&gt;.nregs</code>"
-  
+        </span></dt><dd><p>Maximum size supported by size class.</p></dd><dt><a name="arenas.bin.i.nregs"/><span class="term">
+          <quote><code class="mallctl">arenas.bin.&lt;i&gt;.nregs</code></quote>
           (<span class="type">uint32_t</span>)
           <code class="literal">r-</code>
-        </span></dt><dd><p>Number of regions per page run.</p></dd><dt><a name="arenas.bin.i.run_size"></a><span class="term">
-          
-    "<code class="mallctl">arenas.bin.&lt;i&gt;.run_size</code>"
-  
+        </span></dt><dd><p>Number of regions per page run.</p></dd><dt><a name="arenas.bin.i.run_size"/><span class="term">
+          <quote><code class="mallctl">arenas.bin.&lt;i&gt;.run_size</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
-        </span></dt><dd><p>Number of bytes per page run.</p></dd><dt><a name="arenas.nlruns"></a><span class="term">
-          
-    "<code class="mallctl">arenas.nlruns</code>"
-  
+        </span></dt><dd><p>Number of bytes per page run.</p></dd><dt><a name="arenas.nlruns"/><span class="term">
+          <quote><code class="mallctl">arenas.nlruns</code></quote>
           (<span class="type">unsigned</span>)
           <code class="literal">r-</code>
-        </span></dt><dd><p>Total number of large size classes.</p></dd><dt><a name="arenas.lrun.i.size"></a><span class="term">
-          
-    "<code class="mallctl">arenas.lrun.&lt;i&gt;.size</code>"
-  
+        </span></dt><dd><p>Total number of large size classes.</p></dd><dt><a name="arenas.lrun.i.size"/><span class="term">
+          <quote><code class="mallctl">arenas.lrun.&lt;i&gt;.size</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p>Maximum size supported by this large size
-        class.</p></dd><dt><a name="arenas.nhchunks"></a><span class="term">
-          
-    "<code class="mallctl">arenas.nhchunks</code>"
-  
+        class.</p></dd><dt><a name="arenas.nhchunks"/><span class="term">
+          <quote><code class="mallctl">arenas.nhchunks</code></quote>
           (<span class="type">unsigned</span>)
           <code class="literal">r-</code>
-        </span></dt><dd><p>Total number of huge size classes.</p></dd><dt><a name="arenas.hchunk.i.size"></a><span class="term">
-          
-    "<code class="mallctl">arenas.hchunk.&lt;i&gt;.size</code>"
-  
+        </span></dt><dd><p>Total number of huge size classes.</p></dd><dt><a name="arenas.hchunk.i.size"/><span class="term">
+          <quote><code class="mallctl">arenas.hchunk.&lt;i&gt;.size</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p>Maximum size supported by this huge size
-        class.</p></dd><dt><a name="arenas.extend"></a><span class="term">
-          
-    "<code class="mallctl">arenas.extend</code>"
-  
+        class.</p></dd><dt><a name="arenas.extend"/><span class="term">
+          <quote><code class="mallctl">arenas.extend</code></quote>
           (<span class="type">unsigned</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p>Extend the array of arenas by appending a new arena,
-        and returning the new arena index.</p></dd><dt><a name="prof.thread_active_init"></a><span class="term">
-          
-    "<code class="mallctl">prof.thread_active_init</code>"
-  
+        and returning the new arena index.</p></dd><dt><a name="prof.thread_active_init"/><span class="term">
+          <quote><code class="mallctl">prof.thread_active_init</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">rw</code>
           [<code class="option">--enable-prof</code>]
-        </span></dt><dd><p>Control the initial setting for <a class="link" href="#thread.prof.active">
-    "<code class="mallctl">thread.prof.active</code>"
-  </a>
-        in newly created threads.  See the <a class="link" href="#opt.prof_thread_active_init">
-    "<code class="mallctl">opt.prof_thread_active_init</code>"
-  </a>
-        option for additional information.</p></dd><dt><a name="prof.active"></a><span class="term">
-          
-    "<code class="mallctl">prof.active</code>"
-  
+        </span></dt><dd><p>Control the initial setting for <a class="link" href="#thread.prof.active"><quote><code class="mallctl">thread.prof.active</code></quote></a>
+        in newly created threads.  See the <a class="link" href="#opt.prof_thread_active_init"><quote><code class="mallctl">opt.prof_thread_active_init</code></quote></a>
+        option for additional information.</p></dd><dt><a name="prof.active"/><span class="term">
+          <quote><code class="mallctl">prof.active</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">rw</code>
           [<code class="option">--enable-prof</code>]
         </span></dt><dd><p>Control whether sampling is currently active.  See the
-        <a class="link" href="#opt.prof_active">
-    "<code class="mallctl">opt.prof_active</code>"
-  </a>
-        option for additional information, as well as the interrelated <a class="link" href="#thread.prof.active">
-    "<code class="mallctl">thread.prof.active</code>"
-  </a>
-        mallctl.</p></dd><dt><a name="prof.dump"></a><span class="term">
-          
-    "<code class="mallctl">prof.dump</code>"
-  
+        <a class="link" href="#opt.prof_active"><quote><code class="mallctl">opt.prof_active</code></quote></a>
+        option for additional information, as well as the interrelated <a class="link" href="#thread.prof.active"><quote><code class="mallctl">thread.prof.active</code></quote></a>
+        mallctl.</p></dd><dt><a name="prof.dump"/><span class="term">
+          <quote><code class="mallctl">prof.dump</code></quote>
           (<span class="type">const char *</span>)
           <code class="literal">-w</code>
           [<code class="option">--enable-prof</code>]
@@ -1237,13 +1057,9 @@ typedef struct {
         is specified, to a file according to the pattern
         <code class="filename">&lt;prefix&gt;.&lt;pid&gt;.&lt;seq&gt;.m&lt;mseq&gt;.heap</code>,
         where <code class="literal">&lt;prefix&gt;</code> is controlled by the
-        <a class="link" href="#opt.prof_prefix">
-    "<code class="mallctl">opt.prof_prefix</code>"
-  </a>
-        option.</p></dd><dt><a name="prof.gdump"></a><span class="term">
-          
-    "<code class="mallctl">prof.gdump</code>"
-  
+        <a class="link" href="#opt.prof_prefix"><quote><code class="mallctl">opt.prof_prefix</code></quote></a>
+        option.</p></dd><dt><a name="prof.gdump"/><span class="term">
+          <quote><code class="mallctl">prof.gdump</code></quote>
           (<span class="type">bool</span>)
           <code class="literal">rw</code>
           [<code class="option">--enable-prof</code>]
@@ -1251,103 +1067,67 @@ typedef struct {
         the total virtual memory exceeds the previous maximum.  Profiles are
         dumped to files named according to the pattern
         <code class="filename">&lt;prefix&gt;.&lt;pid&gt;.&lt;seq&gt;.u&lt;useq&gt;.heap</code>,
-        where <code class="literal">&lt;prefix&gt;</code> is controlled by the <a class="link" href="#opt.prof_prefix">
-    "<code class="mallctl">opt.prof_prefix</code>"
-  </a>
-        option.</p></dd><dt><a name="prof.reset"></a><span class="term">
-          
-    "<code class="mallctl">prof.reset</code>"
-  
+        where <code class="literal">&lt;prefix&gt;</code> is controlled by the <a class="link" href="#opt.prof_prefix"><quote><code class="mallctl">opt.prof_prefix</code></quote></a>
+        option.</p></dd><dt><a name="prof.reset"/><span class="term">
+          <quote><code class="mallctl">prof.reset</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">-w</code>
           [<code class="option">--enable-prof</code>]
         </span></dt><dd><p>Reset all memory profile statistics, and optionally
-        update the sample rate (see <a class="link" href="#opt.lg_prof_sample">
-    "<code class="mallctl">opt.lg_prof_sample</code>"
-  </a>
-        and <a class="link" href="#prof.lg_sample">
-    "<code class="mallctl">prof.lg_sample</code>"
-  </a>).
-        </p></dd><dt><a name="prof.lg_sample"></a><span class="term">
-          
-    "<code class="mallctl">prof.lg_sample</code>"
-  
+        update the sample rate (see <a class="link" href="#opt.lg_prof_sample"><quote><code class="mallctl">opt.lg_prof_sample</code></quote></a>
+        and <a class="link" href="#prof.lg_sample"><quote><code class="mallctl">prof.lg_sample</code></quote></a>).
+        </p></dd><dt><a name="prof.lg_sample"/><span class="term">
+          <quote><code class="mallctl">prof.lg_sample</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-prof</code>]
-        </span></dt><dd><p>Get the current sample rate (see <a class="link" href="#opt.lg_prof_sample">
-    "<code class="mallctl">opt.lg_prof_sample</code>"
-  </a>).
-        </p></dd><dt><a name="prof.interval"></a><span class="term">
-          
-    "<code class="mallctl">prof.interval</code>"
-  
+        </span></dt><dd><p>Get the current sample rate (see <a class="link" href="#opt.lg_prof_sample"><quote><code class="mallctl">opt.lg_prof_sample</code></quote></a>).
+        </p></dd><dt><a name="prof.interval"/><span class="term">
+          <quote><code class="mallctl">prof.interval</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-prof</code>]
         </span></dt><dd><p>Average number of bytes allocated between
-        inverval-based profile dumps.  See the
-        <a class="link" href="#opt.lg_prof_interval">
-    "<code class="mallctl">opt.lg_prof_interval</code>"
-  </a>
-        option for additional information.</p></dd><dt><a name="stats.cactive"></a><span class="term">
-          
-    "<code class="mallctl">stats.cactive</code>"
-  
+        interval-based profile dumps.  See the
+        <a class="link" href="#opt.lg_prof_interval"><quote><code class="mallctl">opt.lg_prof_interval</code></quote></a>
+        option for additional information.</p></dd><dt><a name="stats.cactive"/><span class="term">
+          <quote><code class="mallctl">stats.cactive</code></quote>
           (<span class="type">size_t *</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Pointer to a counter that contains an approximate count
         of the current number of bytes in active pages.  The estimate may be
         high, but never low, because each arena rounds up when computing its
-        contribution to the counter.  Note that the <a class="link" href="#epoch">
-    "<code class="mallctl">epoch</code>"
-  </a> mallctl has no bearing
+        contribution to the counter.  Note that the <a class="link" href="#epoch"><quote><code class="mallctl">epoch</code></quote></a> mallctl has no bearing
         on this counter.  Furthermore, counter consistency is maintained via
         atomic operations, so it is necessary to use an atomic operation in
         order to guarantee a consistent read when dereferencing the pointer.
-        </p></dd><dt><a name="stats.allocated"></a><span class="term">
-          
-    "<code class="mallctl">stats.allocated</code>"
-  
+        </p></dd><dt><a name="stats.allocated"/><span class="term">
+          <quote><code class="mallctl">stats.allocated</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Total number of bytes allocated by the
-        application.</p></dd><dt><a name="stats.active"></a><span class="term">
-          
-    "<code class="mallctl">stats.active</code>"
-  
+        application.</p></dd><dt><a name="stats.active"/><span class="term">
+          <quote><code class="mallctl">stats.active</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Total number of bytes in active pages allocated by the
         application.  This is a multiple of the page size, and greater than or
-        equal to <a class="link" href="#stats.allocated">
-    "<code class="mallctl">stats.allocated</code>"
-  </a>.
+        equal to <a class="link" href="#stats.allocated"><quote><code class="mallctl">stats.allocated</code></quote></a>.
         This does not include <a class="link" href="#stats.arenas.i.pdirty">
-        
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.pdirty</code>"
-  </a>, nor pages
-        entirely devoted to allocator metadata.</p></dd><dt><a name="stats.metadata"></a><span class="term">
-          
-    "<code class="mallctl">stats.metadata</code>"
-  
+        <quote><code class="mallctl">stats.arenas.&lt;i&gt;.pdirty</code></quote></a>, nor pages
+        entirely devoted to allocator metadata.</p></dd><dt><a name="stats.metadata"/><span class="term">
+          <quote><code class="mallctl">stats.metadata</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Total number of bytes dedicated to metadata, which
         comprise base allocations used for bootstrap-sensitive internal
-        allocator data structures, arena chunk headers (see <a class="link" href="#stats.arenas.i.metadata.mapped">
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.metadata.mapped</code>"
-  </a>),
-        and internal allocations (see <a class="link" href="#stats.arenas.i.metadata.allocated">
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.metadata.allocated</code>"
-  </a>).</p></dd><dt><a name="stats.resident"></a><span class="term">
-          
-    "<code class="mallctl">stats.resident</code>"
-  
+        allocator data structures, arena chunk headers (see <a class="link" href="#stats.arenas.i.metadata.mapped"><quote><code class="mallctl">stats.arenas.&lt;i&gt;.metadata.mapped</code></quote></a>),
+        and internal allocations (see <a class="link" href="#stats.arenas.i.metadata.allocated"><quote><code class="mallctl">stats.arenas.&lt;i&gt;.metadata.allocated</code></quote></a>).</p></dd><dt><a name="stats.resident"/><span class="term">
+          <quote><code class="mallctl">stats.resident</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
@@ -1357,436 +1137,393 @@ typedef struct {
         pages.  This is a maximum rather than precise because pages may not
         actually be physically resident if they correspond to demand-zeroed
         virtual memory that has not yet been touched.  This is a multiple of the
-        page size, and is larger than <a class="link" href="#stats.active">
-    "<code class="mallctl">stats.active</code>"
-  </a>.</p></dd><dt><a name="stats.mapped"></a><span class="term">
-          
-    "<code class="mallctl">stats.mapped</code>"
-  
+        page size, and is larger than <a class="link" href="#stats.active"><quote><code class="mallctl">stats.active</code></quote></a>.</p></dd><dt><a name="stats.mapped"/><span class="term">
+          <quote><code class="mallctl">stats.mapped</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Total number of bytes in active chunks mapped by the
         allocator.  This is a multiple of the chunk size, and is larger than
-        <a class="link" href="#stats.active">
-    "<code class="mallctl">stats.active</code>"
-  </a>.
+        <a class="link" href="#stats.active"><quote><code class="mallctl">stats.active</code></quote></a>.
         This does not include inactive chunks, even those that contain unused
         dirty pages, which means that there is no strict ordering between this
-        and <a class="link" href="#stats.resident">
-    "<code class="mallctl">stats.resident</code>"
-  </a>.</p></dd><dt><a name="stats.arenas.i.dss"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.dss</code>"
-  
+        and <a class="link" href="#stats.resident"><quote><code class="mallctl">stats.resident</code></quote></a>.</p></dd><dt><a name="stats.retained"/><span class="term">
+          <quote><code class="mallctl">stats.retained</code></quote>
+          (<span class="type">size_t</span>)
+          <code class="literal">r-</code>
+          [<code class="option">--enable-stats</code>]
+        </span></dt><dd><p>Total number of bytes in virtual memory mappings that
+        were retained rather than being returned to the operating system via
+        e.g. <span class="citerefentry"><span class="refentrytitle">munmap</span>(2)</span>.  Retained virtual memory is
+        typically untouched, decommitted, or purged, so it has no strongly
+        associated physical memory (see <a class="link" href="#arena.i.chunk_hooks">chunk hooks</a> for details).  Retained
+        memory is excluded from mapped memory statistics, e.g. <a class="link" href="#stats.mapped"><quote><code class="mallctl">stats.mapped</code></quote></a>.
+        </p></dd><dt><a name="stats.arenas.i.dss"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.dss</code></quote>
           (<span class="type">const char *</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p>dss (<span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span>) allocation precedence as
-        related to <span class="citerefentry"><span class="refentrytitle">mmap</span>(2)</span> allocation.  See <a class="link" href="#opt.dss">
-    "<code class="mallctl">opt.dss</code>"
-  </a> for details.
-        </p></dd><dt><a name="stats.arenas.i.lg_dirty_mult"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.lg_dirty_mult</code>"
-  
+        related to <span class="citerefentry"><span class="refentrytitle">mmap</span>(2)</span> allocation.  See <a class="link" href="#opt.dss"><quote><code class="mallctl">opt.dss</code></quote></a> for details.
+        </p></dd><dt><a name="stats.arenas.i.lg_dirty_mult"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.lg_dirty_mult</code></quote>
           (<span class="type">ssize_t</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p>Minimum ratio (log base 2) of active to dirty pages.
-        See <a class="link" href="#opt.lg_dirty_mult">
-    "<code class="mallctl">opt.lg_dirty_mult</code>"
-  </a>
-        for details.</p></dd><dt><a name="stats.arenas.i.nthreads"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.nthreads</code>"
-  
+        See <a class="link" href="#opt.lg_dirty_mult"><quote><code class="mallctl">opt.lg_dirty_mult</code></quote></a>
+        for details.</p></dd><dt><a name="stats.arenas.i.decay_time"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.decay_time</code></quote>
+          (<span class="type">ssize_t</span>)
+          <code class="literal">r-</code>
+        </span></dt><dd><p>Approximate time in seconds from the creation of a set
+        of unused dirty pages until an equivalent set of unused dirty pages is
+        purged and/or reused.  See <a class="link" href="#opt.decay_time"><quote><code class="mallctl">opt.decay_time</code></quote></a>
+        for details.</p></dd><dt><a name="stats.arenas.i.nthreads"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.nthreads</code></quote>
           (<span class="type">unsigned</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p>Number of threads currently assigned to
-        arena.</p></dd><dt><a name="stats.arenas.i.pactive"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.pactive</code>"
-  
+        arena.</p></dd><dt><a name="stats.arenas.i.pactive"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.pactive</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
-        </span></dt><dd><p>Number of pages in active runs.</p></dd><dt><a name="stats.arenas.i.pdirty"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.pdirty</code>"
-  
+        </span></dt><dd><p>Number of pages in active runs.</p></dd><dt><a name="stats.arenas.i.pdirty"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.pdirty</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
         </span></dt><dd><p>Number of pages within unused runs that are potentially
-        dirty, and for which <code class="function">madvise</code>(<em class="parameter"><code>...</code></em>, 
-        <em class="parameter"><code><code class="constant">MADV_DONTNEED</code></code></em>) or
-        similar has not been called.</p></dd><dt><a name="stats.arenas.i.mapped"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.mapped</code>"
-  
+        dirty, and for which <code class="function">madvise<em class="parameter"><code>...</code></em>
+        <em class="parameter"><code><code class="constant">MADV_DONTNEED</code></code></em></code> or
+        similar has not been called.</p></dd><dt><a name="stats.arenas.i.mapped"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.mapped</code></quote>
+          (<span class="type">size_t</span>)
+          <code class="literal">r-</code>
+          [<code class="option">--enable-stats</code>]
+        </span></dt><dd><p>Number of mapped bytes.</p></dd><dt><a name="stats.arenas.i.retained"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.retained</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
-        </span></dt><dd><p>Number of mapped bytes.</p></dd><dt><a name="stats.arenas.i.metadata.mapped"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.metadata.mapped</code>"
-  
+        </span></dt><dd><p>Number of retained bytes.  See <a class="link" href="#stats.retained"><quote><code class="mallctl">stats.retained</code></quote></a> for
+        details.</p></dd><dt><a name="stats.arenas.i.metadata.mapped"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.metadata.mapped</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Number of mapped bytes in arena chunk headers, which
-        track the states of the non-metadata pages.</p></dd><dt><a name="stats.arenas.i.metadata.allocated"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.metadata.allocated</code>"
-  
+        track the states of the non-metadata pages.</p></dd><dt><a name="stats.arenas.i.metadata.allocated"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.metadata.allocated</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Number of bytes dedicated to internal allocations.
         Internal allocations differ from application-originated allocations in
         that they are for internal use, and that they are omitted from heap
-        profiles.  This statistic is reported separately from <a class="link" href="#stats.metadata">
-    "<code class="mallctl">stats.metadata</code>"
-  </a> and
-        <a class="link" href="#stats.arenas.i.metadata.mapped">
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.metadata.mapped</code>"
-  </a>
-        because it overlaps with e.g. the <a class="link" href="#stats.allocated">
-    "<code class="mallctl">stats.allocated</code>"
-  </a> and
-        <a class="link" href="#stats.active">
-    "<code class="mallctl">stats.active</code>"
-  </a>
+        profiles.  This statistic is reported separately from <a class="link" href="#stats.metadata"><quote><code class="mallctl">stats.metadata</code></quote></a> and
+        <a class="link" href="#stats.arenas.i.metadata.mapped"><quote><code class="mallctl">stats.arenas.&lt;i&gt;.metadata.mapped</code></quote></a>
+        because it overlaps with e.g. the <a class="link" href="#stats.allocated"><quote><code class="mallctl">stats.allocated</code></quote></a> and
+        <a class="link" href="#stats.active"><quote><code class="mallctl">stats.active</code></quote></a>
         statistics, whereas the other metadata statistics do
-        not.</p></dd><dt><a name="stats.arenas.i.npurge"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.npurge</code>"
-  
+        not.</p></dd><dt><a name="stats.arenas.i.npurge"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.npurge</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Number of dirty page purge sweeps performed.
-        </p></dd><dt><a name="stats.arenas.i.nmadvise"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.nmadvise</code>"
-  
+        </p></dd><dt><a name="stats.arenas.i.nmadvise"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.nmadvise</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
-        </span></dt><dd><p>Number of <code class="function">madvise</code>(<em class="parameter"><code>...</code></em>, 
-        <em class="parameter"><code><code class="constant">MADV_DONTNEED</code></code></em>) or
-        similar calls made to purge dirty pages.</p></dd><dt><a name="stats.arenas.i.purged"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.purged</code>"
-  
+        </span></dt><dd><p>Number of <code class="function">madvise<em class="parameter"><code>...</code></em>
+        <em class="parameter"><code><code class="constant">MADV_DONTNEED</code></code></em></code> or
+        similar calls made to purge dirty pages.</p></dd><dt><a name="stats.arenas.i.purged"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.purged</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
-        </span></dt><dd><p>Number of pages purged.</p></dd><dt><a name="stats.arenas.i.small.allocated"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.small.allocated</code>"
-  
+        </span></dt><dd><p>Number of pages purged.</p></dd><dt><a name="stats.arenas.i.small.allocated"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.small.allocated</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Number of bytes currently allocated by small objects.
-        </p></dd><dt><a name="stats.arenas.i.small.nmalloc"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.small.nmalloc</code>"
-  
+        </p></dd><dt><a name="stats.arenas.i.small.nmalloc"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.small.nmalloc</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of allocation requests served by
-        small bins.</p></dd><dt><a name="stats.arenas.i.small.ndalloc"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.small.ndalloc</code>"
-  
+        small bins.</p></dd><dt><a name="stats.arenas.i.small.ndalloc"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.small.ndalloc</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of small objects returned to bins.
-        </p></dd><dt><a name="stats.arenas.i.small.nrequests"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.small.nrequests</code>"
-  
+        </p></dd><dt><a name="stats.arenas.i.small.nrequests"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.small.nrequests</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of small allocation requests.
-        </p></dd><dt><a name="stats.arenas.i.large.allocated"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.large.allocated</code>"
-  
+        </p></dd><dt><a name="stats.arenas.i.large.allocated"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.large.allocated</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Number of bytes currently allocated by large objects.
-        </p></dd><dt><a name="stats.arenas.i.large.nmalloc"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.large.nmalloc</code>"
-  
+        </p></dd><dt><a name="stats.arenas.i.large.nmalloc"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.large.nmalloc</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of large allocation requests served
-        directly by the arena.</p></dd><dt><a name="stats.arenas.i.large.ndalloc"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.large.ndalloc</code>"
-  
+        directly by the arena.</p></dd><dt><a name="stats.arenas.i.large.ndalloc"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.large.ndalloc</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of large deallocation requests served
-        directly by the arena.</p></dd><dt><a name="stats.arenas.i.large.nrequests"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.large.nrequests</code>"
-  
+        directly by the arena.</p></dd><dt><a name="stats.arenas.i.large.nrequests"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.large.nrequests</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of large allocation requests.
-        </p></dd><dt><a name="stats.arenas.i.huge.allocated"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.huge.allocated</code>"
-  
+        </p></dd><dt><a name="stats.arenas.i.huge.allocated"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.huge.allocated</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Number of bytes currently allocated by huge objects.
-        </p></dd><dt><a name="stats.arenas.i.huge.nmalloc"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.huge.nmalloc</code>"
-  
+        </p></dd><dt><a name="stats.arenas.i.huge.nmalloc"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.huge.nmalloc</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of huge allocation requests served
-        directly by the arena.</p></dd><dt><a name="stats.arenas.i.huge.ndalloc"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.huge.ndalloc</code>"
-  
+        directly by the arena.</p></dd><dt><a name="stats.arenas.i.huge.ndalloc"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.huge.ndalloc</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of huge deallocation requests served
-        directly by the arena.</p></dd><dt><a name="stats.arenas.i.huge.nrequests"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.huge.nrequests</code>"
-  
+        directly by the arena.</p></dd><dt><a name="stats.arenas.i.huge.nrequests"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.huge.nrequests</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of huge allocation requests.
-        </p></dd><dt><a name="stats.arenas.i.bins.j.nmalloc"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nmalloc</code>"
-  
+        </p></dd><dt><a name="stats.arenas.i.bins.j.nmalloc"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nmalloc</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of allocations served by bin.
-        </p></dd><dt><a name="stats.arenas.i.bins.j.ndalloc"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.ndalloc</code>"
-  
+        </p></dd><dt><a name="stats.arenas.i.bins.j.ndalloc"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.ndalloc</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of allocations returned to bin.
-        </p></dd><dt><a name="stats.arenas.i.bins.j.nrequests"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nrequests</code>"
-  
+        </p></dd><dt><a name="stats.arenas.i.bins.j.nrequests"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nrequests</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of allocation
-        requests.</p></dd><dt><a name="stats.arenas.i.bins.j.curregs"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.curregs</code>"
-  
+        requests.</p></dd><dt><a name="stats.arenas.i.bins.j.curregs"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.curregs</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Current number of regions for this size
-        class.</p></dd><dt><a name="stats.arenas.i.bins.j.nfills"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nfills</code>"
-  
+        class.</p></dd><dt><a name="stats.arenas.i.bins.j.nfills"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nfills</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code> <code class="option">--enable-tcache</code>]
-        </span></dt><dd><p>Cumulative number of tcache fills.</p></dd><dt><a name="stats.arenas.i.bins.j.nflushes"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nflushes</code>"
-  
+        </span></dt><dd><p>Cumulative number of tcache fills.</p></dd><dt><a name="stats.arenas.i.bins.j.nflushes"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nflushes</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code> <code class="option">--enable-tcache</code>]
-        </span></dt><dd><p>Cumulative number of tcache flushes.</p></dd><dt><a name="stats.arenas.i.bins.j.nruns"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nruns</code>"
-  
+        </span></dt><dd><p>Cumulative number of tcache flushes.</p></dd><dt><a name="stats.arenas.i.bins.j.nruns"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nruns</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
-        </span></dt><dd><p>Cumulative number of runs created.</p></dd><dt><a name="stats.arenas.i.bins.j.nreruns"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nreruns</code>"
-  
+        </span></dt><dd><p>Cumulative number of runs created.</p></dd><dt><a name="stats.arenas.i.bins.j.nreruns"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nreruns</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of times the current run from which
-        to allocate changed.</p></dd><dt><a name="stats.arenas.i.bins.j.curruns"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.curruns</code>"
-  
+        to allocate changed.</p></dd><dt><a name="stats.arenas.i.bins.j.curruns"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.curruns</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
-        </span></dt><dd><p>Current number of runs.</p></dd><dt><a name="stats.arenas.i.lruns.j.nmalloc"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.lruns.&lt;j&gt;.nmalloc</code>"
-  
+        </span></dt><dd><p>Current number of runs.</p></dd><dt><a name="stats.arenas.i.lruns.j.nmalloc"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.lruns.&lt;j&gt;.nmalloc</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of allocation requests for this size
-        class served directly by the arena.</p></dd><dt><a name="stats.arenas.i.lruns.j.ndalloc"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.lruns.&lt;j&gt;.ndalloc</code>"
-  
+        class served directly by the arena.</p></dd><dt><a name="stats.arenas.i.lruns.j.ndalloc"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.lruns.&lt;j&gt;.ndalloc</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of deallocation requests for this
-        size class served directly by the arena.</p></dd><dt><a name="stats.arenas.i.lruns.j.nrequests"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.lruns.&lt;j&gt;.nrequests</code>"
-  
+        size class served directly by the arena.</p></dd><dt><a name="stats.arenas.i.lruns.j.nrequests"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.lruns.&lt;j&gt;.nrequests</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of allocation requests for this size
-        class.</p></dd><dt><a name="stats.arenas.i.lruns.j.curruns"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.lruns.&lt;j&gt;.curruns</code>"
-  
+        class.</p></dd><dt><a name="stats.arenas.i.lruns.j.curruns"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.lruns.&lt;j&gt;.curruns</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Current number of runs for this size class.
-        </p></dd><dt><a name="stats.arenas.i.hchunks.j.nmalloc"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.hchunks.&lt;j&gt;.nmalloc</code>"
-  
+        </p></dd><dt><a name="stats.arenas.i.hchunks.j.nmalloc"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.hchunks.&lt;j&gt;.nmalloc</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of allocation requests for this size
-        class served directly by the arena.</p></dd><dt><a name="stats.arenas.i.hchunks.j.ndalloc"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.hchunks.&lt;j&gt;.ndalloc</code>"
-  
+        class served directly by the arena.</p></dd><dt><a name="stats.arenas.i.hchunks.j.ndalloc"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.hchunks.&lt;j&gt;.ndalloc</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of deallocation requests for this
-        size class served directly by the arena.</p></dd><dt><a name="stats.arenas.i.hchunks.j.nrequests"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.hchunks.&lt;j&gt;.nrequests</code>"
-  
+        size class served directly by the arena.</p></dd><dt><a name="stats.arenas.i.hchunks.j.nrequests"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.hchunks.&lt;j&gt;.nrequests</code></quote>
           (<span class="type">uint64_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Cumulative number of allocation requests for this size
-        class.</p></dd><dt><a name="stats.arenas.i.hchunks.j.curhchunks"></a><span class="term">
-          
-    "<code class="mallctl">stats.arenas.&lt;i&gt;.hchunks.&lt;j&gt;.curhchunks</code>"
-  
+        class.</p></dd><dt><a name="stats.arenas.i.hchunks.j.curhchunks"/><span class="term">
+          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.hchunks.&lt;j&gt;.curhchunks</code></quote>
           (<span class="type">size_t</span>)
           <code class="literal">r-</code>
           [<code class="option">--enable-stats</code>]
         </span></dt><dd><p>Current number of huge allocations for this size class.
-        </p></dd></dl></div></div><div class="refsect1"><a name="debugging_malloc_problems"></a><h2>DEBUGGING MALLOC PROBLEMS</h2><p>When debugging, it is a good idea to configure/build jemalloc with
+        </p></dd></dl></div></div><div class="refsect1"><a name="heap_profile_format"/><h2>HEAP PROFILE FORMAT</h2><p>Although the heap profiling functionality was originally designed to
+    be compatible with the
+    <span class="command"><strong>pprof</strong></span> command that is developed as part of the <a class="ulink" href="http://code.google.com/p/gperftools/" target="_top">gperftools
+    package</a>, the addition of per thread heap profiling functionality
+    required a different heap profile format.  The <span class="command"><strong>jeprof</strong></span>
+    command is derived from <span class="command"><strong>pprof</strong></span>, with enhancements to
+    support the heap profile format described here.</p><p>In the following hypothetical heap profile, <code class="constant">[...]</code>
+    indicates elision for the sake of compactness.  </p><pre class="programlisting">
+heap_v2/524288
+  t*: 28106: 56637512 [0: 0]
+  [...]
+  t3: 352: 16777344 [0: 0]
+  [...]
+  t99: 17754: 29341640 [0: 0]
+  [...]
+@ 0x5f86da8 0x5f5a1dc [...] 0x29e4d4e 0xa200316 0xabb2988 [...]
+  t*: 13: 6688 [0: 0]
+  t3: 12: 6496 [0: ]
+  t99: 1: 192 [0: 0]
+[...]
+
+MAPPED_LIBRARIES:
+[...]</pre><p> The following matches the above heap profile, but most
+tokens are replaced with <code class="constant">&lt;description&gt;</code> to indicate
+descriptions of the corresponding fields.  </p><pre class="programlisting">
+&lt;heap_profile_format_version&gt;/&lt;mean_sample_interval&gt;
+  &lt;aggregate&gt;: &lt;curobjs&gt;: &lt;curbytes&gt; [&lt;cumobjs&gt;: &lt;cumbytes&gt;]
+  [...]
+  &lt;thread_3_aggregate&gt;: &lt;curobjs&gt;: &lt;curbytes&gt;[&lt;cumobjs&gt;: &lt;cumbytes&gt;]
+  [...]
+  &lt;thread_99_aggregate&gt;: &lt;curobjs&gt;: &lt;curbytes&gt;[&lt;cumobjs&gt;: &lt;cumbytes&gt;]
+  [...]
+@ &lt;top_frame&gt; &lt;frame&gt; [...] &lt;frame&gt; &lt;frame&gt; &lt;frame&gt; [...]
+  &lt;backtrace_aggregate&gt;: &lt;curobjs&gt;: &lt;curbytes&gt; [&lt;cumobjs&gt;: &lt;cumbytes&gt;]
+  &lt;backtrace_thread_3&gt;: &lt;curobjs&gt;: &lt;curbytes&gt; [&lt;cumobjs&gt;: &lt;cumbytes&gt;]
+  &lt;backtrace_thread_99&gt;: &lt;curobjs&gt;: &lt;curbytes&gt; [&lt;cumobjs&gt;: &lt;cumbytes&gt;]
+[...]
+
+MAPPED_LIBRARIES:
+&lt;/proc/&lt;pid&gt;/maps&gt;</pre></div><div class="refsect1"><a name="debugging_malloc_problems"/><h2>DEBUGGING MALLOC PROBLEMS</h2><p>When debugging, it is a good idea to configure/build jemalloc with
     the <code class="option">--enable-debug</code> and <code class="option">--enable-fill</code>
     options, and recompile the program with suitable options and symbols for
     debugger support.  When so configured, jemalloc incorporates a wide variety
     of run-time assertions that catch application errors such as double-free,
-    write-after-free, etc.</p><p>Programs often accidentally depend on &#8220;uninitialized&#8221;
+    write-after-free, etc.</p><p>Programs often accidentally depend on <span class="quote">“<span class="quote">uninitialized</span>�</span>
     memory actually being filled with zero bytes.  Junk filling
-    (see the <a class="link" href="#opt.junk">
-    "<code class="mallctl">opt.junk</code>"
-  </a>
+    (see the <a class="link" href="#opt.junk"><quote><code class="mallctl">opt.junk</code></quote></a>
     option) tends to expose such bugs in the form of obviously incorrect
     results and/or coredumps.  Conversely, zero
-    filling (see the <a class="link" href="#opt.zero">
-    "<code class="mallctl">opt.zero</code>"
-  </a> option) eliminates
+    filling (see the <a class="link" href="#opt.zero"><quote><code class="mallctl">opt.zero</code></quote></a> option) eliminates
     the symptoms of such bugs.  Between these two options, it is usually
     possible to quickly detect, diagnose, and eliminate such bugs.</p><p>This implementation does not provide much detail about the problems
     it detects, because the performance impact for storing such information
     would be prohibitive.  However, jemalloc does integrate with the most
     excellent <a class="ulink" href="http://valgrind.org/" target="_top">Valgrind</a> tool if the
-    <code class="option">--enable-valgrind</code> configuration option is enabled.</p></div><div class="refsect1"><a name="diagnostic_messages"></a><h2>DIAGNOSTIC MESSAGES</h2><p>If any of the memory allocation/deallocation functions detect an
+    <code class="option">--enable-valgrind</code> configuration option is enabled.</p></div><div class="refsect1"><a name="diagnostic_messages"/><h2>DIAGNOSTIC MESSAGES</h2><p>If any of the memory allocation/deallocation functions detect an
     error or warning condition, a message will be printed to file descriptor
     <code class="constant">STDERR_FILENO</code>.  Errors will result in the process
-    dumping core.  If the <a class="link" href="#opt.abort">
-    "<code class="mallctl">opt.abort</code>"
-  </a> option is set, most
+    dumping core.  If the <a class="link" href="#opt.abort"><quote><code class="mallctl">opt.abort</code></quote></a> option is set, most
     warnings are treated as errors.</p><p>The <code class="varname">malloc_message</code> variable allows the programmer
     to override the function which emits the text strings forming the errors
     and warnings if for some reason the <code class="constant">STDERR_FILENO</code> file
     descriptor is not suitable for this.
-    <code class="function">malloc_message</code>(<em class="parameter"><code></code></em>) takes the
+    <code class="function">malloc_message()</code> takes the
     <em class="parameter"><code>cbopaque</code></em> pointer argument that is
     <code class="constant">NULL</code> unless overridden by the arguments in a call to
-    <code class="function">malloc_stats_print</code>(<em class="parameter"><code></code></em>), followed by a string
+    <code class="function">malloc_stats_print()</code>, followed by a string
     pointer.  Please note that doing anything which tries to allocate memory in
     this function is likely to result in a crash or deadlock.</p><p>All messages are prefixed by
-    &#8220;<code class="computeroutput">&lt;jemalloc&gt;: </code>&#8221;.</p></div><div class="refsect1"><a name="return_values"></a><h2>RETURN VALUES</h2><div class="refsect2"><a name="idp46949776"></a><h3>Standard API</h3><p>The <code class="function">malloc</code>(<em class="parameter"><code></code></em>) and
-      <code class="function">calloc</code>(<em class="parameter"><code></code></em>) functions return a pointer to the
+    <span class="quote">“<span class="quote"><code class="computeroutput">&lt;jemalloc&gt;: </code></span>�</span>.</p></div><div class="refsect1"><a name="return_values"/><h2>RETURN VALUES</h2><div class="refsect2"><a name="idm45291897515152"/><h3>Standard API</h3><p>The <code class="function">malloc()</code> and
+      <code class="function">calloc()</code> functions return a pointer to the
       allocated memory if successful; otherwise a <code class="constant">NULL</code>
       pointer is returned and <code class="varname">errno</code> is set to
-      <span class="errorname">ENOMEM</span>.</p><p>The <code class="function">posix_memalign</code>(<em class="parameter"><code></code></em>) function
+      <span class="errorname">ENOMEM</span>.</p><p>The <code class="function">posix_memalign()</code> function
       returns the value 0 if successful; otherwise it returns an error value.
-      The <code class="function">posix_memalign</code>(<em class="parameter"><code></code></em>) function will fail
+      The <code class="function">posix_memalign()</code> function will fail
       if:
         </p><div class="variablelist"><dl class="variablelist"><dt><span class="term"><span class="errorname">EINVAL</span></span></dt><dd><p>The <em class="parameter"><code>alignment</code></em> parameter is
             not a power of 2 at least as large as
             <code class="code">sizeof(<span class="type">void *</span>)</code>.
             </p></dd><dt><span class="term"><span class="errorname">ENOMEM</span></span></dt><dd><p>Memory allocation error.</p></dd></dl></div><p>
-      </p><p>The <code class="function">aligned_alloc</code>(<em class="parameter"><code></code></em>) function returns
+      </p><p>The <code class="function">aligned_alloc()</code> function returns
       a pointer to the allocated memory if successful; otherwise a
       <code class="constant">NULL</code> pointer is returned and
       <code class="varname">errno</code> is set.  The
-      <code class="function">aligned_alloc</code>(<em class="parameter"><code></code></em>) function will fail if:
+      <code class="function">aligned_alloc()</code> function will fail if:
         </p><div class="variablelist"><dl class="variablelist"><dt><span class="term"><span class="errorname">EINVAL</span></span></dt><dd><p>The <em class="parameter"><code>alignment</code></em> parameter is
             not a power of 2.
             </p></dd><dt><span class="term"><span class="errorname">ENOMEM</span></span></dt><dd><p>Memory allocation error.</p></dd></dl></div><p>
-      </p><p>The <code class="function">realloc</code>(<em class="parameter"><code></code></em>) function returns a
+      </p><p>The <code class="function">realloc()</code> function returns a
       pointer, possibly identical to <em class="parameter"><code>ptr</code></em>, to the
       allocated memory if successful; otherwise a <code class="constant">NULL</code>
       pointer is returned, and <code class="varname">errno</code> is set to
       <span class="errorname">ENOMEM</span> if the error was the result of an
-      allocation failure.  The <code class="function">realloc</code>(<em class="parameter"><code></code></em>)
+      allocation failure.  The <code class="function">realloc()</code>
       function always leaves the original buffer intact when an error occurs.
-      </p><p>The <code class="function">free</code>(<em class="parameter"><code></code></em>) function returns no
-      value.</p></div><div class="refsect2"><a name="idp46974576"></a><h3>Non-standard API</h3><p>The <code class="function">mallocx</code>(<em class="parameter"><code></code></em>) and
-      <code class="function">rallocx</code>(<em class="parameter"><code></code></em>) functions return a pointer to
+      </p><p>The <code class="function">free()</code> function returns no
+      value.</p></div><div class="refsect2"><a name="idm45291897493664"/><h3>Non-standard API</h3><p>The <code class="function">mallocx()</code> and
+      <code class="function">rallocx()</code> functions return a pointer to
       the allocated memory if successful; otherwise a <code class="constant">NULL</code>
       pointer is returned to indicate insufficient contiguous memory was
-      available to service the allocation request.  </p><p>The <code class="function">xallocx</code>(<em class="parameter"><code></code></em>) function returns the
+      available to service the allocation request.  </p><p>The <code class="function">xallocx()</code> function returns the
       real size of the resulting resized allocation pointed to by
       <em class="parameter"><code>ptr</code></em>, which is a value less than
       <em class="parameter"><code>size</code></em> if the allocation could not be adequately
-      grown in place.  </p><p>The <code class="function">sallocx</code>(<em class="parameter"><code></code></em>) function returns the
+      grown in place.  </p><p>The <code class="function">sallocx()</code> function returns the
       real size of the allocation pointed to by <em class="parameter"><code>ptr</code></em>.
-      </p><p>The <code class="function">nallocx</code>(<em class="parameter"><code></code></em>) returns the real size
+      </p><p>The <code class="function">nallocx()</code> returns the real size
       that would result from a successful equivalent
-      <code class="function">mallocx</code>(<em class="parameter"><code></code></em>) function call, or zero if
-      insufficient memory is available to perform the size computation.  </p><p>The <code class="function">mallctl</code>(<em class="parameter"><code></code></em>),
-      <code class="function">mallctlnametomib</code>(<em class="parameter"><code></code></em>), and
-      <code class="function">mallctlbymib</code>(<em class="parameter"><code></code></em>) functions return 0 on
+      <code class="function">mallocx()</code> function call, or zero if
+      insufficient memory is available to perform the size computation.  </p><p>The <code class="function">mallctl()</code>,
+      <code class="function">mallctlnametomib()</code>, and
+      <code class="function">mallctlbymib()</code> functions return 0 on
       success; otherwise they return an error value.  The functions will fail
       if:
         </p><div class="variablelist"><dl class="variablelist"><dt><span class="term"><span class="errorname">EINVAL</span></span></dt><dd><p><em class="parameter"><code>newp</code></em> is not
@@ -1798,28 +1535,28 @@ typedef struct {
             value.</p></dd><dt><span class="term"><span class="errorname">EPERM</span></span></dt><dd><p>Attempt to read or write void value, or attempt to
             write read-only value.</p></dd><dt><span class="term"><span class="errorname">EAGAIN</span></span></dt><dd><p>A memory allocation failure
             occurred.</p></dd><dt><span class="term"><span class="errorname">EFAULT</span></span></dt><dd><p>An interface with side effects failed in some way
-            not directly related to <code class="function">mallctl*</code>(<em class="parameter"><code></code></em>)
+            not directly related to <code class="function">mallctl*()</code>
             read/write processing.</p></dd></dl></div><p>
-      </p><p>The <code class="function">malloc_usable_size</code>(<em class="parameter"><code></code></em>) function
+      </p><p>The <code class="function">malloc_usable_size()</code> function
       returns the usable size of the allocation pointed to by
-      <em class="parameter"><code>ptr</code></em>.  </p></div></div><div class="refsect1"><a name="environment"></a><h2>ENVIRONMENT</h2><p>The following environment variable affects the execution of the
+      <em class="parameter"><code>ptr</code></em>.  </p></div></div><div class="refsect1"><a name="environment"/><h2>ENVIRONMENT</h2><p>The following environment variable affects the execution of the
     allocation functions:
       </p><div class="variablelist"><dl class="variablelist"><dt><span class="term"><code class="envar">MALLOC_CONF</code></span></dt><dd><p>If the environment variable
           <code class="envar">MALLOC_CONF</code> is set, the characters it contains
           will be interpreted as options.</p></dd></dl></div><p>
-    </p></div><div class="refsect1"><a name="examples"></a><h2>EXAMPLES</h2><p>To dump core whenever a problem occurs:
+    </p></div><div class="refsect1"><a name="examples"/><h2>EXAMPLES</h2><p>To dump core whenever a problem occurs:
       </p><pre class="screen">ln -s 'abort:true' /etc/malloc.conf</pre><p>
     </p><p>To specify in the source a chunk size that is 16 MiB:
       </p><pre class="programlisting">
-malloc_conf = "lg_chunk:24";</pre></div><div class="refsect1"><a name="see_also"></a><h2>SEE ALSO</h2><p><span class="citerefentry"><span class="refentrytitle">madvise</span>(2)</span>,
+malloc_conf = "lg_chunk:24";</pre></div><div class="refsect1"><a name="see_also"/><h2>SEE ALSO</h2><p><span class="citerefentry"><span class="refentrytitle">madvise</span>(2)</span>,
     <span class="citerefentry"><span class="refentrytitle">mmap</span>(2)</span>,
     <span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span>,
     <span class="citerefentry"><span class="refentrytitle">utrace</span>(2)</span>,
     <span class="citerefentry"><span class="refentrytitle">alloca</span>(3)</span>,
     <span class="citerefentry"><span class="refentrytitle">atexit</span>(3)</span>,
-    <span class="citerefentry"><span class="refentrytitle">getpagesize</span>(3)</span></p></div><div class="refsect1"><a name="standards"></a><h2>STANDARDS</h2><p>The <code class="function">malloc</code>(<em class="parameter"><code></code></em>),
-    <code class="function">calloc</code>(<em class="parameter"><code></code></em>),
-    <code class="function">realloc</code>(<em class="parameter"><code></code></em>), and
-    <code class="function">free</code>(<em class="parameter"><code></code></em>) functions conform to ISO/IEC
-    9899:1990 (&#8220;ISO C90&#8221;).</p><p>The <code class="function">posix_memalign</code>(<em class="parameter"><code></code></em>) function conforms
-    to IEEE Std 1003.1-2001 (&#8220;POSIX.1&#8221;).</p></div></div></body></html>
+    <span class="citerefentry"><span class="refentrytitle">getpagesize</span>(3)</span></p></div><div class="refsect1"><a name="standards"/><h2>STANDARDS</h2><p>The <code class="function">malloc()</code>,
+    <code class="function">calloc()</code>,
+    <code class="function">realloc()</code>, and
+    <code class="function">free()</code> functions conform to ISO/IEC
+    9899:1990 (<span class="quote">“<span class="quote">ISO C90</span>�</span>).</p><p>The <code class="function">posix_memalign()</code> function conforms
+    to IEEE Std 1003.1-2001 (<span class="quote">“<span class="quote">POSIX.1</span>�</span>).</p></div></div></body></html>
\ No newline at end of file
diff --git a/deps/jemalloc/doc/jemalloc.xml.in b/deps/jemalloc/doc/jemalloc.xml.in
index 8fc774b18..d9c83452d 100644
--- a/deps/jemalloc/doc/jemalloc.xml.in
+++ b/deps/jemalloc/doc/jemalloc.xml.in
@@ -52,7 +52,7 @@
     <title>LIBRARY</title>
     <para>This manual describes jemalloc @jemalloc_version@.  More information
     can be found at the <ulink
-    url="http://www.canonware.com/jemalloc/">jemalloc website</ulink>.</para>
+    url="http://jemalloc.net/">jemalloc website</ulink>.</para>
   </refsect1>
   <refsynopsisdiv>
     <title>SYNOPSIS</title>
@@ -180,20 +180,20 @@
     <refsect2>
       <title>Standard API</title>
 
-      <para>The <function>malloc<parameter/></function> function allocates
+      <para>The <function>malloc()</function> function allocates
       <parameter>size</parameter> bytes of uninitialized memory.  The allocated
       space is suitably aligned (after possible pointer coercion) for storage
       of any type of object.</para>
 
-      <para>The <function>calloc<parameter/></function> function allocates
+      <para>The <function>calloc()</function> function allocates
       space for <parameter>number</parameter> objects, each
       <parameter>size</parameter> bytes in length.  The result is identical to
-      calling <function>malloc<parameter/></function> with an argument of
+      calling <function>malloc()</function> with an argument of
       <parameter>number</parameter> * <parameter>size</parameter>, with the
       exception that the allocated memory is explicitly initialized to zero
       bytes.</para>
 
-      <para>The <function>posix_memalign<parameter/></function> function
+      <para>The <function>posix_memalign()</function> function
       allocates <parameter>size</parameter> bytes of memory such that the
       allocation's base address is a multiple of
       <parameter>alignment</parameter>, and returns the allocation in the value
@@ -201,7 +201,7 @@
       <parameter>alignment</parameter> must be a power of 2 at least as large as
       <code language="C">sizeof(<type>void *</type>)</code>.</para>
 
-      <para>The <function>aligned_alloc<parameter/></function> function
+      <para>The <function>aligned_alloc()</function> function
       allocates <parameter>size</parameter> bytes of memory such that the
       allocation's base address is a multiple of
       <parameter>alignment</parameter>.  The requested
@@ -209,7 +209,7 @@
       undefined if <parameter>size</parameter> is not an integral multiple of
       <parameter>alignment</parameter>.</para>
 
-      <para>The <function>realloc<parameter/></function> function changes the
+      <para>The <function>realloc()</function> function changes the
       size of the previously allocated memory referenced by
       <parameter>ptr</parameter> to <parameter>size</parameter> bytes.  The
       contents of the memory are unchanged up to the lesser of the new and old
@@ -217,26 +217,26 @@
       portion of the memory are undefined.  Upon success, the memory referenced
       by <parameter>ptr</parameter> is freed and a pointer to the newly
       allocated memory is returned.  Note that
-      <function>realloc<parameter/></function> may move the memory allocation,
+      <function>realloc()</function> may move the memory allocation,
       resulting in a different return value than <parameter>ptr</parameter>.
       If <parameter>ptr</parameter> is <constant>NULL</constant>, the
-      <function>realloc<parameter/></function> function behaves identically to
-      <function>malloc<parameter/></function> for the specified size.</para>
+      <function>realloc()</function> function behaves identically to
+      <function>malloc()</function> for the specified size.</para>
 
-      <para>The <function>free<parameter/></function> function causes the
+      <para>The <function>free()</function> function causes the
       allocated memory referenced by <parameter>ptr</parameter> to be made
       available for future allocations.  If <parameter>ptr</parameter> is
       <constant>NULL</constant>, no action occurs.</para>
     </refsect2>
     <refsect2>
       <title>Non-standard API</title>
-      <para>The <function>mallocx<parameter/></function>,
-      <function>rallocx<parameter/></function>,
-      <function>xallocx<parameter/></function>,
-      <function>sallocx<parameter/></function>,
-      <function>dallocx<parameter/></function>,
-      <function>sdallocx<parameter/></function>, and
-      <function>nallocx<parameter/></function> functions all have a
+      <para>The <function>mallocx()</function>,
+      <function>rallocx()</function>,
+      <function>xallocx()</function>,
+      <function>sallocx()</function>,
+      <function>dallocx()</function>,
+      <function>sdallocx()</function>, and
+      <function>nallocx()</function> functions all have a
       <parameter>flags</parameter> argument that can be used to specify
       options.  The functions only check the options that are contextually
       relevant.  Use bitwise or (<code language="C">|</code>) operations to
@@ -307,21 +307,19 @@
         </variablelist>
       </para>
 
-      <para>The <function>mallocx<parameter/></function> function allocates at
+      <para>The <function>mallocx()</function> function allocates at
       least <parameter>size</parameter> bytes of memory, and returns a pointer
       to the base address of the allocation.  Behavior is undefined if
-      <parameter>size</parameter> is <constant>0</constant>, or if request size
-      overflows due to size class and/or alignment constraints.</para>
+      <parameter>size</parameter> is <constant>0</constant>.</para>
 
-      <para>The <function>rallocx<parameter/></function> function resizes the
+      <para>The <function>rallocx()</function> function resizes the
       allocation at <parameter>ptr</parameter> to be at least
       <parameter>size</parameter> bytes, and returns a pointer to the base
       address of the resulting allocation, which may or may not have moved from
       its original location.  Behavior is undefined if
-      <parameter>size</parameter> is <constant>0</constant>, or if request size
-      overflows due to size class and/or alignment constraints.</para>
+      <parameter>size</parameter> is <constant>0</constant>.</para>
 
-      <para>The <function>xallocx<parameter/></function> function resizes the
+      <para>The <function>xallocx()</function> function resizes the
       allocation at <parameter>ptr</parameter> in place to be at least
       <parameter>size</parameter> bytes, and returns the real size of the
       allocation.  If <parameter>extra</parameter> is non-zero, an attempt is
@@ -334,32 +332,32 @@
       language="C">(<parameter>size</parameter> + <parameter>extra</parameter>
       &gt; <constant>SIZE_T_MAX</constant>)</code>.</para>
 
-      <para>The <function>sallocx<parameter/></function> function returns the
+      <para>The <function>sallocx()</function> function returns the
       real size of the allocation at <parameter>ptr</parameter>.</para>
 
-      <para>The <function>dallocx<parameter/></function> function causes the
+      <para>The <function>dallocx()</function> function causes the
       memory referenced by <parameter>ptr</parameter> to be made available for
       future allocations.</para>
 
-      <para>The <function>sdallocx<parameter/></function> function is an
-      extension of <function>dallocx<parameter/></function> with a
+      <para>The <function>sdallocx()</function> function is an
+      extension of <function>dallocx()</function> with a
       <parameter>size</parameter> parameter to allow the caller to pass in the
       allocation size as an optimization.  The minimum valid input size is the
       original requested size of the allocation, and the maximum valid input
       size is the corresponding value returned by
-      <function>nallocx<parameter/></function> or
-      <function>sallocx<parameter/></function>.</para>
+      <function>nallocx()</function> or
+      <function>sallocx()</function>.</para>
 
-      <para>The <function>nallocx<parameter/></function> function allocates no
+      <para>The <function>nallocx()</function> function allocates no
       memory, but it performs the same size computation as the
-      <function>mallocx<parameter/></function> function, and returns the real
+      <function>mallocx()</function> function, and returns the real
       size of the allocation that would result from the equivalent
-      <function>mallocx<parameter/></function> function call.  Behavior is
-      undefined if <parameter>size</parameter> is <constant>0</constant>, or if
-      request size overflows due to size class and/or alignment
-      constraints.</para>
+      <function>mallocx()</function> function call, or
+      <constant>0</constant> if the inputs exceed the maximum supported size
+      class and/or alignment.  Behavior is undefined if
+      <parameter>size</parameter> is <constant>0</constant>.</para>
 
-      <para>The <function>mallctl<parameter/></function> function provides a
+      <para>The <function>mallctl()</function> function provides a
       general interface for introspecting the memory allocator, as well as
       setting modifiable parameters and triggering actions.  The
       period-separated <parameter>name</parameter> argument specifies a
@@ -374,12 +372,12 @@
       <parameter>newlen</parameter>; otherwise pass <constant>NULL</constant>
       and <constant>0</constant>.</para>
 
-      <para>The <function>mallctlnametomib<parameter/></function> function
+      <para>The <function>mallctlnametomib()</function> function
       provides a way to avoid repeated name lookups for applications that
       repeatedly query the same portion of the namespace, by translating a name
-      to a &ldquo;Management Information Base&rdquo; (MIB) that can be passed
-      repeatedly to <function>mallctlbymib<parameter/></function>.  Upon
-      successful return from <function>mallctlnametomib<parameter/></function>,
+      to a <quote>Management Information Base</quote> (MIB) that can be passed
+      repeatedly to <function>mallctlbymib()</function>.  Upon
+      successful return from <function>mallctlnametomib()</function>,
       <parameter>mibp</parameter> contains an array of
       <parameter>*miblenp</parameter> integers, where
       <parameter>*miblenp</parameter> is the lesser of the number of components
@@ -408,43 +406,44 @@ for (i = 0; i < nbins; i++) {
 
 	mib[2] = i;
 	len = sizeof(bin_size);
-	mallctlbymib(mib, miblen, &bin_size, &len, NULL, 0);
+	mallctlbymib(mib, miblen, (void *)&bin_size, &len, NULL, 0);
 	/* Do something with bin_size... */
 }]]></programlisting></para>
 
-      <para>The <function>malloc_stats_print<parameter/></function> function
-      writes human-readable summary statistics via the
-      <parameter>write_cb</parameter> callback function pointer and
-      <parameter>cbopaque</parameter> data passed to
-      <parameter>write_cb</parameter>, or
-      <function>malloc_message<parameter/></function> if
-      <parameter>write_cb</parameter> is <constant>NULL</constant>.  This
-      function can be called repeatedly.  General information that never
-      changes during execution can be omitted by specifying "g" as a character
+      <para>The <function>malloc_stats_print()</function> function writes
+      summary statistics via the <parameter>write_cb</parameter> callback
+      function pointer and <parameter>cbopaque</parameter> data passed to
+      <parameter>write_cb</parameter>, or <function>malloc_message()</function>
+      if <parameter>write_cb</parameter> is <constant>NULL</constant>.  The
+      statistics are presented in human-readable form unless <quote>J</quote> is
+      specified as a character within the <parameter>opts</parameter> string, in
+      which case the statistics are presented in <ulink
+      url="http://www.json.org/">JSON format</ulink>.  This function can be
+      called repeatedly.  General information that never changes during
+      execution can be omitted by specifying <quote>g</quote> as a character
       within the <parameter>opts</parameter> string.  Note that
-      <function>malloc_message<parameter/></function> uses the
-      <function>mallctl*<parameter/></function> functions internally, so
-      inconsistent statistics can be reported if multiple threads use these
-      functions simultaneously.  If <option>--enable-stats</option> is
-      specified during configuration, &ldquo;m&rdquo; and &ldquo;a&rdquo; can
-      be specified to omit merged arena and per arena statistics, respectively;
-      &ldquo;b&rdquo;, &ldquo;l&rdquo;, and &ldquo;h&rdquo; can be specified to
-      omit per size class statistics for bins, large objects, and huge objects,
-      respectively.  Unrecognized characters are silently ignored.  Note that
-      thread caching may prevent some statistics from being completely up to
-      date, since extra locking would be required to merge counters that track
-      thread cache operations.
-      </para>
-
-      <para>The <function>malloc_usable_size<parameter/></function> function
+      <function>malloc_message()</function> uses the
+      <function>mallctl*()</function> functions internally, so inconsistent
+      statistics can be reported if multiple threads use these functions
+      simultaneously.  If <option>--enable-stats</option> is specified during
+      configuration, <quote>m</quote> and <quote>a</quote> can be specified to
+      omit merged arena and per arena statistics, respectively;
+      <quote>b</quote>, <quote>l</quote>, and <quote>h</quote> can be specified
+      to omit per size class statistics for bins, large objects, and huge
+      objects, respectively.  Unrecognized characters are silently ignored.
+      Note that thread caching may prevent some statistics from being completely
+      up to date, since extra locking would be required to merge counters that
+      track thread cache operations.</para>
+
+      <para>The <function>malloc_usable_size()</function> function
       returns the usable size of the allocation pointed to by
       <parameter>ptr</parameter>.  The return value may be larger than the size
       that was requested during allocation.  The
-      <function>malloc_usable_size<parameter/></function> function is not a
-      mechanism for in-place <function>realloc<parameter/></function>; rather
+      <function>malloc_usable_size()</function> function is not a
+      mechanism for in-place <function>realloc()</function>; rather
       it is provided solely as a tool for introspection purposes.  Any
       discrepancy between the requested allocation size and the size reported
-      by <function>malloc_usable_size<parameter/></function> should not be
+      by <function>malloc_usable_size()</function> should not be
       depended on, since such behavior is entirely implementation-dependent.
       </para>
     </refsect2>
@@ -455,19 +454,20 @@ for (i = 0; i < nbins; i++) {
     routines, the allocator initializes its internals based in part on various
     options that can be specified at compile- or run-time.</para>
 
-    <para>The string pointed to by the global variable
-    <varname>malloc_conf</varname>, the &ldquo;name&rdquo; of the file
-    referenced by the symbolic link named <filename
-    class="symlink">/etc/malloc.conf</filename>, and the value of the
+    <para>The string specified via <option>--with-malloc-conf</option>, the
+    string pointed to by the global variable <varname>malloc_conf</varname>, the
+    <quote>name</quote> of the file referenced by the symbolic link named
+    <filename class="symlink">/etc/malloc.conf</filename>, and the value of the
     environment variable <envar>MALLOC_CONF</envar>, will be interpreted, in
     that order, from left to right as options.  Note that
     <varname>malloc_conf</varname> may be read before
-    <function>main<parameter/></function> is entered, so the declaration of
+    <function>main()</function> is entered, so the declaration of
     <varname>malloc_conf</varname> should specify an initializer that contains
-    the final value to be read by jemalloc.  <varname>malloc_conf</varname> is
-    a compile-time setting, whereas <filename
-    class="symlink">/etc/malloc.conf</filename> and <envar>MALLOC_CONF</envar>
-    can be safely set any time prior to program invocation.</para>
+    the final value to be read by jemalloc.  <option>--with-malloc-conf</option>
+    and <varname>malloc_conf</varname> are compile-time mechanisms, whereas
+    <filename class="symlink">/etc/malloc.conf</filename> and
+    <envar>MALLOC_CONF</envar> can be safely set any time prior to program
+    invocation.</para>
 
     <para>An options string is a comma-separated list of option:value pairs.
     There is one key corresponding to each <link
@@ -517,23 +517,18 @@ for (i = 0; i < nbins; i++) {
     common case, but it increases memory usage and fragmentation, since a
     bounded number of objects can remain allocated in each thread cache.</para>
 
-    <para>Memory is conceptually broken into equal-sized chunks, where the
-    chunk size is a power of two that is greater than the page size.  Chunks
-    are always aligned to multiples of the chunk size.  This alignment makes it
-    possible to find metadata for user objects very quickly.</para>
-
-    <para>User objects are broken into three categories according to size:
-    small, large, and huge.  Small and large objects are managed entirely by
-    arenas; huge objects are additionally aggregated in a single data structure
-    that is shared by all threads.  Huge objects are typically used by
-    applications infrequently enough that this single data structure is not a
-    scalability issue.</para>
-
-    <para>Each chunk that is managed by an arena tracks its contents as runs of
+    <para>Memory is conceptually broken into equal-sized chunks, where the chunk
+    size is a power of two that is greater than the page size.  Chunks are
+    always aligned to multiples of the chunk size.  This alignment makes it
+    possible to find metadata for user objects very quickly.  User objects are
+    broken into three categories according to size: small, large, and huge.
+    Multiple small and large objects can reside within a single chunk, whereas
+    huge objects each have one or more chunks backing them.  Each chunk that
+    contains small and/or large objects tracks its contents as runs of
     contiguous pages (unused, backing a set of small objects, or backing one
-    large object).  The combination of chunk alignment and chunk page maps
-    makes it possible to determine all metadata regarding small and large
-    allocations in constant time.</para>
+    large object).  The combination of chunk alignment and chunk page maps makes
+    it possible to determine all metadata regarding small and large allocations
+    in constant time.</para>
 
     <para>Small objects are managed in groups by page runs.  Each run maintains
     a bitmap to track which regions are in use.  Allocation requests that are no
@@ -546,8 +541,8 @@ for (i = 0; i < nbins; i++) {
     are smaller than four times the page size, large size classes are smaller
     than the chunk size (see the <link
     linkend="opt.lg_chunk"><mallctl>opt.lg_chunk</mallctl></link> option), and
-    huge size classes extend from the chunk size up to one size class less than
-    the full address space size.</para>
+    huge size classes extend from the chunk size up to the largest size class
+    that does not exceed <constant>PTRDIFF_MAX</constant>.</para>
 
     <para>Allocations are packed tightly together, which can be an issue for
     multi-threaded applications.  If you need to assure that allocations do not
@@ -555,14 +550,14 @@ for (i = 0; i < nbins; i++) {
     nearest multiple of the cacheline size, or specify cacheline alignment when
     allocating.</para>
 
-    <para>The <function>realloc<parameter/></function>,
-    <function>rallocx<parameter/></function>, and
-    <function>xallocx<parameter/></function> functions may resize allocations
+    <para>The <function>realloc()</function>,
+    <function>rallocx()</function>, and
+    <function>xallocx()</function> functions may resize allocations
     without moving them under limited circumstances.  Unlike the
-    <function>*allocx<parameter/></function> API, the standard API does not
+    <function>*allocx()</function> API, the standard API does not
     officially round up the usable size of an allocation to the nearest size
     class, so technically it is necessary to call
-    <function>realloc<parameter/></function> to grow e.g. a 9-byte allocation to
+    <function>realloc()</function> to grow e.g. a 9-byte allocation to
     16 bytes, or shrink a 16-byte allocation to 9 bytes.  Growth and shrinkage
     trivially succeeds in place as long as the pre-size and post-size both round
     up to the same size class.  No other API guarantees are made regarding
@@ -665,7 +660,7 @@ for (i = 0; i < nbins; i++) {
           <entry>[1280 KiB, 1536 KiB, 1792 KiB]</entry>
         </row>
         <row>
-          <entry morerows="6">Huge</entry>
+          <entry morerows="8">Huge</entry>
           <entry>256 KiB</entry>
           <entry>[2 MiB]</entry>
         </row>
@@ -693,6 +688,14 @@ for (i = 0; i < nbins; i++) {
           <entry>...</entry>
           <entry>...</entry>
         </row>
+        <row>
+          <entry>512 PiB</entry>
+          <entry>[2560 PiB, 3 EiB, 3584 PiB, 4 EiB]</entry>
+        </row>
+        <row>
+          <entry>1 EiB</entry>
+          <entry>[5 EiB, 6 EiB, 7 EiB]</entry>
+        </row>
       </tbody>
       </tgroup>
     </table>
@@ -700,7 +703,7 @@ for (i = 0; i < nbins; i++) {
   <refsect1 id="mallctl_namespace">
     <title>MALLCTL NAMESPACE</title>
     <para>The following names are defined in the namespace accessible via the
-    <function>mallctl*<parameter/></function> functions.  Value types are
+    <function>mallctl*()</function> functions.  Value types are
     specified in parentheses, their readable/writable statuses are encoded as
     <literal>rw</literal>, <literal>r-</literal>, <literal>-w</literal>, or
     <literal>--</literal>, and required build configuration flags follow, if
@@ -731,7 +734,7 @@ for (i = 0; i < nbins; i++) {
           <literal>rw</literal>
         </term>
         <listitem><para>If a value is passed in, refresh the data from which
-        the <function>mallctl*<parameter/></function> functions report values,
+        the <function>mallctl*()</function> functions report values,
         and increment the epoch.  Return the current epoch.  This is useful for
         detecting whether another thread caused a refresh.</para></listitem>
       </varlistentry>
@@ -776,6 +779,17 @@ for (i = 0; i < nbins; i++) {
         during build configuration.</para></listitem>
       </varlistentry>
 
+      <varlistentry id="config.malloc_conf">
+        <term>
+          <mallctl>config.malloc_conf</mallctl>
+          (<type>const char *</type>)
+          <literal>r-</literal>
+        </term>
+        <listitem><para>Embedded configure-time-specified run-time options
+        string, empty unless <option>--with-malloc-conf</option> was specified
+        during build configuration.</para></listitem>
+      </varlistentry>
+
       <varlistentry id="config.munmap">
         <term>
           <mallctl>config.munmap</mallctl>
@@ -904,12 +918,12 @@ for (i = 0; i < nbins; i++) {
         settings are supported if
         <citerefentry><refentrytitle>sbrk</refentrytitle>
         <manvolnum>2</manvolnum></citerefentry> is supported by the operating
-        system: &ldquo;disabled&rdquo;, &ldquo;primary&rdquo;, and
-        &ldquo;secondary&rdquo;; otherwise only &ldquo;disabled&rdquo; is
-        supported.  The default is &ldquo;secondary&rdquo; if
+        system: <quote>disabled</quote>, <quote>primary</quote>, and
+        <quote>secondary</quote>; otherwise only <quote>disabled</quote> is
+        supported.  The default is <quote>secondary</quote> if
         <citerefentry><refentrytitle>sbrk</refentrytitle>
         <manvolnum>2</manvolnum></citerefentry> is supported by the operating
-        system; &ldquo;disabled&rdquo; otherwise.
+        system; <quote>disabled</quote> otherwise.
         </para></listitem>
       </varlistentry>
 
@@ -929,7 +943,7 @@ for (i = 0; i < nbins; i++) {
       <varlistentry id="opt.narenas">
         <term>
           <mallctl>opt.narenas</mallctl>
-          (<type>size_t</type>)
+          (<type>unsigned</type>)
           <literal>r-</literal>
         </term>
         <listitem><para>Maximum number of arenas to use for automatic
@@ -937,6 +951,20 @@ for (i = 0; i < nbins; i++) {
         number of CPUs, or one if there is a single CPU.</para></listitem>
       </varlistentry>
 
+      <varlistentry id="opt.purge">
+        <term>
+          <mallctl>opt.purge</mallctl>
+          (<type>const char *</type>)
+          <literal>r-</literal>
+        </term>
+        <listitem><para>Purge mode is &ldquo;ratio&rdquo; (default) or
+        &ldquo;decay&rdquo;.  See <link
+        linkend="opt.lg_dirty_mult"><mallctl>opt.lg_dirty_mult</mallctl></link>
+        for details of the ratio mode.  See <link
+        linkend="opt.decay_time"><mallctl>opt.decay_time</mallctl></link> for
+        details of the decay mode.</para></listitem>
+      </varlistentry>
+
       <varlistentry id="opt.lg_dirty_mult">
         <term>
           <mallctl>opt.lg_dirty_mult</mallctl>
@@ -959,6 +987,26 @@ for (i = 0; i < nbins; i++) {
         for related dynamic control options.</para></listitem>
       </varlistentry>
 
+      <varlistentry id="opt.decay_time">
+        <term>
+          <mallctl>opt.decay_time</mallctl>
+          (<type>ssize_t</type>)
+          <literal>r-</literal>
+        </term>
+        <listitem><para>Approximate time in seconds from the creation of a set
+        of unused dirty pages until an equivalent set of unused dirty pages is
+        purged and/or reused.  The pages are incrementally purged according to a
+        sigmoidal decay curve that starts and ends with zero purge rate.  A
+        decay time of 0 causes all unused dirty pages to be purged immediately
+        upon creation.  A decay time of -1 disables purging.  The default decay
+        time is 10 seconds.  See <link
+        linkend="arenas.decay_time"><mallctl>arenas.decay_time</mallctl></link>
+        and <link
+        linkend="arena.i.decay_time"><mallctl>arena.&lt;i&gt;.decay_time</mallctl></link>
+        for related dynamic control options.
+        </para></listitem>
+      </varlistentry>
+
       <varlistentry id="opt.stats_print">
         <term>
           <mallctl>opt.stats_print</mallctl>
@@ -966,19 +1014,19 @@ for (i = 0; i < nbins; i++) {
           <literal>r-</literal>
         </term>
         <listitem><para>Enable/disable statistics printing at exit.  If
-        enabled, the <function>malloc_stats_print<parameter/></function>
+        enabled, the <function>malloc_stats_print()</function>
         function is called at program exit via an
         <citerefentry><refentrytitle>atexit</refentrytitle>
         <manvolnum>3</manvolnum></citerefentry> function.  If
         <option>--enable-stats</option> is specified during configuration, this
         has the potential to cause deadlock for a multi-threaded process that
         exits while one or more threads are executing in the memory allocation
-        functions.  Furthermore, <function>atexit<parameter/></function> may
+        functions.  Furthermore, <function>atexit()</function> may
         allocate memory during application initialization and then deadlock
         internally when jemalloc in turn calls
-        <function>atexit<parameter/></function>, so this option is not
-        univerally usable (though the application can register its own
-        <function>atexit<parameter/></function> function with equivalent
+        <function>atexit()</function>, so this option is not
+        universally usable (though the application can register its own
+        <function>atexit()</function> function with equivalent
         functionality).  Therefore, this option should only be used with care;
         it is primarily intended as a performance tuning aid during application
         development.  This option is disabled by default.</para></listitem>
@@ -991,15 +1039,16 @@ for (i = 0; i < nbins; i++) {
           <literal>r-</literal>
           [<option>--enable-fill</option>]
         </term>
-        <listitem><para>Junk filling.  If set to "alloc", each byte of
-        uninitialized allocated memory will be initialized to
-        <literal>0xa5</literal>.  If set to "free", all deallocated memory will
-        be initialized to <literal>0x5a</literal>.  If set to "true", both
-        allocated and deallocated memory will be initialized, and if set to
-        "false", junk filling be disabled entirely.  This is intended for
-        debugging and will impact performance negatively.  This option is
-        "false" by default unless <option>--enable-debug</option> is specified
-        during configuration, in which case it is "true" by default unless
+        <listitem><para>Junk filling.  If set to <quote>alloc</quote>, each byte
+        of uninitialized allocated memory will be initialized to
+        <literal>0xa5</literal>.  If set to <quote>free</quote>, all deallocated
+        memory will be initialized to <literal>0x5a</literal>.  If set to
+        <quote>true</quote>, both allocated and deallocated memory will be
+        initialized, and if set to <quote>false</quote>, junk filling be
+        disabled entirely.  This is intended for debugging and will impact
+        performance negatively.  This option is <quote>false</quote> by default
+        unless <option>--enable-debug</option> is specified during
+        configuration, in which case it is <quote>true</quote> by default unless
         running inside <ulink
         url="http://valgrind.org/">Valgrind</ulink>.</para></listitem>
       </varlistentry>
@@ -1054,8 +1103,8 @@ for (i = 0; i < nbins; i++) {
         <listitem><para>Zero filling enabled/disabled.  If enabled, each byte
         of uninitialized allocated memory will be initialized to 0.  Note that
         this initialization only happens once for each byte, so
-        <function>realloc<parameter/></function> and
-        <function>rallocx<parameter/></function> calls do not zero memory that
+        <function>realloc()</function> and
+        <function>rallocx()</function> calls do not zero memory that
         was previously allocated.  This is intended for debugging and will
         impact performance negatively.  This option is disabled by default.
         </para></listitem>
@@ -1150,7 +1199,8 @@ malloc_conf = "xmalloc:true";]]></programlisting>
         the <command>jeprof</command> command, which is based on the
         <command>pprof</command> that is developed as part of the <ulink
         url="http://code.google.com/p/gperftools/">gperftools
-        package</ulink>.</para></listitem>
+        package</ulink>.  See <link linkend="heap_profile_format">HEAP PROFILE
+        FORMAT</link> for heap profile format documentation.</para></listitem>
       </varlistentry>
 
       <varlistentry id="opt.prof_prefix">
@@ -1277,11 +1327,11 @@ malloc_conf = "xmalloc:true";]]></programlisting>
         <filename>&lt;prefix&gt;.&lt;pid&gt;.&lt;seq&gt;.f.heap</filename>,
         where <literal>&lt;prefix&gt;</literal> is controlled by the <link
         linkend="opt.prof_prefix"><mallctl>opt.prof_prefix</mallctl></link>
-        option.  Note that <function>atexit<parameter/></function> may allocate
+        option.  Note that <function>atexit()</function> may allocate
         memory during application initialization and then deadlock internally
-        when jemalloc in turn calls <function>atexit<parameter/></function>, so
-        this option is not univerally usable (though the application can
-        register its own <function>atexit<parameter/></function> function with
+        when jemalloc in turn calls <function>atexit()</function>, so
+        this option is not universally usable (though the application can
+        register its own <function>atexit()</function> function with
         equivalent functionality).  This option is disabled by
         default.</para></listitem>
       </varlistentry>
@@ -1340,7 +1390,7 @@ malloc_conf = "xmalloc:true";]]></programlisting>
         <link
         linkend="thread.allocated"><mallctl>thread.allocated</mallctl></link>
         mallctl.  This is useful for avoiding the overhead of repeated
-        <function>mallctl*<parameter/></function> calls.</para></listitem>
+        <function>mallctl*()</function> calls.</para></listitem>
       </varlistentry>
 
       <varlistentry id="thread.deallocated">
@@ -1367,7 +1417,7 @@ malloc_conf = "xmalloc:true";]]></programlisting>
         <link
         linkend="thread.deallocated"><mallctl>thread.deallocated</mallctl></link>
         mallctl.  This is useful for avoiding the overhead of repeated
-        <function>mallctl*<parameter/></function> calls.</para></listitem>
+        <function>mallctl*()</function> calls.</para></listitem>
       </varlistentry>
 
       <varlistentry id="thread.tcache.enabled">
@@ -1418,8 +1468,8 @@ malloc_conf = "xmalloc:true";]]></programlisting>
         can cause asynchronous string deallocation.  Furthermore, each
         invocation of this interface can only read or write; simultaneous
         read/write is not supported due to string lifetime limitations.  The
-        name string must nil-terminated and comprised only of characters in the
-        sets recognized
+        name string must be nil-terminated and comprised only of characters in
+        the sets recognized
         by <citerefentry><refentrytitle>isgraph</refentrytitle>
         <manvolnum>3</manvolnum></citerefentry> and
         <citerefentry><refentrytitle>isblank</refentrytitle>
@@ -1467,7 +1517,7 @@ malloc_conf = "xmalloc:true";]]></programlisting>
         <listitem><para>Flush the specified thread-specific cache (tcache).  The
         same considerations apply to this interface as to <link
         linkend="thread.tcache.flush"><mallctl>thread.tcache.flush</mallctl></link>,
-        except that the tcache will never be automatically be discarded.
+        except that the tcache will never be automatically discarded.
         </para></listitem>
       </varlistentry>
 
@@ -1489,12 +1539,44 @@ malloc_conf = "xmalloc:true";]]></programlisting>
           (<type>void</type>)
           <literal>--</literal>
         </term>
-        <listitem><para>Purge unused dirty pages for arena &lt;i&gt;, or for
+        <listitem><para>Purge all unused dirty pages for arena &lt;i&gt;, or for
         all arenas if &lt;i&gt; equals <link
         linkend="arenas.narenas"><mallctl>arenas.narenas</mallctl></link>.
         </para></listitem>
       </varlistentry>
 
+      <varlistentry id="arena.i.decay">
+        <term>
+          <mallctl>arena.&lt;i&gt;.decay</mallctl>
+          (<type>void</type>)
+          <literal>--</literal>
+        </term>
+        <listitem><para>Trigger decay-based purging of unused dirty pages for
+        arena &lt;i&gt;, or for all arenas if &lt;i&gt; equals <link
+        linkend="arenas.narenas"><mallctl>arenas.narenas</mallctl></link>.
+        The proportion of unused dirty pages to be purged depends on the current
+        time; see <link
+        linkend="opt.decay_time"><mallctl>opt.decay_time</mallctl></link> for
+        details.</para></listitem>
+      </varlistentry>
+
+      <varlistentry id="arena.i.reset">
+        <term>
+          <mallctl>arena.&lt;i&gt;.reset</mallctl>
+          (<type>void</type>)
+          <literal>--</literal>
+        </term>
+        <listitem><para>Discard all of the arena's extant allocations.  This
+        interface can only be used with arenas created via <link
+        linkend="arenas.extend"><mallctl>arenas.extend</mallctl></link>.  None
+        of the arena's discarded/cached allocations may accessed afterward.  As
+        part of this requirement, all thread caches which were used to
+        allocate/deallocate in conjunction with the arena must be flushed
+        beforehand.  This interface cannot be used if running inside Valgrind,
+        nor if the <link linkend="opt.quarantine">quarantine</link> size is
+        non-zero.</para></listitem>
+      </varlistentry>
+
       <varlistentry id="arena.i.dss">
         <term>
           <mallctl>arena.&lt;i&gt;.dss</mallctl>
@@ -1523,6 +1605,22 @@ malloc_conf = "xmalloc:true";]]></programlisting>
         for additional information.</para></listitem>
       </varlistentry>
 
+      <varlistentry id="arena.i.decay_time">
+        <term>
+          <mallctl>arena.&lt;i&gt;.decay_time</mallctl>
+          (<type>ssize_t</type>)
+          <literal>rw</literal>
+        </term>
+        <listitem><para>Current per-arena approximate time in seconds from the
+        creation of a set of unused dirty pages until an equivalent set of
+        unused dirty pages is purged and/or reused.  Each time this interface is
+        set, all currently unused dirty pages are considered to have fully
+        decayed, which causes immediate purging of all unused dirty pages unless
+        the decay time is set to -1 (i.e. purging disabled).  See <link
+        linkend="opt.decay_time"><mallctl>opt.decay_time</mallctl></link> for
+        additional information.</para></listitem>
+      </varlistentry>
+
       <varlistentry id="arena.i.chunk_hooks">
         <term>
           <mallctl>arena.&lt;i&gt;.chunk_hooks</mallctl>
@@ -1757,6 +1855,21 @@ typedef struct {
         for additional information.</para></listitem>
       </varlistentry>
 
+      <varlistentry id="arenas.decay_time">
+        <term>
+          <mallctl>arenas.decay_time</mallctl>
+          (<type>ssize_t</type>)
+          <literal>rw</literal>
+        </term>
+        <listitem><para>Current default per-arena approximate time in seconds
+        from the creation of a set of unused dirty pages until an equivalent set
+        of unused dirty pages is purged and/or reused, used to initialize <link
+        linkend="arena.i.decay_time"><mallctl>arena.&lt;i&gt;.decay_time</mallctl></link>
+        during arena creation.  See <link
+        linkend="opt.decay_time"><mallctl>opt.decay_time</mallctl></link> for
+        additional information.</para></listitem>
+      </varlistentry>
+
       <varlistentry id="arenas.quantum">
         <term>
           <mallctl>arenas.quantum</mallctl>
@@ -1976,7 +2089,7 @@ typedef struct {
           [<option>--enable-prof</option>]
         </term>
         <listitem><para>Average number of bytes allocated between
-        inverval-based profile dumps.  See the
+        interval-based profile dumps.  See the
         <link
         linkend="opt.lg_prof_interval"><mallctl>opt.lg_prof_interval</mallctl></link>
         option for additional information.</para></listitem>
@@ -2075,6 +2188,25 @@ typedef struct {
         linkend="stats.resident"><mallctl>stats.resident</mallctl></link>.</para></listitem>
       </varlistentry>
 
+      <varlistentry id="stats.retained">
+        <term>
+          <mallctl>stats.retained</mallctl>
+          (<type>size_t</type>)
+          <literal>r-</literal>
+          [<option>--enable-stats</option>]
+        </term>
+        <listitem><para>Total number of bytes in virtual memory mappings that
+        were retained rather than being returned to the operating system via
+        e.g. <citerefentry><refentrytitle>munmap</refentrytitle>
+        <manvolnum>2</manvolnum></citerefentry>.  Retained virtual memory is
+        typically untouched, decommitted, or purged, so it has no strongly
+        associated physical memory (see <link
+        linkend="arena.i.chunk_hooks">chunk hooks</link> for details).  Retained
+        memory is excluded from mapped memory statistics, e.g. <link
+        linkend="stats.mapped"><mallctl>stats.mapped</mallctl></link>.
+        </para></listitem>
+      </varlistentry>
+
       <varlistentry id="stats.arenas.i.dss">
         <term>
           <mallctl>stats.arenas.&lt;i&gt;.dss</mallctl>
@@ -2101,6 +2233,19 @@ typedef struct {
         for details.</para></listitem>
       </varlistentry>
 
+      <varlistentry id="stats.arenas.i.decay_time">
+        <term>
+          <mallctl>stats.arenas.&lt;i&gt;.decay_time</mallctl>
+          (<type>ssize_t</type>)
+          <literal>r-</literal>
+        </term>
+        <listitem><para>Approximate time in seconds from the creation of a set
+        of unused dirty pages until an equivalent set of unused dirty pages is
+        purged and/or reused.  See <link
+        linkend="opt.decay_time"><mallctl>opt.decay_time</mallctl></link>
+        for details.</para></listitem>
+      </varlistentry>
+
       <varlistentry id="stats.arenas.i.nthreads">
         <term>
           <mallctl>stats.arenas.&lt;i&gt;.nthreads</mallctl>
@@ -2142,6 +2287,18 @@ typedef struct {
         <listitem><para>Number of mapped bytes.</para></listitem>
       </varlistentry>
 
+      <varlistentry id="stats.arenas.i.retained">
+        <term>
+          <mallctl>stats.arenas.&lt;i&gt;.retained</mallctl>
+          (<type>size_t</type>)
+          <literal>r-</literal>
+          [<option>--enable-stats</option>]
+        </term>
+        <listitem><para>Number of retained bytes.  See <link
+        linkend="stats.retained"><mallctl>stats.retained</mallctl></link> for
+        details.</para></listitem>
+      </varlistentry>
+
       <varlistentry id="stats.arenas.i.metadata.mapped">
         <term>
           <mallctl>stats.arenas.&lt;i&gt;.metadata.mapped</mallctl>
@@ -2523,6 +2680,53 @@ typedef struct {
       </varlistentry>
     </variablelist>
   </refsect1>
+  <refsect1 id="heap_profile_format">
+    <title>HEAP PROFILE FORMAT</title>
+    <para>Although the heap profiling functionality was originally designed to
+    be compatible with the
+    <command>pprof</command> command that is developed as part of the <ulink
+    url="http://code.google.com/p/gperftools/">gperftools
+    package</ulink>, the addition of per thread heap profiling functionality
+    required a different heap profile format.  The <command>jeprof</command>
+    command is derived from <command>pprof</command>, with enhancements to
+    support the heap profile format described here.</para>
+
+    <para>In the following hypothetical heap profile, <constant>[...]</constant>
+    indicates elision for the sake of compactness.  <programlisting><![CDATA[
+heap_v2/524288
+  t*: 28106: 56637512 [0: 0]
+  [...]
+  t3: 352: 16777344 [0: 0]
+  [...]
+  t99: 17754: 29341640 [0: 0]
+  [...]
+@ 0x5f86da8 0x5f5a1dc [...] 0x29e4d4e 0xa200316 0xabb2988 [...]
+  t*: 13: 6688 [0: 0]
+  t3: 12: 6496 [0: ]
+  t99: 1: 192 [0: 0]
+[...]
+
+MAPPED_LIBRARIES:
+[...]]]></programlisting> The following matches the above heap profile, but most
+tokens are replaced with <constant>&lt;description&gt;</constant> to indicate
+descriptions of the corresponding fields.  <programlisting><![CDATA[
+<heap_profile_format_version>/<mean_sample_interval>
+  <aggregate>: <curobjs>: <curbytes> [<cumobjs>: <cumbytes>]
+  [...]
+  <thread_3_aggregate>: <curobjs>: <curbytes>[<cumobjs>: <cumbytes>]
+  [...]
+  <thread_99_aggregate>: <curobjs>: <curbytes>[<cumobjs>: <cumbytes>]
+  [...]
+@ <top_frame> <frame> [...] <frame> <frame> <frame> [...]
+  <backtrace_aggregate>: <curobjs>: <curbytes> [<cumobjs>: <cumbytes>]
+  <backtrace_thread_3>: <curobjs>: <curbytes> [<cumobjs>: <cumbytes>]
+  <backtrace_thread_99>: <curobjs>: <curbytes> [<cumobjs>: <cumbytes>]
+[...]
+
+MAPPED_LIBRARIES:
+</proc/<pid>/maps>]]></programlisting></para>
+  </refsect1>
+
   <refsect1 id="debugging_malloc_problems">
     <title>DEBUGGING MALLOC PROBLEMS</title>
     <para>When debugging, it is a good idea to configure/build jemalloc with
@@ -2532,7 +2736,7 @@ typedef struct {
     of run-time assertions that catch application errors such as double-free,
     write-after-free, etc.</para>
 
-    <para>Programs often accidentally depend on &ldquo;uninitialized&rdquo;
+    <para>Programs often accidentally depend on <quote>uninitialized</quote>
     memory actually being filled with zero bytes.  Junk filling
     (see the <link linkend="opt.junk"><mallctl>opt.junk</mallctl></link>
     option) tends to expose such bugs in the form of obviously incorrect
@@ -2561,29 +2765,29 @@ typedef struct {
     to override the function which emits the text strings forming the errors
     and warnings if for some reason the <constant>STDERR_FILENO</constant> file
     descriptor is not suitable for this.
-    <function>malloc_message<parameter/></function> takes the
+    <function>malloc_message()</function> takes the
     <parameter>cbopaque</parameter> pointer argument that is
     <constant>NULL</constant> unless overridden by the arguments in a call to
-    <function>malloc_stats_print<parameter/></function>, followed by a string
+    <function>malloc_stats_print()</function>, followed by a string
     pointer.  Please note that doing anything which tries to allocate memory in
     this function is likely to result in a crash or deadlock.</para>
 
     <para>All messages are prefixed by
-    &ldquo;<computeroutput>&lt;jemalloc&gt;: </computeroutput>&rdquo;.</para>
+    <quote><computeroutput>&lt;jemalloc&gt;: </computeroutput></quote>.</para>
   </refsect1>
   <refsect1 id="return_values">
     <title>RETURN VALUES</title>
     <refsect2>
       <title>Standard API</title>
-      <para>The <function>malloc<parameter/></function> and
-      <function>calloc<parameter/></function> functions return a pointer to the
+      <para>The <function>malloc()</function> and
+      <function>calloc()</function> functions return a pointer to the
       allocated memory if successful; otherwise a <constant>NULL</constant>
       pointer is returned and <varname>errno</varname> is set to
       <errorname>ENOMEM</errorname>.</para>
 
-      <para>The <function>posix_memalign<parameter/></function> function
+      <para>The <function>posix_memalign()</function> function
       returns the value 0 if successful; otherwise it returns an error value.
-      The <function>posix_memalign<parameter/></function> function will fail
+      The <function>posix_memalign()</function> function will fail
       if:
         <variablelist>
           <varlistentry>
@@ -2602,11 +2806,11 @@ typedef struct {
         </variablelist>
       </para>
 
-      <para>The <function>aligned_alloc<parameter/></function> function returns
+      <para>The <function>aligned_alloc()</function> function returns
       a pointer to the allocated memory if successful; otherwise a
       <constant>NULL</constant> pointer is returned and
       <varname>errno</varname> is set.  The
-      <function>aligned_alloc<parameter/></function> function will fail if:
+      <function>aligned_alloc()</function> function will fail if:
         <variablelist>
           <varlistentry>
             <term><errorname>EINVAL</errorname></term>
@@ -2623,44 +2827,44 @@ typedef struct {
         </variablelist>
       </para>
 
-      <para>The <function>realloc<parameter/></function> function returns a
+      <para>The <function>realloc()</function> function returns a
       pointer, possibly identical to <parameter>ptr</parameter>, to the
       allocated memory if successful; otherwise a <constant>NULL</constant>
       pointer is returned, and <varname>errno</varname> is set to
       <errorname>ENOMEM</errorname> if the error was the result of an
-      allocation failure.  The <function>realloc<parameter/></function>
+      allocation failure.  The <function>realloc()</function>
       function always leaves the original buffer intact when an error occurs.
       </para>
 
-      <para>The <function>free<parameter/></function> function returns no
+      <para>The <function>free()</function> function returns no
       value.</para>
     </refsect2>
     <refsect2>
       <title>Non-standard API</title>
-      <para>The <function>mallocx<parameter/></function> and
-      <function>rallocx<parameter/></function> functions return a pointer to
+      <para>The <function>mallocx()</function> and
+      <function>rallocx()</function> functions return a pointer to
       the allocated memory if successful; otherwise a <constant>NULL</constant>
       pointer is returned to indicate insufficient contiguous memory was
       available to service the allocation request.  </para>
 
-      <para>The <function>xallocx<parameter/></function> function returns the
+      <para>The <function>xallocx()</function> function returns the
       real size of the resulting resized allocation pointed to by
       <parameter>ptr</parameter>, which is a value less than
       <parameter>size</parameter> if the allocation could not be adequately
       grown in place.  </para>
 
-      <para>The <function>sallocx<parameter/></function> function returns the
+      <para>The <function>sallocx()</function> function returns the
       real size of the allocation pointed to by <parameter>ptr</parameter>.
       </para>
 
-      <para>The <function>nallocx<parameter/></function> returns the real size
+      <para>The <function>nallocx()</function> returns the real size
       that would result from a successful equivalent
-      <function>mallocx<parameter/></function> function call, or zero if
+      <function>mallocx()</function> function call, or zero if
       insufficient memory is available to perform the size computation.  </para>
 
-      <para>The <function>mallctl<parameter/></function>,
-      <function>mallctlnametomib<parameter/></function>, and
-      <function>mallctlbymib<parameter/></function> functions return 0 on
+      <para>The <function>mallctl()</function>,
+      <function>mallctlnametomib()</function>, and
+      <function>mallctlbymib()</function> functions return 0 on
       success; otherwise they return an error value.  The functions will fail
       if:
         <variablelist>
@@ -2696,13 +2900,13 @@ typedef struct {
             <term><errorname>EFAULT</errorname></term>
 
             <listitem><para>An interface with side effects failed in some way
-            not directly related to <function>mallctl*<parameter/></function>
+            not directly related to <function>mallctl*()</function>
             read/write processing.</para></listitem>
           </varlistentry>
         </variablelist>
       </para>
 
-      <para>The <function>malloc_usable_size<parameter/></function> function
+      <para>The <function>malloc_usable_size()</function> function
       returns the usable size of the allocation pointed to by
       <parameter>ptr</parameter>.  </para>
     </refsect2>
@@ -2750,13 +2954,13 @@ malloc_conf = "lg_chunk:24";]]></programlisting></para>
   </refsect1>
   <refsect1 id="standards">
     <title>STANDARDS</title>
-    <para>The <function>malloc<parameter/></function>,
-    <function>calloc<parameter/></function>,
-    <function>realloc<parameter/></function>, and
-    <function>free<parameter/></function> functions conform to ISO/IEC
-    9899:1990 (&ldquo;ISO C90&rdquo;).</para>
-
-    <para>The <function>posix_memalign<parameter/></function> function conforms
-    to IEEE Std 1003.1-2001 (&ldquo;POSIX.1&rdquo;).</para>
+    <para>The <function>malloc()</function>,
+    <function>calloc()</function>,
+    <function>realloc()</function>, and
+    <function>free()</function> functions conform to ISO/IEC
+    9899:1990 (<quote>ISO C90</quote>).</para>
+
+    <para>The <function>posix_memalign()</function> function conforms
+    to IEEE Std 1003.1-2001 (<quote>POSIX.1</quote>).</para>
   </refsect1>
 </refentry>
diff --git a/deps/jemalloc/doc/stylesheet.xsl b/deps/jemalloc/doc/stylesheet.xsl
index 4e334a86f..619365d82 100644
--- a/deps/jemalloc/doc/stylesheet.xsl
+++ b/deps/jemalloc/doc/stylesheet.xsl
@@ -1,7 +1,10 @@
 <xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version="1.0">
   <xsl:param name="funcsynopsis.style">ansi</xsl:param>
-  <xsl:param name="function.parens" select="1"/>
+  <xsl:param name="function.parens" select="0"/>
+  <xsl:template match="function">
+    <xsl:call-template name="inline.monoseq"/>
+  </xsl:template>
   <xsl:template match="mallctl">
-    "<xsl:call-template name="inline.monoseq"/>"
+    <quote><xsl:call-template name="inline.monoseq"/></quote>
   </xsl:template>
 </xsl:stylesheet>
diff --git a/deps/jemalloc/include/jemalloc/internal/arena.h b/deps/jemalloc/include/jemalloc/internal/arena.h
index 12c617979..ce4e6029e 100644
--- a/deps/jemalloc/include/jemalloc/internal/arena.h
+++ b/deps/jemalloc/include/jemalloc/internal/arena.h
@@ -23,14 +23,29 @@
  */
 #define	LG_DIRTY_MULT_DEFAULT	3
 
+typedef enum {
+	purge_mode_ratio = 0,
+	purge_mode_decay = 1,
+
+	purge_mode_limit = 2
+} purge_mode_t;
+#define	PURGE_DEFAULT		purge_mode_ratio
+/* Default decay time in seconds. */
+#define	DECAY_TIME_DEFAULT	10
+/* Number of event ticks between time checks. */
+#define	DECAY_NTICKS_PER_UPDATE	1000
+
 typedef struct arena_runs_dirty_link_s arena_runs_dirty_link_t;
+typedef struct arena_avail_links_s arena_avail_links_t;
 typedef struct arena_run_s arena_run_t;
 typedef struct arena_chunk_map_bits_s arena_chunk_map_bits_t;
 typedef struct arena_chunk_map_misc_s arena_chunk_map_misc_t;
 typedef struct arena_chunk_s arena_chunk_t;
 typedef struct arena_bin_info_s arena_bin_info_t;
+typedef struct arena_decay_s arena_decay_t;
 typedef struct arena_bin_s arena_bin_t;
 typedef struct arena_s arena_t;
+typedef struct arena_tdata_s arena_tdata_t;
 
 #endif /* JEMALLOC_H_TYPES */
 /******************************************************************************/
@@ -140,13 +155,13 @@ struct arena_runs_dirty_link_s {
  */
 struct arena_chunk_map_misc_s {
 	/*
-	 * Linkage for run trees.  There are two disjoint uses:
+	 * Linkage for run heaps.  There are two disjoint uses:
 	 *
-	 * 1) arena_t's runs_avail tree.
+	 * 1) arena_t's runs_avail heaps.
 	 * 2) arena_run_t conceptually uses this linkage for in-use non-full
 	 *    runs, rather than directly embedding linkage.
 	 */
-	rb_node(arena_chunk_map_misc_t)		rb_link;
+	phn(arena_chunk_map_misc_t)		ph_link;
 
 	union {
 		/* Linkage for list of dirty runs. */
@@ -154,16 +169,15 @@ struct arena_chunk_map_misc_s {
 
 		/* Profile counters, used for large object runs. */
 		union {
-			void				*prof_tctx_pun;
-			prof_tctx_t			*prof_tctx;
+			void			*prof_tctx_pun;
+			prof_tctx_t		*prof_tctx;
 		};
 
 		/* Small region run metadata. */
 		arena_run_t			run;
 	};
 };
-typedef rb_tree(arena_chunk_map_misc_t) arena_avail_tree_t;
-typedef rb_tree(arena_chunk_map_misc_t) arena_run_tree_t;
+typedef ph(arena_chunk_map_misc_t) arena_run_heap_t;
 #endif /* JEMALLOC_ARENA_STRUCTS_A */
 
 #ifdef JEMALLOC_ARENA_STRUCTS_B
@@ -176,6 +190,14 @@ struct arena_chunk_s {
 	 */
 	extent_node_t		node;
 
+	/*
+	 * True if memory could be backed by transparent huge pages.  This is
+	 * only directly relevant to Linux, since it is the only supported
+	 * platform on which jemalloc interacts with explicit transparent huge
+	 * page controls.
+	 */
+	bool			hugepage;
+
 	/*
 	 * Map of pages within chunk that keeps track of free/large/small.  The
 	 * first map_bias entries are omitted, since the chunk header does not
@@ -220,28 +242,71 @@ struct arena_chunk_s {
  */
 struct arena_bin_info_s {
 	/* Size of regions in a run for this bin's size class. */
-	size_t		reg_size;
+	size_t			reg_size;
 
 	/* Redzone size. */
-	size_t		redzone_size;
+	size_t			redzone_size;
 
 	/* Interval between regions (reg_size + (redzone_size << 1)). */
-	size_t		reg_interval;
+	size_t			reg_interval;
 
 	/* Total size of a run for this bin's size class. */
-	size_t		run_size;
+	size_t			run_size;
 
 	/* Total number of regions in a run for this bin's size class. */
-	uint32_t	nregs;
+	uint32_t		nregs;
 
 	/*
 	 * Metadata used to manipulate bitmaps for runs associated with this
 	 * bin.
 	 */
-	bitmap_info_t	bitmap_info;
+	bitmap_info_t		bitmap_info;
 
 	/* Offset of first region in a run for this bin's size class. */
-	uint32_t	reg0_offset;
+	uint32_t		reg0_offset;
+};
+
+struct arena_decay_s {
+	/*
+	 * Approximate time in seconds from the creation of a set of unused
+	 * dirty pages until an equivalent set of unused dirty pages is purged
+	 * and/or reused.
+	 */
+	ssize_t			time;
+	/* time / SMOOTHSTEP_NSTEPS. */
+	nstime_t		interval;
+	/*
+	 * Time at which the current decay interval logically started.  We do
+	 * not actually advance to a new epoch until sometime after it starts
+	 * because of scheduling and computation delays, and it is even possible
+	 * to completely skip epochs.  In all cases, during epoch advancement we
+	 * merge all relevant activity into the most recently recorded epoch.
+	 */
+	nstime_t		epoch;
+	/* Deadline randomness generator. */
+	uint64_t		jitter_state;
+	/*
+	 * Deadline for current epoch.  This is the sum of interval and per
+	 * epoch jitter which is a uniform random variable in [0..interval).
+	 * Epochs always advance by precise multiples of interval, but we
+	 * randomize the deadline to reduce the likelihood of arenas purging in
+	 * lockstep.
+	 */
+	nstime_t		deadline;
+	/*
+	 * Number of dirty pages at beginning of current epoch.  During epoch
+	 * advancement we use the delta between arena->decay.ndirty and
+	 * arena->ndirty to determine how many dirty pages, if any, were
+	 * generated.
+	 */
+	size_t			ndirty;
+	/*
+	 * Trailing log of how many unused dirty pages were generated during
+	 * each of the past SMOOTHSTEP_NSTEPS decay epochs, where the last
+	 * element is the most recent epoch.  Corresponding epoch times are
+	 * relative to epoch.
+	 */
+	size_t			backlog[SMOOTHSTEP_NSTEPS];
 };
 
 struct arena_bin_s {
@@ -251,25 +316,25 @@ struct arena_bin_s {
 	 * which may be acquired while holding one or more bin locks, but not
 	 * vise versa.
 	 */
-	malloc_mutex_t	lock;
+	malloc_mutex_t		lock;
 
 	/*
 	 * Current run being used to service allocations of this bin's size
 	 * class.
 	 */
-	arena_run_t	*runcur;
+	arena_run_t		*runcur;
 
 	/*
-	 * Tree of non-full runs.  This tree is used when looking for an
+	 * Heap of non-full runs.  This heap is used when looking for an
 	 * existing run when runcur is no longer usable.  We choose the
 	 * non-full run that is lowest in memory; this policy tends to keep
 	 * objects packed well, and it can also help reduce the number of
 	 * almost-empty chunks.
 	 */
-	arena_run_tree_t runs;
+	arena_run_heap_t	runs;
 
 	/* Bin statistics. */
-	malloc_bin_stats_t stats;
+	malloc_bin_stats_t	stats;
 };
 
 struct arena_s {
@@ -277,15 +342,23 @@ struct arena_s {
 	unsigned		ind;
 
 	/*
-	 * Number of threads currently assigned to this arena.  This field is
-	 * protected by arenas_lock.
+	 * Number of threads currently assigned to this arena, synchronized via
+	 * atomic operations.  Each thread has two distinct assignments, one for
+	 * application-serving allocation, and the other for internal metadata
+	 * allocation.  Internal metadata must not be allocated from arenas
+	 * created via the arenas.extend mallctl, because the arena.<i>.reset
+	 * mallctl indiscriminately discards all allocations for the affected
+	 * arena.
+	 *
+	 *   0: Application allocation.
+	 *   1: Internal metadata allocation.
 	 */
-	unsigned		nthreads;
+	unsigned		nthreads[2];
 
 	/*
 	 * There are three classes of arena operations from a locking
 	 * perspective:
-	 * 1) Thread assignment (modifies nthreads) is protected by arenas_lock.
+	 * 1) Thread assignment (modifies nthreads) is synchronized via atomics.
 	 * 2) Bin-related operations are protected by bin locks.
 	 * 3) Chunk- and run-related operations are protected by this mutex.
 	 */
@@ -305,10 +378,16 @@ struct arena_s {
 	 * PRNG state for cache index randomization of large allocation base
 	 * pointers.
 	 */
-	uint64_t		offset_state;
+	size_t			offset_state;
 
 	dss_prec_t		dss_prec;
 
+	/* Extant arena chunks. */
+	ql_head(extent_node_t)	achunks;
+
+	/* Extent serial number generator state. */
+	size_t			extent_sn_next;
+
 	/*
 	 * In order to avoid rapid chunk allocation/deallocation when an arena
 	 * oscillates right on the cusp of needing a new chunk, cache the most
@@ -324,7 +403,7 @@ struct arena_s {
 	/* Minimum ratio (log base 2) of nactive:ndirty. */
 	ssize_t			lg_dirty_mult;
 
-	/* True if a thread is currently executing arena_purge(). */
+	/* True if a thread is currently executing arena_purge_to_limit(). */
 	bool			purging;
 
 	/* Number of pages in active runs and huge regions. */
@@ -338,12 +417,6 @@ struct arena_s {
 	 */
 	size_t			ndirty;
 
-	/*
-	 * Size/address-ordered tree of this arena's available runs.  The tree
-	 * is used for first-best-fit run allocation.
-	 */
-	arena_avail_tree_t	runs_avail;
-
 	/*
 	 * Unused dirty memory this arena manages.  Dirty memory is conceptually
 	 * tracked as an arbitrarily interleaved LRU of dirty runs and cached
@@ -375,6 +448,9 @@ struct arena_s {
 	arena_runs_dirty_link_t	runs_dirty;
 	extent_node_t		chunks_cache;
 
+	/* Decay-based purging state. */
+	arena_decay_t		decay;
+
 	/* Extant huge allocations. */
 	ql_head(extent_node_t)	huge;
 	/* Synchronizes all huge allocation/update/deallocation. */
@@ -387,9 +463,9 @@ struct arena_s {
 	 * orderings are needed, which is why there are two trees with the same
 	 * contents.
 	 */
-	extent_tree_t		chunks_szad_cached;
+	extent_tree_t		chunks_szsnad_cached;
 	extent_tree_t		chunks_ad_cached;
-	extent_tree_t		chunks_szad_retained;
+	extent_tree_t		chunks_szsnad_retained;
 	extent_tree_t		chunks_ad_retained;
 
 	malloc_mutex_t		chunks_mtx;
@@ -402,6 +478,19 @@ struct arena_s {
 
 	/* bins is used to store trees of free regions. */
 	arena_bin_t		bins[NBINS];
+
+	/*
+	 * Size-segregated address-ordered heaps of this arena's available runs,
+	 * used for first-best-fit run allocation.  Runs are quantized, i.e.
+	 * they reside in the last heap which corresponds to a size class less
+	 * than or equal to the run size.
+	 */
+	arena_run_heap_t	runs_avail[NPSIZES];
+};
+
+/* Used in conjunction with tsd for fast arena-related context lookup. */
+struct arena_tdata_s {
+	ticker_t		decay_ticker;
 };
 #endif /* JEMALLOC_ARENA_STRUCTS_B */
 
@@ -417,7 +506,10 @@ static const size_t	large_pad =
 #endif
     ;
 
+extern purge_mode_t	opt_purge;
+extern const char	*purge_mode_names[];
 extern ssize_t		opt_lg_dirty_mult;
+extern ssize_t		opt_decay_time;
 
 extern arena_bin_info_t	arena_bin_info[NBINS];
 
@@ -428,27 +520,37 @@ extern size_t		large_maxclass; /* Max large size class. */
 extern unsigned		nlclasses; /* Number of large size classes. */
 extern unsigned		nhclasses; /* Number of huge size classes. */
 
+#ifdef JEMALLOC_JET
+typedef size_t (run_quantize_t)(size_t);
+extern run_quantize_t *run_quantize_floor;
+extern run_quantize_t *run_quantize_ceil;
+#endif
 void	arena_chunk_cache_maybe_insert(arena_t *arena, extent_node_t *node,
     bool cache);
 void	arena_chunk_cache_maybe_remove(arena_t *arena, extent_node_t *node,
     bool cache);
-extent_node_t	*arena_node_alloc(arena_t *arena);
-void	arena_node_dalloc(arena_t *arena, extent_node_t *node);
-void	*arena_chunk_alloc_huge(arena_t *arena, size_t usize, size_t alignment,
-    bool *zero);
-void	arena_chunk_dalloc_huge(arena_t *arena, void *chunk, size_t usize);
-void	arena_chunk_ralloc_huge_similar(arena_t *arena, void *chunk,
-    size_t oldsize, size_t usize);
-void	arena_chunk_ralloc_huge_shrink(arena_t *arena, void *chunk,
-    size_t oldsize, size_t usize);
-bool	arena_chunk_ralloc_huge_expand(arena_t *arena, void *chunk,
-    size_t oldsize, size_t usize, bool *zero);
-ssize_t	arena_lg_dirty_mult_get(arena_t *arena);
-bool	arena_lg_dirty_mult_set(arena_t *arena, ssize_t lg_dirty_mult);
-void	arena_maybe_purge(arena_t *arena);
-void	arena_purge_all(arena_t *arena);
-void	arena_tcache_fill_small(arena_t *arena, tcache_bin_t *tbin,
-    szind_t binind, uint64_t prof_accumbytes);
+extent_node_t	*arena_node_alloc(tsdn_t *tsdn, arena_t *arena);
+void	arena_node_dalloc(tsdn_t *tsdn, arena_t *arena, extent_node_t *node);
+void	*arena_chunk_alloc_huge(tsdn_t *tsdn, arena_t *arena, size_t usize,
+    size_t alignment, size_t *sn, bool *zero);
+void	arena_chunk_dalloc_huge(tsdn_t *tsdn, arena_t *arena, void *chunk,
+    size_t usize, size_t sn);
+void	arena_chunk_ralloc_huge_similar(tsdn_t *tsdn, arena_t *arena,
+    void *chunk, size_t oldsize, size_t usize);
+void	arena_chunk_ralloc_huge_shrink(tsdn_t *tsdn, arena_t *arena,
+    void *chunk, size_t oldsize, size_t usize, size_t sn);
+bool	arena_chunk_ralloc_huge_expand(tsdn_t *tsdn, arena_t *arena,
+    void *chunk, size_t oldsize, size_t usize, bool *zero);
+ssize_t	arena_lg_dirty_mult_get(tsdn_t *tsdn, arena_t *arena);
+bool	arena_lg_dirty_mult_set(tsdn_t *tsdn, arena_t *arena,
+    ssize_t lg_dirty_mult);
+ssize_t	arena_decay_time_get(tsdn_t *tsdn, arena_t *arena);
+bool	arena_decay_time_set(tsdn_t *tsdn, arena_t *arena, ssize_t decay_time);
+void	arena_purge(tsdn_t *tsdn, arena_t *arena, bool all);
+void	arena_maybe_purge(tsdn_t *tsdn, arena_t *arena);
+void	arena_reset(tsd_t *tsd, arena_t *arena);
+void	arena_tcache_fill_small(tsdn_t *tsdn, arena_t *arena,
+    tcache_bin_t *tbin, szind_t binind, uint64_t prof_accumbytes);
 void	arena_alloc_junk_small(void *ptr, arena_bin_info_t *bin_info,
     bool zero);
 #ifdef JEMALLOC_JET
@@ -461,75 +563,100 @@ extern arena_dalloc_junk_small_t *arena_dalloc_junk_small;
 void	arena_dalloc_junk_small(void *ptr, arena_bin_info_t *bin_info);
 #endif
 void	arena_quarantine_junk_small(void *ptr, size_t usize);
-void	*arena_malloc_small(arena_t *arena, size_t size, bool zero);
-void	*arena_malloc_large(arena_t *arena, size_t size, bool zero);
-void	*arena_palloc(tsd_t *tsd, arena_t *arena, size_t usize,
+void	*arena_malloc_large(tsdn_t *tsdn, arena_t *arena, szind_t ind,
+    bool zero);
+void	*arena_malloc_hard(tsdn_t *tsdn, arena_t *arena, size_t size,
+    szind_t ind, bool zero);
+void	*arena_palloc(tsdn_t *tsdn, arena_t *arena, size_t usize,
     size_t alignment, bool zero, tcache_t *tcache);
-void	arena_prof_promoted(const void *ptr, size_t size);
-void	arena_dalloc_bin_junked_locked(arena_t *arena, arena_chunk_t *chunk,
-    void *ptr, arena_chunk_map_bits_t *bitselm);
-void	arena_dalloc_bin(arena_t *arena, arena_chunk_t *chunk, void *ptr,
-    size_t pageind, arena_chunk_map_bits_t *bitselm);
-void	arena_dalloc_small(arena_t *arena, arena_chunk_t *chunk, void *ptr,
-    size_t pageind);
+void	arena_prof_promoted(tsdn_t *tsdn, const void *ptr, size_t size);
+void	arena_dalloc_bin_junked_locked(tsdn_t *tsdn, arena_t *arena,
+    arena_chunk_t *chunk, void *ptr, arena_chunk_map_bits_t *bitselm);
+void	arena_dalloc_bin(tsdn_t *tsdn, arena_t *arena, arena_chunk_t *chunk,
+    void *ptr, size_t pageind, arena_chunk_map_bits_t *bitselm);
+void	arena_dalloc_small(tsdn_t *tsdn, arena_t *arena, arena_chunk_t *chunk,
+    void *ptr, size_t pageind);
 #ifdef JEMALLOC_JET
 typedef void (arena_dalloc_junk_large_t)(void *, size_t);
 extern arena_dalloc_junk_large_t *arena_dalloc_junk_large;
 #else
 void	arena_dalloc_junk_large(void *ptr, size_t usize);
 #endif
-void	arena_dalloc_large_junked_locked(arena_t *arena, arena_chunk_t *chunk,
+void	arena_dalloc_large_junked_locked(tsdn_t *tsdn, arena_t *arena,
+    arena_chunk_t *chunk, void *ptr);
+void	arena_dalloc_large(tsdn_t *tsdn, arena_t *arena, arena_chunk_t *chunk,
     void *ptr);
-void	arena_dalloc_large(arena_t *arena, arena_chunk_t *chunk, void *ptr);
 #ifdef JEMALLOC_JET
 typedef void (arena_ralloc_junk_large_t)(void *, size_t, size_t);
 extern arena_ralloc_junk_large_t *arena_ralloc_junk_large;
 #endif
-bool	arena_ralloc_no_move(void *ptr, size_t oldsize, size_t size,
-    size_t extra, bool zero);
+bool	arena_ralloc_no_move(tsdn_t *tsdn, void *ptr, size_t oldsize,
+    size_t size, size_t extra, bool zero);
 void	*arena_ralloc(tsd_t *tsd, arena_t *arena, void *ptr, size_t oldsize,
     size_t size, size_t alignment, bool zero, tcache_t *tcache);
-dss_prec_t	arena_dss_prec_get(arena_t *arena);
-bool	arena_dss_prec_set(arena_t *arena, dss_prec_t dss_prec);
+dss_prec_t	arena_dss_prec_get(tsdn_t *tsdn, arena_t *arena);
+bool	arena_dss_prec_set(tsdn_t *tsdn, arena_t *arena, dss_prec_t dss_prec);
 ssize_t	arena_lg_dirty_mult_default_get(void);
 bool	arena_lg_dirty_mult_default_set(ssize_t lg_dirty_mult);
-void	arena_stats_merge(arena_t *arena, const char **dss,
-    ssize_t *lg_dirty_mult, size_t *nactive, size_t *ndirty,
-    arena_stats_t *astats, malloc_bin_stats_t *bstats,
-    malloc_large_stats_t *lstats, malloc_huge_stats_t *hstats);
-arena_t	*arena_new(unsigned ind);
-bool	arena_boot(void);
-void	arena_prefork(arena_t *arena);
-void	arena_postfork_parent(arena_t *arena);
-void	arena_postfork_child(arena_t *arena);
+ssize_t	arena_decay_time_default_get(void);
+bool	arena_decay_time_default_set(ssize_t decay_time);
+void	arena_basic_stats_merge(tsdn_t *tsdn, arena_t *arena,
+    unsigned *nthreads, const char **dss, ssize_t *lg_dirty_mult,
+    ssize_t *decay_time, size_t *nactive, size_t *ndirty);
+void	arena_stats_merge(tsdn_t *tsdn, arena_t *arena, unsigned *nthreads,
+    const char **dss, ssize_t *lg_dirty_mult, ssize_t *decay_time,
+    size_t *nactive, size_t *ndirty, arena_stats_t *astats,
+    malloc_bin_stats_t *bstats, malloc_large_stats_t *lstats,
+    malloc_huge_stats_t *hstats);
+unsigned	arena_nthreads_get(arena_t *arena, bool internal);
+void	arena_nthreads_inc(arena_t *arena, bool internal);
+void	arena_nthreads_dec(arena_t *arena, bool internal);
+size_t	arena_extent_sn_next(arena_t *arena);
+arena_t	*arena_new(tsdn_t *tsdn, unsigned ind);
+void	arena_boot(void);
+void	arena_prefork0(tsdn_t *tsdn, arena_t *arena);
+void	arena_prefork1(tsdn_t *tsdn, arena_t *arena);
+void	arena_prefork2(tsdn_t *tsdn, arena_t *arena);
+void	arena_prefork3(tsdn_t *tsdn, arena_t *arena);
+void	arena_postfork_parent(tsdn_t *tsdn, arena_t *arena);
+void	arena_postfork_child(tsdn_t *tsdn, arena_t *arena);
 
 #endif /* JEMALLOC_H_EXTERNS */
 /******************************************************************************/
 #ifdef JEMALLOC_H_INLINES
 
 #ifndef JEMALLOC_ENABLE_INLINE
-arena_chunk_map_bits_t	*arena_bitselm_get(arena_chunk_t *chunk,
+arena_chunk_map_bits_t	*arena_bitselm_get_mutable(arena_chunk_t *chunk,
     size_t pageind);
-arena_chunk_map_misc_t	*arena_miscelm_get(arena_chunk_t *chunk,
+const arena_chunk_map_bits_t	*arena_bitselm_get_const(
+    const arena_chunk_t *chunk, size_t pageind);
+arena_chunk_map_misc_t	*arena_miscelm_get_mutable(arena_chunk_t *chunk,
     size_t pageind);
-size_t	arena_miscelm_to_pageind(arena_chunk_map_misc_t *miscelm);
-void	*arena_miscelm_to_rpages(arena_chunk_map_misc_t *miscelm);
+const arena_chunk_map_misc_t	*arena_miscelm_get_const(
+    const arena_chunk_t *chunk, size_t pageind);
+size_t	arena_miscelm_to_pageind(const arena_chunk_map_misc_t *miscelm);
+void	*arena_miscelm_to_rpages(const arena_chunk_map_misc_t *miscelm);
 arena_chunk_map_misc_t	*arena_rd_to_miscelm(arena_runs_dirty_link_t *rd);
 arena_chunk_map_misc_t	*arena_run_to_miscelm(arena_run_t *run);
-size_t	*arena_mapbitsp_get(arena_chunk_t *chunk, size_t pageind);
-size_t	arena_mapbitsp_read(size_t *mapbitsp);
-size_t	arena_mapbits_get(arena_chunk_t *chunk, size_t pageind);
+size_t	*arena_mapbitsp_get_mutable(arena_chunk_t *chunk, size_t pageind);
+const size_t	*arena_mapbitsp_get_const(const arena_chunk_t *chunk,
+    size_t pageind);
+size_t	arena_mapbitsp_read(const size_t *mapbitsp);
+size_t	arena_mapbits_get(const arena_chunk_t *chunk, size_t pageind);
 size_t	arena_mapbits_size_decode(size_t mapbits);
-size_t	arena_mapbits_unallocated_size_get(arena_chunk_t *chunk,
+size_t	arena_mapbits_unallocated_size_get(const arena_chunk_t *chunk,
+    size_t pageind);
+size_t	arena_mapbits_large_size_get(const arena_chunk_t *chunk,
+    size_t pageind);
+size_t	arena_mapbits_small_runind_get(const arena_chunk_t *chunk,
     size_t pageind);
-size_t	arena_mapbits_large_size_get(arena_chunk_t *chunk, size_t pageind);
-size_t	arena_mapbits_small_runind_get(arena_chunk_t *chunk, size_t pageind);
-szind_t	arena_mapbits_binind_get(arena_chunk_t *chunk, size_t pageind);
-size_t	arena_mapbits_dirty_get(arena_chunk_t *chunk, size_t pageind);
-size_t	arena_mapbits_unzeroed_get(arena_chunk_t *chunk, size_t pageind);
-size_t	arena_mapbits_decommitted_get(arena_chunk_t *chunk, size_t pageind);
-size_t	arena_mapbits_large_get(arena_chunk_t *chunk, size_t pageind);
-size_t	arena_mapbits_allocated_get(arena_chunk_t *chunk, size_t pageind);
+szind_t	arena_mapbits_binind_get(const arena_chunk_t *chunk, size_t pageind);
+size_t	arena_mapbits_dirty_get(const arena_chunk_t *chunk, size_t pageind);
+size_t	arena_mapbits_unzeroed_get(const arena_chunk_t *chunk, size_t pageind);
+size_t	arena_mapbits_decommitted_get(const arena_chunk_t *chunk,
+    size_t pageind);
+size_t	arena_mapbits_large_get(const arena_chunk_t *chunk, size_t pageind);
+size_t	arena_mapbits_allocated_get(const arena_chunk_t *chunk, size_t pageind);
 void	arena_mapbitsp_write(size_t *mapbitsp, size_t mapbits);
 size_t	arena_mapbits_size_encode(size_t size);
 void	arena_mapbits_unallocated_set(arena_chunk_t *chunk, size_t pageind,
@@ -549,27 +676,31 @@ void	arena_metadata_allocated_sub(arena_t *arena, size_t size);
 size_t	arena_metadata_allocated_get(arena_t *arena);
 bool	arena_prof_accum_impl(arena_t *arena, uint64_t accumbytes);
 bool	arena_prof_accum_locked(arena_t *arena, uint64_t accumbytes);
-bool	arena_prof_accum(arena_t *arena, uint64_t accumbytes);
+bool	arena_prof_accum(tsdn_t *tsdn, arena_t *arena, uint64_t accumbytes);
 szind_t	arena_ptr_small_binind_get(const void *ptr, size_t mapbits);
 szind_t	arena_bin_index(arena_t *arena, arena_bin_t *bin);
-unsigned	arena_run_regind(arena_run_t *run, arena_bin_info_t *bin_info,
+size_t	arena_run_regind(arena_run_t *run, arena_bin_info_t *bin_info,
     const void *ptr);
-prof_tctx_t	*arena_prof_tctx_get(const void *ptr);
-void	arena_prof_tctx_set(const void *ptr, size_t usize, prof_tctx_t *tctx);
-void	arena_prof_tctx_reset(const void *ptr, size_t usize,
+prof_tctx_t	*arena_prof_tctx_get(tsdn_t *tsdn, const void *ptr);
+void	arena_prof_tctx_set(tsdn_t *tsdn, const void *ptr, size_t usize,
+    prof_tctx_t *tctx);
+void	arena_prof_tctx_reset(tsdn_t *tsdn, const void *ptr, size_t usize,
     const void *old_ptr, prof_tctx_t *old_tctx);
-void	*arena_malloc(tsd_t *tsd, arena_t *arena, size_t size, bool zero,
-    tcache_t *tcache);
+void	arena_decay_ticks(tsdn_t *tsdn, arena_t *arena, unsigned nticks);
+void	arena_decay_tick(tsdn_t *tsdn, arena_t *arena);
+void	*arena_malloc(tsdn_t *tsdn, arena_t *arena, size_t size, szind_t ind,
+    bool zero, tcache_t *tcache, bool slow_path);
 arena_t	*arena_aalloc(const void *ptr);
-size_t	arena_salloc(const void *ptr, bool demote);
-void	arena_dalloc(tsd_t *tsd, void *ptr, tcache_t *tcache);
-void	arena_sdalloc(tsd_t *tsd, void *ptr, size_t size, tcache_t *tcache);
+size_t	arena_salloc(tsdn_t *tsdn, const void *ptr, bool demote);
+void	arena_dalloc(tsdn_t *tsdn, void *ptr, tcache_t *tcache, bool slow_path);
+void	arena_sdalloc(tsdn_t *tsdn, void *ptr, size_t size, tcache_t *tcache,
+    bool slow_path);
 #endif
 
 #if (defined(JEMALLOC_ENABLE_INLINE) || defined(JEMALLOC_ARENA_C_))
 #  ifdef JEMALLOC_ARENA_INLINE_A
 JEMALLOC_ALWAYS_INLINE arena_chunk_map_bits_t *
-arena_bitselm_get(arena_chunk_t *chunk, size_t pageind)
+arena_bitselm_get_mutable(arena_chunk_t *chunk, size_t pageind)
 {
 
 	assert(pageind >= map_bias);
@@ -578,8 +709,15 @@ arena_bitselm_get(arena_chunk_t *chunk, size_t pageind)
 	return (&chunk->map_bits[pageind-map_bias]);
 }
 
+JEMALLOC_ALWAYS_INLINE const arena_chunk_map_bits_t *
+arena_bitselm_get_const(const arena_chunk_t *chunk, size_t pageind)
+{
+
+	return (arena_bitselm_get_mutable((arena_chunk_t *)chunk, pageind));
+}
+
 JEMALLOC_ALWAYS_INLINE arena_chunk_map_misc_t *
-arena_miscelm_get(arena_chunk_t *chunk, size_t pageind)
+arena_miscelm_get_mutable(arena_chunk_t *chunk, size_t pageind)
 {
 
 	assert(pageind >= map_bias);
@@ -589,8 +727,15 @@ arena_miscelm_get(arena_chunk_t *chunk, size_t pageind)
 	    (uintptr_t)map_misc_offset) + pageind-map_bias);
 }
 
+JEMALLOC_ALWAYS_INLINE const arena_chunk_map_misc_t *
+arena_miscelm_get_const(const arena_chunk_t *chunk, size_t pageind)
+{
+
+	return (arena_miscelm_get_mutable((arena_chunk_t *)chunk, pageind));
+}
+
 JEMALLOC_ALWAYS_INLINE size_t
-arena_miscelm_to_pageind(arena_chunk_map_misc_t *miscelm)
+arena_miscelm_to_pageind(const arena_chunk_map_misc_t *miscelm)
 {
 	arena_chunk_t *chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(miscelm);
 	size_t pageind = ((uintptr_t)miscelm - ((uintptr_t)chunk +
@@ -603,7 +748,7 @@ arena_miscelm_to_pageind(arena_chunk_map_misc_t *miscelm)
 }
 
 JEMALLOC_ALWAYS_INLINE void *
-arena_miscelm_to_rpages(arena_chunk_map_misc_t *miscelm)
+arena_miscelm_to_rpages(const arena_chunk_map_misc_t *miscelm)
 {
 	arena_chunk_t *chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(miscelm);
 	size_t pageind = arena_miscelm_to_pageind(miscelm);
@@ -636,24 +781,31 @@ arena_run_to_miscelm(arena_run_t *run)
 }
 
 JEMALLOC_ALWAYS_INLINE size_t *
-arena_mapbitsp_get(arena_chunk_t *chunk, size_t pageind)
+arena_mapbitsp_get_mutable(arena_chunk_t *chunk, size_t pageind)
 {
 
-	return (&arena_bitselm_get(chunk, pageind)->bits);
+	return (&arena_bitselm_get_mutable(chunk, pageind)->bits);
+}
+
+JEMALLOC_ALWAYS_INLINE const size_t *
+arena_mapbitsp_get_const(const arena_chunk_t *chunk, size_t pageind)
+{
+
+	return (arena_mapbitsp_get_mutable((arena_chunk_t *)chunk, pageind));
 }
 
 JEMALLOC_ALWAYS_INLINE size_t
-arena_mapbitsp_read(size_t *mapbitsp)
+arena_mapbitsp_read(const size_t *mapbitsp)
 {
 
 	return (*mapbitsp);
 }
 
 JEMALLOC_ALWAYS_INLINE size_t
-arena_mapbits_get(arena_chunk_t *chunk, size_t pageind)
+arena_mapbits_get(const arena_chunk_t *chunk, size_t pageind)
 {
 
-	return (arena_mapbitsp_read(arena_mapbitsp_get(chunk, pageind)));
+	return (arena_mapbitsp_read(arena_mapbitsp_get_const(chunk, pageind)));
 }
 
 JEMALLOC_ALWAYS_INLINE size_t
@@ -673,7 +825,7 @@ arena_mapbits_size_decode(size_t mapbits)
 }
 
 JEMALLOC_ALWAYS_INLINE size_t
-arena_mapbits_unallocated_size_get(arena_chunk_t *chunk, size_t pageind)
+arena_mapbits_unallocated_size_get(const arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
 
@@ -683,7 +835,7 @@ arena_mapbits_unallocated_size_get(arena_chunk_t *chunk, size_t pageind)
 }
 
 JEMALLOC_ALWAYS_INLINE size_t
-arena_mapbits_large_size_get(arena_chunk_t *chunk, size_t pageind)
+arena_mapbits_large_size_get(const arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
 
@@ -694,7 +846,7 @@ arena_mapbits_large_size_get(arena_chunk_t *chunk, size_t pageind)
 }
 
 JEMALLOC_ALWAYS_INLINE size_t
-arena_mapbits_small_runind_get(arena_chunk_t *chunk, size_t pageind)
+arena_mapbits_small_runind_get(const arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
 
@@ -705,7 +857,7 @@ arena_mapbits_small_runind_get(arena_chunk_t *chunk, size_t pageind)
 }
 
 JEMALLOC_ALWAYS_INLINE szind_t
-arena_mapbits_binind_get(arena_chunk_t *chunk, size_t pageind)
+arena_mapbits_binind_get(const arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
 	szind_t binind;
@@ -717,7 +869,7 @@ arena_mapbits_binind_get(arena_chunk_t *chunk, size_t pageind)
 }
 
 JEMALLOC_ALWAYS_INLINE size_t
-arena_mapbits_dirty_get(arena_chunk_t *chunk, size_t pageind)
+arena_mapbits_dirty_get(const arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
 
@@ -728,7 +880,7 @@ arena_mapbits_dirty_get(arena_chunk_t *chunk, size_t pageind)
 }
 
 JEMALLOC_ALWAYS_INLINE size_t
-arena_mapbits_unzeroed_get(arena_chunk_t *chunk, size_t pageind)
+arena_mapbits_unzeroed_get(const arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
 
@@ -739,7 +891,7 @@ arena_mapbits_unzeroed_get(arena_chunk_t *chunk, size_t pageind)
 }
 
 JEMALLOC_ALWAYS_INLINE size_t
-arena_mapbits_decommitted_get(arena_chunk_t *chunk, size_t pageind)
+arena_mapbits_decommitted_get(const arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
 
@@ -750,7 +902,7 @@ arena_mapbits_decommitted_get(arena_chunk_t *chunk, size_t pageind)
 }
 
 JEMALLOC_ALWAYS_INLINE size_t
-arena_mapbits_large_get(arena_chunk_t *chunk, size_t pageind)
+arena_mapbits_large_get(const arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
 
@@ -759,7 +911,7 @@ arena_mapbits_large_get(arena_chunk_t *chunk, size_t pageind)
 }
 
 JEMALLOC_ALWAYS_INLINE size_t
-arena_mapbits_allocated_get(arena_chunk_t *chunk, size_t pageind)
+arena_mapbits_allocated_get(const arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
 
@@ -795,7 +947,7 @@ JEMALLOC_ALWAYS_INLINE void
 arena_mapbits_unallocated_set(arena_chunk_t *chunk, size_t pageind, size_t size,
     size_t flags)
 {
-	size_t *mapbitsp = arena_mapbitsp_get(chunk, pageind);
+	size_t *mapbitsp = arena_mapbitsp_get_mutable(chunk, pageind);
 
 	assert((size & PAGE_MASK) == 0);
 	assert((flags & CHUNK_MAP_FLAGS_MASK) == flags);
@@ -809,7 +961,7 @@ JEMALLOC_ALWAYS_INLINE void
 arena_mapbits_unallocated_size_set(arena_chunk_t *chunk, size_t pageind,
     size_t size)
 {
-	size_t *mapbitsp = arena_mapbitsp_get(chunk, pageind);
+	size_t *mapbitsp = arena_mapbitsp_get_mutable(chunk, pageind);
 	size_t mapbits = arena_mapbitsp_read(mapbitsp);
 
 	assert((size & PAGE_MASK) == 0);
@@ -821,7 +973,7 @@ arena_mapbits_unallocated_size_set(arena_chunk_t *chunk, size_t pageind,
 JEMALLOC_ALWAYS_INLINE void
 arena_mapbits_internal_set(arena_chunk_t *chunk, size_t pageind, size_t flags)
 {
-	size_t *mapbitsp = arena_mapbitsp_get(chunk, pageind);
+	size_t *mapbitsp = arena_mapbitsp_get_mutable(chunk, pageind);
 
 	assert((flags & CHUNK_MAP_UNZEROED) == flags);
 	arena_mapbitsp_write(mapbitsp, flags);
@@ -831,7 +983,7 @@ JEMALLOC_ALWAYS_INLINE void
 arena_mapbits_large_set(arena_chunk_t *chunk, size_t pageind, size_t size,
     size_t flags)
 {
-	size_t *mapbitsp = arena_mapbitsp_get(chunk, pageind);
+	size_t *mapbitsp = arena_mapbitsp_get_mutable(chunk, pageind);
 
 	assert((size & PAGE_MASK) == 0);
 	assert((flags & CHUNK_MAP_FLAGS_MASK) == flags);
@@ -846,7 +998,7 @@ JEMALLOC_ALWAYS_INLINE void
 arena_mapbits_large_binind_set(arena_chunk_t *chunk, size_t pageind,
     szind_t binind)
 {
-	size_t *mapbitsp = arena_mapbitsp_get(chunk, pageind);
+	size_t *mapbitsp = arena_mapbitsp_get_mutable(chunk, pageind);
 	size_t mapbits = arena_mapbitsp_read(mapbitsp);
 
 	assert(binind <= BININD_INVALID);
@@ -860,7 +1012,7 @@ JEMALLOC_ALWAYS_INLINE void
 arena_mapbits_small_set(arena_chunk_t *chunk, size_t pageind, size_t runind,
     szind_t binind, size_t flags)
 {
-	size_t *mapbitsp = arena_mapbitsp_get(chunk, pageind);
+	size_t *mapbitsp = arena_mapbitsp_get_mutable(chunk, pageind);
 
 	assert(binind < BININD_INVALID);
 	assert(pageind - runind >= map_bias);
@@ -917,7 +1069,7 @@ arena_prof_accum_locked(arena_t *arena, uint64_t accumbytes)
 }
 
 JEMALLOC_INLINE bool
-arena_prof_accum(arena_t *arena, uint64_t accumbytes)
+arena_prof_accum(tsdn_t *tsdn, arena_t *arena, uint64_t accumbytes)
 {
 
 	cassert(config_prof);
@@ -928,9 +1080,9 @@ arena_prof_accum(arena_t *arena, uint64_t accumbytes)
 	{
 		bool ret;
 
-		malloc_mutex_lock(&arena->lock);
+		malloc_mutex_lock(tsdn, &arena->lock);
 		ret = arena_prof_accum_impl(arena, accumbytes);
-		malloc_mutex_unlock(&arena->lock);
+		malloc_mutex_unlock(tsdn, &arena->lock);
 		return (ret);
 	}
 }
@@ -948,12 +1100,12 @@ arena_ptr_small_binind_get(const void *ptr, size_t mapbits)
 		size_t pageind;
 		size_t actual_mapbits;
 		size_t rpages_ind;
-		arena_run_t *run;
+		const arena_run_t *run;
 		arena_bin_t *bin;
 		szind_t run_binind, actual_binind;
 		arena_bin_info_t *bin_info;
-		arena_chunk_map_misc_t *miscelm;
-		void *rpages;
+		const arena_chunk_map_misc_t *miscelm;
+		const void *rpages;
 
 		assert(binind != BININD_INVALID);
 		assert(binind < NBINS);
@@ -966,11 +1118,11 @@ arena_ptr_small_binind_get(const void *ptr, size_t mapbits)
 		assert(arena_mapbits_allocated_get(chunk, pageind) != 0);
 		rpages_ind = pageind - arena_mapbits_small_runind_get(chunk,
 		    pageind);
-		miscelm = arena_miscelm_get(chunk, rpages_ind);
+		miscelm = arena_miscelm_get_const(chunk, rpages_ind);
 		run = &miscelm->run;
 		run_binind = run->binind;
 		bin = &arena->bins[run_binind];
-		actual_binind = bin - arena->bins;
+		actual_binind = (szind_t)(bin - arena->bins);
 		assert(run_binind == actual_binind);
 		bin_info = &arena_bin_info[actual_binind];
 		rpages = arena_miscelm_to_rpages(miscelm);
@@ -987,16 +1139,15 @@ arena_ptr_small_binind_get(const void *ptr, size_t mapbits)
 JEMALLOC_INLINE szind_t
 arena_bin_index(arena_t *arena, arena_bin_t *bin)
 {
-	szind_t binind = bin - arena->bins;
+	szind_t binind = (szind_t)(bin - arena->bins);
 	assert(binind < NBINS);
 	return (binind);
 }
 
-JEMALLOC_INLINE unsigned
+JEMALLOC_INLINE size_t
 arena_run_regind(arena_run_t *run, arena_bin_info_t *bin_info, const void *ptr)
 {
-	unsigned shift, diff, regind;
-	size_t interval;
+	size_t diff, interval, shift, regind;
 	arena_chunk_map_misc_t *miscelm = arena_run_to_miscelm(run);
 	void *rpages = arena_miscelm_to_rpages(miscelm);
 
@@ -1011,12 +1162,12 @@ arena_run_regind(arena_run_t *run, arena_bin_info_t *bin_info, const void *ptr)
 	 * Avoid doing division with a variable divisor if possible.  Using
 	 * actual division here can reduce allocator throughput by over 20%!
 	 */
-	diff = (unsigned)((uintptr_t)ptr - (uintptr_t)rpages -
+	diff = (size_t)((uintptr_t)ptr - (uintptr_t)rpages -
 	    bin_info->reg0_offset);
 
 	/* Rescale (factor powers of 2 out of the numerator and denominator). */
 	interval = bin_info->reg_interval;
-	shift = jemalloc_ffs(interval) - 1;
+	shift = ffs_zu(interval) - 1;
 	diff >>= shift;
 	interval >>= shift;
 
@@ -1038,9 +1189,9 @@ arena_run_regind(arena_run_t *run, arena_bin_info_t *bin_info, const void *ptr)
 		 * divide by 0, and 1 and 2 are both powers of two, which are
 		 * handled above.
 		 */
-#define	SIZE_INV_SHIFT	((sizeof(unsigned) << 3) - LG_RUN_MAXREGS)
-#define	SIZE_INV(s)	(((1U << SIZE_INV_SHIFT) / (s)) + 1)
-		static const unsigned interval_invs[] = {
+#define	SIZE_INV_SHIFT	((sizeof(size_t) << 3) - LG_RUN_MAXREGS)
+#define	SIZE_INV(s)	(((ZU(1) << SIZE_INV_SHIFT) / (s)) + 1)
+		static const size_t interval_invs[] = {
 		    SIZE_INV(3),
 		    SIZE_INV(4), SIZE_INV(5), SIZE_INV(6), SIZE_INV(7),
 		    SIZE_INV(8), SIZE_INV(9), SIZE_INV(10), SIZE_INV(11),
@@ -1051,8 +1202,8 @@ arena_run_regind(arena_run_t *run, arena_bin_info_t *bin_info, const void *ptr)
 		    SIZE_INV(28), SIZE_INV(29), SIZE_INV(30), SIZE_INV(31)
 		};
 
-		if (likely(interval <= ((sizeof(interval_invs) /
-		    sizeof(unsigned)) + 2))) {
+		if (likely(interval <= ((sizeof(interval_invs) / sizeof(size_t))
+		    + 2))) {
 			regind = (diff * interval_invs[interval - 3]) >>
 			    SIZE_INV_SHIFT;
 		} else
@@ -1067,7 +1218,7 @@ arena_run_regind(arena_run_t *run, arena_bin_info_t *bin_info, const void *ptr)
 }
 
 JEMALLOC_INLINE prof_tctx_t *
-arena_prof_tctx_get(const void *ptr)
+arena_prof_tctx_get(tsdn_t *tsdn, const void *ptr)
 {
 	prof_tctx_t *ret;
 	arena_chunk_t *chunk;
@@ -1083,18 +1234,19 @@ arena_prof_tctx_get(const void *ptr)
 		if (likely((mapbits & CHUNK_MAP_LARGE) == 0))
 			ret = (prof_tctx_t *)(uintptr_t)1U;
 		else {
-			arena_chunk_map_misc_t *elm = arena_miscelm_get(chunk,
-			    pageind);
+			arena_chunk_map_misc_t *elm =
+			    arena_miscelm_get_mutable(chunk, pageind);
 			ret = atomic_read_p(&elm->prof_tctx_pun);
 		}
 	} else
-		ret = huge_prof_tctx_get(ptr);
+		ret = huge_prof_tctx_get(tsdn, ptr);
 
 	return (ret);
 }
 
 JEMALLOC_INLINE void
-arena_prof_tctx_set(const void *ptr, size_t usize, prof_tctx_t *tctx)
+arena_prof_tctx_set(tsdn_t *tsdn, const void *ptr, size_t usize,
+    prof_tctx_t *tctx)
 {
 	arena_chunk_t *chunk;
 
@@ -1113,7 +1265,7 @@ arena_prof_tctx_set(const void *ptr, size_t usize, prof_tctx_t *tctx)
 
 			assert(arena_mapbits_large_get(chunk, pageind) != 0);
 
-			elm = arena_miscelm_get(chunk, pageind);
+			elm = arena_miscelm_get_mutable(chunk, pageind);
 			atomic_write_p(&elm->prof_tctx_pun, tctx);
 		} else {
 			/*
@@ -1125,12 +1277,12 @@ arena_prof_tctx_set(const void *ptr, size_t usize, prof_tctx_t *tctx)
 			assert(arena_mapbits_large_get(chunk, pageind) == 0);
 		}
 	} else
-		huge_prof_tctx_set(ptr, tctx);
+		huge_prof_tctx_set(tsdn, ptr, tctx);
 }
 
 JEMALLOC_INLINE void
-arena_prof_tctx_reset(const void *ptr, size_t usize, const void *old_ptr,
-    prof_tctx_t *old_tctx)
+arena_prof_tctx_reset(tsdn_t *tsdn, const void *ptr, size_t usize,
+    const void *old_ptr, prof_tctx_t *old_tctx)
 {
 
 	cassert(config_prof);
@@ -1149,43 +1301,59 @@ arena_prof_tctx_reset(const void *ptr, size_t usize, const void *old_ptr,
 			    0);
 			assert(arena_mapbits_large_get(chunk, pageind) != 0);
 
-			elm = arena_miscelm_get(chunk, pageind);
+			elm = arena_miscelm_get_mutable(chunk, pageind);
 			atomic_write_p(&elm->prof_tctx_pun,
 			    (prof_tctx_t *)(uintptr_t)1U);
 		} else
-			huge_prof_tctx_reset(ptr);
+			huge_prof_tctx_reset(tsdn, ptr);
 	}
 }
 
+JEMALLOC_ALWAYS_INLINE void
+arena_decay_ticks(tsdn_t *tsdn, arena_t *arena, unsigned nticks)
+{
+	tsd_t *tsd;
+	ticker_t *decay_ticker;
+
+	if (unlikely(tsdn_null(tsdn)))
+		return;
+	tsd = tsdn_tsd(tsdn);
+	decay_ticker = decay_ticker_get(tsd, arena->ind);
+	if (unlikely(decay_ticker == NULL))
+		return;
+	if (unlikely(ticker_ticks(decay_ticker, nticks)))
+		arena_purge(tsdn, arena, false);
+}
+
+JEMALLOC_ALWAYS_INLINE void
+arena_decay_tick(tsdn_t *tsdn, arena_t *arena)
+{
+
+	arena_decay_ticks(tsdn, arena, 1);
+}
+
 JEMALLOC_ALWAYS_INLINE void *
-arena_malloc(tsd_t *tsd, arena_t *arena, size_t size, bool zero,
-    tcache_t *tcache)
+arena_malloc(tsdn_t *tsdn, arena_t *arena, size_t size, szind_t ind, bool zero,
+    tcache_t *tcache, bool slow_path)
 {
 
+	assert(!tsdn_null(tsdn) || tcache == NULL);
 	assert(size != 0);
 
-	arena = arena_choose(tsd, arena);
-	if (unlikely(arena == NULL))
-		return (NULL);
+	if (likely(tcache != NULL)) {
+		if (likely(size <= SMALL_MAXCLASS)) {
+			return (tcache_alloc_small(tsdn_tsd(tsdn), arena,
+			    tcache, size, ind, zero, slow_path));
+		}
+		if (likely(size <= tcache_maxclass)) {
+			return (tcache_alloc_large(tsdn_tsd(tsdn), arena,
+			    tcache, size, ind, zero, slow_path));
+		}
+		/* (size > tcache_maxclass) case falls through. */
+		assert(size > tcache_maxclass);
+	}
 
-	if (likely(size <= SMALL_MAXCLASS)) {
-		if (likely(tcache != NULL)) {
-			return (tcache_alloc_small(tsd, arena, tcache, size,
-			    zero));
-		} else
-			return (arena_malloc_small(arena, size, zero));
-	} else if (likely(size <= large_maxclass)) {
-		/*
-		 * Initialize tcache after checking size in order to avoid
-		 * infinite recursion during tcache initialization.
-		 */
-		if (likely(tcache != NULL) && size <= tcache_maxclass) {
-			return (tcache_alloc_large(tsd, arena, tcache, size,
-			    zero));
-		} else
-			return (arena_malloc_large(arena, size, zero));
-	} else
-		return (huge_malloc(tsd, arena, size, zero, tcache));
+	return (arena_malloc_hard(tsdn, arena, size, ind, zero));
 }
 
 JEMALLOC_ALWAYS_INLINE arena_t *
@@ -1202,7 +1370,7 @@ arena_aalloc(const void *ptr)
 
 /* Return the size of the allocation pointed to by ptr. */
 JEMALLOC_ALWAYS_INLINE size_t
-arena_salloc(const void *ptr, bool demote)
+arena_salloc(tsdn_t *tsdn, const void *ptr, bool demote)
 {
 	size_t ret;
 	arena_chunk_t *chunk;
@@ -1245,17 +1413,18 @@ arena_salloc(const void *ptr, bool demote)
 			ret = index2size(binind);
 		}
 	} else
-		ret = huge_salloc(ptr);
+		ret = huge_salloc(tsdn, ptr);
 
 	return (ret);
 }
 
 JEMALLOC_ALWAYS_INLINE void
-arena_dalloc(tsd_t *tsd, void *ptr, tcache_t *tcache)
+arena_dalloc(tsdn_t *tsdn, void *ptr, tcache_t *tcache, bool slow_path)
 {
 	arena_chunk_t *chunk;
 	size_t pageind, mapbits;
 
+	assert(!tsdn_null(tsdn) || tcache == NULL);
 	assert(ptr != NULL);
 
 	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(ptr);
@@ -1268,10 +1437,12 @@ arena_dalloc(tsd_t *tsd, void *ptr, tcache_t *tcache)
 			if (likely(tcache != NULL)) {
 				szind_t binind = arena_ptr_small_binind_get(ptr,
 				    mapbits);
-				tcache_dalloc_small(tsd, tcache, ptr, binind);
+				tcache_dalloc_small(tsdn_tsd(tsdn), tcache, ptr,
+				    binind, slow_path);
 			} else {
-				arena_dalloc_small(extent_node_arena_get(
-				    &chunk->node), chunk, ptr, pageind);
+				arena_dalloc_small(tsdn,
+				    extent_node_arena_get(&chunk->node), chunk,
+				    ptr, pageind);
 			}
 		} else {
 			size_t size = arena_mapbits_large_size_get(chunk,
@@ -1282,28 +1453,33 @@ arena_dalloc(tsd_t *tsd, void *ptr, tcache_t *tcache)
 
 			if (likely(tcache != NULL) && size - large_pad <=
 			    tcache_maxclass) {
-				tcache_dalloc_large(tsd, tcache, ptr, size -
-				    large_pad);
+				tcache_dalloc_large(tsdn_tsd(tsdn), tcache, ptr,
+				    size - large_pad, slow_path);
 			} else {
-				arena_dalloc_large(extent_node_arena_get(
-				    &chunk->node), chunk, ptr);
+				arena_dalloc_large(tsdn,
+				    extent_node_arena_get(&chunk->node), chunk,
+				    ptr);
 			}
 		}
 	} else
-		huge_dalloc(tsd, ptr, tcache);
+		huge_dalloc(tsdn, ptr);
 }
 
 JEMALLOC_ALWAYS_INLINE void
-arena_sdalloc(tsd_t *tsd, void *ptr, size_t size, tcache_t *tcache)
+arena_sdalloc(tsdn_t *tsdn, void *ptr, size_t size, tcache_t *tcache,
+    bool slow_path)
 {
 	arena_chunk_t *chunk;
 
+	assert(!tsdn_null(tsdn) || tcache == NULL);
+
 	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(ptr);
 	if (likely(chunk != ptr)) {
 		if (config_prof && opt_prof) {
 			size_t pageind = ((uintptr_t)ptr - (uintptr_t)chunk) >>
 			    LG_PAGE;
-			assert(arena_mapbits_allocated_get(chunk, pageind) != 0);
+			assert(arena_mapbits_allocated_get(chunk, pageind) !=
+			    0);
 			if (arena_mapbits_large_get(chunk, pageind) != 0) {
 				/*
 				 * Make sure to use promoted size, not request
@@ -1313,32 +1489,36 @@ arena_sdalloc(tsd_t *tsd, void *ptr, size_t size, tcache_t *tcache)
 				    pageind) - large_pad;
 			}
 		}
-		assert(s2u(size) == s2u(arena_salloc(ptr, false)));
+		assert(s2u(size) == s2u(arena_salloc(tsdn, ptr, false)));
 
 		if (likely(size <= SMALL_MAXCLASS)) {
 			/* Small allocation. */
 			if (likely(tcache != NULL)) {
 				szind_t binind = size2index(size);
-				tcache_dalloc_small(tsd, tcache, ptr, binind);
+				tcache_dalloc_small(tsdn_tsd(tsdn), tcache, ptr,
+				    binind, slow_path);
 			} else {
 				size_t pageind = ((uintptr_t)ptr -
 				    (uintptr_t)chunk) >> LG_PAGE;
-				arena_dalloc_small(extent_node_arena_get(
-				    &chunk->node), chunk, ptr, pageind);
+				arena_dalloc_small(tsdn,
+				    extent_node_arena_get(&chunk->node), chunk,
+				    ptr, pageind);
 			}
 		} else {
 			assert(config_cache_oblivious || ((uintptr_t)ptr &
 			    PAGE_MASK) == 0);
 
-			if (likely(tcache != NULL) && size <= tcache_maxclass)
-				tcache_dalloc_large(tsd, tcache, ptr, size);
-			else {
-				arena_dalloc_large(extent_node_arena_get(
-				    &chunk->node), chunk, ptr);
+			if (likely(tcache != NULL) && size <= tcache_maxclass) {
+				tcache_dalloc_large(tsdn_tsd(tsdn), tcache, ptr,
+				    size, slow_path);
+			} else {
+				arena_dalloc_large(tsdn,
+				    extent_node_arena_get(&chunk->node), chunk,
+				    ptr);
 			}
 		}
 	} else
-		huge_dalloc(tsd, ptr, tcache);
+		huge_dalloc(tsdn, ptr);
 }
 #  endif /* JEMALLOC_ARENA_INLINE_B */
 #endif
diff --git a/deps/jemalloc/include/jemalloc/internal/assert.h b/deps/jemalloc/include/jemalloc/internal/assert.h
new file mode 100644
index 000000000..6f8f7eb93
--- /dev/null
+++ b/deps/jemalloc/include/jemalloc/internal/assert.h
@@ -0,0 +1,45 @@
+/*
+ * Define a custom assert() in order to reduce the chances of deadlock during
+ * assertion failure.
+ */
+#ifndef assert
+#define	assert(e) do {							\
+	if (unlikely(config_debug && !(e))) {				\
+		malloc_printf(						\
+		    "<jemalloc>: %s:%d: Failed assertion: \"%s\"\n",	\
+		    __FILE__, __LINE__, #e);				\
+		abort();						\
+	}								\
+} while (0)
+#endif
+
+#ifndef not_reached
+#define	not_reached() do {						\
+	if (config_debug) {						\
+		malloc_printf(						\
+		    "<jemalloc>: %s:%d: Unreachable code reached\n",	\
+		    __FILE__, __LINE__);				\
+		abort();						\
+	}								\
+	unreachable();							\
+} while (0)
+#endif
+
+#ifndef not_implemented
+#define	not_implemented() do {						\
+	if (config_debug) {						\
+		malloc_printf("<jemalloc>: %s:%d: Not implemented\n",	\
+		    __FILE__, __LINE__);				\
+		abort();						\
+	}								\
+} while (0)
+#endif
+
+#ifndef assert_not_implemented
+#define	assert_not_implemented(e) do {					\
+	if (unlikely(config_debug && !(e)))				\
+		not_implemented();					\
+} while (0)
+#endif
+
+
diff --git a/deps/jemalloc/include/jemalloc/internal/atomic.h b/deps/jemalloc/include/jemalloc/internal/atomic.h
index a9aad35d1..3f15ea149 100644
--- a/deps/jemalloc/include/jemalloc/internal/atomic.h
+++ b/deps/jemalloc/include/jemalloc/internal/atomic.h
@@ -28,8 +28,8 @@
  * callers.
  *
  *   <t> atomic_read_<t>(<t> *p) { return (*p); }
- *   <t> atomic_add_<t>(<t> *p, <t> x) { return (*p + x); }
- *   <t> atomic_sub_<t>(<t> *p, <t> x) { return (*p - x); }
+ *   <t> atomic_add_<t>(<t> *p, <t> x) { return (*p += x); }
+ *   <t> atomic_sub_<t>(<t> *p, <t> x) { return (*p -= x); }
  *   bool atomic_cas_<t>(<t> *p, <t> c, <t> s)
  *   {
  *     if (*p != c)
diff --git a/deps/jemalloc/include/jemalloc/internal/base.h b/deps/jemalloc/include/jemalloc/internal/base.h
index 39e46ee44..d6b81e162 100644
--- a/deps/jemalloc/include/jemalloc/internal/base.h
+++ b/deps/jemalloc/include/jemalloc/internal/base.h
@@ -9,12 +9,13 @@
 /******************************************************************************/
 #ifdef JEMALLOC_H_EXTERNS
 
-void	*base_alloc(size_t size);
-void	base_stats_get(size_t *allocated, size_t *resident, size_t *mapped);
+void	*base_alloc(tsdn_t *tsdn, size_t size);
+void	base_stats_get(tsdn_t *tsdn, size_t *allocated, size_t *resident,
+    size_t *mapped);
 bool	base_boot(void);
-void	base_prefork(void);
-void	base_postfork_parent(void);
-void	base_postfork_child(void);
+void	base_prefork(tsdn_t *tsdn);
+void	base_postfork_parent(tsdn_t *tsdn);
+void	base_postfork_child(tsdn_t *tsdn);
 
 #endif /* JEMALLOC_H_EXTERNS */
 /******************************************************************************/
diff --git a/deps/jemalloc/include/jemalloc/internal/bitmap.h b/deps/jemalloc/include/jemalloc/internal/bitmap.h
index fcc6005c7..36f38b59c 100644
--- a/deps/jemalloc/include/jemalloc/internal/bitmap.h
+++ b/deps/jemalloc/include/jemalloc/internal/bitmap.h
@@ -15,6 +15,15 @@ typedef unsigned long bitmap_t;
 #define	BITMAP_GROUP_NBITS		(ZU(1) << LG_BITMAP_GROUP_NBITS)
 #define	BITMAP_GROUP_NBITS_MASK		(BITMAP_GROUP_NBITS-1)
 
+/*
+ * Do some analysis on how big the bitmap is before we use a tree.  For a brute
+ * force linear search, if we would have to call ffs_lu() more than 2^3 times,
+ * use a tree instead.
+ */
+#if LG_BITMAP_MAXBITS - LG_BITMAP_GROUP_NBITS > 3
+#  define USE_TREE
+#endif
+
 /* Number of groups required to store a given number of bits. */
 #define	BITMAP_BITS2GROUPS(nbits)					\
     ((nbits + BITMAP_GROUP_NBITS_MASK) >> LG_BITMAP_GROUP_NBITS)
@@ -48,6 +57,8 @@ typedef unsigned long bitmap_t;
 /*
  * Maximum number of groups required to support LG_BITMAP_MAXBITS.
  */
+#ifdef USE_TREE
+
 #if LG_BITMAP_MAXBITS <= LG_BITMAP_GROUP_NBITS
 #  define BITMAP_GROUPS_MAX	BITMAP_GROUPS_1_LEVEL(BITMAP_MAXBITS)
 #elif LG_BITMAP_MAXBITS <= LG_BITMAP_GROUP_NBITS * 2
@@ -65,6 +76,12 @@ typedef unsigned long bitmap_t;
     (LG_BITMAP_MAXBITS / LG_SIZEOF_BITMAP)				\
     + !!(LG_BITMAP_MAXBITS % LG_SIZEOF_BITMAP)
 
+#else /* USE_TREE */
+
+#define	BITMAP_GROUPS_MAX BITMAP_BITS2GROUPS(BITMAP_MAXBITS)
+
+#endif /* USE_TREE */
+
 #endif /* JEMALLOC_H_TYPES */
 /******************************************************************************/
 #ifdef JEMALLOC_H_STRUCTS
@@ -78,6 +95,7 @@ struct bitmap_info_s {
 	/* Logical number of bits in bitmap (stored at bottom level). */
 	size_t nbits;
 
+#ifdef USE_TREE
 	/* Number of levels necessary for nbits. */
 	unsigned nlevels;
 
@@ -86,6 +104,10 @@ struct bitmap_info_s {
 	 * bottom to top (e.g. the bottom level is stored in levels[0]).
 	 */
 	bitmap_level_t levels[BITMAP_MAX_LEVELS+1];
+#else /* USE_TREE */
+	/* Number of groups necessary for nbits. */
+	size_t ngroups;
+#endif /* USE_TREE */
 };
 
 #endif /* JEMALLOC_H_STRUCTS */
@@ -93,9 +115,8 @@ struct bitmap_info_s {
 #ifdef JEMALLOC_H_EXTERNS
 
 void	bitmap_info_init(bitmap_info_t *binfo, size_t nbits);
-size_t	bitmap_info_ngroups(const bitmap_info_t *binfo);
-size_t	bitmap_size(size_t nbits);
 void	bitmap_init(bitmap_t *bitmap, const bitmap_info_t *binfo);
+size_t	bitmap_size(const bitmap_info_t *binfo);
 
 #endif /* JEMALLOC_H_EXTERNS */
 /******************************************************************************/
@@ -113,10 +134,20 @@ void	bitmap_unset(bitmap_t *bitmap, const bitmap_info_t *binfo, size_t bit);
 JEMALLOC_INLINE bool
 bitmap_full(bitmap_t *bitmap, const bitmap_info_t *binfo)
 {
-	unsigned rgoff = binfo->levels[binfo->nlevels].group_offset - 1;
+#ifdef USE_TREE
+	size_t rgoff = binfo->levels[binfo->nlevels].group_offset - 1;
 	bitmap_t rg = bitmap[rgoff];
 	/* The bitmap is full iff the root group is 0. */
 	return (rg == 0);
+#else
+	size_t i;
+
+	for (i = 0; i < binfo->ngroups; i++) {
+		if (bitmap[i] != 0)
+			return (false);
+	}
+	return (true);
+#endif
 }
 
 JEMALLOC_INLINE bool
@@ -128,7 +159,7 @@ bitmap_get(bitmap_t *bitmap, const bitmap_info_t *binfo, size_t bit)
 	assert(bit < binfo->nbits);
 	goff = bit >> LG_BITMAP_GROUP_NBITS;
 	g = bitmap[goff];
-	return (!(g & (1LU << (bit & BITMAP_GROUP_NBITS_MASK))));
+	return (!(g & (ZU(1) << (bit & BITMAP_GROUP_NBITS_MASK))));
 }
 
 JEMALLOC_INLINE void
@@ -143,10 +174,11 @@ bitmap_set(bitmap_t *bitmap, const bitmap_info_t *binfo, size_t bit)
 	goff = bit >> LG_BITMAP_GROUP_NBITS;
 	gp = &bitmap[goff];
 	g = *gp;
-	assert(g & (1LU << (bit & BITMAP_GROUP_NBITS_MASK)));
-	g ^= 1LU << (bit & BITMAP_GROUP_NBITS_MASK);
+	assert(g & (ZU(1) << (bit & BITMAP_GROUP_NBITS_MASK)));
+	g ^= ZU(1) << (bit & BITMAP_GROUP_NBITS_MASK);
 	*gp = g;
 	assert(bitmap_get(bitmap, binfo, bit));
+#ifdef USE_TREE
 	/* Propagate group state transitions up the tree. */
 	if (g == 0) {
 		unsigned i;
@@ -155,13 +187,14 @@ bitmap_set(bitmap_t *bitmap, const bitmap_info_t *binfo, size_t bit)
 			goff = bit >> LG_BITMAP_GROUP_NBITS;
 			gp = &bitmap[binfo->levels[i].group_offset + goff];
 			g = *gp;
-			assert(g & (1LU << (bit & BITMAP_GROUP_NBITS_MASK)));
-			g ^= 1LU << (bit & BITMAP_GROUP_NBITS_MASK);
+			assert(g & (ZU(1) << (bit & BITMAP_GROUP_NBITS_MASK)));
+			g ^= ZU(1) << (bit & BITMAP_GROUP_NBITS_MASK);
 			*gp = g;
 			if (g != 0)
 				break;
 		}
 	}
+#endif
 }
 
 /* sfu: set first unset. */
@@ -174,15 +207,24 @@ bitmap_sfu(bitmap_t *bitmap, const bitmap_info_t *binfo)
 
 	assert(!bitmap_full(bitmap, binfo));
 
+#ifdef USE_TREE
 	i = binfo->nlevels - 1;
 	g = bitmap[binfo->levels[i].group_offset];
-	bit = jemalloc_ffsl(g) - 1;
+	bit = ffs_lu(g) - 1;
 	while (i > 0) {
 		i--;
 		g = bitmap[binfo->levels[i].group_offset + bit];
-		bit = (bit << LG_BITMAP_GROUP_NBITS) + (jemalloc_ffsl(g) - 1);
+		bit = (bit << LG_BITMAP_GROUP_NBITS) + (ffs_lu(g) - 1);
 	}
-
+#else
+	i = 0;
+	g = bitmap[0];
+	while ((bit = ffs_lu(g)) == 0) {
+		i++;
+		g = bitmap[i];
+	}
+	bit = (i << LG_BITMAP_GROUP_NBITS) + (bit - 1);
+#endif
 	bitmap_set(bitmap, binfo, bit);
 	return (bit);
 }
@@ -193,7 +235,7 @@ bitmap_unset(bitmap_t *bitmap, const bitmap_info_t *binfo, size_t bit)
 	size_t goff;
 	bitmap_t *gp;
 	bitmap_t g;
-	bool propagate;
+	UNUSED bool propagate;
 
 	assert(bit < binfo->nbits);
 	assert(bitmap_get(bitmap, binfo, bit));
@@ -201,10 +243,11 @@ bitmap_unset(bitmap_t *bitmap, const bitmap_info_t *binfo, size_t bit)
 	gp = &bitmap[goff];
 	g = *gp;
 	propagate = (g == 0);
-	assert((g & (1LU << (bit & BITMAP_GROUP_NBITS_MASK))) == 0);
-	g ^= 1LU << (bit & BITMAP_GROUP_NBITS_MASK);
+	assert((g & (ZU(1) << (bit & BITMAP_GROUP_NBITS_MASK))) == 0);
+	g ^= ZU(1) << (bit & BITMAP_GROUP_NBITS_MASK);
 	*gp = g;
 	assert(!bitmap_get(bitmap, binfo, bit));
+#ifdef USE_TREE
 	/* Propagate group state transitions up the tree. */
 	if (propagate) {
 		unsigned i;
@@ -214,14 +257,15 @@ bitmap_unset(bitmap_t *bitmap, const bitmap_info_t *binfo, size_t bit)
 			gp = &bitmap[binfo->levels[i].group_offset + goff];
 			g = *gp;
 			propagate = (g == 0);
-			assert((g & (1LU << (bit & BITMAP_GROUP_NBITS_MASK)))
+			assert((g & (ZU(1) << (bit & BITMAP_GROUP_NBITS_MASK)))
 			    == 0);
-			g ^= 1LU << (bit & BITMAP_GROUP_NBITS_MASK);
+			g ^= ZU(1) << (bit & BITMAP_GROUP_NBITS_MASK);
 			*gp = g;
 			if (!propagate)
 				break;
 		}
 	}
+#endif /* USE_TREE */
 }
 
 #endif
diff --git a/deps/jemalloc/include/jemalloc/internal/chunk.h b/deps/jemalloc/include/jemalloc/internal/chunk.h
index 5d1938353..50b9904b0 100644
--- a/deps/jemalloc/include/jemalloc/internal/chunk.h
+++ b/deps/jemalloc/include/jemalloc/internal/chunk.h
@@ -48,32 +48,30 @@ extern size_t		chunk_npages;
 
 extern const chunk_hooks_t	chunk_hooks_default;
 
-chunk_hooks_t	chunk_hooks_get(arena_t *arena);
-chunk_hooks_t	chunk_hooks_set(arena_t *arena,
+chunk_hooks_t	chunk_hooks_get(tsdn_t *tsdn, arena_t *arena);
+chunk_hooks_t	chunk_hooks_set(tsdn_t *tsdn, arena_t *arena,
     const chunk_hooks_t *chunk_hooks);
 
-bool	chunk_register(const void *chunk, const extent_node_t *node);
+bool	chunk_register(tsdn_t *tsdn, const void *chunk,
+    const extent_node_t *node);
 void	chunk_deregister(const void *chunk, const extent_node_t *node);
 void	*chunk_alloc_base(size_t size);
-void	*chunk_alloc_cache(arena_t *arena, chunk_hooks_t *chunk_hooks,
-    void *new_addr, size_t size, size_t alignment, bool *zero,
-    bool dalloc_node);
-void	*chunk_alloc_wrapper(arena_t *arena, chunk_hooks_t *chunk_hooks,
-    void *new_addr, size_t size, size_t alignment, bool *zero, bool *commit);
-void	chunk_dalloc_cache(arena_t *arena, chunk_hooks_t *chunk_hooks,
-    void *chunk, size_t size, bool committed);
-void	chunk_dalloc_arena(arena_t *arena, chunk_hooks_t *chunk_hooks,
-    void *chunk, size_t size, bool zeroed, bool committed);
-void	chunk_dalloc_wrapper(arena_t *arena, chunk_hooks_t *chunk_hooks,
-    void *chunk, size_t size, bool committed);
-bool	chunk_purge_arena(arena_t *arena, void *chunk, size_t offset,
+void	*chunk_alloc_cache(tsdn_t *tsdn, arena_t *arena,
+    chunk_hooks_t *chunk_hooks, void *new_addr, size_t size, size_t alignment,
+    size_t *sn, bool *zero, bool *commit, bool dalloc_node);
+void	*chunk_alloc_wrapper(tsdn_t *tsdn, arena_t *arena,
+    chunk_hooks_t *chunk_hooks, void *new_addr, size_t size, size_t alignment,
+    size_t *sn, bool *zero, bool *commit);
+void	chunk_dalloc_cache(tsdn_t *tsdn, arena_t *arena,
+    chunk_hooks_t *chunk_hooks, void *chunk, size_t size, size_t sn,
+    bool committed);
+void	chunk_dalloc_wrapper(tsdn_t *tsdn, arena_t *arena,
+    chunk_hooks_t *chunk_hooks, void *chunk, size_t size, size_t sn,
+    bool zeroed, bool committed);
+bool	chunk_purge_wrapper(tsdn_t *tsdn, arena_t *arena,
+    chunk_hooks_t *chunk_hooks, void *chunk, size_t size, size_t offset,
     size_t length);
-bool	chunk_purge_wrapper(arena_t *arena, chunk_hooks_t *chunk_hooks,
-    void *chunk, size_t size, size_t offset, size_t length);
 bool	chunk_boot(void);
-void	chunk_prefork(void);
-void	chunk_postfork_parent(void);
-void	chunk_postfork_child(void);
 
 #endif /* JEMALLOC_H_EXTERNS */
 /******************************************************************************/
diff --git a/deps/jemalloc/include/jemalloc/internal/chunk_dss.h b/deps/jemalloc/include/jemalloc/internal/chunk_dss.h
index 388f46be0..da8511ba0 100644
--- a/deps/jemalloc/include/jemalloc/internal/chunk_dss.h
+++ b/deps/jemalloc/include/jemalloc/internal/chunk_dss.h
@@ -23,13 +23,11 @@ extern const char *dss_prec_names[];
 
 dss_prec_t	chunk_dss_prec_get(void);
 bool	chunk_dss_prec_set(dss_prec_t dss_prec);
-void	*chunk_alloc_dss(arena_t *arena, void *new_addr, size_t size,
-    size_t alignment, bool *zero, bool *commit);
+void	*chunk_alloc_dss(tsdn_t *tsdn, arena_t *arena, void *new_addr,
+    size_t size, size_t alignment, bool *zero, bool *commit);
 bool	chunk_in_dss(void *chunk);
-bool	chunk_dss_boot(void);
-void	chunk_dss_prefork(void);
-void	chunk_dss_postfork_parent(void);
-void	chunk_dss_postfork_child(void);
+bool	chunk_dss_mergeable(void *chunk_a, void *chunk_b);
+void	chunk_dss_boot(void);
 
 #endif /* JEMALLOC_H_EXTERNS */
 /******************************************************************************/
diff --git a/deps/jemalloc/include/jemalloc/internal/chunk_mmap.h b/deps/jemalloc/include/jemalloc/internal/chunk_mmap.h
index 7d8014c58..6f2d0ac2e 100644
--- a/deps/jemalloc/include/jemalloc/internal/chunk_mmap.h
+++ b/deps/jemalloc/include/jemalloc/internal/chunk_mmap.h
@@ -9,8 +9,8 @@
 /******************************************************************************/
 #ifdef JEMALLOC_H_EXTERNS
 
-void	*chunk_alloc_mmap(size_t size, size_t alignment, bool *zero,
-    bool *commit);
+void	*chunk_alloc_mmap(void *new_addr, size_t size, size_t alignment,
+    bool *zero, bool *commit);
 bool	chunk_dalloc_mmap(void *chunk, size_t size);
 
 #endif /* JEMALLOC_H_EXTERNS */
diff --git a/deps/jemalloc/include/jemalloc/internal/ckh.h b/deps/jemalloc/include/jemalloc/internal/ckh.h
index 75c1c979f..f75ad90b7 100644
--- a/deps/jemalloc/include/jemalloc/internal/ckh.h
+++ b/deps/jemalloc/include/jemalloc/internal/ckh.h
@@ -40,9 +40,7 @@ struct ckh_s {
 #endif
 
 	/* Used for pseudo-random number generation. */
-#define	CKH_A		1103515241
-#define	CKH_C		12347
-	uint32_t	prng_state;
+	uint64_t	prng_state;
 
 	/* Total number of items. */
 	size_t		count;
@@ -74,7 +72,7 @@ bool	ckh_iter(ckh_t *ckh, size_t *tabind, void **key, void **data);
 bool	ckh_insert(tsd_t *tsd, ckh_t *ckh, const void *key, const void *data);
 bool	ckh_remove(tsd_t *tsd, ckh_t *ckh, const void *searchkey, void **key,
     void **data);
-bool	ckh_search(ckh_t *ckh, const void *seachkey, void **key, void **data);
+bool	ckh_search(ckh_t *ckh, const void *searchkey, void **key, void **data);
 void	ckh_string_hash(const void *key, size_t r_hash[2]);
 bool	ckh_string_keycomp(const void *k1, const void *k2);
 void	ckh_pointer_hash(const void *key, size_t r_hash[2]);
diff --git a/deps/jemalloc/include/jemalloc/internal/ctl.h b/deps/jemalloc/include/jemalloc/internal/ctl.h
index 751c14b5b..af0f6d7c5 100644
--- a/deps/jemalloc/include/jemalloc/internal/ctl.h
+++ b/deps/jemalloc/include/jemalloc/internal/ctl.h
@@ -21,13 +21,14 @@ struct ctl_named_node_s {
 	/* If (nchildren == 0), this is a terminal node. */
 	unsigned		nchildren;
 	const			ctl_node_t *children;
-	int			(*ctl)(const size_t *, size_t, void *, size_t *,
-	    void *, size_t);
+	int			(*ctl)(tsd_t *, const size_t *, size_t, void *,
+	    size_t *, void *, size_t);
 };
 
 struct ctl_indexed_node_s {
 	struct ctl_node_s	node;
-	const ctl_named_node_t	*(*index)(const size_t *, size_t, size_t);
+	const ctl_named_node_t	*(*index)(tsdn_t *, const size_t *, size_t,
+	    size_t);
 };
 
 struct ctl_arena_stats_s {
@@ -35,8 +36,12 @@ struct ctl_arena_stats_s {
 	unsigned		nthreads;
 	const char		*dss;
 	ssize_t			lg_dirty_mult;
+	ssize_t			decay_time;
 	size_t			pactive;
 	size_t			pdirty;
+
+	/* The remainder are only populated if config_stats is true. */
+
 	arena_stats_t		astats;
 
 	/* Aggregate stats for small size classes, based on bin stats. */
@@ -56,6 +61,7 @@ struct ctl_stats_s {
 	size_t			metadata;
 	size_t			resident;
 	size_t			mapped;
+	size_t			retained;
 	unsigned		narenas;
 	ctl_arena_stats_t	*arenas;	/* (narenas + 1) elements. */
 };
@@ -64,16 +70,17 @@ struct ctl_stats_s {
 /******************************************************************************/
 #ifdef JEMALLOC_H_EXTERNS
 
-int	ctl_byname(const char *name, void *oldp, size_t *oldlenp, void *newp,
-    size_t newlen);
-int	ctl_nametomib(const char *name, size_t *mibp, size_t *miblenp);
-
-int	ctl_bymib(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
+int	ctl_byname(tsd_t *tsd, const char *name, void *oldp, size_t *oldlenp,
     void *newp, size_t newlen);
+int	ctl_nametomib(tsdn_t *tsdn, const char *name, size_t *mibp,
+    size_t *miblenp);
+
+int	ctl_bymib(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen);
 bool	ctl_boot(void);
-void	ctl_prefork(void);
-void	ctl_postfork_parent(void);
-void	ctl_postfork_child(void);
+void	ctl_prefork(tsdn_t *tsdn);
+void	ctl_postfork_parent(tsdn_t *tsdn);
+void	ctl_postfork_child(tsdn_t *tsdn);
 
 #define	xmallctl(name, oldp, oldlenp, newp, newlen) do {		\
 	if (je_mallctl(name, oldp, oldlenp, newp, newlen)		\
diff --git a/deps/jemalloc/include/jemalloc/internal/extent.h b/deps/jemalloc/include/jemalloc/internal/extent.h
index 386d50ef4..168ffe643 100644
--- a/deps/jemalloc/include/jemalloc/internal/extent.h
+++ b/deps/jemalloc/include/jemalloc/internal/extent.h
@@ -18,6 +18,20 @@ struct extent_node_s {
 	/* Total region size. */
 	size_t			en_size;
 
+	/*
+	 * Serial number (potentially non-unique).
+	 *
+	 * In principle serial numbers can wrap around on 32-bit systems if
+	 * JEMALLOC_MUNMAP is defined, but as long as comparison functions fall
+	 * back on address comparison for equal serial numbers, stable (if
+	 * imperfect) ordering is maintained.
+	 *
+	 * Serial numbers may not be unique even in the absence of wrap-around,
+	 * e.g. when splitting an extent and assigning the same serial number to
+	 * both resulting adjacent extents.
+	 */
+	size_t			en_sn;
+
 	/*
 	 * The zeroed flag is used by chunk recycling code to track whether
 	 * memory is zero-filled.
@@ -45,10 +59,10 @@ struct extent_node_s {
 	qr(extent_node_t)	cc_link;
 
 	union {
-		/* Linkage for the size/address-ordered tree. */
-		rb_node(extent_node_t)	szad_link;
+		/* Linkage for the size/sn/address-ordered tree. */
+		rb_node(extent_node_t)	szsnad_link;
 
-		/* Linkage for arena's huge and node_cache lists. */
+		/* Linkage for arena's achunks, huge, and node_cache lists. */
 		ql_elm(extent_node_t)	ql_link;
 	};
 
@@ -61,7 +75,7 @@ typedef rb_tree(extent_node_t) extent_tree_t;
 /******************************************************************************/
 #ifdef JEMALLOC_H_EXTERNS
 
-rb_proto(, extent_tree_szad_, extent_tree_t, extent_node_t)
+rb_proto(, extent_tree_szsnad_, extent_tree_t, extent_node_t)
 
 rb_proto(, extent_tree_ad_, extent_tree_t, extent_node_t)
 
@@ -73,6 +87,7 @@ rb_proto(, extent_tree_ad_, extent_tree_t, extent_node_t)
 arena_t	*extent_node_arena_get(const extent_node_t *node);
 void	*extent_node_addr_get(const extent_node_t *node);
 size_t	extent_node_size_get(const extent_node_t *node);
+size_t	extent_node_sn_get(const extent_node_t *node);
 bool	extent_node_zeroed_get(const extent_node_t *node);
 bool	extent_node_committed_get(const extent_node_t *node);
 bool	extent_node_achunk_get(const extent_node_t *node);
@@ -80,12 +95,13 @@ prof_tctx_t	*extent_node_prof_tctx_get(const extent_node_t *node);
 void	extent_node_arena_set(extent_node_t *node, arena_t *arena);
 void	extent_node_addr_set(extent_node_t *node, void *addr);
 void	extent_node_size_set(extent_node_t *node, size_t size);
+void	extent_node_sn_set(extent_node_t *node, size_t sn);
 void	extent_node_zeroed_set(extent_node_t *node, bool zeroed);
 void	extent_node_committed_set(extent_node_t *node, bool committed);
 void	extent_node_achunk_set(extent_node_t *node, bool achunk);
 void	extent_node_prof_tctx_set(extent_node_t *node, prof_tctx_t *tctx);
 void	extent_node_init(extent_node_t *node, arena_t *arena, void *addr,
-    size_t size, bool zeroed, bool committed);
+    size_t size, size_t sn, bool zeroed, bool committed);
 void	extent_node_dirty_linkage_init(extent_node_t *node);
 void	extent_node_dirty_insert(extent_node_t *node,
     arena_runs_dirty_link_t *runs_dirty, extent_node_t *chunks_dirty);
@@ -114,6 +130,13 @@ extent_node_size_get(const extent_node_t *node)
 	return (node->en_size);
 }
 
+JEMALLOC_INLINE size_t
+extent_node_sn_get(const extent_node_t *node)
+{
+
+	return (node->en_sn);
+}
+
 JEMALLOC_INLINE bool
 extent_node_zeroed_get(const extent_node_t *node)
 {
@@ -164,6 +187,13 @@ extent_node_size_set(extent_node_t *node, size_t size)
 	node->en_size = size;
 }
 
+JEMALLOC_INLINE void
+extent_node_sn_set(extent_node_t *node, size_t sn)
+{
+
+	node->en_sn = sn;
+}
+
 JEMALLOC_INLINE void
 extent_node_zeroed_set(extent_node_t *node, bool zeroed)
 {
@@ -194,12 +224,13 @@ extent_node_prof_tctx_set(extent_node_t *node, prof_tctx_t *tctx)
 
 JEMALLOC_INLINE void
 extent_node_init(extent_node_t *node, arena_t *arena, void *addr, size_t size,
-    bool zeroed, bool committed)
+    size_t sn, bool zeroed, bool committed)
 {
 
 	extent_node_arena_set(node, arena);
 	extent_node_addr_set(node, addr);
 	extent_node_size_set(node, size);
+	extent_node_sn_set(node, sn);
 	extent_node_zeroed_set(node, zeroed);
 	extent_node_committed_set(node, committed);
 	extent_node_achunk_set(node, false);
diff --git a/deps/jemalloc/include/jemalloc/internal/hash.h b/deps/jemalloc/include/jemalloc/internal/hash.h
index bcead337a..1ff2d9a05 100644
--- a/deps/jemalloc/include/jemalloc/internal/hash.h
+++ b/deps/jemalloc/include/jemalloc/internal/hash.h
@@ -1,6 +1,6 @@
 /*
  * The following hash function is based on MurmurHash3, placed into the public
- * domain by Austin Appleby.  See http://code.google.com/p/smhasher/ for
+ * domain by Austin Appleby.  See https://github.com/aappleby/smhasher for
  * details.
  */
 /******************************************************************************/
@@ -49,6 +49,14 @@ JEMALLOC_INLINE uint32_t
 hash_get_block_32(const uint32_t *p, int i)
 {
 
+	/* Handle unaligned read. */
+	if (unlikely((uintptr_t)p & (sizeof(uint32_t)-1)) != 0) {
+		uint32_t ret;
+
+		memcpy(&ret, (uint8_t *)(p + i), sizeof(uint32_t));
+		return (ret);
+	}
+
 	return (p[i]);
 }
 
@@ -56,6 +64,14 @@ JEMALLOC_INLINE uint64_t
 hash_get_block_64(const uint64_t *p, int i)
 {
 
+	/* Handle unaligned read. */
+	if (unlikely((uintptr_t)p & (sizeof(uint64_t)-1)) != 0) {
+		uint64_t ret;
+
+		memcpy(&ret, (uint8_t *)(p + i), sizeof(uint64_t));
+		return (ret);
+	}
+
 	return (p[i]);
 }
 
@@ -321,13 +337,18 @@ hash_x64_128(const void *key, const int len, const uint32_t seed,
 JEMALLOC_INLINE void
 hash(const void *key, size_t len, const uint32_t seed, size_t r_hash[2])
 {
+
+	assert(len <= INT_MAX); /* Unfortunate implementation limitation. */
+
 #if (LG_SIZEOF_PTR == 3 && !defined(JEMALLOC_BIG_ENDIAN))
-	hash_x64_128(key, len, seed, (uint64_t *)r_hash);
+	hash_x64_128(key, (int)len, seed, (uint64_t *)r_hash);
 #else
-	uint64_t hashes[2];
-	hash_x86_128(key, len, seed, hashes);
-	r_hash[0] = (size_t)hashes[0];
-	r_hash[1] = (size_t)hashes[1];
+	{
+		uint64_t hashes[2];
+		hash_x86_128(key, (int)len, seed, hashes);
+		r_hash[0] = (size_t)hashes[0];
+		r_hash[1] = (size_t)hashes[1];
+	}
 #endif
 }
 #endif
diff --git a/deps/jemalloc/include/jemalloc/internal/huge.h b/deps/jemalloc/include/jemalloc/internal/huge.h
index ece7af980..22184d9bb 100644
--- a/deps/jemalloc/include/jemalloc/internal/huge.h
+++ b/deps/jemalloc/include/jemalloc/internal/huge.h
@@ -9,24 +9,23 @@
 /******************************************************************************/
 #ifdef JEMALLOC_H_EXTERNS
 
-void	*huge_malloc(tsd_t *tsd, arena_t *arena, size_t size, bool zero,
-    tcache_t *tcache);
-void	*huge_palloc(tsd_t *tsd, arena_t *arena, size_t size, size_t alignment,
-    bool zero, tcache_t *tcache);
-bool	huge_ralloc_no_move(void *ptr, size_t oldsize, size_t usize_min,
-    size_t usize_max, bool zero);
+void	*huge_malloc(tsdn_t *tsdn, arena_t *arena, size_t usize, bool zero);
+void	*huge_palloc(tsdn_t *tsdn, arena_t *arena, size_t usize,
+    size_t alignment, bool zero);
+bool	huge_ralloc_no_move(tsdn_t *tsdn, void *ptr, size_t oldsize,
+    size_t usize_min, size_t usize_max, bool zero);
 void	*huge_ralloc(tsd_t *tsd, arena_t *arena, void *ptr, size_t oldsize,
     size_t usize, size_t alignment, bool zero, tcache_t *tcache);
 #ifdef JEMALLOC_JET
 typedef void (huge_dalloc_junk_t)(void *, size_t);
 extern huge_dalloc_junk_t *huge_dalloc_junk;
 #endif
-void	huge_dalloc(tsd_t *tsd, void *ptr, tcache_t *tcache);
+void	huge_dalloc(tsdn_t *tsdn, void *ptr);
 arena_t	*huge_aalloc(const void *ptr);
-size_t	huge_salloc(const void *ptr);
-prof_tctx_t	*huge_prof_tctx_get(const void *ptr);
-void	huge_prof_tctx_set(const void *ptr, prof_tctx_t *tctx);
-void	huge_prof_tctx_reset(const void *ptr);
+size_t	huge_salloc(tsdn_t *tsdn, const void *ptr);
+prof_tctx_t	*huge_prof_tctx_get(tsdn_t *tsdn, const void *ptr);
+void	huge_prof_tctx_set(tsdn_t *tsdn, const void *ptr, prof_tctx_t *tctx);
+void	huge_prof_tctx_reset(tsdn_t *tsdn, const void *ptr);
 
 #endif /* JEMALLOC_H_EXTERNS */
 /******************************************************************************/
diff --git a/deps/jemalloc/include/jemalloc/internal/jemalloc_internal.h.in b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal.h.in
index 8536a3eda..e7ace7d8c 100644
--- a/deps/jemalloc/include/jemalloc/internal/jemalloc_internal.h.in
+++ b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal.h.in
@@ -49,6 +49,7 @@ static const bool config_lazy_lock =
     false
 #endif
     ;
+static const char * const config_malloc_conf = JEMALLOC_CONFIG_MALLOC_CONF;
 static const bool config_prof =
 #ifdef JEMALLOC_PROF
     true
@@ -160,7 +161,10 @@ static const bool config_cache_oblivious =
 #include <malloc/malloc.h>
 #endif
 
+#include "jemalloc/internal/ph.h"
+#ifndef __PGI
 #define	RB_COMPACT
+#endif
 #include "jemalloc/internal/rb.h"
 #include "jemalloc/internal/qr.h"
 #include "jemalloc/internal/ql.h"
@@ -183,6 +187,9 @@ static const bool config_cache_oblivious =
 
 #include "jemalloc/internal/jemalloc_internal_macros.h"
 
+/* Page size index type. */
+typedef unsigned pszind_t;
+
 /* Size class index type. */
 typedef unsigned szind_t;
 
@@ -232,7 +239,7 @@ typedef unsigned szind_t;
 #  ifdef __alpha__
 #    define LG_QUANTUM		4
 #  endif
-#  if (defined(__sparc64__) || defined(__sparcv9))
+#  if (defined(__sparc64__) || defined(__sparcv9) || defined(__sparc_v9__))
 #    define LG_QUANTUM		4
 #  endif
 #  if (defined(__amd64__) || defined(__x86_64__) || defined(_M_X64))
@@ -256,6 +263,9 @@ typedef unsigned szind_t;
 #  ifdef __powerpc__
 #    define LG_QUANTUM		4
 #  endif
+#  ifdef __riscv__
+#    define LG_QUANTUM		4
+#  endif
 #  ifdef __s390__
 #    define LG_QUANTUM		4
 #  endif
@@ -317,13 +327,17 @@ typedef unsigned szind_t;
 #define	PAGE		((size_t)(1U << LG_PAGE))
 #define	PAGE_MASK	((size_t)(PAGE - 1))
 
+/* Return the page base address for the page containing address a. */
+#define	PAGE_ADDR2BASE(a)						\
+	((void *)((uintptr_t)(a) & ~PAGE_MASK))
+
 /* Return the smallest pagesize multiple that is >= s. */
 #define	PAGE_CEILING(s)							\
 	(((s) + PAGE_MASK) & ~PAGE_MASK)
 
 /* Return the nearest aligned address at or below a. */
 #define	ALIGNMENT_ADDR2BASE(a, alignment)				\
-	((void *)((uintptr_t)(a) & (-(alignment))))
+	((void *)((uintptr_t)(a) & ((~(alignment)) + 1)))
 
 /* Return the offset between a and the nearest aligned address at or below a. */
 #define	ALIGNMENT_ADDR2OFFSET(a, alignment)				\
@@ -331,7 +345,7 @@ typedef unsigned szind_t;
 
 /* Return the smallest alignment multiple that is >= s. */
 #define	ALIGNMENT_CEILING(s, alignment)					\
-	(((s) + (alignment - 1)) & (-(alignment)))
+	(((s) + (alignment - 1)) & ((~(alignment)) + 1))
 
 /* Declare a variable-length array. */
 #if __STDC_VERSION__ < 199901L
@@ -351,14 +365,19 @@ typedef unsigned szind_t;
 #  define VARIABLE_ARRAY(type, name, count) type name[(count)]
 #endif
 
+#include "jemalloc/internal/nstime.h"
 #include "jemalloc/internal/valgrind.h"
 #include "jemalloc/internal/util.h"
 #include "jemalloc/internal/atomic.h"
+#include "jemalloc/internal/spin.h"
 #include "jemalloc/internal/prng.h"
+#include "jemalloc/internal/ticker.h"
 #include "jemalloc/internal/ckh.h"
 #include "jemalloc/internal/size_classes.h"
+#include "jemalloc/internal/smoothstep.h"
 #include "jemalloc/internal/stats.h"
 #include "jemalloc/internal/ctl.h"
+#include "jemalloc/internal/witness.h"
 #include "jemalloc/internal/mutex.h"
 #include "jemalloc/internal/tsd.h"
 #include "jemalloc/internal/mb.h"
@@ -379,14 +398,19 @@ typedef unsigned szind_t;
 /******************************************************************************/
 #define	JEMALLOC_H_STRUCTS
 
+#include "jemalloc/internal/nstime.h"
 #include "jemalloc/internal/valgrind.h"
 #include "jemalloc/internal/util.h"
 #include "jemalloc/internal/atomic.h"
+#include "jemalloc/internal/spin.h"
 #include "jemalloc/internal/prng.h"
+#include "jemalloc/internal/ticker.h"
 #include "jemalloc/internal/ckh.h"
 #include "jemalloc/internal/size_classes.h"
+#include "jemalloc/internal/smoothstep.h"
 #include "jemalloc/internal/stats.h"
 #include "jemalloc/internal/ctl.h"
+#include "jemalloc/internal/witness.h"
 #include "jemalloc/internal/mutex.h"
 #include "jemalloc/internal/mb.h"
 #include "jemalloc/internal/bitmap.h"
@@ -422,13 +446,27 @@ extern bool	opt_redzone;
 extern bool	opt_utrace;
 extern bool	opt_xmalloc;
 extern bool	opt_zero;
-extern size_t	opt_narenas;
+extern unsigned	opt_narenas;
 
 extern bool	in_valgrind;
 
 /* Number of CPUs. */
-extern unsigned		ncpus;
+extern unsigned	ncpus;
+
+/* Number of arenas used for automatic multiplexing of threads and arenas. */
+extern unsigned	narenas_auto;
 
+/*
+ * Arenas that are used to service external requests.  Not all elements of the
+ * arenas array are necessarily used; arenas are created lazily as needed.
+ */
+extern arena_t	**arenas;
+
+/*
+ * pind2sz_tab encodes the same information as could be computed by
+ * pind2sz_compute().
+ */
+extern size_t const	pind2sz_tab[NPSIZES];
 /*
  * index2size_tab encodes the same information as could be computed (at
  * unacceptable cost in some code paths) by index2size_compute().
@@ -447,31 +485,35 @@ void	a0dalloc(void *ptr);
 void	*bootstrap_malloc(size_t size);
 void	*bootstrap_calloc(size_t num, size_t size);
 void	bootstrap_free(void *ptr);
-arena_t	*arenas_extend(unsigned ind);
-arena_t	*arena_init(unsigned ind);
 unsigned	narenas_total_get(void);
-arena_t	*arena_get_hard(tsd_t *tsd, unsigned ind, bool init_if_missing);
-arena_t	*arena_choose_hard(tsd_t *tsd);
+arena_t	*arena_init(tsdn_t *tsdn, unsigned ind);
+arena_tdata_t	*arena_tdata_get_hard(tsd_t *tsd, unsigned ind);
+arena_t	*arena_choose_hard(tsd_t *tsd, bool internal);
 void	arena_migrate(tsd_t *tsd, unsigned oldind, unsigned newind);
-unsigned	arena_nbound(unsigned ind);
 void	thread_allocated_cleanup(tsd_t *tsd);
 void	thread_deallocated_cleanup(tsd_t *tsd);
+void	iarena_cleanup(tsd_t *tsd);
 void	arena_cleanup(tsd_t *tsd);
-void	arenas_cache_cleanup(tsd_t *tsd);
-void	narenas_cache_cleanup(tsd_t *tsd);
-void	arenas_cache_bypass_cleanup(tsd_t *tsd);
+void	arenas_tdata_cleanup(tsd_t *tsd);
+void	narenas_tdata_cleanup(tsd_t *tsd);
+void	arenas_tdata_bypass_cleanup(tsd_t *tsd);
 void	jemalloc_prefork(void);
 void	jemalloc_postfork_parent(void);
 void	jemalloc_postfork_child(void);
 
+#include "jemalloc/internal/nstime.h"
 #include "jemalloc/internal/valgrind.h"
 #include "jemalloc/internal/util.h"
 #include "jemalloc/internal/atomic.h"
+#include "jemalloc/internal/spin.h"
 #include "jemalloc/internal/prng.h"
+#include "jemalloc/internal/ticker.h"
 #include "jemalloc/internal/ckh.h"
 #include "jemalloc/internal/size_classes.h"
+#include "jemalloc/internal/smoothstep.h"
 #include "jemalloc/internal/stats.h"
 #include "jemalloc/internal/ctl.h"
+#include "jemalloc/internal/witness.h"
 #include "jemalloc/internal/mutex.h"
 #include "jemalloc/internal/mb.h"
 #include "jemalloc/internal/bitmap.h"
@@ -492,16 +534,21 @@ void	jemalloc_postfork_child(void);
 /******************************************************************************/
 #define	JEMALLOC_H_INLINES
 
+#include "jemalloc/internal/nstime.h"
 #include "jemalloc/internal/valgrind.h"
 #include "jemalloc/internal/util.h"
 #include "jemalloc/internal/atomic.h"
+#include "jemalloc/internal/spin.h"
 #include "jemalloc/internal/prng.h"
+#include "jemalloc/internal/ticker.h"
 #include "jemalloc/internal/ckh.h"
 #include "jemalloc/internal/size_classes.h"
+#include "jemalloc/internal/smoothstep.h"
 #include "jemalloc/internal/stats.h"
 #include "jemalloc/internal/ctl.h"
-#include "jemalloc/internal/mutex.h"
 #include "jemalloc/internal/tsd.h"
+#include "jemalloc/internal/witness.h"
+#include "jemalloc/internal/mutex.h"
 #include "jemalloc/internal/mb.h"
 #include "jemalloc/internal/extent.h"
 #include "jemalloc/internal/base.h"
@@ -511,6 +558,11 @@ void	jemalloc_postfork_child(void);
 #include "jemalloc/internal/huge.h"
 
 #ifndef JEMALLOC_ENABLE_INLINE
+pszind_t	psz2ind(size_t psz);
+size_t	pind2sz_compute(pszind_t pind);
+size_t	pind2sz_lookup(pszind_t pind);
+size_t	pind2sz(pszind_t pind);
+size_t	psz2u(size_t psz);
 szind_t	size2index_compute(size_t size);
 szind_t	size2index_lookup(size_t size);
 szind_t	size2index(size_t size);
@@ -521,39 +573,121 @@ size_t	s2u_compute(size_t size);
 size_t	s2u_lookup(size_t size);
 size_t	s2u(size_t size);
 size_t	sa2u(size_t size, size_t alignment);
+arena_t	*arena_choose_impl(tsd_t *tsd, arena_t *arena, bool internal);
 arena_t	*arena_choose(tsd_t *tsd, arena_t *arena);
-arena_t	*arena_get(tsd_t *tsd, unsigned ind, bool init_if_missing,
+arena_t	*arena_ichoose(tsd_t *tsd, arena_t *arena);
+arena_tdata_t	*arena_tdata_get(tsd_t *tsd, unsigned ind,
     bool refresh_if_missing);
+arena_t	*arena_get(tsdn_t *tsdn, unsigned ind, bool init_if_missing);
+ticker_t	*decay_ticker_get(tsd_t *tsd, unsigned ind);
 #endif
 
 #if (defined(JEMALLOC_ENABLE_INLINE) || defined(JEMALLOC_C_))
+JEMALLOC_INLINE pszind_t
+psz2ind(size_t psz)
+{
+
+	if (unlikely(psz > HUGE_MAXCLASS))
+		return (NPSIZES);
+	{
+		pszind_t x = lg_floor((psz<<1)-1);
+		pszind_t shift = (x < LG_SIZE_CLASS_GROUP + LG_PAGE) ? 0 : x -
+		    (LG_SIZE_CLASS_GROUP + LG_PAGE);
+		pszind_t grp = shift << LG_SIZE_CLASS_GROUP;
+
+		pszind_t lg_delta = (x < LG_SIZE_CLASS_GROUP + LG_PAGE + 1) ?
+		    LG_PAGE : x - LG_SIZE_CLASS_GROUP - 1;
+
+		size_t delta_inverse_mask = ZI(-1) << lg_delta;
+		pszind_t mod = ((((psz-1) & delta_inverse_mask) >> lg_delta)) &
+		    ((ZU(1) << LG_SIZE_CLASS_GROUP) - 1);
+
+		pszind_t ind = grp + mod;
+		return (ind);
+	}
+}
+
+JEMALLOC_INLINE size_t
+pind2sz_compute(pszind_t pind)
+{
+
+	{
+		size_t grp = pind >> LG_SIZE_CLASS_GROUP;
+		size_t mod = pind & ((ZU(1) << LG_SIZE_CLASS_GROUP) - 1);
+
+		size_t grp_size_mask = ~((!!grp)-1);
+		size_t grp_size = ((ZU(1) << (LG_PAGE +
+		    (LG_SIZE_CLASS_GROUP-1))) << grp) & grp_size_mask;
+
+		size_t shift = (grp == 0) ? 1 : grp;
+		size_t lg_delta = shift + (LG_PAGE-1);
+		size_t mod_size = (mod+1) << lg_delta;
+
+		size_t sz = grp_size + mod_size;
+		return (sz);
+	}
+}
+
+JEMALLOC_INLINE size_t
+pind2sz_lookup(pszind_t pind)
+{
+	size_t ret = (size_t)pind2sz_tab[pind];
+	assert(ret == pind2sz_compute(pind));
+	return (ret);
+}
+
+JEMALLOC_INLINE size_t
+pind2sz(pszind_t pind)
+{
+
+	assert(pind < NPSIZES);
+	return (pind2sz_lookup(pind));
+}
+
+JEMALLOC_INLINE size_t
+psz2u(size_t psz)
+{
+
+	if (unlikely(psz > HUGE_MAXCLASS))
+		return (0);
+	{
+		size_t x = lg_floor((psz<<1)-1);
+		size_t lg_delta = (x < LG_SIZE_CLASS_GROUP + LG_PAGE + 1) ?
+		    LG_PAGE : x - LG_SIZE_CLASS_GROUP - 1;
+		size_t delta = ZU(1) << lg_delta;
+		size_t delta_mask = delta - 1;
+		size_t usize = (psz + delta_mask) & ~delta_mask;
+		return (usize);
+	}
+}
+
 JEMALLOC_INLINE szind_t
 size2index_compute(size_t size)
 {
 
+	if (unlikely(size > HUGE_MAXCLASS))
+		return (NSIZES);
 #if (NTBINS != 0)
 	if (size <= (ZU(1) << LG_TINY_MAXCLASS)) {
-		size_t lg_tmin = LG_TINY_MAXCLASS - NTBINS + 1;
-		size_t lg_ceil = lg_floor(pow2_ceil(size));
+		szind_t lg_tmin = LG_TINY_MAXCLASS - NTBINS + 1;
+		szind_t lg_ceil = lg_floor(pow2_ceil_zu(size));
 		return (lg_ceil < lg_tmin ? 0 : lg_ceil - lg_tmin);
 	}
 #endif
 	{
-		size_t x = unlikely(ZI(size) < 0) ? ((size<<1) ?
-		    (ZU(1)<<(LG_SIZEOF_PTR+3)) : ((ZU(1)<<(LG_SIZEOF_PTR+3))-1))
-		    : lg_floor((size<<1)-1);
-		size_t shift = (x < LG_SIZE_CLASS_GROUP + LG_QUANTUM) ? 0 :
+		szind_t x = lg_floor((size<<1)-1);
+		szind_t shift = (x < LG_SIZE_CLASS_GROUP + LG_QUANTUM) ? 0 :
 		    x - (LG_SIZE_CLASS_GROUP + LG_QUANTUM);
-		size_t grp = shift << LG_SIZE_CLASS_GROUP;
+		szind_t grp = shift << LG_SIZE_CLASS_GROUP;
 
-		size_t lg_delta = (x < LG_SIZE_CLASS_GROUP + LG_QUANTUM + 1)
+		szind_t lg_delta = (x < LG_SIZE_CLASS_GROUP + LG_QUANTUM + 1)
 		    ? LG_QUANTUM : x - LG_SIZE_CLASS_GROUP - 1;
 
 		size_t delta_inverse_mask = ZI(-1) << lg_delta;
-		size_t mod = ((((size-1) & delta_inverse_mask) >> lg_delta)) &
+		szind_t mod = ((((size-1) & delta_inverse_mask) >> lg_delta)) &
 		    ((ZU(1) << LG_SIZE_CLASS_GROUP) - 1);
 
-		size_t index = NTBINS + grp + mod;
+		szind_t index = NTBINS + grp + mod;
 		return (index);
 	}
 }
@@ -564,8 +698,7 @@ size2index_lookup(size_t size)
 
 	assert(size <= LOOKUP_MAXCLASS);
 	{
-		size_t ret = ((size_t)(size2index_tab[(size-1) >>
-		    LG_TINY_MIN]));
+		szind_t ret = (size2index_tab[(size-1) >> LG_TINY_MIN]);
 		assert(ret == size2index_compute(size));
 		return (ret);
 	}
@@ -628,18 +761,18 @@ JEMALLOC_ALWAYS_INLINE size_t
 s2u_compute(size_t size)
 {
 
+	if (unlikely(size > HUGE_MAXCLASS))
+		return (0);
 #if (NTBINS > 0)
 	if (size <= (ZU(1) << LG_TINY_MAXCLASS)) {
 		size_t lg_tmin = LG_TINY_MAXCLASS - NTBINS + 1;
-		size_t lg_ceil = lg_floor(pow2_ceil(size));
+		size_t lg_ceil = lg_floor(pow2_ceil_zu(size));
 		return (lg_ceil < lg_tmin ? (ZU(1) << lg_tmin) :
 		    (ZU(1) << lg_ceil));
 	}
 #endif
 	{
-		size_t x = unlikely(ZI(size) < 0) ? ((size<<1) ?
-		    (ZU(1)<<(LG_SIZEOF_PTR+3)) : ((ZU(1)<<(LG_SIZEOF_PTR+3))-1))
-		    : lg_floor((size<<1)-1);
+		size_t x = lg_floor((size<<1)-1);
 		size_t lg_delta = (x < LG_SIZE_CLASS_GROUP + LG_QUANTUM + 1)
 		    ?  LG_QUANTUM : x - LG_SIZE_CLASS_GROUP - 1;
 		size_t delta = ZU(1) << lg_delta;
@@ -723,17 +856,16 @@ sa2u(size_t size, size_t alignment)
 			return (usize);
 	}
 
-	/* Huge size class.  Beware of size_t overflow. */
+	/* Huge size class.  Beware of overflow. */
+
+	if (unlikely(alignment > HUGE_MAXCLASS))
+		return (0);
 
 	/*
 	 * We can't achieve subchunk alignment, so round up alignment to the
 	 * minimum that can actually be supported.
 	 */
 	alignment = CHUNK_CEILING(alignment);
-	if (alignment == 0) {
-		/* size_t overflow. */
-		return (0);
-	}
 
 	/* Make sure result is a huge size class. */
 	if (size <= chunksize)
@@ -759,45 +891,84 @@ sa2u(size_t size, size_t alignment)
 
 /* Choose an arena based on a per-thread value. */
 JEMALLOC_INLINE arena_t *
-arena_choose(tsd_t *tsd, arena_t *arena)
+arena_choose_impl(tsd_t *tsd, arena_t *arena, bool internal)
 {
 	arena_t *ret;
 
 	if (arena != NULL)
 		return (arena);
 
-	if (unlikely((ret = tsd_arena_get(tsd)) == NULL))
-		ret = arena_choose_hard(tsd);
+	ret = internal ? tsd_iarena_get(tsd) : tsd_arena_get(tsd);
+	if (unlikely(ret == NULL))
+		ret = arena_choose_hard(tsd, internal);
 
 	return (ret);
 }
 
 JEMALLOC_INLINE arena_t *
-arena_get(tsd_t *tsd, unsigned ind, bool init_if_missing,
-    bool refresh_if_missing)
+arena_choose(tsd_t *tsd, arena_t *arena)
 {
-	arena_t *arena;
-	arena_t **arenas_cache = tsd_arenas_cache_get(tsd);
 
-	/* init_if_missing requires refresh_if_missing. */
-	assert(!init_if_missing || refresh_if_missing);
+	return (arena_choose_impl(tsd, arena, false));
+}
+
+JEMALLOC_INLINE arena_t *
+arena_ichoose(tsd_t *tsd, arena_t *arena)
+{
 
-	if (unlikely(arenas_cache == NULL)) {
-		/* arenas_cache hasn't been initialized yet. */
-		return (arena_get_hard(tsd, ind, init_if_missing));
+	return (arena_choose_impl(tsd, arena, true));
+}
+
+JEMALLOC_INLINE arena_tdata_t *
+arena_tdata_get(tsd_t *tsd, unsigned ind, bool refresh_if_missing)
+{
+	arena_tdata_t *tdata;
+	arena_tdata_t *arenas_tdata = tsd_arenas_tdata_get(tsd);
+
+	if (unlikely(arenas_tdata == NULL)) {
+		/* arenas_tdata hasn't been initialized yet. */
+		return (arena_tdata_get_hard(tsd, ind));
 	}
-	if (unlikely(ind >= tsd_narenas_cache_get(tsd))) {
+	if (unlikely(ind >= tsd_narenas_tdata_get(tsd))) {
 		/*
-		 * ind is invalid, cache is old (too small), or arena to be
+		 * ind is invalid, cache is old (too small), or tdata to be
 		 * initialized.
 		 */
-		return (refresh_if_missing ? arena_get_hard(tsd, ind,
-		    init_if_missing) : NULL);
+		return (refresh_if_missing ? arena_tdata_get_hard(tsd, ind) :
+		    NULL);
 	}
-	arena = arenas_cache[ind];
-	if (likely(arena != NULL) || !refresh_if_missing)
-		return (arena);
-	return (arena_get_hard(tsd, ind, init_if_missing));
+
+	tdata = &arenas_tdata[ind];
+	if (likely(tdata != NULL) || !refresh_if_missing)
+		return (tdata);
+	return (arena_tdata_get_hard(tsd, ind));
+}
+
+JEMALLOC_INLINE arena_t *
+arena_get(tsdn_t *tsdn, unsigned ind, bool init_if_missing)
+{
+	arena_t *ret;
+
+	assert(ind <= MALLOCX_ARENA_MAX);
+
+	ret = arenas[ind];
+	if (unlikely(ret == NULL)) {
+		ret = atomic_read_p((void *)&arenas[ind]);
+		if (init_if_missing && unlikely(ret == NULL))
+			ret = arena_init(tsdn, ind);
+	}
+	return (ret);
+}
+
+JEMALLOC_INLINE ticker_t *
+decay_ticker_get(tsd_t *tsd, unsigned ind)
+{
+	arena_tdata_t *tdata;
+
+	tdata = arena_tdata_get(tsd, ind, true);
+	if (unlikely(tdata == NULL))
+		return (NULL);
+	return (&tdata->decay_ticker);
 }
 #endif
 
@@ -818,27 +989,27 @@ arena_get(tsd_t *tsd, unsigned ind, bool init_if_missing,
 
 #ifndef JEMALLOC_ENABLE_INLINE
 arena_t	*iaalloc(const void *ptr);
-size_t	isalloc(const void *ptr, bool demote);
-void	*iallocztm(tsd_t *tsd, size_t size, bool zero, tcache_t *tcache,
-    bool is_metadata, arena_t *arena);
-void	*imalloct(tsd_t *tsd, size_t size, tcache_t *tcache, arena_t *arena);
-void	*imalloc(tsd_t *tsd, size_t size);
-void	*icalloct(tsd_t *tsd, size_t size, tcache_t *tcache, arena_t *arena);
-void	*icalloc(tsd_t *tsd, size_t size);
-void	*ipallocztm(tsd_t *tsd, size_t usize, size_t alignment, bool zero,
+size_t	isalloc(tsdn_t *tsdn, const void *ptr, bool demote);
+void	*iallocztm(tsdn_t *tsdn, size_t size, szind_t ind, bool zero,
+    tcache_t *tcache, bool is_metadata, arena_t *arena, bool slow_path);
+void	*ialloc(tsd_t *tsd, size_t size, szind_t ind, bool zero,
+    bool slow_path);
+void	*ipallocztm(tsdn_t *tsdn, size_t usize, size_t alignment, bool zero,
     tcache_t *tcache, bool is_metadata, arena_t *arena);
-void	*ipalloct(tsd_t *tsd, size_t usize, size_t alignment, bool zero,
+void	*ipalloct(tsdn_t *tsdn, size_t usize, size_t alignment, bool zero,
     tcache_t *tcache, arena_t *arena);
 void	*ipalloc(tsd_t *tsd, size_t usize, size_t alignment, bool zero);
-size_t	ivsalloc(const void *ptr, bool demote);
+size_t	ivsalloc(tsdn_t *tsdn, const void *ptr, bool demote);
 size_t	u2rz(size_t usize);
-size_t	p2rz(const void *ptr);
-void	idalloctm(tsd_t *tsd, void *ptr, tcache_t *tcache, bool is_metadata);
-void	idalloct(tsd_t *tsd, void *ptr, tcache_t *tcache);
+size_t	p2rz(tsdn_t *tsdn, const void *ptr);
+void	idalloctm(tsdn_t *tsdn, void *ptr, tcache_t *tcache, bool is_metadata,
+    bool slow_path);
 void	idalloc(tsd_t *tsd, void *ptr);
-void	iqalloc(tsd_t *tsd, void *ptr, tcache_t *tcache);
-void	isdalloct(tsd_t *tsd, void *ptr, size_t size, tcache_t *tcache);
-void	isqalloc(tsd_t *tsd, void *ptr, size_t size, tcache_t *tcache);
+void	iqalloc(tsd_t *tsd, void *ptr, tcache_t *tcache, bool slow_path);
+void	isdalloct(tsdn_t *tsdn, void *ptr, size_t size, tcache_t *tcache,
+    bool slow_path);
+void	isqalloc(tsd_t *tsd, void *ptr, size_t size, tcache_t *tcache,
+    bool slow_path);
 void	*iralloct_realign(tsd_t *tsd, void *ptr, size_t oldsize, size_t size,
     size_t extra, size_t alignment, bool zero, tcache_t *tcache,
     arena_t *arena);
@@ -846,8 +1017,8 @@ void	*iralloct(tsd_t *tsd, void *ptr, size_t oldsize, size_t size,
     size_t alignment, bool zero, tcache_t *tcache, arena_t *arena);
 void	*iralloc(tsd_t *tsd, void *ptr, size_t oldsize, size_t size,
     size_t alignment, bool zero);
-bool	ixalloc(void *ptr, size_t oldsize, size_t size, size_t extra,
-    size_t alignment, bool zero);
+bool	ixalloc(tsdn_t *tsdn, void *ptr, size_t oldsize, size_t size,
+    size_t extra, size_t alignment, bool zero);
 #endif
 
 #if (defined(JEMALLOC_ENABLE_INLINE) || defined(JEMALLOC_C_))
@@ -862,100 +1033,85 @@ iaalloc(const void *ptr)
 
 /*
  * Typical usage:
+ *   tsdn_t *tsdn = [...]
  *   void *ptr = [...]
- *   size_t sz = isalloc(ptr, config_prof);
+ *   size_t sz = isalloc(tsdn, ptr, config_prof);
  */
 JEMALLOC_ALWAYS_INLINE size_t
-isalloc(const void *ptr, bool demote)
+isalloc(tsdn_t *tsdn, const void *ptr, bool demote)
 {
 
 	assert(ptr != NULL);
 	/* Demotion only makes sense if config_prof is true. */
 	assert(config_prof || !demote);
 
-	return (arena_salloc(ptr, demote));
+	return (arena_salloc(tsdn, ptr, demote));
 }
 
 JEMALLOC_ALWAYS_INLINE void *
-iallocztm(tsd_t *tsd, size_t size, bool zero, tcache_t *tcache, bool is_metadata,
-    arena_t *arena)
+iallocztm(tsdn_t *tsdn, size_t size, szind_t ind, bool zero, tcache_t *tcache,
+    bool is_metadata, arena_t *arena, bool slow_path)
 {
 	void *ret;
 
 	assert(size != 0);
+	assert(!is_metadata || tcache == NULL);
+	assert(!is_metadata || arena == NULL || arena->ind < narenas_auto);
 
-	ret = arena_malloc(tsd, arena, size, zero, tcache);
+	ret = arena_malloc(tsdn, arena, size, ind, zero, tcache, slow_path);
 	if (config_stats && is_metadata && likely(ret != NULL)) {
-		arena_metadata_allocated_add(iaalloc(ret), isalloc(ret,
-		    config_prof));
+		arena_metadata_allocated_add(iaalloc(ret),
+		    isalloc(tsdn, ret, config_prof));
 	}
 	return (ret);
 }
 
 JEMALLOC_ALWAYS_INLINE void *
-imalloct(tsd_t *tsd, size_t size, tcache_t *tcache, arena_t *arena)
+ialloc(tsd_t *tsd, size_t size, szind_t ind, bool zero, bool slow_path)
 {
 
-	return (iallocztm(tsd, size, false, tcache, false, arena));
+	return (iallocztm(tsd_tsdn(tsd), size, ind, zero, tcache_get(tsd, true),
+	    false, NULL, slow_path));
 }
 
 JEMALLOC_ALWAYS_INLINE void *
-imalloc(tsd_t *tsd, size_t size)
-{
-
-	return (iallocztm(tsd, size, false, tcache_get(tsd, true), false, NULL));
-}
-
-JEMALLOC_ALWAYS_INLINE void *
-icalloct(tsd_t *tsd, size_t size, tcache_t *tcache, arena_t *arena)
-{
-
-	return (iallocztm(tsd, size, true, tcache, false, arena));
-}
-
-JEMALLOC_ALWAYS_INLINE void *
-icalloc(tsd_t *tsd, size_t size)
-{
-
-	return (iallocztm(tsd, size, true, tcache_get(tsd, true), false, NULL));
-}
-
-JEMALLOC_ALWAYS_INLINE void *
-ipallocztm(tsd_t *tsd, size_t usize, size_t alignment, bool zero,
+ipallocztm(tsdn_t *tsdn, size_t usize, size_t alignment, bool zero,
     tcache_t *tcache, bool is_metadata, arena_t *arena)
 {
 	void *ret;
 
 	assert(usize != 0);
 	assert(usize == sa2u(usize, alignment));
+	assert(!is_metadata || tcache == NULL);
+	assert(!is_metadata || arena == NULL || arena->ind < narenas_auto);
 
-	ret = arena_palloc(tsd, arena, usize, alignment, zero, tcache);
+	ret = arena_palloc(tsdn, arena, usize, alignment, zero, tcache);
 	assert(ALIGNMENT_ADDR2BASE(ret, alignment) == ret);
 	if (config_stats && is_metadata && likely(ret != NULL)) {
-		arena_metadata_allocated_add(iaalloc(ret), isalloc(ret,
+		arena_metadata_allocated_add(iaalloc(ret), isalloc(tsdn, ret,
 		    config_prof));
 	}
 	return (ret);
 }
 
 JEMALLOC_ALWAYS_INLINE void *
-ipalloct(tsd_t *tsd, size_t usize, size_t alignment, bool zero,
+ipalloct(tsdn_t *tsdn, size_t usize, size_t alignment, bool zero,
     tcache_t *tcache, arena_t *arena)
 {
 
-	return (ipallocztm(tsd, usize, alignment, zero, tcache, false, arena));
+	return (ipallocztm(tsdn, usize, alignment, zero, tcache, false, arena));
 }
 
 JEMALLOC_ALWAYS_INLINE void *
 ipalloc(tsd_t *tsd, size_t usize, size_t alignment, bool zero)
 {
 
-	return (ipallocztm(tsd, usize, alignment, zero, tcache_get(tsd,
-	    NULL), false, NULL));
+	return (ipallocztm(tsd_tsdn(tsd), usize, alignment, zero,
+	    tcache_get(tsd, true), false, NULL));
 }
 
 JEMALLOC_ALWAYS_INLINE size_t
-ivsalloc(const void *ptr, bool demote)
+ivsalloc(tsdn_t *tsdn, const void *ptr, bool demote)
 {
 	extent_node_t *node;
 
@@ -967,7 +1123,7 @@ ivsalloc(const void *ptr, bool demote)
 	assert(extent_node_addr_get(node) == ptr ||
 	    extent_node_achunk_get(node));
 
-	return (isalloc(ptr, demote));
+	return (isalloc(tsdn, ptr, demote));
 }
 
 JEMALLOC_INLINE size_t
@@ -985,65 +1141,62 @@ u2rz(size_t usize)
 }
 
 JEMALLOC_INLINE size_t
-p2rz(const void *ptr)
+p2rz(tsdn_t *tsdn, const void *ptr)
 {
-	size_t usize = isalloc(ptr, false);
+	size_t usize = isalloc(tsdn, ptr, false);
 
 	return (u2rz(usize));
 }
 
 JEMALLOC_ALWAYS_INLINE void
-idalloctm(tsd_t *tsd, void *ptr, tcache_t *tcache, bool is_metadata)
+idalloctm(tsdn_t *tsdn, void *ptr, tcache_t *tcache, bool is_metadata,
+    bool slow_path)
 {
 
 	assert(ptr != NULL);
+	assert(!is_metadata || tcache == NULL);
+	assert(!is_metadata || iaalloc(ptr)->ind < narenas_auto);
 	if (config_stats && is_metadata) {
-		arena_metadata_allocated_sub(iaalloc(ptr), isalloc(ptr,
+		arena_metadata_allocated_sub(iaalloc(ptr), isalloc(tsdn, ptr,
 		    config_prof));
 	}
 
-	arena_dalloc(tsd, ptr, tcache);
-}
-
-JEMALLOC_ALWAYS_INLINE void
-idalloct(tsd_t *tsd, void *ptr, tcache_t *tcache)
-{
-
-	idalloctm(tsd, ptr, tcache, false);
+	arena_dalloc(tsdn, ptr, tcache, slow_path);
 }
 
 JEMALLOC_ALWAYS_INLINE void
 idalloc(tsd_t *tsd, void *ptr)
 {
 
-	idalloctm(tsd, ptr, tcache_get(tsd, false), false);
+	idalloctm(tsd_tsdn(tsd), ptr, tcache_get(tsd, false), false, true);
 }
 
 JEMALLOC_ALWAYS_INLINE void
-iqalloc(tsd_t *tsd, void *ptr, tcache_t *tcache)
+iqalloc(tsd_t *tsd, void *ptr, tcache_t *tcache, bool slow_path)
 {
 
-	if (config_fill && unlikely(opt_quarantine))
+	if (slow_path && config_fill && unlikely(opt_quarantine))
 		quarantine(tsd, ptr);
 	else
-		idalloctm(tsd, ptr, tcache, false);
+		idalloctm(tsd_tsdn(tsd), ptr, tcache, false, slow_path);
 }
 
 JEMALLOC_ALWAYS_INLINE void
-isdalloct(tsd_t *tsd, void *ptr, size_t size, tcache_t *tcache)
+isdalloct(tsdn_t *tsdn, void *ptr, size_t size, tcache_t *tcache,
+    bool slow_path)
 {
 
-	arena_sdalloc(tsd, ptr, size, tcache);
+	arena_sdalloc(tsdn, ptr, size, tcache, slow_path);
 }
 
 JEMALLOC_ALWAYS_INLINE void
-isqalloc(tsd_t *tsd, void *ptr, size_t size, tcache_t *tcache)
+isqalloc(tsd_t *tsd, void *ptr, size_t size, tcache_t *tcache, bool slow_path)
 {
 
-	if (config_fill && unlikely(opt_quarantine))
+	if (slow_path && config_fill && unlikely(opt_quarantine))
 		quarantine(tsd, ptr);
 	else
-		isdalloct(tsd, ptr, size, tcache);
+		isdalloct(tsd_tsdn(tsd), ptr, size, tcache, slow_path);
 }
 
 JEMALLOC_ALWAYS_INLINE void *
@@ -1054,17 +1207,18 @@ iralloct_realign(tsd_t *tsd, void *ptr, size_t oldsize, size_t size,
 	size_t usize, copysize;
 
 	usize = sa2u(size + extra, alignment);
-	if (usize == 0)
+	if (unlikely(usize == 0 || usize > HUGE_MAXCLASS))
 		return (NULL);
-	p = ipalloct(tsd, usize, alignment, zero, tcache, arena);
+	p = ipalloct(tsd_tsdn(tsd), usize, alignment, zero, tcache, arena);
 	if (p == NULL) {
 		if (extra == 0)
 			return (NULL);
 		/* Try again, without extra this time. */
 		usize = sa2u(size, alignment);
-		if (usize == 0)
+		if (unlikely(usize == 0 || usize > HUGE_MAXCLASS))
 			return (NULL);
-		p = ipalloct(tsd, usize, alignment, zero, tcache, arena);
+		p = ipalloct(tsd_tsdn(tsd), usize, alignment, zero, tcache,
+		    arena);
 		if (p == NULL)
 			return (NULL);
 	}
@@ -1074,7 +1228,7 @@ iralloct_realign(tsd_t *tsd, void *ptr, size_t oldsize, size_t size,
 	 */
 	copysize = (size < oldsize) ? size : oldsize;
 	memcpy(p, ptr, copysize);
-	isqalloc(tsd, ptr, oldsize, tcache);
+	isqalloc(tsd, ptr, oldsize, tcache, true);
 	return (p);
 }
 
@@ -1110,8 +1264,8 @@ iralloc(tsd_t *tsd, void *ptr, size_t oldsize, size_t size, size_t alignment,
 }
 
 JEMALLOC_ALWAYS_INLINE bool
-ixalloc(void *ptr, size_t oldsize, size_t size, size_t extra, size_t alignment,
-    bool zero)
+ixalloc(tsdn_t *tsdn, void *ptr, size_t oldsize, size_t size, size_t extra,
+    size_t alignment, bool zero)
 {
 
 	assert(ptr != NULL);
@@ -1123,7 +1277,7 @@ ixalloc(void *ptr, size_t oldsize, size_t size, size_t extra, size_t alignment,
 		return (true);
 	}
 
-	return (arena_ralloc_no_move(ptr, oldsize, size, extra, zero));
+	return (arena_ralloc_no_move(tsdn, ptr, oldsize, size, extra, zero));
 }
 #endif
 
diff --git a/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_decls.h b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_decls.h
index a601d6ebb..c907d9109 100644
--- a/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_decls.h
+++ b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_decls.h
@@ -17,7 +17,18 @@
 #    include <sys/uio.h>
 #  endif
 #  include <pthread.h>
+#  ifdef JEMALLOC_OS_UNFAIR_LOCK
+#    include <os/lock.h>
+#  endif
+#  ifdef JEMALLOC_GLIBC_MALLOC_HOOK
+#    include <sched.h>
+#  endif
 #  include <errno.h>
+#  include <sys/time.h>
+#  include <time.h>
+#  ifdef JEMALLOC_HAVE_MACH_ABSOLUTE_TIME
+#    include <mach/mach_time.h>
+#  endif
 #endif
 #include <sys/types.h>
 
diff --git a/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_defs.h.in b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_defs.h.in
index b0f8caaf8..def4ba550 100644
--- a/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_defs.h.in
+++ b/deps/jemalloc/include/jemalloc/internal/jemalloc_internal_defs.h.in
@@ -56,9 +56,9 @@
 #undef JEMALLOC_HAVE_BUILTIN_CLZ
 
 /*
- * Defined if madvise(2) is available.
+ * Defined if os_unfair_lock_*() functions are available, as provided by Darwin.
  */
-#undef JEMALLOC_HAVE_MADVISE
+#undef JEMALLOC_OS_UNFAIR_LOCK
 
 /*
  * Defined if OSSpin*() functions are available, as provided by Darwin, and
@@ -66,6 +66,9 @@
  */
 #undef JEMALLOC_OSSPIN
 
+/* Defined if syscall(2) is usable. */
+#undef JEMALLOC_USE_SYSCALL
+
 /*
  * Defined if secure_getenv(3) is available.
  */
@@ -76,6 +79,24 @@
  */
 #undef JEMALLOC_HAVE_ISSETUGID
 
+/* Defined if pthread_atfork(3) is available. */
+#undef JEMALLOC_HAVE_PTHREAD_ATFORK
+
+/*
+ * Defined if clock_gettime(CLOCK_MONOTONIC_COARSE, ...) is available.
+ */
+#undef JEMALLOC_HAVE_CLOCK_MONOTONIC_COARSE
+
+/*
+ * Defined if clock_gettime(CLOCK_MONOTONIC, ...) is available.
+ */
+#undef JEMALLOC_HAVE_CLOCK_MONOTONIC
+
+/*
+ * Defined if mach_absolute_time() is available.
+ */
+#undef JEMALLOC_HAVE_MACH_ABSOLUTE_TIME
+
 /*
  * Defined if _malloc_thread_cleanup() exists.  At least in the case of
  * FreeBSD, pthread_key_create() allocates, which if used during malloc
@@ -189,9 +210,16 @@
 #undef JEMALLOC_TLS
 
 /*
- * ffs()/ffsl() functions to use for bitmapping.  Don't use these directly;
- * instead, use jemalloc_ffs() or jemalloc_ffsl() from util.h.
+ * Used to mark unreachable code to quiet "end of non-void" compiler warnings.
+ * Don't use this directly; instead use unreachable() from util.h
  */
+#undef JEMALLOC_INTERNAL_UNREACHABLE
+
+/*
+ * ffs*() functions to use for bitmapping.  Don't use these directly; instead,
+ * use ffs_*() from util.h.
+ */
+#undef JEMALLOC_INTERNAL_FFSLL
 #undef JEMALLOC_INTERNAL_FFSL
 #undef JEMALLOC_INTERNAL_FFS
 
@@ -213,18 +241,35 @@
 #undef JEMALLOC_ZONE
 #undef JEMALLOC_ZONE_VERSION
 
+/*
+ * Methods for determining whether the OS overcommits.
+ * JEMALLOC_PROC_SYS_VM_OVERCOMMIT_MEMORY: Linux's
+ *                                         /proc/sys/vm.overcommit_memory file.
+ * JEMALLOC_SYSCTL_VM_OVERCOMMIT: FreeBSD's vm.overcommit sysctl.
+ */
+#undef JEMALLOC_SYSCTL_VM_OVERCOMMIT
+#undef JEMALLOC_PROC_SYS_VM_OVERCOMMIT_MEMORY
+
+/* Defined if madvise(2) is available. */
+#undef JEMALLOC_HAVE_MADVISE
+
 /*
  * Methods for purging unused pages differ between operating systems.
  *
- *   madvise(..., MADV_DONTNEED) : On Linux, this immediately discards pages,
- *                                 such that new pages will be demand-zeroed if
- *                                 the address region is later touched.
- *   madvise(..., MADV_FREE) : On FreeBSD and Darwin, this marks pages as being
- *                             unused, such that they will be discarded rather
- *                             than swapped out.
+ *   madvise(..., MADV_FREE) : This marks pages as being unused, such that they
+ *                             will be discarded rather than swapped out.
+ *   madvise(..., MADV_DONTNEED) : This immediately discards pages, such that
+ *                                 new pages will be demand-zeroed if the
+ *                                 address region is later touched.
  */
-#undef JEMALLOC_PURGE_MADVISE_DONTNEED
 #undef JEMALLOC_PURGE_MADVISE_FREE
+#undef JEMALLOC_PURGE_MADVISE_DONTNEED
+
+/*
+ * Defined if transparent huge pages are supported via the MADV_[NO]HUGEPAGE
+ * arguments to madvise(2).
+ */
+#undef JEMALLOC_THP
 
 /* Define if operating system has alloca.h header. */
 #undef JEMALLOC_HAS_ALLOCA_H
@@ -241,6 +286,9 @@
 /* sizeof(long) == 2^LG_SIZEOF_LONG. */
 #undef LG_SIZEOF_LONG
 
+/* sizeof(long long) == 2^LG_SIZEOF_LONG_LONG. */
+#undef LG_SIZEOF_LONG_LONG
+
 /* sizeof(intmax_t) == 2^LG_SIZEOF_INTMAX_T. */
 #undef LG_SIZEOF_INTMAX_T
 
@@ -259,4 +307,7 @@
  */
 #undef JEMALLOC_EXPORT
 
+/* config.malloc_conf options string. */
+#undef JEMALLOC_CONFIG_MALLOC_CONF
+
 #endif /* JEMALLOC_INTERNAL_DEFS_H_ */
diff --git a/deps/jemalloc/include/jemalloc/internal/mb.h b/deps/jemalloc/include/jemalloc/internal/mb.h
index 3cfa78729..5384728fd 100644
--- a/deps/jemalloc/include/jemalloc/internal/mb.h
+++ b/deps/jemalloc/include/jemalloc/internal/mb.h
@@ -42,7 +42,7 @@ mb_write(void)
 	    : /* Inputs. */
 	    : "memory" /* Clobbers. */
 	    );
-#else
+#  else
 	/*
 	 * This is hopefully enough to keep the compiler from reordering
 	 * instructions around this one.
@@ -52,7 +52,7 @@ mb_write(void)
 	    : /* Inputs. */
 	    : "memory" /* Clobbers. */
 	    );
-#endif
+#  endif
 }
 #elif (defined(__amd64__) || defined(__x86_64__))
 JEMALLOC_INLINE void
@@ -104,9 +104,9 @@ mb_write(void)
 {
 	malloc_mutex_t mtx;
 
-	malloc_mutex_init(&mtx);
-	malloc_mutex_lock(&mtx);
-	malloc_mutex_unlock(&mtx);
+	malloc_mutex_init(&mtx, "mb", WITNESS_RANK_OMIT);
+	malloc_mutex_lock(TSDN_NULL, &mtx);
+	malloc_mutex_unlock(TSDN_NULL, &mtx);
 }
 #endif
 #endif
diff --git a/deps/jemalloc/include/jemalloc/internal/mutex.h b/deps/jemalloc/include/jemalloc/internal/mutex.h
index f051f2917..b442d2d4e 100644
--- a/deps/jemalloc/include/jemalloc/internal/mutex.h
+++ b/deps/jemalloc/include/jemalloc/internal/mutex.h
@@ -5,18 +5,25 @@ typedef struct malloc_mutex_s malloc_mutex_t;
 
 #ifdef _WIN32
 #  define MALLOC_MUTEX_INITIALIZER
+#elif (defined(JEMALLOC_OS_UNFAIR_LOCK))
+#  define MALLOC_MUTEX_INITIALIZER					\
+     {OS_UNFAIR_LOCK_INIT, WITNESS_INITIALIZER(WITNESS_RANK_OMIT)}
 #elif (defined(JEMALLOC_OSSPIN))
-#  define MALLOC_MUTEX_INITIALIZER {0}
+#  define MALLOC_MUTEX_INITIALIZER {0, WITNESS_INITIALIZER(WITNESS_RANK_OMIT)}
 #elif (defined(JEMALLOC_MUTEX_INIT_CB))
-#  define MALLOC_MUTEX_INITIALIZER {PTHREAD_MUTEX_INITIALIZER, NULL}
+#  define MALLOC_MUTEX_INITIALIZER					\
+    {PTHREAD_MUTEX_INITIALIZER, NULL, WITNESS_INITIALIZER(WITNESS_RANK_OMIT)}
 #else
 #  if (defined(JEMALLOC_HAVE_PTHREAD_MUTEX_ADAPTIVE_NP) &&		\
        defined(PTHREAD_ADAPTIVE_MUTEX_INITIALIZER_NP))
 #    define MALLOC_MUTEX_TYPE PTHREAD_MUTEX_ADAPTIVE_NP
-#    define MALLOC_MUTEX_INITIALIZER {PTHREAD_ADAPTIVE_MUTEX_INITIALIZER_NP}
+#    define MALLOC_MUTEX_INITIALIZER					\
+       {PTHREAD_ADAPTIVE_MUTEX_INITIALIZER_NP,				\
+        WITNESS_INITIALIZER(WITNESS_RANK_OMIT)}
 #  else
 #    define MALLOC_MUTEX_TYPE PTHREAD_MUTEX_DEFAULT
-#    define MALLOC_MUTEX_INITIALIZER {PTHREAD_MUTEX_INITIALIZER}
+#    define MALLOC_MUTEX_INITIALIZER					\
+       {PTHREAD_MUTEX_INITIALIZER, WITNESS_INITIALIZER(WITNESS_RANK_OMIT)}
 #  endif
 #endif
 
@@ -31,6 +38,8 @@ struct malloc_mutex_s {
 #  else
 	CRITICAL_SECTION	lock;
 #  endif
+#elif (defined(JEMALLOC_OS_UNFAIR_LOCK))
+	os_unfair_lock		lock;
 #elif (defined(JEMALLOC_OSSPIN))
 	OSSpinLock		lock;
 #elif (defined(JEMALLOC_MUTEX_INIT_CB))
@@ -39,6 +48,7 @@ struct malloc_mutex_s {
 #else
 	pthread_mutex_t		lock;
 #endif
+	witness_t		witness;
 };
 
 #endif /* JEMALLOC_H_STRUCTS */
@@ -52,52 +62,62 @@ extern bool isthreaded;
 #  define isthreaded true
 #endif
 
-bool	malloc_mutex_init(malloc_mutex_t *mutex);
-void	malloc_mutex_prefork(malloc_mutex_t *mutex);
-void	malloc_mutex_postfork_parent(malloc_mutex_t *mutex);
-void	malloc_mutex_postfork_child(malloc_mutex_t *mutex);
-bool	mutex_boot(void);
+bool	malloc_mutex_init(malloc_mutex_t *mutex, const char *name,
+    witness_rank_t rank);
+void	malloc_mutex_prefork(tsdn_t *tsdn, malloc_mutex_t *mutex);
+void	malloc_mutex_postfork_parent(tsdn_t *tsdn, malloc_mutex_t *mutex);
+void	malloc_mutex_postfork_child(tsdn_t *tsdn, malloc_mutex_t *mutex);
+bool	malloc_mutex_boot(void);
 
 #endif /* JEMALLOC_H_EXTERNS */
 /******************************************************************************/
 #ifdef JEMALLOC_H_INLINES
 
 #ifndef JEMALLOC_ENABLE_INLINE
-void	malloc_mutex_lock(malloc_mutex_t *mutex);
-void	malloc_mutex_unlock(malloc_mutex_t *mutex);
+void	malloc_mutex_lock(tsdn_t *tsdn, malloc_mutex_t *mutex);
+void	malloc_mutex_unlock(tsdn_t *tsdn, malloc_mutex_t *mutex);
+void	malloc_mutex_assert_owner(tsdn_t *tsdn, malloc_mutex_t *mutex);
+void	malloc_mutex_assert_not_owner(tsdn_t *tsdn, malloc_mutex_t *mutex);
 #endif
 
 #if (defined(JEMALLOC_ENABLE_INLINE) || defined(JEMALLOC_MUTEX_C_))
 JEMALLOC_INLINE void
-malloc_mutex_lock(malloc_mutex_t *mutex)
+malloc_mutex_lock(tsdn_t *tsdn, malloc_mutex_t *mutex)
 {
 
 	if (isthreaded) {
+		witness_assert_not_owner(tsdn, &mutex->witness);
 #ifdef _WIN32
 #  if _WIN32_WINNT >= 0x0600
 		AcquireSRWLockExclusive(&mutex->lock);
 #  else
 		EnterCriticalSection(&mutex->lock);
 #  endif
+#elif (defined(JEMALLOC_OS_UNFAIR_LOCK))
+		os_unfair_lock_lock(&mutex->lock);
 #elif (defined(JEMALLOC_OSSPIN))
 		OSSpinLockLock(&mutex->lock);
 #else
 		pthread_mutex_lock(&mutex->lock);
 #endif
+		witness_lock(tsdn, &mutex->witness);
 	}
 }
 
 JEMALLOC_INLINE void
-malloc_mutex_unlock(malloc_mutex_t *mutex)
+malloc_mutex_unlock(tsdn_t *tsdn, malloc_mutex_t *mutex)
 {
 
 	if (isthreaded) {
+		witness_unlock(tsdn, &mutex->witness);
 #ifdef _WIN32
 #  if _WIN32_WINNT >= 0x0600
 		ReleaseSRWLockExclusive(&mutex->lock);
 #  else
 		LeaveCriticalSection(&mutex->lock);
 #  endif
+#elif (defined(JEMALLOC_OS_UNFAIR_LOCK))
+		os_unfair_lock_unlock(&mutex->lock);
 #elif (defined(JEMALLOC_OSSPIN))
 		OSSpinLockUnlock(&mutex->lock);
 #else
@@ -105,6 +125,22 @@ malloc_mutex_unlock(malloc_mutex_t *mutex)
 #endif
 	}
 }
+
+JEMALLOC_INLINE void
+malloc_mutex_assert_owner(tsdn_t *tsdn, malloc_mutex_t *mutex)
+{
+
+	if (isthreaded)
+		witness_assert_owner(tsdn, &mutex->witness);
+}
+
+JEMALLOC_INLINE void
+malloc_mutex_assert_not_owner(tsdn_t *tsdn, malloc_mutex_t *mutex)
+{
+
+	if (isthreaded)
+		witness_assert_not_owner(tsdn, &mutex->witness);
+}
 #endif
 
 #endif /* JEMALLOC_H_INLINES */
diff --git a/deps/jemalloc/include/jemalloc/internal/nstime.h b/deps/jemalloc/include/jemalloc/internal/nstime.h
new file mode 100644
index 000000000..93b27dc80
--- /dev/null
+++ b/deps/jemalloc/include/jemalloc/internal/nstime.h
@@ -0,0 +1,48 @@
+/******************************************************************************/
+#ifdef JEMALLOC_H_TYPES
+
+typedef struct nstime_s nstime_t;
+
+/* Maximum supported number of seconds (~584 years). */
+#define	NSTIME_SEC_MAX	KQU(18446744072)
+
+#endif /* JEMALLOC_H_TYPES */
+/******************************************************************************/
+#ifdef JEMALLOC_H_STRUCTS
+
+struct nstime_s {
+	uint64_t	ns;
+};
+
+#endif /* JEMALLOC_H_STRUCTS */
+/******************************************************************************/
+#ifdef JEMALLOC_H_EXTERNS
+
+void	nstime_init(nstime_t *time, uint64_t ns);
+void	nstime_init2(nstime_t *time, uint64_t sec, uint64_t nsec);
+uint64_t	nstime_ns(const nstime_t *time);
+uint64_t	nstime_sec(const nstime_t *time);
+uint64_t	nstime_nsec(const nstime_t *time);
+void	nstime_copy(nstime_t *time, const nstime_t *source);
+int	nstime_compare(const nstime_t *a, const nstime_t *b);
+void	nstime_add(nstime_t *time, const nstime_t *addend);
+void	nstime_subtract(nstime_t *time, const nstime_t *subtrahend);
+void	nstime_imultiply(nstime_t *time, uint64_t multiplier);
+void	nstime_idivide(nstime_t *time, uint64_t divisor);
+uint64_t	nstime_divide(const nstime_t *time, const nstime_t *divisor);
+#ifdef JEMALLOC_JET
+typedef bool (nstime_monotonic_t)(void);
+extern nstime_monotonic_t *nstime_monotonic;
+typedef bool (nstime_update_t)(nstime_t *);
+extern nstime_update_t *nstime_update;
+#else
+bool	nstime_monotonic(void);
+bool	nstime_update(nstime_t *time);
+#endif
+
+#endif /* JEMALLOC_H_EXTERNS */
+/******************************************************************************/
+#ifdef JEMALLOC_H_INLINES
+
+#endif /* JEMALLOC_H_INLINES */
+/******************************************************************************/
diff --git a/deps/jemalloc/include/jemalloc/internal/pages.h b/deps/jemalloc/include/jemalloc/internal/pages.h
index da7eb9686..4ae9f156a 100644
--- a/deps/jemalloc/include/jemalloc/internal/pages.h
+++ b/deps/jemalloc/include/jemalloc/internal/pages.h
@@ -9,13 +9,16 @@
 /******************************************************************************/
 #ifdef JEMALLOC_H_EXTERNS
 
-void	*pages_map(void *addr, size_t size);
+void	*pages_map(void *addr, size_t size, bool *commit);
 void	pages_unmap(void *addr, size_t size);
 void	*pages_trim(void *addr, size_t alloc_size, size_t leadsize,
-    size_t size);
+    size_t size, bool *commit);
 bool	pages_commit(void *addr, size_t size);
 bool	pages_decommit(void *addr, size_t size);
 bool	pages_purge(void *addr, size_t size);
+bool	pages_huge(void *addr, size_t size);
+bool	pages_nohuge(void *addr, size_t size);
+void	pages_boot(void);
 
 #endif /* JEMALLOC_H_EXTERNS */
 /******************************************************************************/
diff --git a/deps/jemalloc/include/jemalloc/internal/ph.h b/deps/jemalloc/include/jemalloc/internal/ph.h
new file mode 100644
index 000000000..4f91c333f
--- /dev/null
+++ b/deps/jemalloc/include/jemalloc/internal/ph.h
@@ -0,0 +1,345 @@
+/*
+ * A Pairing Heap implementation.
+ *
+ * "The Pairing Heap: A New Form of Self-Adjusting Heap"
+ * https://www.cs.cmu.edu/~sleator/papers/pairing-heaps.pdf
+ *
+ * With auxiliary twopass list, described in a follow on paper.
+ *
+ * "Pairing Heaps: Experiments and Analysis"
+ * http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.106.2988&rep=rep1&type=pdf
+ *
+ *******************************************************************************
+ */
+
+#ifndef PH_H_
+#define	PH_H_
+
+/* Node structure. */
+#define	phn(a_type)							\
+struct {								\
+	a_type	*phn_prev;						\
+	a_type	*phn_next;						\
+	a_type	*phn_lchild;						\
+}
+
+/* Root structure. */
+#define	ph(a_type)							\
+struct {								\
+	a_type	*ph_root;						\
+}
+
+/* Internal utility macros. */
+#define	phn_lchild_get(a_type, a_field, a_phn)				\
+	(a_phn->a_field.phn_lchild)
+#define	phn_lchild_set(a_type, a_field, a_phn, a_lchild) do {		\
+	a_phn->a_field.phn_lchild = a_lchild;				\
+} while (0)
+
+#define	phn_next_get(a_type, a_field, a_phn)				\
+	(a_phn->a_field.phn_next)
+#define	phn_prev_set(a_type, a_field, a_phn, a_prev) do {		\
+	a_phn->a_field.phn_prev = a_prev;				\
+} while (0)
+
+#define	phn_prev_get(a_type, a_field, a_phn)				\
+	(a_phn->a_field.phn_prev)
+#define	phn_next_set(a_type, a_field, a_phn, a_next) do {		\
+	a_phn->a_field.phn_next = a_next;				\
+} while (0)
+
+#define	phn_merge_ordered(a_type, a_field, a_phn0, a_phn1, a_cmp) do {	\
+	a_type *phn0child;						\
+									\
+	assert(a_phn0 != NULL);						\
+	assert(a_phn1 != NULL);						\
+	assert(a_cmp(a_phn0, a_phn1) <= 0);				\
+									\
+	phn_prev_set(a_type, a_field, a_phn1, a_phn0);			\
+	phn0child = phn_lchild_get(a_type, a_field, a_phn0);		\
+	phn_next_set(a_type, a_field, a_phn1, phn0child);		\
+	if (phn0child != NULL)						\
+		phn_prev_set(a_type, a_field, phn0child, a_phn1);	\
+	phn_lchild_set(a_type, a_field, a_phn0, a_phn1);		\
+} while (0)
+
+#define	phn_merge(a_type, a_field, a_phn0, a_phn1, a_cmp, r_phn) do {	\
+	if (a_phn0 == NULL)						\
+		r_phn = a_phn1;						\
+	else if (a_phn1 == NULL)					\
+		r_phn = a_phn0;						\
+	else if (a_cmp(a_phn0, a_phn1) < 0) {				\
+		phn_merge_ordered(a_type, a_field, a_phn0, a_phn1,	\
+		    a_cmp);						\
+		r_phn = a_phn0;						\
+	} else {							\
+		phn_merge_ordered(a_type, a_field, a_phn1, a_phn0,	\
+		    a_cmp);						\
+		r_phn = a_phn1;						\
+	}								\
+} while (0)
+
+#define	ph_merge_siblings(a_type, a_field, a_phn, a_cmp, r_phn) do {	\
+	a_type *head = NULL;						\
+	a_type *tail = NULL;						\
+	a_type *phn0 = a_phn;						\
+	a_type *phn1 = phn_next_get(a_type, a_field, phn0);		\
+									\
+	/*								\
+	 * Multipass merge, wherein the first two elements of a FIFO	\
+	 * are repeatedly merged, and each result is appended to the	\
+	 * singly linked FIFO, until the FIFO contains only a single	\
+	 * element.  We start with a sibling list but no reference to	\
+	 * its tail, so we do a single pass over the sibling list to	\
+	 * populate the FIFO.						\
+	 */								\
+	if (phn1 != NULL) {						\
+		a_type *phnrest = phn_next_get(a_type, a_field, phn1);	\
+		if (phnrest != NULL)					\
+			phn_prev_set(a_type, a_field, phnrest, NULL);	\
+		phn_prev_set(a_type, a_field, phn0, NULL);		\
+		phn_next_set(a_type, a_field, phn0, NULL);		\
+		phn_prev_set(a_type, a_field, phn1, NULL);		\
+		phn_next_set(a_type, a_field, phn1, NULL);		\
+		phn_merge(a_type, a_field, phn0, phn1, a_cmp, phn0);	\
+		head = tail = phn0;					\
+		phn0 = phnrest;						\
+		while (phn0 != NULL) {					\
+			phn1 = phn_next_get(a_type, a_field, phn0);	\
+			if (phn1 != NULL) {				\
+				phnrest = phn_next_get(a_type, a_field,	\
+				    phn1);				\
+				if (phnrest != NULL) {			\
+					phn_prev_set(a_type, a_field,	\
+					    phnrest, NULL);		\
+				}					\
+				phn_prev_set(a_type, a_field, phn0,	\
+				    NULL);				\
+				phn_next_set(a_type, a_field, phn0,	\
+				    NULL);				\
+				phn_prev_set(a_type, a_field, phn1,	\
+				    NULL);				\
+				phn_next_set(a_type, a_field, phn1,	\
+				    NULL);				\
+				phn_merge(a_type, a_field, phn0, phn1,	\
+				    a_cmp, phn0);			\
+				phn_next_set(a_type, a_field, tail,	\
+				    phn0);				\
+				tail = phn0;				\
+				phn0 = phnrest;				\
+			} else {					\
+				phn_next_set(a_type, a_field, tail,	\
+				    phn0);				\
+				tail = phn0;				\
+				phn0 = NULL;				\
+			}						\
+		}							\
+		phn0 = head;						\
+		phn1 = phn_next_get(a_type, a_field, phn0);		\
+		if (phn1 != NULL) {					\
+			while (true) {					\
+				head = phn_next_get(a_type, a_field,	\
+				    phn1);				\
+				assert(phn_prev_get(a_type, a_field,	\
+				    phn0) == NULL);			\
+				phn_next_set(a_type, a_field, phn0,	\
+				    NULL);				\
+				assert(phn_prev_get(a_type, a_field,	\
+				    phn1) == NULL);			\
+				phn_next_set(a_type, a_field, phn1,	\
+				    NULL);				\
+				phn_merge(a_type, a_field, phn0, phn1,	\
+				    a_cmp, phn0);			\
+				if (head == NULL)			\
+					break;				\
+				phn_next_set(a_type, a_field, tail,	\
+				    phn0);				\
+				tail = phn0;				\
+				phn0 = head;				\
+				phn1 = phn_next_get(a_type, a_field,	\
+				    phn0);				\
+			}						\
+		}							\
+	}								\
+	r_phn = phn0;							\
+} while (0)
+
+#define	ph_merge_aux(a_type, a_field, a_ph, a_cmp) do {			\
+	a_type *phn = phn_next_get(a_type, a_field, a_ph->ph_root);	\
+	if (phn != NULL) {						\
+		phn_prev_set(a_type, a_field, a_ph->ph_root, NULL);	\
+		phn_next_set(a_type, a_field, a_ph->ph_root, NULL);	\
+		phn_prev_set(a_type, a_field, phn, NULL);		\
+		ph_merge_siblings(a_type, a_field, phn, a_cmp, phn);	\
+		assert(phn_next_get(a_type, a_field, phn) == NULL);	\
+		phn_merge(a_type, a_field, a_ph->ph_root, phn, a_cmp,	\
+		    a_ph->ph_root);					\
+	}								\
+} while (0)
+
+#define	ph_merge_children(a_type, a_field, a_phn, a_cmp, r_phn) do {	\
+	a_type *lchild = phn_lchild_get(a_type, a_field, a_phn);	\
+	if (lchild == NULL)						\
+		r_phn = NULL;						\
+	else {								\
+		ph_merge_siblings(a_type, a_field, lchild, a_cmp,	\
+		    r_phn);						\
+	}								\
+} while (0)
+
+/*
+ * The ph_proto() macro generates function prototypes that correspond to the
+ * functions generated by an equivalently parameterized call to ph_gen().
+ */
+#define	ph_proto(a_attr, a_prefix, a_ph_type, a_type)			\
+a_attr void	a_prefix##new(a_ph_type *ph);				\
+a_attr bool	a_prefix##empty(a_ph_type *ph);				\
+a_attr a_type	*a_prefix##first(a_ph_type *ph);			\
+a_attr void	a_prefix##insert(a_ph_type *ph, a_type *phn);		\
+a_attr a_type	*a_prefix##remove_first(a_ph_type *ph);			\
+a_attr void	a_prefix##remove(a_ph_type *ph, a_type *phn);
+
+/*
+ * The ph_gen() macro generates a type-specific pairing heap implementation,
+ * based on the above cpp macros.
+ */
+#define	ph_gen(a_attr, a_prefix, a_ph_type, a_type, a_field, a_cmp)	\
+a_attr void								\
+a_prefix##new(a_ph_type *ph)						\
+{									\
+									\
+	memset(ph, 0, sizeof(ph(a_type)));				\
+}									\
+a_attr bool								\
+a_prefix##empty(a_ph_type *ph)						\
+{									\
+									\
+	return (ph->ph_root == NULL);					\
+}									\
+a_attr a_type *								\
+a_prefix##first(a_ph_type *ph)						\
+{									\
+									\
+	if (ph->ph_root == NULL)					\
+		return (NULL);						\
+	ph_merge_aux(a_type, a_field, ph, a_cmp);			\
+	return (ph->ph_root);						\
+}									\
+a_attr void								\
+a_prefix##insert(a_ph_type *ph, a_type *phn)				\
+{									\
+									\
+	memset(&phn->a_field, 0, sizeof(phn(a_type)));			\
+									\
+	/*								\
+	 * Treat the root as an aux list during insertion, and lazily	\
+	 * merge during a_prefix##remove_first().  For elements that	\
+	 * are inserted, then removed via a_prefix##remove() before the	\
+	 * aux list is ever processed, this makes insert/remove		\
+	 * constant-time, whereas eager merging would make insert	\
+	 * O(log n).							\
+	 */								\
+	if (ph->ph_root == NULL)					\
+		ph->ph_root = phn;					\
+	else {								\
+		phn_next_set(a_type, a_field, phn, phn_next_get(a_type,	\
+		    a_field, ph->ph_root));				\
+		if (phn_next_get(a_type, a_field, ph->ph_root) !=	\
+		    NULL) {						\
+			phn_prev_set(a_type, a_field,			\
+			    phn_next_get(a_type, a_field, ph->ph_root),	\
+			    phn);					\
+		}							\
+		phn_prev_set(a_type, a_field, phn, ph->ph_root);	\
+		phn_next_set(a_type, a_field, ph->ph_root, phn);	\
+	}								\
+}									\
+a_attr a_type *								\
+a_prefix##remove_first(a_ph_type *ph)					\
+{									\
+	a_type *ret;							\
+									\
+	if (ph->ph_root == NULL)					\
+		return (NULL);						\
+	ph_merge_aux(a_type, a_field, ph, a_cmp);			\
+									\
+	ret = ph->ph_root;						\
+									\
+	ph_merge_children(a_type, a_field, ph->ph_root, a_cmp,		\
+	    ph->ph_root);						\
+									\
+	return (ret);							\
+}									\
+a_attr void								\
+a_prefix##remove(a_ph_type *ph, a_type *phn)				\
+{									\
+	a_type *replace, *parent;					\
+									\
+	/*								\
+	 * We can delete from aux list without merging it, but we need	\
+	 * to merge if we are dealing with the root node.		\
+	 */								\
+	if (ph->ph_root == phn) {					\
+		ph_merge_aux(a_type, a_field, ph, a_cmp);		\
+		if (ph->ph_root == phn) {				\
+			ph_merge_children(a_type, a_field, ph->ph_root,	\
+			    a_cmp, ph->ph_root);			\
+			return;						\
+		}							\
+	}								\
+									\
+	/* Get parent (if phn is leftmost child) before mutating. */	\
+	if ((parent = phn_prev_get(a_type, a_field, phn)) != NULL) {	\
+		if (phn_lchild_get(a_type, a_field, parent) != phn)	\
+			parent = NULL;					\
+	}								\
+	/* Find a possible replacement node, and link to parent. */	\
+	ph_merge_children(a_type, a_field, phn, a_cmp, replace);	\
+	/* Set next/prev for sibling linked list. */			\
+	if (replace != NULL) {						\
+		if (parent != NULL) {					\
+			phn_prev_set(a_type, a_field, replace, parent);	\
+			phn_lchild_set(a_type, a_field, parent,		\
+			    replace);					\
+		} else {						\
+			phn_prev_set(a_type, a_field, replace,		\
+			    phn_prev_get(a_type, a_field, phn));	\
+			if (phn_prev_get(a_type, a_field, phn) !=	\
+			    NULL) {					\
+				phn_next_set(a_type, a_field,		\
+				    phn_prev_get(a_type, a_field, phn),	\
+				    replace);				\
+			}						\
+		}							\
+		phn_next_set(a_type, a_field, replace,			\
+		    phn_next_get(a_type, a_field, phn));		\
+		if (phn_next_get(a_type, a_field, phn) != NULL) {	\
+			phn_prev_set(a_type, a_field,			\
+			    phn_next_get(a_type, a_field, phn),		\
+			    replace);					\
+		}							\
+	} else {							\
+		if (parent != NULL) {					\
+			a_type *next = phn_next_get(a_type, a_field,	\
+			    phn);					\
+			phn_lchild_set(a_type, a_field, parent, next);	\
+			if (next != NULL) {				\
+				phn_prev_set(a_type, a_field, next,	\
+				    parent);				\
+			}						\
+		} else {						\
+			assert(phn_prev_get(a_type, a_field, phn) !=	\
+			    NULL);					\
+			phn_next_set(a_type, a_field,			\
+			    phn_prev_get(a_type, a_field, phn),		\
+			    phn_next_get(a_type, a_field, phn));	\
+		}							\
+		if (phn_next_get(a_type, a_field, phn) != NULL) {	\
+			phn_prev_set(a_type, a_field,			\
+			    phn_next_get(a_type, a_field, phn),		\
+			    phn_prev_get(a_type, a_field, phn));	\
+		}							\
+	}								\
+}
+
+#endif /* PH_H_ */
diff --git a/deps/jemalloc/include/jemalloc/internal/private_symbols.txt b/deps/jemalloc/include/jemalloc/internal/private_symbols.txt
index a90021aa6..c1c6c4090 100644
--- a/deps/jemalloc/include/jemalloc/internal/private_symbols.txt
+++ b/deps/jemalloc/include/jemalloc/internal/private_symbols.txt
@@ -3,12 +3,15 @@ a0get
 a0malloc
 arena_aalloc
 arena_alloc_junk_small
+arena_basic_stats_merge
 arena_bin_index
 arena_bin_info
-arena_bitselm_get
+arena_bitselm_get_const
+arena_bitselm_get_mutable
 arena_boot
 arena_choose
 arena_choose_hard
+arena_choose_impl
 arena_chunk_alloc_huge
 arena_chunk_cache_maybe_insert
 arena_chunk_cache_maybe_remove
@@ -25,18 +28,25 @@ arena_dalloc_junk_small
 arena_dalloc_large
 arena_dalloc_large_junked_locked
 arena_dalloc_small
+arena_decay_tick
+arena_decay_ticks
+arena_decay_time_default_get
+arena_decay_time_default_set
+arena_decay_time_get
+arena_decay_time_set
 arena_dss_prec_get
 arena_dss_prec_set
+arena_extent_sn_next
 arena_get
-arena_get_hard
+arena_ichoose
 arena_init
 arena_lg_dirty_mult_default_get
 arena_lg_dirty_mult_default_set
 arena_lg_dirty_mult_get
 arena_lg_dirty_mult_set
 arena_malloc
+arena_malloc_hard
 arena_malloc_large
-arena_malloc_small
 arena_mapbits_allocated_get
 arena_mapbits_binind_get
 arena_mapbits_decommitted_get
@@ -47,9 +57,6 @@ arena_mapbits_large_binind_set
 arena_mapbits_large_get
 arena_mapbits_large_set
 arena_mapbits_large_size_get
-arena_mapbitsp_get
-arena_mapbitsp_read
-arena_mapbitsp_write
 arena_mapbits_size_decode
 arena_mapbits_size_encode
 arena_mapbits_small_runind_get
@@ -58,23 +65,33 @@ arena_mapbits_unallocated_set
 arena_mapbits_unallocated_size_get
 arena_mapbits_unallocated_size_set
 arena_mapbits_unzeroed_get
+arena_mapbitsp_get_const
+arena_mapbitsp_get_mutable
+arena_mapbitsp_read
+arena_mapbitsp_write
 arena_maxrun
 arena_maybe_purge
 arena_metadata_allocated_add
 arena_metadata_allocated_get
 arena_metadata_allocated_sub
 arena_migrate
-arena_miscelm_get
+arena_miscelm_get_const
+arena_miscelm_get_mutable
 arena_miscelm_to_pageind
 arena_miscelm_to_rpages
-arena_nbound
 arena_new
 arena_node_alloc
 arena_node_dalloc
+arena_nthreads_dec
+arena_nthreads_get
+arena_nthreads_inc
 arena_palloc
 arena_postfork_child
 arena_postfork_parent
-arena_prefork
+arena_prefork0
+arena_prefork1
+arena_prefork2
+arena_prefork3
 arena_prof_accum
 arena_prof_accum_impl
 arena_prof_accum_locked
@@ -83,21 +100,25 @@ arena_prof_tctx_get
 arena_prof_tctx_reset
 arena_prof_tctx_set
 arena_ptr_small_binind_get
-arena_purge_all
+arena_purge
 arena_quarantine_junk_small
 arena_ralloc
 arena_ralloc_junk_large
 arena_ralloc_no_move
 arena_rd_to_miscelm
 arena_redzone_corruption
+arena_reset
 arena_run_regind
 arena_run_to_miscelm
 arena_salloc
-arenas_cache_bypass_cleanup
-arenas_cache_cleanup
 arena_sdalloc
 arena_stats_merge
 arena_tcache_fill_small
+arena_tdata_get
+arena_tdata_get_hard
+arenas
+arenas_tdata_bypass_cleanup
+arenas_tdata_cleanup
 atomic_add_p
 atomic_add_u
 atomic_add_uint32
@@ -113,6 +134,11 @@ atomic_sub_u
 atomic_sub_uint32
 atomic_sub_uint64
 atomic_sub_z
+atomic_write_p
+atomic_write_u
+atomic_write_uint32
+atomic_write_uint64
+atomic_write_z
 base_alloc
 base_boot
 base_postfork_child
@@ -122,7 +148,6 @@ base_stats_get
 bitmap_full
 bitmap_get
 bitmap_info_init
-bitmap_info_ngroups
 bitmap_init
 bitmap_set
 bitmap_sfu
@@ -139,32 +164,25 @@ chunk_alloc_dss
 chunk_alloc_mmap
 chunk_alloc_wrapper
 chunk_boot
-chunk_dalloc_arena
 chunk_dalloc_cache
 chunk_dalloc_mmap
 chunk_dalloc_wrapper
 chunk_deregister
 chunk_dss_boot
-chunk_dss_postfork_child
-chunk_dss_postfork_parent
+chunk_dss_mergeable
 chunk_dss_prec_get
 chunk_dss_prec_set
-chunk_dss_prefork
 chunk_hooks_default
 chunk_hooks_get
 chunk_hooks_set
 chunk_in_dss
 chunk_lookup
 chunk_npages
-chunk_postfork_child
-chunk_postfork_parent
-chunk_prefork
-chunk_purge_arena
 chunk_purge_wrapper
 chunk_register
+chunks_rtree
 chunksize
 chunksize_mask
-chunks_rtree
 ckh_count
 ckh_delete
 ckh_insert
@@ -183,6 +201,7 @@ ctl_nametomib
 ctl_postfork_child
 ctl_postfork_parent
 ctl_prefork
+decay_ticker_get
 dss_prec_names
 extent_node_achunk_get
 extent_node_achunk_set
@@ -190,6 +209,8 @@ extent_node_addr_get
 extent_node_addr_set
 extent_node_arena_get
 extent_node_arena_set
+extent_node_committed_get
+extent_node_committed_set
 extent_node_dirty_insert
 extent_node_dirty_linkage_init
 extent_node_dirty_remove
@@ -198,8 +219,12 @@ extent_node_prof_tctx_get
 extent_node_prof_tctx_set
 extent_node_size_get
 extent_node_size_set
+extent_node_sn_get
+extent_node_sn_set
 extent_node_zeroed_get
 extent_node_zeroed_set
+extent_tree_ad_destroy
+extent_tree_ad_destroy_recurse
 extent_tree_ad_empty
 extent_tree_ad_first
 extent_tree_ad_insert
@@ -217,23 +242,31 @@ extent_tree_ad_reverse_iter
 extent_tree_ad_reverse_iter_recurse
 extent_tree_ad_reverse_iter_start
 extent_tree_ad_search
-extent_tree_szad_empty
-extent_tree_szad_first
-extent_tree_szad_insert
-extent_tree_szad_iter
-extent_tree_szad_iter_recurse
-extent_tree_szad_iter_start
-extent_tree_szad_last
-extent_tree_szad_new
-extent_tree_szad_next
-extent_tree_szad_nsearch
-extent_tree_szad_prev
-extent_tree_szad_psearch
-extent_tree_szad_remove
-extent_tree_szad_reverse_iter
-extent_tree_szad_reverse_iter_recurse
-extent_tree_szad_reverse_iter_start
-extent_tree_szad_search
+extent_tree_szsnad_destroy
+extent_tree_szsnad_destroy_recurse
+extent_tree_szsnad_empty
+extent_tree_szsnad_first
+extent_tree_szsnad_insert
+extent_tree_szsnad_iter
+extent_tree_szsnad_iter_recurse
+extent_tree_szsnad_iter_start
+extent_tree_szsnad_last
+extent_tree_szsnad_new
+extent_tree_szsnad_next
+extent_tree_szsnad_nsearch
+extent_tree_szsnad_prev
+extent_tree_szsnad_psearch
+extent_tree_szsnad_remove
+extent_tree_szsnad_reverse_iter
+extent_tree_szsnad_reverse_iter_recurse
+extent_tree_szsnad_reverse_iter_start
+extent_tree_szsnad_search
+ffs_llu
+ffs_lu
+ffs_u
+ffs_u32
+ffs_u64
+ffs_zu
 get_errno
 hash
 hash_fmix_32
@@ -257,19 +290,16 @@ huge_ralloc
 huge_ralloc_no_move
 huge_salloc
 iaalloc
+ialloc
 iallocztm
-icalloc
-icalloct
+iarena_cleanup
 idalloc
-idalloct
 idalloctm
-imalloc
-imalloct
+in_valgrind
 index2size
 index2size_compute
 index2size_lookup
 index2size_tab
-in_valgrind
 ipalloc
 ipalloct
 ipallocztm
@@ -288,7 +318,11 @@ jemalloc_postfork_parent
 jemalloc_prefork
 large_maxclass
 lg_floor
+lg_prof_sample
 malloc_cprintf
+malloc_mutex_assert_not_owner
+malloc_mutex_assert_owner
+malloc_mutex_boot
 malloc_mutex_init
 malloc_mutex_lock
 malloc_mutex_postfork_child
@@ -310,12 +344,29 @@ malloc_write
 map_bias
 map_misc_offset
 mb_write
-mutex_boot
-narenas_cache_cleanup
+narenas_auto
+narenas_tdata_cleanup
 narenas_total_get
 ncpus
 nhbins
+nhclasses
+nlclasses
+nstime_add
+nstime_compare
+nstime_copy
+nstime_divide
+nstime_idivide
+nstime_imultiply
+nstime_init
+nstime_init2
+nstime_monotonic
+nstime_ns
+nstime_nsec
+nstime_sec
+nstime_subtract
+nstime_update
 opt_abort
+opt_decay_time
 opt_dss
 opt_junk
 opt_junk_alloc
@@ -334,6 +385,7 @@ opt_prof_gdump
 opt_prof_leak
 opt_prof_prefix
 opt_prof_thread_active_init
+opt_purge
 opt_quarantine
 opt_redzone
 opt_stats_print
@@ -342,13 +394,32 @@ opt_utrace
 opt_xmalloc
 opt_zero
 p2rz
+pages_boot
 pages_commit
 pages_decommit
+pages_huge
 pages_map
+pages_nohuge
 pages_purge
 pages_trim
 pages_unmap
-pow2_ceil
+pind2sz
+pind2sz_compute
+pind2sz_lookup
+pind2sz_tab
+pow2_ceil_u32
+pow2_ceil_u64
+pow2_ceil_zu
+prng_lg_range_u32
+prng_lg_range_u64
+prng_lg_range_zu
+prng_range_u32
+prng_range_u64
+prng_range_zu
+prng_state_next_u32
+prng_state_next_u64
+prng_state_next_zu
+prof_active
 prof_active_get
 prof_active_get_unlocked
 prof_active_set
@@ -358,6 +429,7 @@ prof_backtrace
 prof_boot0
 prof_boot1
 prof_boot2
+prof_bt_count
 prof_dump_header
 prof_dump_open
 prof_free
@@ -375,7 +447,8 @@ prof_malloc_sample_object
 prof_mdump
 prof_postfork_child
 prof_postfork_parent
-prof_prefork
+prof_prefork0
+prof_prefork1
 prof_realloc
 prof_reset
 prof_sample_accum_update
@@ -384,6 +457,7 @@ prof_tctx_get
 prof_tctx_reset
 prof_tctx_set
 prof_tdata_cleanup
+prof_tdata_count
 prof_tdata_get
 prof_tdata_init
 prof_tdata_reinit
@@ -393,11 +467,13 @@ prof_thread_active_init_set
 prof_thread_active_set
 prof_thread_name_get
 prof_thread_name_set
+psz2ind
+psz2u
+purge_mode_names
 quarantine
 quarantine_alloc_hook
 quarantine_alloc_hook_work
 quarantine_cleanup
-register_zone
 rtree_child_read
 rtree_child_read_hard
 rtree_child_tryread
@@ -413,6 +489,8 @@ rtree_subtree_read_hard
 rtree_subtree_tryread
 rtree_val_read
 rtree_val_write
+run_quantize_ceil
+run_quantize_floor
 s2u
 s2u_compute
 s2u_lookup
@@ -422,6 +500,8 @@ size2index
 size2index_compute
 size2index_lookup
 size2index_tab
+spin_adaptive
+spin_init
 stats_cactive
 stats_cactive_add
 stats_cactive_get
@@ -431,8 +511,6 @@ tcache_alloc_easy
 tcache_alloc_large
 tcache_alloc_small
 tcache_alloc_small_hard
-tcache_arena_associate
-tcache_arena_dissociate
 tcache_arena_reassociate
 tcache_bin_flush_large
 tcache_bin_flush_small
@@ -451,49 +529,103 @@ tcache_flush
 tcache_get
 tcache_get_hard
 tcache_maxclass
-tcaches
 tcache_salloc
+tcache_stats_merge
+tcaches
 tcaches_create
 tcaches_destroy
 tcaches_flush
 tcaches_get
-tcache_stats_merge
 thread_allocated_cleanup
 thread_deallocated_cleanup
+ticker_copy
+ticker_init
+ticker_read
+ticker_tick
+ticker_ticks
 tsd_arena_get
 tsd_arena_set
+tsd_arenap_get
+tsd_arenas_tdata_bypass_get
+tsd_arenas_tdata_bypass_set
+tsd_arenas_tdata_bypassp_get
+tsd_arenas_tdata_get
+tsd_arenas_tdata_set
+tsd_arenas_tdatap_get
 tsd_boot
 tsd_boot0
 tsd_boot1
 tsd_booted
+tsd_booted_get
 tsd_cleanup
 tsd_cleanup_wrapper
 tsd_fetch
+tsd_fetch_impl
 tsd_get
-tsd_wrapper_get
-tsd_wrapper_set
+tsd_get_allocates
+tsd_iarena_get
+tsd_iarena_set
+tsd_iarenap_get
 tsd_initialized
 tsd_init_check_recursion
 tsd_init_finish
 tsd_init_head
+tsd_narenas_tdata_get
+tsd_narenas_tdata_set
+tsd_narenas_tdatap_get
+tsd_wrapper_get
+tsd_wrapper_set
 tsd_nominal
+tsd_prof_tdata_get
+tsd_prof_tdata_set
+tsd_prof_tdatap_get
 tsd_quarantine_get
 tsd_quarantine_set
+tsd_quarantinep_get
 tsd_set
 tsd_tcache_enabled_get
 tsd_tcache_enabled_set
+tsd_tcache_enabledp_get
 tsd_tcache_get
 tsd_tcache_set
-tsd_tls
-tsd_tsd
-tsd_prof_tdata_get
-tsd_prof_tdata_set
+tsd_tcachep_get
 tsd_thread_allocated_get
 tsd_thread_allocated_set
+tsd_thread_allocatedp_get
 tsd_thread_deallocated_get
 tsd_thread_deallocated_set
+tsd_thread_deallocatedp_get
+tsd_tls
+tsd_tsd
+tsd_tsdn
+tsd_witness_fork_get
+tsd_witness_fork_set
+tsd_witness_forkp_get
+tsd_witnesses_get
+tsd_witnesses_set
+tsd_witnessesp_get
+tsdn_fetch
+tsdn_null
+tsdn_tsd
 u2rz
 valgrind_freelike_block
 valgrind_make_mem_defined
 valgrind_make_mem_noaccess
 valgrind_make_mem_undefined
+witness_assert_lockless
+witness_assert_not_owner
+witness_assert_owner
+witness_fork_cleanup
+witness_init
+witness_lock
+witness_lock_error
+witness_lockless_error
+witness_not_owner_error
+witness_owner
+witness_owner_error
+witness_postfork_child
+witness_postfork_parent
+witness_prefork
+witness_unlock
+witnesses_cleanup
+zone_register
diff --git a/deps/jemalloc/include/jemalloc/internal/prng.h b/deps/jemalloc/include/jemalloc/internal/prng.h
index 216d0ef47..c2bda19c6 100644
--- a/deps/jemalloc/include/jemalloc/internal/prng.h
+++ b/deps/jemalloc/include/jemalloc/internal/prng.h
@@ -18,31 +18,13 @@
  * proportional to bit position.  For example, the lowest bit has a cycle of 2,
  * the next has a cycle of 4, etc.  For this reason, we prefer to use the upper
  * bits.
- *
- * Macro parameters:
- *   uint32_t r          : Result.
- *   unsigned lg_range   : (0..32], number of least significant bits to return.
- *   uint32_t state      : Seed value.
- *   const uint32_t a, c : See above discussion.
  */
-#define	prng32(r, lg_range, state, a, c) do {				\
-	assert((lg_range) > 0);						\
-	assert((lg_range) <= 32);					\
-									\
-	r = (state * (a)) + (c);					\
-	state = r;							\
-	r >>= (32 - (lg_range));					\
-} while (false)
-
-/* Same as prng32(), but 64 bits of pseudo-randomness, using uint64_t. */
-#define	prng64(r, lg_range, state, a, c) do {				\
-	assert((lg_range) > 0);						\
-	assert((lg_range) <= 64);					\
-									\
-	r = (state * (a)) + (c);					\
-	state = r;							\
-	r >>= (64 - (lg_range));					\
-} while (false)
+
+#define	PRNG_A_32	UINT32_C(1103515241)
+#define	PRNG_C_32	UINT32_C(12347)
+
+#define	PRNG_A_64	UINT64_C(6364136223846793005)
+#define	PRNG_C_64	UINT64_C(1442695040888963407)
 
 #endif /* JEMALLOC_H_TYPES */
 /******************************************************************************/
@@ -56,5 +38,170 @@
 /******************************************************************************/
 #ifdef JEMALLOC_H_INLINES
 
+#ifndef JEMALLOC_ENABLE_INLINE
+uint32_t	prng_state_next_u32(uint32_t state);
+uint64_t	prng_state_next_u64(uint64_t state);
+size_t	prng_state_next_zu(size_t state);
+
+uint32_t	prng_lg_range_u32(uint32_t *state, unsigned lg_range,
+    bool atomic);
+uint64_t	prng_lg_range_u64(uint64_t *state, unsigned lg_range);
+size_t	prng_lg_range_zu(size_t *state, unsigned lg_range, bool atomic);
+
+uint32_t	prng_range_u32(uint32_t *state, uint32_t range, bool atomic);
+uint64_t	prng_range_u64(uint64_t *state, uint64_t range);
+size_t	prng_range_zu(size_t *state, size_t range, bool atomic);
+#endif
+
+#if (defined(JEMALLOC_ENABLE_INLINE) || defined(JEMALLOC_PRNG_C_))
+JEMALLOC_ALWAYS_INLINE uint32_t
+prng_state_next_u32(uint32_t state)
+{
+
+	return ((state * PRNG_A_32) + PRNG_C_32);
+}
+
+JEMALLOC_ALWAYS_INLINE uint64_t
+prng_state_next_u64(uint64_t state)
+{
+
+	return ((state * PRNG_A_64) + PRNG_C_64);
+}
+
+JEMALLOC_ALWAYS_INLINE size_t
+prng_state_next_zu(size_t state)
+{
+
+#if LG_SIZEOF_PTR == 2
+	return ((state * PRNG_A_32) + PRNG_C_32);
+#elif LG_SIZEOF_PTR == 3
+	return ((state * PRNG_A_64) + PRNG_C_64);
+#else
+#error Unsupported pointer size
+#endif
+}
+
+JEMALLOC_ALWAYS_INLINE uint32_t
+prng_lg_range_u32(uint32_t *state, unsigned lg_range, bool atomic)
+{
+	uint32_t ret, state1;
+
+	assert(lg_range > 0);
+	assert(lg_range <= 32);
+
+	if (atomic) {
+		uint32_t state0;
+
+		do {
+			state0 = atomic_read_uint32(state);
+			state1 = prng_state_next_u32(state0);
+		} while (atomic_cas_uint32(state, state0, state1));
+	} else {
+		state1 = prng_state_next_u32(*state);
+		*state = state1;
+	}
+	ret = state1 >> (32 - lg_range);
+
+	return (ret);
+}
+
+/* 64-bit atomic operations cannot be supported on all relevant platforms. */
+JEMALLOC_ALWAYS_INLINE uint64_t
+prng_lg_range_u64(uint64_t *state, unsigned lg_range)
+{
+	uint64_t ret, state1;
+
+	assert(lg_range > 0);
+	assert(lg_range <= 64);
+
+	state1 = prng_state_next_u64(*state);
+	*state = state1;
+	ret = state1 >> (64 - lg_range);
+
+	return (ret);
+}
+
+JEMALLOC_ALWAYS_INLINE size_t
+prng_lg_range_zu(size_t *state, unsigned lg_range, bool atomic)
+{
+	size_t ret, state1;
+
+	assert(lg_range > 0);
+	assert(lg_range <= ZU(1) << (3 + LG_SIZEOF_PTR));
+
+	if (atomic) {
+		size_t state0;
+
+		do {
+			state0 = atomic_read_z(state);
+			state1 = prng_state_next_zu(state0);
+		} while (atomic_cas_z(state, state0, state1));
+	} else {
+		state1 = prng_state_next_zu(*state);
+		*state = state1;
+	}
+	ret = state1 >> ((ZU(1) << (3 + LG_SIZEOF_PTR)) - lg_range);
+
+	return (ret);
+}
+
+JEMALLOC_ALWAYS_INLINE uint32_t
+prng_range_u32(uint32_t *state, uint32_t range, bool atomic)
+{
+	uint32_t ret;
+	unsigned lg_range;
+
+	assert(range > 1);
+
+	/* Compute the ceiling of lg(range). */
+	lg_range = ffs_u32(pow2_ceil_u32(range)) - 1;
+
+	/* Generate a result in [0..range) via repeated trial. */
+	do {
+		ret = prng_lg_range_u32(state, lg_range, atomic);
+	} while (ret >= range);
+
+	return (ret);
+}
+
+JEMALLOC_ALWAYS_INLINE uint64_t
+prng_range_u64(uint64_t *state, uint64_t range)
+{
+	uint64_t ret;
+	unsigned lg_range;
+
+	assert(range > 1);
+
+	/* Compute the ceiling of lg(range). */
+	lg_range = ffs_u64(pow2_ceil_u64(range)) - 1;
+
+	/* Generate a result in [0..range) via repeated trial. */
+	do {
+		ret = prng_lg_range_u64(state, lg_range);
+	} while (ret >= range);
+
+	return (ret);
+}
+
+JEMALLOC_ALWAYS_INLINE size_t
+prng_range_zu(size_t *state, size_t range, bool atomic)
+{
+	size_t ret;
+	unsigned lg_range;
+
+	assert(range > 1);
+
+	/* Compute the ceiling of lg(range). */
+	lg_range = ffs_u64(pow2_ceil_u64(range)) - 1;
+
+	/* Generate a result in [0..range) via repeated trial. */
+	do {
+		ret = prng_lg_range_zu(state, lg_range, atomic);
+	} while (ret >= range);
+
+	return (ret);
+}
+#endif
+
 #endif /* JEMALLOC_H_INLINES */
 /******************************************************************************/
diff --git a/deps/jemalloc/include/jemalloc/internal/prof.h b/deps/jemalloc/include/jemalloc/internal/prof.h
index e5198c3e8..8293b71ed 100644
--- a/deps/jemalloc/include/jemalloc/internal/prof.h
+++ b/deps/jemalloc/include/jemalloc/internal/prof.h
@@ -281,7 +281,7 @@ extern uint64_t	prof_interval;
 extern size_t	lg_prof_sample;
 
 void	prof_alloc_rollback(tsd_t *tsd, prof_tctx_t *tctx, bool updated);
-void	prof_malloc_sample_object(const void *ptr, size_t usize,
+void	prof_malloc_sample_object(tsdn_t *tsdn, const void *ptr, size_t usize,
     prof_tctx_t *tctx);
 void	prof_free_sampled_object(tsd_t *tsd, size_t usize, prof_tctx_t *tctx);
 void	bt_init(prof_bt_t *bt, void **vec);
@@ -293,32 +293,33 @@ size_t	prof_bt_count(void);
 const prof_cnt_t *prof_cnt_all(void);
 typedef int (prof_dump_open_t)(bool, const char *);
 extern prof_dump_open_t *prof_dump_open;
-typedef bool (prof_dump_header_t)(bool, const prof_cnt_t *);
+typedef bool (prof_dump_header_t)(tsdn_t *, bool, const prof_cnt_t *);
 extern prof_dump_header_t *prof_dump_header;
 #endif
-void	prof_idump(void);
-bool	prof_mdump(const char *filename);
-void	prof_gdump(void);
+void	prof_idump(tsdn_t *tsdn);
+bool	prof_mdump(tsd_t *tsd, const char *filename);
+void	prof_gdump(tsdn_t *tsdn);
 prof_tdata_t	*prof_tdata_init(tsd_t *tsd);
 prof_tdata_t	*prof_tdata_reinit(tsd_t *tsd, prof_tdata_t *tdata);
 void	prof_reset(tsd_t *tsd, size_t lg_sample);
 void	prof_tdata_cleanup(tsd_t *tsd);
-const char	*prof_thread_name_get(void);
-bool	prof_active_get(void);
-bool	prof_active_set(bool active);
+bool	prof_active_get(tsdn_t *tsdn);
+bool	prof_active_set(tsdn_t *tsdn, bool active);
+const char	*prof_thread_name_get(tsd_t *tsd);
 int	prof_thread_name_set(tsd_t *tsd, const char *thread_name);
-bool	prof_thread_active_get(void);
-bool	prof_thread_active_set(bool active);
-bool	prof_thread_active_init_get(void);
-bool	prof_thread_active_init_set(bool active_init);
-bool	prof_gdump_get(void);
-bool	prof_gdump_set(bool active);
+bool	prof_thread_active_get(tsd_t *tsd);
+bool	prof_thread_active_set(tsd_t *tsd, bool active);
+bool	prof_thread_active_init_get(tsdn_t *tsdn);
+bool	prof_thread_active_init_set(tsdn_t *tsdn, bool active_init);
+bool	prof_gdump_get(tsdn_t *tsdn);
+bool	prof_gdump_set(tsdn_t *tsdn, bool active);
 void	prof_boot0(void);
 void	prof_boot1(void);
-bool	prof_boot2(void);
-void	prof_prefork(void);
-void	prof_postfork_parent(void);
-void	prof_postfork_child(void);
+bool	prof_boot2(tsd_t *tsd);
+void	prof_prefork0(tsdn_t *tsdn);
+void	prof_prefork1(tsdn_t *tsdn);
+void	prof_postfork_parent(tsdn_t *tsdn);
+void	prof_postfork_child(tsdn_t *tsdn);
 void	prof_sample_threshold_update(prof_tdata_t *tdata);
 
 #endif /* JEMALLOC_H_EXTERNS */
@@ -329,17 +330,17 @@ void	prof_sample_threshold_update(prof_tdata_t *tdata);
 bool	prof_active_get_unlocked(void);
 bool	prof_gdump_get_unlocked(void);
 prof_tdata_t	*prof_tdata_get(tsd_t *tsd, bool create);
+prof_tctx_t	*prof_tctx_get(tsdn_t *tsdn, const void *ptr);
+void	prof_tctx_set(tsdn_t *tsdn, const void *ptr, size_t usize,
+    prof_tctx_t *tctx);
+void	prof_tctx_reset(tsdn_t *tsdn, const void *ptr, size_t usize,
+    const void *old_ptr, prof_tctx_t *tctx);
 bool	prof_sample_accum_update(tsd_t *tsd, size_t usize, bool commit,
     prof_tdata_t **tdata_out);
 prof_tctx_t	*prof_alloc_prep(tsd_t *tsd, size_t usize, bool prof_active,
     bool update);
-prof_tctx_t	*prof_tctx_get(const void *ptr);
-void	prof_tctx_set(const void *ptr, size_t usize, prof_tctx_t *tctx);
-void	prof_tctx_reset(const void *ptr, size_t usize, const void *old_ptr,
-    prof_tctx_t *tctx);
-void	prof_malloc_sample_object(const void *ptr, size_t usize,
+void	prof_malloc(tsdn_t *tsdn, const void *ptr, size_t usize,
     prof_tctx_t *tctx);
-void	prof_malloc(const void *ptr, size_t usize, prof_tctx_t *tctx);
 void	prof_realloc(tsd_t *tsd, const void *ptr, size_t usize,
     prof_tctx_t *tctx, bool prof_active, bool updated, const void *old_ptr,
     size_t old_usize, prof_tctx_t *old_tctx);
@@ -397,34 +398,34 @@ prof_tdata_get(tsd_t *tsd, bool create)
 }
 
 JEMALLOC_ALWAYS_INLINE prof_tctx_t *
-prof_tctx_get(const void *ptr)
+prof_tctx_get(tsdn_t *tsdn, const void *ptr)
 {
 
 	cassert(config_prof);
 	assert(ptr != NULL);
 
-	return (arena_prof_tctx_get(ptr));
+	return (arena_prof_tctx_get(tsdn, ptr));
 }
 
 JEMALLOC_ALWAYS_INLINE void
-prof_tctx_set(const void *ptr, size_t usize, prof_tctx_t *tctx)
+prof_tctx_set(tsdn_t *tsdn, const void *ptr, size_t usize, prof_tctx_t *tctx)
 {
 
 	cassert(config_prof);
 	assert(ptr != NULL);
 
-	arena_prof_tctx_set(ptr, usize, tctx);
+	arena_prof_tctx_set(tsdn, ptr, usize, tctx);
 }
 
 JEMALLOC_ALWAYS_INLINE void
-prof_tctx_reset(const void *ptr, size_t usize, const void *old_ptr,
+prof_tctx_reset(tsdn_t *tsdn, const void *ptr, size_t usize, const void *old_ptr,
     prof_tctx_t *old_tctx)
 {
 
 	cassert(config_prof);
 	assert(ptr != NULL);
 
-	arena_prof_tctx_reset(ptr, usize, old_ptr, old_tctx);
+	arena_prof_tctx_reset(tsdn, ptr, usize, old_ptr, old_tctx);
 }
 
 JEMALLOC_ALWAYS_INLINE bool
@@ -436,16 +437,16 @@ prof_sample_accum_update(tsd_t *tsd, size_t usize, bool update,
 	cassert(config_prof);
 
 	tdata = prof_tdata_get(tsd, true);
-	if ((uintptr_t)tdata <= (uintptr_t)PROF_TDATA_STATE_MAX)
+	if (unlikely((uintptr_t)tdata <= (uintptr_t)PROF_TDATA_STATE_MAX))
 		tdata = NULL;
 
 	if (tdata_out != NULL)
 		*tdata_out = tdata;
 
-	if (tdata == NULL)
+	if (unlikely(tdata == NULL))
 		return (true);
 
-	if (tdata->bytes_until_sample >= usize) {
+	if (likely(tdata->bytes_until_sample >= usize)) {
 		if (update)
 			tdata->bytes_until_sample -= usize;
 		return (true);
@@ -479,17 +480,17 @@ prof_alloc_prep(tsd_t *tsd, size_t usize, bool prof_active, bool update)
 }
 
 JEMALLOC_ALWAYS_INLINE void
-prof_malloc(const void *ptr, size_t usize, prof_tctx_t *tctx)
+prof_malloc(tsdn_t *tsdn, const void *ptr, size_t usize, prof_tctx_t *tctx)
 {
 
 	cassert(config_prof);
 	assert(ptr != NULL);
-	assert(usize == isalloc(ptr, true));
+	assert(usize == isalloc(tsdn, ptr, true));
 
 	if (unlikely((uintptr_t)tctx > (uintptr_t)1U))
-		prof_malloc_sample_object(ptr, usize, tctx);
+		prof_malloc_sample_object(tsdn, ptr, usize, tctx);
 	else
-		prof_tctx_set(ptr, usize, (prof_tctx_t *)(uintptr_t)1U);
+		prof_tctx_set(tsdn, ptr, usize, (prof_tctx_t *)(uintptr_t)1U);
 }
 
 JEMALLOC_ALWAYS_INLINE void
@@ -503,7 +504,7 @@ prof_realloc(tsd_t *tsd, const void *ptr, size_t usize, prof_tctx_t *tctx,
 	assert(ptr != NULL || (uintptr_t)tctx <= (uintptr_t)1U);
 
 	if (prof_active && !updated && ptr != NULL) {
-		assert(usize == isalloc(ptr, true));
+		assert(usize == isalloc(tsd_tsdn(tsd), ptr, true));
 		if (prof_sample_accum_update(tsd, usize, true, NULL)) {
 			/*
 			 * Don't sample.  The usize passed to prof_alloc_prep()
@@ -512,6 +513,7 @@ prof_realloc(tsd_t *tsd, const void *ptr, size_t usize, prof_tctx_t *tctx,
 			 * though its actual usize was insufficient to cross the
 			 * sample threshold.
 			 */
+			prof_alloc_rollback(tsd, tctx, true);
 			tctx = (prof_tctx_t *)(uintptr_t)1U;
 		}
 	}
@@ -520,9 +522,9 @@ prof_realloc(tsd_t *tsd, const void *ptr, size_t usize, prof_tctx_t *tctx,
 	old_sampled = ((uintptr_t)old_tctx > (uintptr_t)1U);
 
 	if (unlikely(sampled))
-		prof_malloc_sample_object(ptr, usize, tctx);
+		prof_malloc_sample_object(tsd_tsdn(tsd), ptr, usize, tctx);
 	else
-		prof_tctx_reset(ptr, usize, old_ptr, old_tctx);
+		prof_tctx_reset(tsd_tsdn(tsd), ptr, usize, old_ptr, old_tctx);
 
 	if (unlikely(old_sampled))
 		prof_free_sampled_object(tsd, old_usize, old_tctx);
@@ -531,10 +533,10 @@ prof_realloc(tsd_t *tsd, const void *ptr, size_t usize, prof_tctx_t *tctx,
 JEMALLOC_ALWAYS_INLINE void
 prof_free(tsd_t *tsd, const void *ptr, size_t usize)
 {
-	prof_tctx_t *tctx = prof_tctx_get(ptr);
+	prof_tctx_t *tctx = prof_tctx_get(tsd_tsdn(tsd), ptr);
 
 	cassert(config_prof);
-	assert(usize == isalloc(ptr, true));
+	assert(usize == isalloc(tsd_tsdn(tsd), ptr, true));
 
 	if (unlikely((uintptr_t)tctx > (uintptr_t)1U))
 		prof_free_sampled_object(tsd, usize, tctx);
diff --git a/deps/jemalloc/include/jemalloc/internal/rb.h b/deps/jemalloc/include/jemalloc/internal/rb.h
index 2ca8e5933..3770342f8 100644
--- a/deps/jemalloc/include/jemalloc/internal/rb.h
+++ b/deps/jemalloc/include/jemalloc/internal/rb.h
@@ -42,7 +42,6 @@ struct {								\
 #define	rb_tree(a_type)							\
 struct {								\
     a_type *rbt_root;							\
-    a_type rbt_nil;							\
 }
 
 /* Left accessors. */
@@ -79,6 +78,15 @@ struct {								\
     (a_node)->a_field.rbn_right_red = (a_type *) (((intptr_t)		\
       (a_node)->a_field.rbn_right_red) & ((ssize_t)-2));		\
 } while (0)
+
+/* Node initializer. */
+#define	rbt_node_new(a_type, a_field, a_rbt, a_node) do {		\
+    /* Bookkeeping bit cannot be used by node pointer. */		\
+    assert(((uintptr_t)(a_node) & 0x1) == 0);				\
+    rbtn_left_set(a_type, a_field, (a_node), NULL);	\
+    rbtn_right_set(a_type, a_field, (a_node), NULL);	\
+    rbtn_red_set(a_type, a_field, (a_node));				\
+} while (0)
 #else
 /* Right accessors. */
 #define	rbtn_right_get(a_type, a_field, a_node)				\
@@ -99,28 +107,26 @@ struct {								\
 #define	rbtn_black_set(a_type, a_field, a_node) do {			\
     (a_node)->a_field.rbn_red = false;					\
 } while (0)
-#endif
 
 /* Node initializer. */
 #define	rbt_node_new(a_type, a_field, a_rbt, a_node) do {		\
-    rbtn_left_set(a_type, a_field, (a_node), &(a_rbt)->rbt_nil);	\
-    rbtn_right_set(a_type, a_field, (a_node), &(a_rbt)->rbt_nil);	\
+    rbtn_left_set(a_type, a_field, (a_node), NULL);	\
+    rbtn_right_set(a_type, a_field, (a_node), NULL);	\
     rbtn_red_set(a_type, a_field, (a_node));				\
 } while (0)
+#endif
 
 /* Tree initializer. */
 #define	rb_new(a_type, a_field, a_rbt) do {				\
-    (a_rbt)->rbt_root = &(a_rbt)->rbt_nil;				\
-    rbt_node_new(a_type, a_field, a_rbt, &(a_rbt)->rbt_nil);		\
-    rbtn_black_set(a_type, a_field, &(a_rbt)->rbt_nil);			\
+    (a_rbt)->rbt_root = NULL;						\
 } while (0)
 
 /* Internal utility macros. */
 #define	rbtn_first(a_type, a_field, a_rbt, a_root, r_node) do {		\
     (r_node) = (a_root);						\
-    if ((r_node) != &(a_rbt)->rbt_nil) {				\
+    if ((r_node) != NULL) {						\
 	for (;								\
-	  rbtn_left_get(a_type, a_field, (r_node)) != &(a_rbt)->rbt_nil;\
+	  rbtn_left_get(a_type, a_field, (r_node)) != NULL;		\
 	  (r_node) = rbtn_left_get(a_type, a_field, (r_node))) {	\
 	}								\
     }									\
@@ -128,10 +134,9 @@ struct {								\
 
 #define	rbtn_last(a_type, a_field, a_rbt, a_root, r_node) do {		\
     (r_node) = (a_root);						\
-    if ((r_node) != &(a_rbt)->rbt_nil) {				\
-	for (; rbtn_right_get(a_type, a_field, (r_node)) !=		\
-	  &(a_rbt)->rbt_nil; (r_node) = rbtn_right_get(a_type, a_field,	\
-	  (r_node))) {							\
+    if ((r_node) != NULL) {						\
+	for (; rbtn_right_get(a_type, a_field, (r_node)) != NULL;	\
+	  (r_node) = rbtn_right_get(a_type, a_field, (r_node))) {	\
 	}								\
     }									\
 } while (0)
@@ -169,11 +174,11 @@ a_prefix##next(a_rbt_type *rbtree, a_type *node);			\
 a_attr a_type *								\
 a_prefix##prev(a_rbt_type *rbtree, a_type *node);			\
 a_attr a_type *								\
-a_prefix##search(a_rbt_type *rbtree, a_type *key);			\
+a_prefix##search(a_rbt_type *rbtree, const a_type *key);		\
 a_attr a_type *								\
-a_prefix##nsearch(a_rbt_type *rbtree, a_type *key);			\
+a_prefix##nsearch(a_rbt_type *rbtree, const a_type *key);		\
 a_attr a_type *								\
-a_prefix##psearch(a_rbt_type *rbtree, a_type *key);			\
+a_prefix##psearch(a_rbt_type *rbtree, const a_type *key);		\
 a_attr void								\
 a_prefix##insert(a_rbt_type *rbtree, a_type *node);			\
 a_attr void								\
@@ -183,7 +188,10 @@ a_prefix##iter(a_rbt_type *rbtree, a_type *start, a_type *(*cb)(	\
   a_rbt_type *, a_type *, void *), void *arg);				\
 a_attr a_type *								\
 a_prefix##reverse_iter(a_rbt_type *rbtree, a_type *start,		\
-  a_type *(*cb)(a_rbt_type *, a_type *, void *), void *arg);
+  a_type *(*cb)(a_rbt_type *, a_type *, void *), void *arg);		\
+a_attr void								\
+a_prefix##destroy(a_rbt_type *rbtree, void (*cb)(a_type *, void *),	\
+  void *arg);
 
 /*
  * The rb_gen() macro generates a type-specific red-black tree implementation,
@@ -254,7 +262,7 @@ a_prefix##reverse_iter(a_rbt_type *rbtree, a_type *start,		\
  *            last/first.
  *
  *   static ex_node_t *
- *   ex_search(ex_t *tree, ex_node_t *key);
+ *   ex_search(ex_t *tree, const ex_node_t *key);
  *       Description: Search for node that matches key.
  *       Args:
  *         tree: Pointer to an initialized red-black tree object.
@@ -262,9 +270,9 @@ a_prefix##reverse_iter(a_rbt_type *rbtree, a_type *start,		\
  *       Ret: Node in tree that matches key, or NULL if no match.
  *
  *   static ex_node_t *
- *   ex_nsearch(ex_t *tree, ex_node_t *key);
+ *   ex_nsearch(ex_t *tree, const ex_node_t *key);
  *   static ex_node_t *
- *   ex_psearch(ex_t *tree, ex_node_t *key);
+ *   ex_psearch(ex_t *tree, const ex_node_t *key);
  *       Description: Search for node that matches key.  If no match is found,
  *                    return what would be key's successor/predecessor, were
  *                    key in tree.
@@ -312,6 +320,20 @@ a_prefix##reverse_iter(a_rbt_type *rbtree, a_type *start,		\
  *         arg  : Opaque pointer passed to cb().
  *       Ret: NULL if iteration completed, or the non-NULL callback return value
  *            that caused termination of the iteration.
+ *
+ *   static void
+ *   ex_destroy(ex_t *tree, void (*cb)(ex_node_t *, void *), void *arg);
+ *       Description: Iterate over the tree with post-order traversal, remove
+ *                    each node, and run the callback if non-null.  This is
+ *                    used for destroying a tree without paying the cost to
+ *                    rebalance it.  The tree must not be otherwise altered
+ *                    during traversal.
+ *       Args:
+ *         tree: Pointer to an initialized red-black tree object.
+ *         cb  : Callback function, which, if non-null, is called for each node
+ *               during iteration.  There is no way to stop iteration once it
+ *               has begun.
+ *         arg : Opaque pointer passed to cb().
  */
 #define	rb_gen(a_attr, a_prefix, a_rbt_type, a_type, a_field, a_cmp)	\
 a_attr void								\
@@ -320,36 +342,30 @@ a_prefix##new(a_rbt_type *rbtree) {					\
 }									\
 a_attr bool								\
 a_prefix##empty(a_rbt_type *rbtree) {					\
-    return (rbtree->rbt_root == &rbtree->rbt_nil);			\
+    return (rbtree->rbt_root == NULL);					\
 }									\
 a_attr a_type *								\
 a_prefix##first(a_rbt_type *rbtree) {					\
     a_type *ret;							\
     rbtn_first(a_type, a_field, rbtree, rbtree->rbt_root, ret);		\
-    if (ret == &rbtree->rbt_nil) {					\
-	ret = NULL;							\
-    }									\
     return (ret);							\
 }									\
 a_attr a_type *								\
 a_prefix##last(a_rbt_type *rbtree) {					\
     a_type *ret;							\
     rbtn_last(a_type, a_field, rbtree, rbtree->rbt_root, ret);		\
-    if (ret == &rbtree->rbt_nil) {					\
-	ret = NULL;							\
-    }									\
     return (ret);							\
 }									\
 a_attr a_type *								\
 a_prefix##next(a_rbt_type *rbtree, a_type *node) {			\
     a_type *ret;							\
-    if (rbtn_right_get(a_type, a_field, node) != &rbtree->rbt_nil) {	\
+    if (rbtn_right_get(a_type, a_field, node) != NULL) {		\
 	rbtn_first(a_type, a_field, rbtree, rbtn_right_get(a_type,	\
 	  a_field, node), ret);						\
     } else {								\
 	a_type *tnode = rbtree->rbt_root;				\
-	assert(tnode != &rbtree->rbt_nil);				\
-	ret = &rbtree->rbt_nil;						\
+	assert(tnode != NULL);						\
+	ret = NULL;							\
 	while (true) {							\
 	    int cmp = (a_cmp)(node, tnode);				\
 	    if (cmp < 0) {						\
@@ -360,24 +376,21 @@ a_prefix##next(a_rbt_type *rbtree, a_type *node) {			\
 	    } else {							\
 		break;							\
 	    }								\
-	    assert(tnode != &rbtree->rbt_nil);				\
+	    assert(tnode != NULL);					\
 	}								\
     }									\
-    if (ret == &rbtree->rbt_nil) {					\
-	ret = (NULL);							\
-    }									\
     return (ret);							\
 }									\
 a_attr a_type *								\
 a_prefix##prev(a_rbt_type *rbtree, a_type *node) {			\
     a_type *ret;							\
-    if (rbtn_left_get(a_type, a_field, node) != &rbtree->rbt_nil) {	\
+    if (rbtn_left_get(a_type, a_field, node) != NULL) {			\
 	rbtn_last(a_type, a_field, rbtree, rbtn_left_get(a_type,	\
 	  a_field, node), ret);						\
     } else {								\
 	a_type *tnode = rbtree->rbt_root;				\
-	assert(tnode != &rbtree->rbt_nil);				\
-	ret = &rbtree->rbt_nil;						\
+	assert(tnode != NULL);						\
+	ret = NULL;							\
 	while (true) {							\
 	    int cmp = (a_cmp)(node, tnode);				\
 	    if (cmp < 0) {						\
@@ -388,20 +401,17 @@ a_prefix##prev(a_rbt_type *rbtree, a_type *node) {			\
 	    } else {							\
 		break;							\
 	    }								\
-	    assert(tnode != &rbtree->rbt_nil);				\
+	    assert(tnode != NULL);					\
 	}								\
     }									\
-    if (ret == &rbtree->rbt_nil) {					\
-	ret = (NULL);							\
-    }									\
     return (ret);							\
 }									\
 a_attr a_type *								\
-a_prefix##search(a_rbt_type *rbtree, a_type *key) {			\
+a_prefix##search(a_rbt_type *rbtree, const a_type *key) {		\
     a_type *ret;							\
     int cmp;								\
     ret = rbtree->rbt_root;						\
-    while (ret != &rbtree->rbt_nil					\
+    while (ret != NULL							\
       && (cmp = (a_cmp)(key, ret)) != 0) {				\
 	if (cmp < 0) {							\
 	    ret = rbtn_left_get(a_type, a_field, ret);			\
@@ -409,17 +419,14 @@ a_prefix##search(a_rbt_type *rbtree, a_type *key) {			\
 	    ret = rbtn_right_get(a_type, a_field, ret);			\
 	}								\
     }									\
-    if (ret == &rbtree->rbt_nil) {					\
-	ret = (NULL);							\
-    }									\
     return (ret);							\
 }									\
 a_attr a_type *								\
-a_prefix##nsearch(a_rbt_type *rbtree, a_type *key) {			\
+a_prefix##nsearch(a_rbt_type *rbtree, const a_type *key) {		\
     a_type *ret;							\
     a_type *tnode = rbtree->rbt_root;					\
-    ret = &rbtree->rbt_nil;						\
-    while (tnode != &rbtree->rbt_nil) {					\
+    ret = NULL;								\
+    while (tnode != NULL) {						\
 	int cmp = (a_cmp)(key, tnode);					\
 	if (cmp < 0) {							\
 	    ret = tnode;						\
@@ -431,17 +438,14 @@ a_prefix##nsearch(a_rbt_type *rbtree, a_type *key) {			\
 	    break;							\
 	}								\
     }									\
-    if (ret == &rbtree->rbt_nil) {					\
-	ret = (NULL);							\
-    }									\
     return (ret);							\
 }									\
 a_attr a_type *								\
-a_prefix##psearch(a_rbt_type *rbtree, a_type *key) {			\
+a_prefix##psearch(a_rbt_type *rbtree, const a_type *key) {		\
     a_type *ret;							\
     a_type *tnode = rbtree->rbt_root;					\
-    ret = &rbtree->rbt_nil;						\
-    while (tnode != &rbtree->rbt_nil) {					\
+    ret = NULL;								\
+    while (tnode != NULL) {						\
 	int cmp = (a_cmp)(key, tnode);					\
 	if (cmp < 0) {							\
 	    tnode = rbtn_left_get(a_type, a_field, tnode);		\
@@ -453,9 +457,6 @@ a_prefix##psearch(a_rbt_type *rbtree, a_type *key) {			\
 	    break;							\
 	}								\
     }									\
-    if (ret == &rbtree->rbt_nil) {					\
-	ret = (NULL);							\
-    }									\
     return (ret);							\
 }									\
 a_attr void								\
@@ -467,7 +468,7 @@ a_prefix##insert(a_rbt_type *rbtree, a_type *node) {			\
     rbt_node_new(a_type, a_field, rbtree, node);			\
     /* Wind. */								\
     path->node = rbtree->rbt_root;					\
-    for (pathp = path; pathp->node != &rbtree->rbt_nil; pathp++) {	\
+    for (pathp = path; pathp->node != NULL; pathp++) {			\
 	int cmp = pathp->cmp = a_cmp(node, pathp->node);		\
 	assert(cmp != 0);						\
 	if (cmp < 0) {							\
@@ -487,7 +488,8 @@ a_prefix##insert(a_rbt_type *rbtree, a_type *node) {			\
 	    rbtn_left_set(a_type, a_field, cnode, left);		\
 	    if (rbtn_red_get(a_type, a_field, left)) {			\
 		a_type *leftleft = rbtn_left_get(a_type, a_field, left);\
-		if (rbtn_red_get(a_type, a_field, leftleft)) {		\
+		if (leftleft != NULL && rbtn_red_get(a_type, a_field,	\
+		  leftleft)) {						\
 		    /* Fix up 4-node. */				\
 		    a_type *tnode;					\
 		    rbtn_black_set(a_type, a_field, leftleft);		\
@@ -502,7 +504,8 @@ a_prefix##insert(a_rbt_type *rbtree, a_type *node) {			\
 	    rbtn_right_set(a_type, a_field, cnode, right);		\
 	    if (rbtn_red_get(a_type, a_field, right)) {			\
 		a_type *left = rbtn_left_get(a_type, a_field, cnode);	\
-		if (rbtn_red_get(a_type, a_field, left)) {		\
+		if (left != NULL && rbtn_red_get(a_type, a_field,	\
+		  left)) {						\
 		    /* Split 4-node. */					\
 		    rbtn_black_set(a_type, a_field, left);		\
 		    rbtn_black_set(a_type, a_field, right);		\
@@ -535,7 +538,7 @@ a_prefix##remove(a_rbt_type *rbtree, a_type *node) {			\
     /* Wind. */								\
     nodep = NULL; /* Silence compiler warning. */			\
     path->node = rbtree->rbt_root;					\
-    for (pathp = path; pathp->node != &rbtree->rbt_nil; pathp++) {	\
+    for (pathp = path; pathp->node != NULL; pathp++) {			\
 	int cmp = pathp->cmp = a_cmp(node, pathp->node);		\
 	if (cmp < 0) {							\
 	    pathp[1].node = rbtn_left_get(a_type, a_field,		\
@@ -547,7 +550,7 @@ a_prefix##remove(a_rbt_type *rbtree, a_type *node) {			\
 	        /* Find node's successor, in preparation for swap. */	\
 		pathp->cmp = 1;						\
 		nodep = pathp;						\
-		for (pathp++; pathp->node != &rbtree->rbt_nil;		\
+		for (pathp++; pathp->node != NULL;			\
 		  pathp++) {						\
 		    pathp->cmp = -1;					\
 		    pathp[1].node = rbtn_left_get(a_type, a_field,	\
@@ -590,7 +593,7 @@ a_prefix##remove(a_rbt_type *rbtree, a_type *node) {			\
 	}								\
     } else {								\
 	a_type *left = rbtn_left_get(a_type, a_field, node);		\
-	if (left != &rbtree->rbt_nil) {					\
+	if (left != NULL) {						\
 	    /* node has no successor, but it has a left child.        */\
 	    /* Splice node out, without losing the left child.        */\
 	    assert(!rbtn_red_get(a_type, a_field, node));		\
@@ -610,33 +613,32 @@ a_prefix##remove(a_rbt_type *rbtree, a_type *node) {			\
 	    return;							\
 	} else if (pathp == path) {					\
 	    /* The tree only contained one node. */			\
-	    rbtree->rbt_root = &rbtree->rbt_nil;			\
+	    rbtree->rbt_root = NULL;					\
 	    return;							\
 	}								\
     }									\
     if (rbtn_red_get(a_type, a_field, pathp->node)) {			\
 	/* Prune red node, which requires no fixup. */			\
 	assert(pathp[-1].cmp < 0);					\
-	rbtn_left_set(a_type, a_field, pathp[-1].node,			\
-	  &rbtree->rbt_nil);						\
+	rbtn_left_set(a_type, a_field, pathp[-1].node, NULL);		\
 	return;								\
     }									\
     /* The node to be pruned is black, so unwind until balance is     */\
     /* restored.                                                      */\
-    pathp->node = &rbtree->rbt_nil;					\
+    pathp->node = NULL;							\
     for (pathp--; (uintptr_t)pathp >= (uintptr_t)path; pathp--) {	\
 	assert(pathp->cmp != 0);					\
 	if (pathp->cmp < 0) {						\
 	    rbtn_left_set(a_type, a_field, pathp->node,			\
 	      pathp[1].node);						\
-	    assert(!rbtn_red_get(a_type, a_field, pathp[1].node));	\
 	    if (rbtn_red_get(a_type, a_field, pathp->node)) {		\
 		a_type *right = rbtn_right_get(a_type, a_field,		\
 		  pathp->node);						\
 		a_type *rightleft = rbtn_left_get(a_type, a_field,	\
 		  right);						\
 		a_type *tnode;						\
-		if (rbtn_red_get(a_type, a_field, rightleft)) {		\
+		if (rightleft != NULL && rbtn_red_get(a_type, a_field,	\
+		  rightleft)) {						\
 		    /* In the following diagrams, ||, //, and \\      */\
 		    /* indicate the path to the removed node.         */\
 		    /*                                                */\
@@ -679,7 +681,8 @@ a_prefix##remove(a_rbt_type *rbtree, a_type *node) {			\
 		  pathp->node);						\
 		a_type *rightleft = rbtn_left_get(a_type, a_field,	\
 		  right);						\
-		if (rbtn_red_get(a_type, a_field, rightleft)) {		\
+		if (rightleft != NULL && rbtn_red_get(a_type, a_field,	\
+		  rightleft)) {						\
 		    /*      ||                                        */\
 		    /*    pathp(b)                                    */\
 		    /*  //        \                                   */\
@@ -733,7 +736,8 @@ a_prefix##remove(a_rbt_type *rbtree, a_type *node) {			\
 		  left);						\
 		a_type *leftrightleft = rbtn_left_get(a_type, a_field,	\
 		  leftright);						\
-		if (rbtn_red_get(a_type, a_field, leftrightleft)) {	\
+		if (leftrightleft != NULL && rbtn_red_get(a_type,	\
+		  a_field, leftrightleft)) {				\
 		    /*      ||                                        */\
 		    /*    pathp(b)                                    */\
 		    /*   /        \\                                  */\
@@ -759,7 +763,7 @@ a_prefix##remove(a_rbt_type *rbtree, a_type *node) {			\
 		    /*   (b)                                          */\
 		    /*   /                                            */\
 		    /* (b)                                            */\
-		    assert(leftright != &rbtree->rbt_nil);		\
+		    assert(leftright != NULL);				\
 		    rbtn_red_set(a_type, a_field, leftright);		\
 		    rbtn_rotate_right(a_type, a_field, pathp->node,	\
 		      tnode);						\
@@ -782,7 +786,8 @@ a_prefix##remove(a_rbt_type *rbtree, a_type *node) {			\
 		return;							\
 	    } else if (rbtn_red_get(a_type, a_field, pathp->node)) {	\
 		a_type *leftleft = rbtn_left_get(a_type, a_field, left);\
-		if (rbtn_red_get(a_type, a_field, leftleft)) {		\
+		if (leftleft != NULL && rbtn_red_get(a_type, a_field,	\
+		  leftleft)) {						\
 		    /*        ||                                      */\
 		    /*      pathp(r)                                  */\
 		    /*     /        \\                                */\
@@ -820,7 +825,8 @@ a_prefix##remove(a_rbt_type *rbtree, a_type *node) {			\
 		}							\
 	    } else {							\
 		a_type *leftleft = rbtn_left_get(a_type, a_field, left);\
-		if (rbtn_red_get(a_type, a_field, leftleft)) {		\
+		if (leftleft != NULL && rbtn_red_get(a_type, a_field,	\
+		  leftleft)) {						\
 		    /*               ||                               */\
 		    /*             pathp(b)                           */\
 		    /*            /        \\                         */\
@@ -866,13 +872,13 @@ a_prefix##remove(a_rbt_type *rbtree, a_type *node) {			\
 a_attr a_type *								\
 a_prefix##iter_recurse(a_rbt_type *rbtree, a_type *node,		\
   a_type *(*cb)(a_rbt_type *, a_type *, void *), void *arg) {		\
-    if (node == &rbtree->rbt_nil) {					\
-	return (&rbtree->rbt_nil);					\
+    if (node == NULL) {							\
+	return (NULL);							\
     } else {								\
 	a_type *ret;							\
 	if ((ret = a_prefix##iter_recurse(rbtree, rbtn_left_get(a_type,	\
-	  a_field, node), cb, arg)) != &rbtree->rbt_nil			\
-	  || (ret = cb(rbtree, node, arg)) != NULL) {			\
+	  a_field, node), cb, arg)) != NULL || (ret = cb(rbtree, node,	\
+	  arg)) != NULL) {						\
 	    return (ret);						\
 	}								\
 	return (a_prefix##iter_recurse(rbtree, rbtn_right_get(a_type,	\
@@ -886,8 +892,8 @@ a_prefix##iter_start(a_rbt_type *rbtree, a_type *start, a_type *node,	\
     if (cmp < 0) {							\
 	a_type *ret;							\
 	if ((ret = a_prefix##iter_start(rbtree, start,			\
-	  rbtn_left_get(a_type, a_field, node), cb, arg)) !=		\
-	  &rbtree->rbt_nil || (ret = cb(rbtree, node, arg)) != NULL) {	\
+	  rbtn_left_get(a_type, a_field, node), cb, arg)) != NULL ||	\
+	  (ret = cb(rbtree, node, arg)) != NULL) {			\
 	    return (ret);						\
 	}								\
 	return (a_prefix##iter_recurse(rbtree, rbtn_right_get(a_type,	\
@@ -914,21 +920,18 @@ a_prefix##iter(a_rbt_type *rbtree, a_type *start, a_type *(*cb)(	\
     } else {								\
 	ret = a_prefix##iter_recurse(rbtree, rbtree->rbt_root, cb, arg);\
     }									\
-    if (ret == &rbtree->rbt_nil) {					\
-	ret = NULL;							\
-    }									\
     return (ret);							\
 }									\
 a_attr a_type *								\
 a_prefix##reverse_iter_recurse(a_rbt_type *rbtree, a_type *node,	\
   a_type *(*cb)(a_rbt_type *, a_type *, void *), void *arg) {		\
-    if (node == &rbtree->rbt_nil) {					\
-	return (&rbtree->rbt_nil);					\
+    if (node == NULL) {							\
+	return (NULL);							\
     } else {								\
 	a_type *ret;							\
 	if ((ret = a_prefix##reverse_iter_recurse(rbtree,		\
-	  rbtn_right_get(a_type, a_field, node), cb, arg)) !=		\
-	  &rbtree->rbt_nil || (ret = cb(rbtree, node, arg)) != NULL) {	\
+	  rbtn_right_get(a_type, a_field, node), cb, arg)) != NULL ||	\
+	  (ret = cb(rbtree, node, arg)) != NULL) {			\
 	    return (ret);						\
 	}								\
 	return (a_prefix##reverse_iter_recurse(rbtree,			\
@@ -943,8 +946,8 @@ a_prefix##reverse_iter_start(a_rbt_type *rbtree, a_type *start,		\
     if (cmp > 0) {							\
 	a_type *ret;							\
 	if ((ret = a_prefix##reverse_iter_start(rbtree, start,		\
-	  rbtn_right_get(a_type, a_field, node), cb, arg)) !=		\
-	  &rbtree->rbt_nil || (ret = cb(rbtree, node, arg)) != NULL) {	\
+	  rbtn_right_get(a_type, a_field, node), cb, arg)) != NULL ||	\
+	  (ret = cb(rbtree, node, arg)) != NULL) {			\
 	    return (ret);						\
 	}								\
 	return (a_prefix##reverse_iter_recurse(rbtree,			\
@@ -972,10 +975,29 @@ a_prefix##reverse_iter(a_rbt_type *rbtree, a_type *start,		\
 	ret = a_prefix##reverse_iter_recurse(rbtree, rbtree->rbt_root,	\
 	  cb, arg);							\
     }									\
-    if (ret == &rbtree->rbt_nil) {					\
-	ret = NULL;							\
-    }									\
     return (ret);							\
+}									\
+a_attr void								\
+a_prefix##destroy_recurse(a_rbt_type *rbtree, a_type *node, void (*cb)(	\
+  a_type *, void *), void *arg) {					\
+    if (node == NULL) {							\
+	return;								\
+    }									\
+    a_prefix##destroy_recurse(rbtree, rbtn_left_get(a_type, a_field,	\
+      node), cb, arg);							\
+    rbtn_left_set(a_type, a_field, (node), NULL);			\
+    a_prefix##destroy_recurse(rbtree, rbtn_right_get(a_type, a_field,	\
+      node), cb, arg);							\
+    rbtn_right_set(a_type, a_field, (node), NULL);			\
+    if (cb) {								\
+	cb(node, arg);							\
+    }									\
+}									\
+a_attr void								\
+a_prefix##destroy(a_rbt_type *rbtree, void (*cb)(a_type *, void *),	\
+  void *arg) {								\
+    a_prefix##destroy_recurse(rbtree, rbtree->rbt_root, cb, arg);	\
+    rbtree->rbt_root = NULL;						\
 }
 
 #endif /* RB_H_ */
diff --git a/deps/jemalloc/include/jemalloc/internal/rtree.h b/deps/jemalloc/include/jemalloc/internal/rtree.h
index 28ae9d1dd..8d0c584da 100644
--- a/deps/jemalloc/include/jemalloc/internal/rtree.h
+++ b/deps/jemalloc/include/jemalloc/internal/rtree.h
@@ -15,9 +15,10 @@ typedef struct rtree_s rtree_t;
  * machine address width.
  */
 #define	LG_RTREE_BITS_PER_LEVEL	4
-#define	RTREE_BITS_PER_LEVEL	(ZU(1) << LG_RTREE_BITS_PER_LEVEL)
+#define	RTREE_BITS_PER_LEVEL	(1U << LG_RTREE_BITS_PER_LEVEL)
+/* Maximum rtree height. */
 #define	RTREE_HEIGHT_MAX						\
-    ((ZU(1) << (LG_SIZEOF_PTR+3)) / RTREE_BITS_PER_LEVEL)
+    ((1U << (LG_SIZEOF_PTR+3)) / RTREE_BITS_PER_LEVEL)
 
 /* Used for two-stage lock-free node initialization. */
 #define	RTREE_NODE_INITIALIZING	((rtree_node_elm_t *)0x1)
@@ -111,22 +112,25 @@ unsigned	rtree_start_level(rtree_t *rtree, uintptr_t key);
 uintptr_t	rtree_subkey(rtree_t *rtree, uintptr_t key, unsigned level);
 
 bool	rtree_node_valid(rtree_node_elm_t *node);
-rtree_node_elm_t	*rtree_child_tryread(rtree_node_elm_t *elm);
+rtree_node_elm_t	*rtree_child_tryread(rtree_node_elm_t *elm,
+    bool dependent);
 rtree_node_elm_t	*rtree_child_read(rtree_t *rtree, rtree_node_elm_t *elm,
-    unsigned level);
+    unsigned level, bool dependent);
 extent_node_t	*rtree_val_read(rtree_t *rtree, rtree_node_elm_t *elm,
     bool dependent);
 void	rtree_val_write(rtree_t *rtree, rtree_node_elm_t *elm,
     const extent_node_t *val);
-rtree_node_elm_t	*rtree_subtree_tryread(rtree_t *rtree, unsigned level);
-rtree_node_elm_t	*rtree_subtree_read(rtree_t *rtree, unsigned level);
+rtree_node_elm_t	*rtree_subtree_tryread(rtree_t *rtree, unsigned level,
+    bool dependent);
+rtree_node_elm_t	*rtree_subtree_read(rtree_t *rtree, unsigned level,
+    bool dependent);
 
 extent_node_t	*rtree_get(rtree_t *rtree, uintptr_t key, bool dependent);
 bool	rtree_set(rtree_t *rtree, uintptr_t key, const extent_node_t *val);
 #endif
 
 #if (defined(JEMALLOC_ENABLE_INLINE) || defined(JEMALLOC_RTREE_C_))
-JEMALLOC_INLINE unsigned
+JEMALLOC_ALWAYS_INLINE unsigned
 rtree_start_level(rtree_t *rtree, uintptr_t key)
 {
 	unsigned start_level;
@@ -140,7 +144,7 @@ rtree_start_level(rtree_t *rtree, uintptr_t key)
 	return (start_level);
 }
 
-JEMALLOC_INLINE uintptr_t
+JEMALLOC_ALWAYS_INLINE uintptr_t
 rtree_subkey(rtree_t *rtree, uintptr_t key, unsigned level)
 {
 
@@ -149,37 +153,40 @@ rtree_subkey(rtree_t *rtree, uintptr_t key, unsigned level)
 	    rtree->levels[level].bits) - 1));
 }
 
-JEMALLOC_INLINE bool
+JEMALLOC_ALWAYS_INLINE bool
 rtree_node_valid(rtree_node_elm_t *node)
 {
 
 	return ((uintptr_t)node > (uintptr_t)RTREE_NODE_INITIALIZING);
 }
 
-JEMALLOC_INLINE rtree_node_elm_t *
-rtree_child_tryread(rtree_node_elm_t *elm)
+JEMALLOC_ALWAYS_INLINE rtree_node_elm_t *
+rtree_child_tryread(rtree_node_elm_t *elm, bool dependent)
 {
 	rtree_node_elm_t *child;
 
 	/* Double-checked read (first read may be stale. */
 	child = elm->child;
-	if (!rtree_node_valid(child))
+	if (!dependent && !rtree_node_valid(child))
 		child = atomic_read_p(&elm->pun);
+	assert(!dependent || child != NULL);
 	return (child);
 }
 
-JEMALLOC_INLINE rtree_node_elm_t *
-rtree_child_read(rtree_t *rtree, rtree_node_elm_t *elm, unsigned level)
+JEMALLOC_ALWAYS_INLINE rtree_node_elm_t *
+rtree_child_read(rtree_t *rtree, rtree_node_elm_t *elm, unsigned level,
+    bool dependent)
 {
 	rtree_node_elm_t *child;
 
-	child = rtree_child_tryread(elm);
-	if (unlikely(!rtree_node_valid(child)))
+	child = rtree_child_tryread(elm, dependent);
+	if (!dependent && unlikely(!rtree_node_valid(child)))
 		child = rtree_child_read_hard(rtree, elm, level);
+	assert(!dependent || child != NULL);
 	return (child);
 }
 
-JEMALLOC_INLINE extent_node_t *
+JEMALLOC_ALWAYS_INLINE extent_node_t *
 rtree_val_read(rtree_t *rtree, rtree_node_elm_t *elm, bool dependent)
 {
 
@@ -208,54 +215,119 @@ rtree_val_write(rtree_t *rtree, rtree_node_elm_t *elm, const extent_node_t *val)
 	atomic_write_p(&elm->pun, val);
 }
 
-JEMALLOC_INLINE rtree_node_elm_t *
-rtree_subtree_tryread(rtree_t *rtree, unsigned level)
+JEMALLOC_ALWAYS_INLINE rtree_node_elm_t *
+rtree_subtree_tryread(rtree_t *rtree, unsigned level, bool dependent)
 {
 	rtree_node_elm_t *subtree;
 
 	/* Double-checked read (first read may be stale. */
 	subtree = rtree->levels[level].subtree;
-	if (!rtree_node_valid(subtree))
+	if (!dependent && unlikely(!rtree_node_valid(subtree)))
 		subtree = atomic_read_p(&rtree->levels[level].subtree_pun);
+	assert(!dependent || subtree != NULL);
 	return (subtree);
 }
 
-JEMALLOC_INLINE rtree_node_elm_t *
-rtree_subtree_read(rtree_t *rtree, unsigned level)
+JEMALLOC_ALWAYS_INLINE rtree_node_elm_t *
+rtree_subtree_read(rtree_t *rtree, unsigned level, bool dependent)
 {
 	rtree_node_elm_t *subtree;
 
-	subtree = rtree_subtree_tryread(rtree, level);
-	if (unlikely(!rtree_node_valid(subtree)))
+	subtree = rtree_subtree_tryread(rtree, level, dependent);
+	if (!dependent && unlikely(!rtree_node_valid(subtree)))
 		subtree = rtree_subtree_read_hard(rtree, level);
+	assert(!dependent || subtree != NULL);
 	return (subtree);
 }
 
-JEMALLOC_INLINE extent_node_t *
+JEMALLOC_ALWAYS_INLINE extent_node_t *
 rtree_get(rtree_t *rtree, uintptr_t key, bool dependent)
 {
 	uintptr_t subkey;
-	unsigned i, start_level;
-	rtree_node_elm_t *node, *child;
+	unsigned start_level;
+	rtree_node_elm_t *node;
 
 	start_level = rtree_start_level(rtree, key);
 
-	for (i = start_level, node = rtree_subtree_tryread(rtree, start_level);
-	    /**/; i++, node = child) {
-		if (!dependent && unlikely(!rtree_node_valid(node)))
-			return (NULL);
-		subkey = rtree_subkey(rtree, key, i);
-		if (i == rtree->height - 1) {
-			/*
-			 * node is a leaf, so it contains values rather than
-			 * child pointers.
-			 */
-			return (rtree_val_read(rtree, &node[subkey],
-			    dependent));
-		}
-		assert(i < rtree->height - 1);
-		child = rtree_child_tryread(&node[subkey]);
+	node = rtree_subtree_tryread(rtree, start_level, dependent);
+#define	RTREE_GET_BIAS	(RTREE_HEIGHT_MAX - rtree->height)
+	switch (start_level + RTREE_GET_BIAS) {
+#define	RTREE_GET_SUBTREE(level)					\
+	case level:							\
+		assert(level < (RTREE_HEIGHT_MAX-1));			\
+		if (!dependent && unlikely(!rtree_node_valid(node)))	\
+			return (NULL);					\
+		subkey = rtree_subkey(rtree, key, level -		\
+		    RTREE_GET_BIAS);					\
+		node = rtree_child_tryread(&node[subkey], dependent);	\
+		/* Fall through. */
+#define	RTREE_GET_LEAF(level)						\
+	case level:							\
+		assert(level == (RTREE_HEIGHT_MAX-1));			\
+		if (!dependent && unlikely(!rtree_node_valid(node)))	\
+			return (NULL);					\
+		subkey = rtree_subkey(rtree, key, level -		\
+		    RTREE_GET_BIAS);					\
+		/*							\
+		 * node is a leaf, so it contains values rather than	\
+		 * child pointers.					\
+		 */							\
+		return (rtree_val_read(rtree, &node[subkey],		\
+		    dependent));
+#if RTREE_HEIGHT_MAX > 1
+	RTREE_GET_SUBTREE(0)
+#endif
+#if RTREE_HEIGHT_MAX > 2
+	RTREE_GET_SUBTREE(1)
+#endif
+#if RTREE_HEIGHT_MAX > 3
+	RTREE_GET_SUBTREE(2)
+#endif
+#if RTREE_HEIGHT_MAX > 4
+	RTREE_GET_SUBTREE(3)
+#endif
+#if RTREE_HEIGHT_MAX > 5
+	RTREE_GET_SUBTREE(4)
+#endif
+#if RTREE_HEIGHT_MAX > 6
+	RTREE_GET_SUBTREE(5)
+#endif
+#if RTREE_HEIGHT_MAX > 7
+	RTREE_GET_SUBTREE(6)
+#endif
+#if RTREE_HEIGHT_MAX > 8
+	RTREE_GET_SUBTREE(7)
+#endif
+#if RTREE_HEIGHT_MAX > 9
+	RTREE_GET_SUBTREE(8)
+#endif
+#if RTREE_HEIGHT_MAX > 10
+	RTREE_GET_SUBTREE(9)
+#endif
+#if RTREE_HEIGHT_MAX > 11
+	RTREE_GET_SUBTREE(10)
+#endif
+#if RTREE_HEIGHT_MAX > 12
+	RTREE_GET_SUBTREE(11)
+#endif
+#if RTREE_HEIGHT_MAX > 13
+	RTREE_GET_SUBTREE(12)
+#endif
+#if RTREE_HEIGHT_MAX > 14
+	RTREE_GET_SUBTREE(13)
+#endif
+#if RTREE_HEIGHT_MAX > 15
+	RTREE_GET_SUBTREE(14)
+#endif
+#if RTREE_HEIGHT_MAX > 16
+#  error Unsupported RTREE_HEIGHT_MAX
+#endif
+	RTREE_GET_LEAF(RTREE_HEIGHT_MAX-1)
+#undef RTREE_GET_SUBTREE
+#undef RTREE_GET_LEAF
+	default: not_reached();
 	}
+#undef RTREE_GET_BIAS
 	not_reached();
 }
 
@@ -268,7 +340,7 @@ rtree_set(rtree_t *rtree, uintptr_t key, const extent_node_t *val)
 
 	start_level = rtree_start_level(rtree, key);
 
-	node = rtree_subtree_read(rtree, start_level);
+	node = rtree_subtree_read(rtree, start_level, false);
 	if (node == NULL)
 		return (true);
 	for (i = start_level; /**/; i++, node = child) {
@@ -282,7 +354,7 @@ rtree_set(rtree_t *rtree, uintptr_t key, const extent_node_t *val)
 			return (false);
 		}
 		assert(i + 1 < rtree->height);
-		child = rtree_child_read(rtree, &node[subkey], i);
+		child = rtree_child_read(rtree, &node[subkey], i, false);
 		if (child == NULL)
 			return (true);
 	}
diff --git a/deps/jemalloc/include/jemalloc/internal/size_classes.sh b/deps/jemalloc/include/jemalloc/internal/size_classes.sh
index fc82036d3..f6fbce4ef 100755
--- a/deps/jemalloc/include/jemalloc/internal/size_classes.sh
+++ b/deps/jemalloc/include/jemalloc/internal/size_classes.sh
@@ -48,6 +48,21 @@ size_class() {
   lg_p=$5
   lg_kmax=$6
 
+  if [ ${lg_delta} -ge ${lg_p} ] ; then
+    psz="yes"
+  else
+    pow2 ${lg_p}; p=${pow2_result}
+    pow2 ${lg_grp}; grp=${pow2_result}
+    pow2 ${lg_delta}; delta=${pow2_result}
+    sz=$((${grp} + ${delta} * ${ndelta}))
+    npgs=$((${sz} / ${p}))
+    if [ ${sz} -eq $((${npgs} * ${p})) ] ; then
+      psz="yes"
+    else
+      psz="no"
+    fi
+  fi
+
   lg ${ndelta}; lg_ndelta=${lg_result}; pow2 ${lg_ndelta}
   if [ ${pow2_result} -lt ${ndelta} ] ; then
     rem="yes"
@@ -74,14 +89,15 @@ size_class() {
   else
     lg_delta_lookup="no"
   fi
-  printf '    SC(%3d, %6d, %8d, %6d, %3s, %2s) \\\n' ${index} ${lg_grp} ${lg_delta} ${ndelta} ${bin} ${lg_delta_lookup}
+  printf '    SC(%3d, %6d, %8d, %6d, %3s, %3s, %2s) \\\n' ${index} ${lg_grp} ${lg_delta} ${ndelta} ${psz} ${bin} ${lg_delta_lookup}
   # Defined upon return:
-  # - lg_delta_lookup (${lg_delta} or "no")
+  # - psz ("yes" or "no")
   # - bin ("yes" or "no")
+  # - lg_delta_lookup (${lg_delta} or "no")
 }
 
 sep_line() {
-  echo "                                               \\"
+  echo "                                                         \\"
 }
 
 size_classes() {
@@ -95,12 +111,13 @@ size_classes() {
   pow2 ${lg_g}; g=${pow2_result}
 
   echo "#define	SIZE_CLASSES \\"
-  echo "  /* index, lg_grp, lg_delta, ndelta, bin, lg_delta_lookup */ \\"
+  echo "  /* index, lg_grp, lg_delta, ndelta, psz, bin, lg_delta_lookup */ \\"
 
   ntbins=0
   nlbins=0
   lg_tiny_maxclass='"NA"'
   nbins=0
+  npsizes=0
 
   # Tiny size classes.
   ndelta=0
@@ -112,6 +129,9 @@ size_classes() {
     if [ ${lg_delta_lookup} != "no" ] ; then
       nlbins=$((${index} + 1))
     fi
+    if [ ${psz} = "yes" ] ; then
+      npsizes=$((${npsizes} + 1))
+    fi
     if [ ${bin} != "no" ] ; then
       nbins=$((${index} + 1))
     fi
@@ -133,19 +153,25 @@ size_classes() {
     index=$((${index} + 1))
     lg_grp=$((${lg_grp} + 1))
     lg_delta=$((${lg_delta} + 1))
+    if [ ${psz} = "yes" ] ; then
+      npsizes=$((${npsizes} + 1))
+    fi
   fi
   while [ ${ndelta} -lt ${g} ] ; do
     size_class ${index} ${lg_grp} ${lg_delta} ${ndelta} ${lg_p} ${lg_kmax}
     index=$((${index} + 1))
     ndelta=$((${ndelta} + 1))
+    if [ ${psz} = "yes" ] ; then
+      npsizes=$((${npsizes} + 1))
+    fi
   done
 
   # All remaining groups.
   lg_grp=$((${lg_grp} + ${lg_g}))
-  while [ ${lg_grp} -lt ${ptr_bits} ] ; do
+  while [ ${lg_grp} -lt $((${ptr_bits} - 1)) ] ; do
     sep_line
     ndelta=1
-    if [ ${lg_grp} -eq $((${ptr_bits} - 1)) ] ; then
+    if [ ${lg_grp} -eq $((${ptr_bits} - 2)) ] ; then
       ndelta_limit=$((${g} - 1))
     else
       ndelta_limit=${g}
@@ -157,6 +183,9 @@ size_classes() {
         # Final written value is correct:
         lookup_maxclass="((((size_t)1) << ${lg_grp}) + (((size_t)${ndelta}) << ${lg_delta}))"
       fi
+      if [ ${psz} = "yes" ] ; then
+        npsizes=$((${npsizes} + 1))
+      fi
       if [ ${bin} != "no" ] ; then
         nbins=$((${index} + 1))
         # Final written value is correct:
@@ -183,6 +212,7 @@ size_classes() {
   # - nlbins
   # - nbins
   # - nsizes
+  # - npsizes
   # - lg_tiny_maxclass
   # - lookup_maxclass
   # - small_maxclass
@@ -200,13 +230,13 @@ cat <<EOF
  * be defined prior to inclusion, and it in turn defines:
  *
  *   LG_SIZE_CLASS_GROUP: Lg of size class count for each size doubling.
- *   SIZE_CLASSES: Complete table of
- *                 SC(index, lg_grp, lg_delta, ndelta, bin, lg_delta_lookup)
- *                 tuples.
+ *   SIZE_CLASSES: Complete table of SC(index, lg_grp, lg_delta, ndelta, psz,
+ *                 bin, lg_delta_lookup) tuples.
  *     index: Size class index.
  *     lg_grp: Lg group base size (no deltas added).
  *     lg_delta: Lg delta to previous size class.
  *     ndelta: Delta multiplier.  size == 1<<lg_grp + ndelta<<lg_delta
+ *     psz: 'yes' if a multiple of the page size, 'no' otherwise.
  *     bin: 'yes' if a small bin size class, 'no' otherwise.
  *     lg_delta_lookup: Same as lg_delta if a lookup table size class, 'no'
  *                      otherwise.
@@ -214,6 +244,7 @@ cat <<EOF
  *   NLBINS: Number of bins supported by the lookup table.
  *   NBINS: Number of small size class bins.
  *   NSIZES: Number of size classes.
+ *   NPSIZES: Number of size classes that are a multiple of (1U << LG_PAGE).
  *   LG_TINY_MAXCLASS: Lg of maximum tiny size class.
  *   LOOKUP_MAXCLASS: Maximum size class included in lookup table.
  *   SMALL_MAXCLASS: Maximum small size class.
@@ -238,6 +269,7 @@ for lg_z in ${lg_zarr} ; do
         echo "#define	NLBINS			${nlbins}"
         echo "#define	NBINS			${nbins}"
         echo "#define	NSIZES			${nsizes}"
+        echo "#define	NPSIZES			${npsizes}"
         echo "#define	LG_TINY_MAXCLASS	${lg_tiny_maxclass}"
         echo "#define	LOOKUP_MAXCLASS		${lookup_maxclass}"
         echo "#define	SMALL_MAXCLASS		${small_maxclass}"
diff --git a/deps/jemalloc/include/jemalloc/internal/smoothstep.h b/deps/jemalloc/include/jemalloc/internal/smoothstep.h
new file mode 100644
index 000000000..c5333ccad
--- /dev/null
+++ b/deps/jemalloc/include/jemalloc/internal/smoothstep.h
@@ -0,0 +1,246 @@
+/*
+ * This file was generated by the following command:
+ *   sh smoothstep.sh smoother 200 24 3 15
+ */
+/******************************************************************************/
+#ifdef JEMALLOC_H_TYPES
+
+/*
+ * This header defines a precomputed table based on the smoothstep family of
+ * sigmoidal curves (https://en.wikipedia.org/wiki/Smoothstep) that grow from 0
+ * to 1 in 0 <= x <= 1.  The table is stored as integer fixed point values so
+ * that floating point math can be avoided.
+ *
+ *                      3     2
+ *   smoothstep(x) = -2x  + 3x
+ *
+ *                       5      4      3
+ *   smootherstep(x) = 6x  - 15x  + 10x
+ *
+ *                          7      6      5      4
+ *   smootheststep(x) = -20x  + 70x  - 84x  + 35x
+ */
+
+#define	SMOOTHSTEP_VARIANT	"smoother"
+#define	SMOOTHSTEP_NSTEPS	200
+#define	SMOOTHSTEP_BFP		24
+#define	SMOOTHSTEP \
+ /* STEP(step, h,                            x,     y) */ \
+    STEP(   1, UINT64_C(0x0000000000000014), 0.005, 0.000001240643750) \
+    STEP(   2, UINT64_C(0x00000000000000a5), 0.010, 0.000009850600000) \
+    STEP(   3, UINT64_C(0x0000000000000229), 0.015, 0.000032995181250) \
+    STEP(   4, UINT64_C(0x0000000000000516), 0.020, 0.000077619200000) \
+    STEP(   5, UINT64_C(0x00000000000009dc), 0.025, 0.000150449218750) \
+    STEP(   6, UINT64_C(0x00000000000010e8), 0.030, 0.000257995800000) \
+    STEP(   7, UINT64_C(0x0000000000001aa4), 0.035, 0.000406555756250) \
+    STEP(   8, UINT64_C(0x0000000000002777), 0.040, 0.000602214400000) \
+    STEP(   9, UINT64_C(0x00000000000037c2), 0.045, 0.000850847793750) \
+    STEP(  10, UINT64_C(0x0000000000004be6), 0.050, 0.001158125000000) \
+    STEP(  11, UINT64_C(0x000000000000643c), 0.055, 0.001529510331250) \
+    STEP(  12, UINT64_C(0x000000000000811f), 0.060, 0.001970265600000) \
+    STEP(  13, UINT64_C(0x000000000000a2e2), 0.065, 0.002485452368750) \
+    STEP(  14, UINT64_C(0x000000000000c9d8), 0.070, 0.003079934200000) \
+    STEP(  15, UINT64_C(0x000000000000f64f), 0.075, 0.003758378906250) \
+    STEP(  16, UINT64_C(0x0000000000012891), 0.080, 0.004525260800000) \
+    STEP(  17, UINT64_C(0x00000000000160e7), 0.085, 0.005384862943750) \
+    STEP(  18, UINT64_C(0x0000000000019f95), 0.090, 0.006341279400000) \
+    STEP(  19, UINT64_C(0x000000000001e4dc), 0.095, 0.007398417481250) \
+    STEP(  20, UINT64_C(0x00000000000230fc), 0.100, 0.008560000000000) \
+    STEP(  21, UINT64_C(0x0000000000028430), 0.105, 0.009829567518750) \
+    STEP(  22, UINT64_C(0x000000000002deb0), 0.110, 0.011210480600000) \
+    STEP(  23, UINT64_C(0x00000000000340b1), 0.115, 0.012705922056250) \
+    STEP(  24, UINT64_C(0x000000000003aa67), 0.120, 0.014318899200000) \
+    STEP(  25, UINT64_C(0x0000000000041c00), 0.125, 0.016052246093750) \
+    STEP(  26, UINT64_C(0x00000000000495a8), 0.130, 0.017908625800000) \
+    STEP(  27, UINT64_C(0x000000000005178b), 0.135, 0.019890532631250) \
+    STEP(  28, UINT64_C(0x000000000005a1cf), 0.140, 0.022000294400000) \
+    STEP(  29, UINT64_C(0x0000000000063498), 0.145, 0.024240074668750) \
+    STEP(  30, UINT64_C(0x000000000006d009), 0.150, 0.026611875000000) \
+    STEP(  31, UINT64_C(0x000000000007743f), 0.155, 0.029117537206250) \
+    STEP(  32, UINT64_C(0x0000000000082157), 0.160, 0.031758745600000) \
+    STEP(  33, UINT64_C(0x000000000008d76b), 0.165, 0.034537029243750) \
+    STEP(  34, UINT64_C(0x0000000000099691), 0.170, 0.037453764200000) \
+    STEP(  35, UINT64_C(0x00000000000a5edf), 0.175, 0.040510175781250) \
+    STEP(  36, UINT64_C(0x00000000000b3067), 0.180, 0.043707340800000) \
+    STEP(  37, UINT64_C(0x00000000000c0b38), 0.185, 0.047046189818750) \
+    STEP(  38, UINT64_C(0x00000000000cef5e), 0.190, 0.050527509400000) \
+    STEP(  39, UINT64_C(0x00000000000ddce6), 0.195, 0.054151944356250) \
+    STEP(  40, UINT64_C(0x00000000000ed3d8), 0.200, 0.057920000000000) \
+    STEP(  41, UINT64_C(0x00000000000fd439), 0.205, 0.061832044393750) \
+    STEP(  42, UINT64_C(0x000000000010de0e), 0.210, 0.065888310600000) \
+    STEP(  43, UINT64_C(0x000000000011f158), 0.215, 0.070088898931250) \
+    STEP(  44, UINT64_C(0x0000000000130e17), 0.220, 0.074433779200000) \
+    STEP(  45, UINT64_C(0x0000000000143448), 0.225, 0.078922792968750) \
+    STEP(  46, UINT64_C(0x00000000001563e7), 0.230, 0.083555655800000) \
+    STEP(  47, UINT64_C(0x0000000000169cec), 0.235, 0.088331959506250) \
+    STEP(  48, UINT64_C(0x000000000017df4f), 0.240, 0.093251174400000) \
+    STEP(  49, UINT64_C(0x0000000000192b04), 0.245, 0.098312651543750) \
+    STEP(  50, UINT64_C(0x00000000001a8000), 0.250, 0.103515625000000) \
+    STEP(  51, UINT64_C(0x00000000001bde32), 0.255, 0.108859214081250) \
+    STEP(  52, UINT64_C(0x00000000001d458b), 0.260, 0.114342425600000) \
+    STEP(  53, UINT64_C(0x00000000001eb5f8), 0.265, 0.119964156118750) \
+    STEP(  54, UINT64_C(0x0000000000202f65), 0.270, 0.125723194200000) \
+    STEP(  55, UINT64_C(0x000000000021b1bb), 0.275, 0.131618222656250) \
+    STEP(  56, UINT64_C(0x0000000000233ce3), 0.280, 0.137647820800000) \
+    STEP(  57, UINT64_C(0x000000000024d0c3), 0.285, 0.143810466693750) \
+    STEP(  58, UINT64_C(0x0000000000266d40), 0.290, 0.150104539400000) \
+    STEP(  59, UINT64_C(0x000000000028123d), 0.295, 0.156528321231250) \
+    STEP(  60, UINT64_C(0x000000000029bf9c), 0.300, 0.163080000000000) \
+    STEP(  61, UINT64_C(0x00000000002b753d), 0.305, 0.169757671268750) \
+    STEP(  62, UINT64_C(0x00000000002d32fe), 0.310, 0.176559340600000) \
+    STEP(  63, UINT64_C(0x00000000002ef8bc), 0.315, 0.183482925806250) \
+    STEP(  64, UINT64_C(0x000000000030c654), 0.320, 0.190526259200000) \
+    STEP(  65, UINT64_C(0x0000000000329b9f), 0.325, 0.197687089843750) \
+    STEP(  66, UINT64_C(0x0000000000347875), 0.330, 0.204963085800000) \
+    STEP(  67, UINT64_C(0x0000000000365cb0), 0.335, 0.212351836381250) \
+    STEP(  68, UINT64_C(0x0000000000384825), 0.340, 0.219850854400000) \
+    STEP(  69, UINT64_C(0x00000000003a3aa8), 0.345, 0.227457578418750) \
+    STEP(  70, UINT64_C(0x00000000003c340f), 0.350, 0.235169375000000) \
+    STEP(  71, UINT64_C(0x00000000003e342b), 0.355, 0.242983540956250) \
+    STEP(  72, UINT64_C(0x0000000000403ace), 0.360, 0.250897305600000) \
+    STEP(  73, UINT64_C(0x00000000004247c8), 0.365, 0.258907832993750) \
+    STEP(  74, UINT64_C(0x0000000000445ae9), 0.370, 0.267012224200000) \
+    STEP(  75, UINT64_C(0x0000000000467400), 0.375, 0.275207519531250) \
+    STEP(  76, UINT64_C(0x00000000004892d8), 0.380, 0.283490700800000) \
+    STEP(  77, UINT64_C(0x00000000004ab740), 0.385, 0.291858693568750) \
+    STEP(  78, UINT64_C(0x00000000004ce102), 0.390, 0.300308369400000) \
+    STEP(  79, UINT64_C(0x00000000004f0fe9), 0.395, 0.308836548106250) \
+    STEP(  80, UINT64_C(0x00000000005143bf), 0.400, 0.317440000000000) \
+    STEP(  81, UINT64_C(0x0000000000537c4d), 0.405, 0.326115448143750) \
+    STEP(  82, UINT64_C(0x000000000055b95b), 0.410, 0.334859570600000) \
+    STEP(  83, UINT64_C(0x000000000057fab1), 0.415, 0.343669002681250) \
+    STEP(  84, UINT64_C(0x00000000005a4015), 0.420, 0.352540339200000) \
+    STEP(  85, UINT64_C(0x00000000005c894e), 0.425, 0.361470136718750) \
+    STEP(  86, UINT64_C(0x00000000005ed622), 0.430, 0.370454915800000) \
+    STEP(  87, UINT64_C(0x0000000000612655), 0.435, 0.379491163256250) \
+    STEP(  88, UINT64_C(0x00000000006379ac), 0.440, 0.388575334400000) \
+    STEP(  89, UINT64_C(0x000000000065cfeb), 0.445, 0.397703855293750) \
+    STEP(  90, UINT64_C(0x00000000006828d6), 0.450, 0.406873125000000) \
+    STEP(  91, UINT64_C(0x00000000006a842f), 0.455, 0.416079517831250) \
+    STEP(  92, UINT64_C(0x00000000006ce1bb), 0.460, 0.425319385600000) \
+    STEP(  93, UINT64_C(0x00000000006f413a), 0.465, 0.434589059868750) \
+    STEP(  94, UINT64_C(0x000000000071a270), 0.470, 0.443884854200000) \
+    STEP(  95, UINT64_C(0x000000000074051d), 0.475, 0.453203066406250) \
+    STEP(  96, UINT64_C(0x0000000000766905), 0.480, 0.462539980800000) \
+    STEP(  97, UINT64_C(0x000000000078cde7), 0.485, 0.471891870443750) \
+    STEP(  98, UINT64_C(0x00000000007b3387), 0.490, 0.481254999400000) \
+    STEP(  99, UINT64_C(0x00000000007d99a4), 0.495, 0.490625624981250) \
+    STEP( 100, UINT64_C(0x0000000000800000), 0.500, 0.500000000000000) \
+    STEP( 101, UINT64_C(0x000000000082665b), 0.505, 0.509374375018750) \
+    STEP( 102, UINT64_C(0x000000000084cc78), 0.510, 0.518745000600000) \
+    STEP( 103, UINT64_C(0x0000000000873218), 0.515, 0.528108129556250) \
+    STEP( 104, UINT64_C(0x00000000008996fa), 0.520, 0.537460019200000) \
+    STEP( 105, UINT64_C(0x00000000008bfae2), 0.525, 0.546796933593750) \
+    STEP( 106, UINT64_C(0x00000000008e5d8f), 0.530, 0.556115145800000) \
+    STEP( 107, UINT64_C(0x000000000090bec5), 0.535, 0.565410940131250) \
+    STEP( 108, UINT64_C(0x0000000000931e44), 0.540, 0.574680614400000) \
+    STEP( 109, UINT64_C(0x0000000000957bd0), 0.545, 0.583920482168750) \
+    STEP( 110, UINT64_C(0x000000000097d729), 0.550, 0.593126875000000) \
+    STEP( 111, UINT64_C(0x00000000009a3014), 0.555, 0.602296144706250) \
+    STEP( 112, UINT64_C(0x00000000009c8653), 0.560, 0.611424665600000) \
+    STEP( 113, UINT64_C(0x00000000009ed9aa), 0.565, 0.620508836743750) \
+    STEP( 114, UINT64_C(0x0000000000a129dd), 0.570, 0.629545084200000) \
+    STEP( 115, UINT64_C(0x0000000000a376b1), 0.575, 0.638529863281250) \
+    STEP( 116, UINT64_C(0x0000000000a5bfea), 0.580, 0.647459660800000) \
+    STEP( 117, UINT64_C(0x0000000000a8054e), 0.585, 0.656330997318750) \
+    STEP( 118, UINT64_C(0x0000000000aa46a4), 0.590, 0.665140429400000) \
+    STEP( 119, UINT64_C(0x0000000000ac83b2), 0.595, 0.673884551856250) \
+    STEP( 120, UINT64_C(0x0000000000aebc40), 0.600, 0.682560000000000) \
+    STEP( 121, UINT64_C(0x0000000000b0f016), 0.605, 0.691163451893750) \
+    STEP( 122, UINT64_C(0x0000000000b31efd), 0.610, 0.699691630600000) \
+    STEP( 123, UINT64_C(0x0000000000b548bf), 0.615, 0.708141306431250) \
+    STEP( 124, UINT64_C(0x0000000000b76d27), 0.620, 0.716509299200000) \
+    STEP( 125, UINT64_C(0x0000000000b98c00), 0.625, 0.724792480468750) \
+    STEP( 126, UINT64_C(0x0000000000bba516), 0.630, 0.732987775800000) \
+    STEP( 127, UINT64_C(0x0000000000bdb837), 0.635, 0.741092167006250) \
+    STEP( 128, UINT64_C(0x0000000000bfc531), 0.640, 0.749102694400000) \
+    STEP( 129, UINT64_C(0x0000000000c1cbd4), 0.645, 0.757016459043750) \
+    STEP( 130, UINT64_C(0x0000000000c3cbf0), 0.650, 0.764830625000000) \
+    STEP( 131, UINT64_C(0x0000000000c5c557), 0.655, 0.772542421581250) \
+    STEP( 132, UINT64_C(0x0000000000c7b7da), 0.660, 0.780149145600000) \
+    STEP( 133, UINT64_C(0x0000000000c9a34f), 0.665, 0.787648163618750) \
+    STEP( 134, UINT64_C(0x0000000000cb878a), 0.670, 0.795036914200000) \
+    STEP( 135, UINT64_C(0x0000000000cd6460), 0.675, 0.802312910156250) \
+    STEP( 136, UINT64_C(0x0000000000cf39ab), 0.680, 0.809473740800000) \
+    STEP( 137, UINT64_C(0x0000000000d10743), 0.685, 0.816517074193750) \
+    STEP( 138, UINT64_C(0x0000000000d2cd01), 0.690, 0.823440659400000) \
+    STEP( 139, UINT64_C(0x0000000000d48ac2), 0.695, 0.830242328731250) \
+    STEP( 140, UINT64_C(0x0000000000d64063), 0.700, 0.836920000000000) \
+    STEP( 141, UINT64_C(0x0000000000d7edc2), 0.705, 0.843471678768750) \
+    STEP( 142, UINT64_C(0x0000000000d992bf), 0.710, 0.849895460600000) \
+    STEP( 143, UINT64_C(0x0000000000db2f3c), 0.715, 0.856189533306250) \
+    STEP( 144, UINT64_C(0x0000000000dcc31c), 0.720, 0.862352179200000) \
+    STEP( 145, UINT64_C(0x0000000000de4e44), 0.725, 0.868381777343750) \
+    STEP( 146, UINT64_C(0x0000000000dfd09a), 0.730, 0.874276805800000) \
+    STEP( 147, UINT64_C(0x0000000000e14a07), 0.735, 0.880035843881250) \
+    STEP( 148, UINT64_C(0x0000000000e2ba74), 0.740, 0.885657574400000) \
+    STEP( 149, UINT64_C(0x0000000000e421cd), 0.745, 0.891140785918750) \
+    STEP( 150, UINT64_C(0x0000000000e58000), 0.750, 0.896484375000000) \
+    STEP( 151, UINT64_C(0x0000000000e6d4fb), 0.755, 0.901687348456250) \
+    STEP( 152, UINT64_C(0x0000000000e820b0), 0.760, 0.906748825600000) \
+    STEP( 153, UINT64_C(0x0000000000e96313), 0.765, 0.911668040493750) \
+    STEP( 154, UINT64_C(0x0000000000ea9c18), 0.770, 0.916444344200000) \
+    STEP( 155, UINT64_C(0x0000000000ebcbb7), 0.775, 0.921077207031250) \
+    STEP( 156, UINT64_C(0x0000000000ecf1e8), 0.780, 0.925566220800000) \
+    STEP( 157, UINT64_C(0x0000000000ee0ea7), 0.785, 0.929911101068750) \
+    STEP( 158, UINT64_C(0x0000000000ef21f1), 0.790, 0.934111689400000) \
+    STEP( 159, UINT64_C(0x0000000000f02bc6), 0.795, 0.938167955606250) \
+    STEP( 160, UINT64_C(0x0000000000f12c27), 0.800, 0.942080000000000) \
+    STEP( 161, UINT64_C(0x0000000000f22319), 0.805, 0.945848055643750) \
+    STEP( 162, UINT64_C(0x0000000000f310a1), 0.810, 0.949472490600000) \
+    STEP( 163, UINT64_C(0x0000000000f3f4c7), 0.815, 0.952953810181250) \
+    STEP( 164, UINT64_C(0x0000000000f4cf98), 0.820, 0.956292659200000) \
+    STEP( 165, UINT64_C(0x0000000000f5a120), 0.825, 0.959489824218750) \
+    STEP( 166, UINT64_C(0x0000000000f6696e), 0.830, 0.962546235800000) \
+    STEP( 167, UINT64_C(0x0000000000f72894), 0.835, 0.965462970756250) \
+    STEP( 168, UINT64_C(0x0000000000f7dea8), 0.840, 0.968241254400000) \
+    STEP( 169, UINT64_C(0x0000000000f88bc0), 0.845, 0.970882462793750) \
+    STEP( 170, UINT64_C(0x0000000000f92ff6), 0.850, 0.973388125000000) \
+    STEP( 171, UINT64_C(0x0000000000f9cb67), 0.855, 0.975759925331250) \
+    STEP( 172, UINT64_C(0x0000000000fa5e30), 0.860, 0.977999705600000) \
+    STEP( 173, UINT64_C(0x0000000000fae874), 0.865, 0.980109467368750) \
+    STEP( 174, UINT64_C(0x0000000000fb6a57), 0.870, 0.982091374200000) \
+    STEP( 175, UINT64_C(0x0000000000fbe400), 0.875, 0.983947753906250) \
+    STEP( 176, UINT64_C(0x0000000000fc5598), 0.880, 0.985681100800000) \
+    STEP( 177, UINT64_C(0x0000000000fcbf4e), 0.885, 0.987294077943750) \
+    STEP( 178, UINT64_C(0x0000000000fd214f), 0.890, 0.988789519400000) \
+    STEP( 179, UINT64_C(0x0000000000fd7bcf), 0.895, 0.990170432481250) \
+    STEP( 180, UINT64_C(0x0000000000fdcf03), 0.900, 0.991440000000000) \
+    STEP( 181, UINT64_C(0x0000000000fe1b23), 0.905, 0.992601582518750) \
+    STEP( 182, UINT64_C(0x0000000000fe606a), 0.910, 0.993658720600000) \
+    STEP( 183, UINT64_C(0x0000000000fe9f18), 0.915, 0.994615137056250) \
+    STEP( 184, UINT64_C(0x0000000000fed76e), 0.920, 0.995474739200000) \
+    STEP( 185, UINT64_C(0x0000000000ff09b0), 0.925, 0.996241621093750) \
+    STEP( 186, UINT64_C(0x0000000000ff3627), 0.930, 0.996920065800000) \
+    STEP( 187, UINT64_C(0x0000000000ff5d1d), 0.935, 0.997514547631250) \
+    STEP( 188, UINT64_C(0x0000000000ff7ee0), 0.940, 0.998029734400000) \
+    STEP( 189, UINT64_C(0x0000000000ff9bc3), 0.945, 0.998470489668750) \
+    STEP( 190, UINT64_C(0x0000000000ffb419), 0.950, 0.998841875000000) \
+    STEP( 191, UINT64_C(0x0000000000ffc83d), 0.955, 0.999149152206250) \
+    STEP( 192, UINT64_C(0x0000000000ffd888), 0.960, 0.999397785600000) \
+    STEP( 193, UINT64_C(0x0000000000ffe55b), 0.965, 0.999593444243750) \
+    STEP( 194, UINT64_C(0x0000000000ffef17), 0.970, 0.999742004200000) \
+    STEP( 195, UINT64_C(0x0000000000fff623), 0.975, 0.999849550781250) \
+    STEP( 196, UINT64_C(0x0000000000fffae9), 0.980, 0.999922380800000) \
+    STEP( 197, UINT64_C(0x0000000000fffdd6), 0.985, 0.999967004818750) \
+    STEP( 198, UINT64_C(0x0000000000ffff5a), 0.990, 0.999990149400000) \
+    STEP( 199, UINT64_C(0x0000000000ffffeb), 0.995, 0.999998759356250) \
+    STEP( 200, UINT64_C(0x0000000001000000), 1.000, 1.000000000000000) \
+
+#endif /* JEMALLOC_H_TYPES */
+/******************************************************************************/
+#ifdef JEMALLOC_H_STRUCTS
+
+
+#endif /* JEMALLOC_H_STRUCTS */
+/******************************************************************************/
+#ifdef JEMALLOC_H_EXTERNS
+
+
+#endif /* JEMALLOC_H_EXTERNS */
+/******************************************************************************/
+#ifdef JEMALLOC_H_INLINES
+
+
+#endif /* JEMALLOC_H_INLINES */
+/******************************************************************************/
diff --git a/deps/jemalloc/include/jemalloc/internal/smoothstep.sh b/deps/jemalloc/include/jemalloc/internal/smoothstep.sh
new file mode 100755
index 000000000..8124693f7
--- /dev/null
+++ b/deps/jemalloc/include/jemalloc/internal/smoothstep.sh
@@ -0,0 +1,115 @@
+#!/bin/sh
+#
+# Generate a discrete lookup table for a sigmoid function in the smoothstep
+# family (https://en.wikipedia.org/wiki/Smoothstep), where the lookup table
+# entries correspond to x in [1/nsteps, 2/nsteps, ..., nsteps/nsteps].  Encode
+# the entries using a binary fixed point representation.
+#
+# Usage: smoothstep.sh <variant> <nsteps> <bfp> <xprec> <yprec>
+#
+#        <variant> is in {smooth, smoother, smoothest}.
+#        <nsteps> must be greater than zero.
+#        <bfp> must be in [0..62]; reasonable values are roughly [10..30].
+#        <xprec> is x decimal precision.
+#        <yprec> is y decimal precision.
+
+#set -x
+
+cmd="sh smoothstep.sh $*"
+variant=$1
+nsteps=$2
+bfp=$3
+xprec=$4
+yprec=$5
+
+case "${variant}" in
+  smooth)
+    ;;
+  smoother)
+    ;;
+  smoothest)
+    ;;
+  *)
+    echo "Unsupported variant"
+    exit 1
+    ;;
+esac
+
+smooth() {
+  step=$1
+  y=`echo ${yprec} k ${step} ${nsteps} / sx _2 lx 3 ^ '*' 3 lx 2 ^ '*' + p | dc | tr -d '\\\\\n' | sed -e 's#^\.#0.#g'`
+  h=`echo ${yprec} k 2 ${bfp} ^ ${y} '*' p | dc | tr -d '\\\\\n' | sed -e 's#^\.#0.#g' | tr '.' ' ' | awk '{print $1}' `
+}
+
+smoother() {
+  step=$1
+  y=`echo ${yprec} k ${step} ${nsteps} / sx 6 lx 5 ^ '*' _15 lx 4 ^ '*' + 10 lx 3 ^ '*' + p | dc | tr -d '\\\\\n' | sed -e 's#^\.#0.#g'`
+  h=`echo ${yprec} k 2 ${bfp} ^ ${y} '*' p | dc | tr -d '\\\\\n' | sed -e 's#^\.#0.#g' | tr '.' ' ' | awk '{print $1}' `
+}
+
+smoothest() {
+  step=$1
+  y=`echo ${yprec} k ${step} ${nsteps} / sx _20 lx 7 ^ '*' 70 lx 6 ^ '*' + _84 lx 5 ^ '*' + 35 lx 4 ^ '*' + p | dc | tr -d '\\\\\n' | sed -e 's#^\.#0.#g'`
+  h=`echo ${yprec} k 2 ${bfp} ^ ${y} '*' p | dc | tr -d '\\\\\n' | sed -e 's#^\.#0.#g' | tr '.' ' ' | awk '{print $1}' `
+}
+
+cat <<EOF
+/*
+ * This file was generated by the following command:
+ *   $cmd
+ */
+/******************************************************************************/
+#ifdef JEMALLOC_H_TYPES
+
+/*
+ * This header defines a precomputed table based on the smoothstep family of
+ * sigmoidal curves (https://en.wikipedia.org/wiki/Smoothstep) that grow from 0
+ * to 1 in 0 <= x <= 1.  The table is stored as integer fixed point values so
+ * that floating point math can be avoided.
+ *
+ *                      3     2
+ *   smoothstep(x) = -2x  + 3x
+ *
+ *                       5      4      3
+ *   smootherstep(x) = 6x  - 15x  + 10x
+ *
+ *                          7      6      5      4
+ *   smootheststep(x) = -20x  + 70x  - 84x  + 35x
+ */
+
+#define	SMOOTHSTEP_VARIANT	"${variant}"
+#define	SMOOTHSTEP_NSTEPS	${nsteps}
+#define	SMOOTHSTEP_BFP		${bfp}
+#define	SMOOTHSTEP \\
+ /* STEP(step, h,                            x,     y) */ \\
+EOF
+
+s=1
+while [ $s -le $nsteps ] ; do
+  $variant ${s}
+  x=`echo ${xprec} k ${s} ${nsteps} / p | dc | tr -d '\\\\\n' | sed -e 's#^\.#0.#g'`
+  printf '    STEP(%4d, UINT64_C(0x%016x), %s, %s) \\\n' ${s} ${h} ${x} ${y}
+
+  s=$((s+1))
+done
+echo
+
+cat <<EOF
+#endif /* JEMALLOC_H_TYPES */
+/******************************************************************************/
+#ifdef JEMALLOC_H_STRUCTS
+
+
+#endif /* JEMALLOC_H_STRUCTS */
+/******************************************************************************/
+#ifdef JEMALLOC_H_EXTERNS
+
+
+#endif /* JEMALLOC_H_EXTERNS */
+/******************************************************************************/
+#ifdef JEMALLOC_H_INLINES
+
+
+#endif /* JEMALLOC_H_INLINES */
+/******************************************************************************/
+EOF
diff --git a/deps/jemalloc/include/jemalloc/internal/spin.h b/deps/jemalloc/include/jemalloc/internal/spin.h
new file mode 100644
index 000000000..9ef5ceb92
--- /dev/null
+++ b/deps/jemalloc/include/jemalloc/internal/spin.h
@@ -0,0 +1,51 @@
+/******************************************************************************/
+#ifdef JEMALLOC_H_TYPES
+
+typedef struct spin_s spin_t;
+
+#endif /* JEMALLOC_H_TYPES */
+/******************************************************************************/
+#ifdef JEMALLOC_H_STRUCTS
+
+struct spin_s {
+	unsigned iteration;
+};
+
+#endif /* JEMALLOC_H_STRUCTS */
+/******************************************************************************/
+#ifdef JEMALLOC_H_EXTERNS
+
+#endif /* JEMALLOC_H_EXTERNS */
+/******************************************************************************/
+#ifdef JEMALLOC_H_INLINES
+
+#ifndef JEMALLOC_ENABLE_INLINE
+void	spin_init(spin_t *spin);
+void	spin_adaptive(spin_t *spin);
+#endif
+
+#if (defined(JEMALLOC_ENABLE_INLINE) || defined(JEMALLOC_SPIN_C_))
+JEMALLOC_INLINE void
+spin_init(spin_t *spin)
+{
+
+	spin->iteration = 0;
+}
+
+JEMALLOC_INLINE void
+spin_adaptive(spin_t *spin)
+{
+	volatile uint64_t i;
+
+	for (i = 0; i < (KQU(1) << spin->iteration); i++)
+		CPU_SPINWAIT;
+
+	if (spin->iteration < 63)
+		spin->iteration++;
+}
+
+#endif
+
+#endif /* JEMALLOC_H_INLINES */
+/******************************************************************************/
+
diff --git a/deps/jemalloc/include/jemalloc/internal/stats.h b/deps/jemalloc/include/jemalloc/internal/stats.h
index c91dba99d..04e7dae14 100644
--- a/deps/jemalloc/include/jemalloc/internal/stats.h
+++ b/deps/jemalloc/include/jemalloc/internal/stats.h
@@ -102,6 +102,14 @@ struct arena_stats_s {
 	/* Number of bytes currently mapped. */
 	size_t		mapped;
 
+	/*
+	 * Number of bytes currently retained as a side effect of munmap() being
+	 * disabled/bypassed.  Retained bytes are technically mapped (though
+	 * always decommitted or purged), but they are excluded from the mapped
+	 * statistic (above).
+	 */
+	size_t		retained;
+
 	/*
 	 * Total number of purge sweeps, total number of madvise calls made,
 	 * and total pages purged in order to keep dirty unused memory under
@@ -168,6 +176,9 @@ JEMALLOC_INLINE void
 stats_cactive_add(size_t size)
 {
 
+	assert(size > 0);
+	assert((size & chunksize_mask) == 0);
+
 	atomic_add_z(&stats_cactive, size);
 }
 
@@ -175,6 +186,9 @@ JEMALLOC_INLINE void
 stats_cactive_sub(size_t size)
 {
 
+	assert(size > 0);
+	assert((size & chunksize_mask) == 0);
+
 	atomic_sub_z(&stats_cactive, size);
 }
 #endif
diff --git a/deps/jemalloc/include/jemalloc/internal/tcache.h b/deps/jemalloc/include/jemalloc/internal/tcache.h
index 5079cd266..01ba062de 100644
--- a/deps/jemalloc/include/jemalloc/internal/tcache.h
+++ b/deps/jemalloc/include/jemalloc/internal/tcache.h
@@ -70,13 +70,20 @@ struct tcache_bin_s {
 	int		low_water;	/* Min # cached since last GC. */
 	unsigned	lg_fill_div;	/* Fill (ncached_max >> lg_fill_div). */
 	unsigned	ncached;	/* # of cached objects. */
+	/*
+	 * To make use of adjacent cacheline prefetch, the items in the avail
+	 * stack goes to higher address for newer allocations.  avail points
+	 * just above the available space, which means that
+	 * avail[-ncached, ... -1] are available items and the lowest item will
+	 * be allocated first.
+	 */
 	void		**avail;	/* Stack of available objects. */
 };
 
 struct tcache_s {
 	ql_elm(tcache_t) link;		/* Used for aggregating stats. */
 	uint64_t	prof_accumbytes;/* Cleared after arena_prof_accum(). */
-	unsigned	ev_cnt;		/* Event count since incremental GC. */
+	ticker_t	gc_ticker;	/* Drives incremental GC. */
 	szind_t		next_gc_bin;	/* Next bin to GC. */
 	tcache_bin_t	tbins[1];	/* Dynamically sized. */
 	/*
@@ -108,7 +115,7 @@ extern tcache_bin_info_t	*tcache_bin_info;
  * Number of tcache bins.  There are NBINS small-object bins, plus 0 or more
  * large-object bins.
  */
-extern size_t	nhbins;
+extern unsigned	nhbins;
 
 /* Maximum cached size class. */
 extern size_t	tcache_maxclass;
@@ -123,27 +130,25 @@ extern size_t	tcache_maxclass;
  */
 extern tcaches_t	*tcaches;
 
-size_t	tcache_salloc(const void *ptr);
+size_t	tcache_salloc(tsdn_t *tsdn, const void *ptr);
 void	tcache_event_hard(tsd_t *tsd, tcache_t *tcache);
-void	*tcache_alloc_small_hard(tsd_t *tsd, arena_t *arena, tcache_t *tcache,
-    tcache_bin_t *tbin, szind_t binind);
+void	*tcache_alloc_small_hard(tsdn_t *tsdn, arena_t *arena, tcache_t *tcache,
+    tcache_bin_t *tbin, szind_t binind, bool *tcache_success);
 void	tcache_bin_flush_small(tsd_t *tsd, tcache_t *tcache, tcache_bin_t *tbin,
     szind_t binind, unsigned rem);
 void	tcache_bin_flush_large(tsd_t *tsd, tcache_bin_t *tbin, szind_t binind,
     unsigned rem, tcache_t *tcache);
-void	tcache_arena_associate(tcache_t *tcache, arena_t *arena);
-void	tcache_arena_reassociate(tcache_t *tcache, arena_t *oldarena,
-    arena_t *newarena);
-void	tcache_arena_dissociate(tcache_t *tcache, arena_t *arena);
+void	tcache_arena_reassociate(tsdn_t *tsdn, tcache_t *tcache,
+    arena_t *oldarena, arena_t *newarena);
 tcache_t *tcache_get_hard(tsd_t *tsd);
-tcache_t *tcache_create(tsd_t *tsd, arena_t *arena);
+tcache_t *tcache_create(tsdn_t *tsdn, arena_t *arena);
 void	tcache_cleanup(tsd_t *tsd);
 void	tcache_enabled_cleanup(tsd_t *tsd);
-void	tcache_stats_merge(tcache_t *tcache, arena_t *arena);
+void	tcache_stats_merge(tsdn_t *tsdn, tcache_t *tcache, arena_t *arena);
 bool	tcaches_create(tsd_t *tsd, unsigned *r_ind);
 void	tcaches_flush(tsd_t *tsd, unsigned ind);
 void	tcaches_destroy(tsd_t *tsd, unsigned ind);
-bool	tcache_boot(void);
+bool	tcache_boot(tsdn_t *tsdn);
 
 #endif /* JEMALLOC_H_EXTERNS */
 /******************************************************************************/
@@ -155,15 +160,15 @@ void	tcache_flush(void);
 bool	tcache_enabled_get(void);
 tcache_t *tcache_get(tsd_t *tsd, bool create);
 void	tcache_enabled_set(bool enabled);
-void	*tcache_alloc_easy(tcache_bin_t *tbin);
+void	*tcache_alloc_easy(tcache_bin_t *tbin, bool *tcache_success);
 void	*tcache_alloc_small(tsd_t *tsd, arena_t *arena, tcache_t *tcache,
-    size_t size, bool zero);
+    size_t size, szind_t ind, bool zero, bool slow_path);
 void	*tcache_alloc_large(tsd_t *tsd, arena_t *arena, tcache_t *tcache,
-    size_t size, bool zero);
+    size_t size, szind_t ind, bool zero, bool slow_path);
 void	tcache_dalloc_small(tsd_t *tsd, tcache_t *tcache, void *ptr,
-    szind_t binind);
+    szind_t binind, bool slow_path);
 void	tcache_dalloc_large(tsd_t *tsd, tcache_t *tcache, void *ptr,
-    size_t size);
+    size_t size, bool slow_path);
 tcache_t	*tcaches_get(tsd_t *tsd, unsigned ind);
 #endif
 
@@ -240,51 +245,74 @@ tcache_event(tsd_t *tsd, tcache_t *tcache)
 	if (TCACHE_GC_INCR == 0)
 		return;
 
-	tcache->ev_cnt++;
-	assert(tcache->ev_cnt <= TCACHE_GC_INCR);
-	if (unlikely(tcache->ev_cnt == TCACHE_GC_INCR))
+	if (unlikely(ticker_tick(&tcache->gc_ticker)))
 		tcache_event_hard(tsd, tcache);
 }
 
 JEMALLOC_ALWAYS_INLINE void *
-tcache_alloc_easy(tcache_bin_t *tbin)
+tcache_alloc_easy(tcache_bin_t *tbin, bool *tcache_success)
 {
 	void *ret;
 
 	if (unlikely(tbin->ncached == 0)) {
 		tbin->low_water = -1;
+		*tcache_success = false;
 		return (NULL);
 	}
+	/*
+	 * tcache_success (instead of ret) should be checked upon the return of
+	 * this function.  We avoid checking (ret == NULL) because there is
+	 * never a null stored on the avail stack (which is unknown to the
+	 * compiler), and eagerly checking ret would cause pipeline stall
+	 * (waiting for the cacheline).
+	 */
+	*tcache_success = true;
+	ret = *(tbin->avail - tbin->ncached);
 	tbin->ncached--;
+
 	if (unlikely((int)tbin->ncached < tbin->low_water))
 		tbin->low_water = tbin->ncached;
-	ret = tbin->avail[tbin->ncached];
+
 	return (ret);
 }
 
 JEMALLOC_ALWAYS_INLINE void *
 tcache_alloc_small(tsd_t *tsd, arena_t *arena, tcache_t *tcache, size_t size,
-    bool zero)
+    szind_t binind, bool zero, bool slow_path)
 {
 	void *ret;
-	szind_t binind;
-	size_t usize;
 	tcache_bin_t *tbin;
+	bool tcache_success;
+	size_t usize JEMALLOC_CC_SILENCE_INIT(0);
 
-	binind = size2index(size);
 	assert(binind < NBINS);
 	tbin = &tcache->tbins[binind];
-	usize = index2size(binind);
-	ret = tcache_alloc_easy(tbin);
-	if (unlikely(ret == NULL)) {
-		ret = tcache_alloc_small_hard(tsd, arena, tcache, tbin, binind);
-		if (ret == NULL)
+	ret = tcache_alloc_easy(tbin, &tcache_success);
+	assert(tcache_success == (ret != NULL));
+	if (unlikely(!tcache_success)) {
+		bool tcache_hard_success;
+		arena = arena_choose(tsd, arena);
+		if (unlikely(arena == NULL))
+			return (NULL);
+
+		ret = tcache_alloc_small_hard(tsd_tsdn(tsd), arena, tcache,
+		    tbin, binind, &tcache_hard_success);
+		if (tcache_hard_success == false)
 			return (NULL);
 	}
-	assert(tcache_salloc(ret) == usize);
+
+	assert(ret);
+	/*
+	 * Only compute usize if required.  The checks in the following if
+	 * statement are all static.
+	 */
+	if (config_prof || (slow_path && config_fill) || unlikely(zero)) {
+		usize = index2size(binind);
+		assert(tcache_salloc(tsd_tsdn(tsd), ret) == usize);
+	}
 
 	if (likely(!zero)) {
-		if (config_fill) {
+		if (slow_path && config_fill) {
 			if (unlikely(opt_junk_alloc)) {
 				arena_alloc_junk_small(ret,
 				    &arena_bin_info[binind], false);
@@ -292,7 +320,7 @@ tcache_alloc_small(tsd_t *tsd, arena_t *arena, tcache_t *tcache, size_t size,
 				memset(ret, 0, usize);
 		}
 	} else {
-		if (config_fill && unlikely(opt_junk_alloc)) {
+		if (slow_path && config_fill && unlikely(opt_junk_alloc)) {
 			arena_alloc_junk_small(ret, &arena_bin_info[binind],
 			    true);
 		}
@@ -309,28 +337,38 @@ tcache_alloc_small(tsd_t *tsd, arena_t *arena, tcache_t *tcache, size_t size,
 
 JEMALLOC_ALWAYS_INLINE void *
 tcache_alloc_large(tsd_t *tsd, arena_t *arena, tcache_t *tcache, size_t size,
-    bool zero)
+    szind_t binind, bool zero, bool slow_path)
 {
 	void *ret;
-	szind_t binind;
-	size_t usize;
 	tcache_bin_t *tbin;
+	bool tcache_success;
 
-	binind = size2index(size);
-	usize = index2size(binind);
-	assert(usize <= tcache_maxclass);
 	assert(binind < nhbins);
 	tbin = &tcache->tbins[binind];
-	ret = tcache_alloc_easy(tbin);
-	if (unlikely(ret == NULL)) {
+	ret = tcache_alloc_easy(tbin, &tcache_success);
+	assert(tcache_success == (ret != NULL));
+	if (unlikely(!tcache_success)) {
 		/*
 		 * Only allocate one large object at a time, because it's quite
 		 * expensive to create one and not use it.
 		 */
-		ret = arena_malloc_large(arena, usize, zero);
+		arena = arena_choose(tsd, arena);
+		if (unlikely(arena == NULL))
+			return (NULL);
+
+		ret = arena_malloc_large(tsd_tsdn(tsd), arena, binind, zero);
 		if (ret == NULL)
 			return (NULL);
 	} else {
+		size_t usize JEMALLOC_CC_SILENCE_INIT(0);
+
+		/* Only compute usize on demand */
+		if (config_prof || (slow_path && config_fill) ||
+		    unlikely(zero)) {
+			usize = index2size(binind);
+			assert(usize <= tcache_maxclass);
+		}
+
 		if (config_prof && usize == LARGE_MINCLASS) {
 			arena_chunk_t *chunk =
 			    (arena_chunk_t *)CHUNK_ADDR2BASE(ret);
@@ -340,10 +378,11 @@ tcache_alloc_large(tsd_t *tsd, arena_t *arena, tcache_t *tcache, size_t size,
 			    BININD_INVALID);
 		}
 		if (likely(!zero)) {
-			if (config_fill) {
-				if (unlikely(opt_junk_alloc))
-					memset(ret, 0xa5, usize);
-				else if (unlikely(opt_zero))
+			if (slow_path && config_fill) {
+				if (unlikely(opt_junk_alloc)) {
+					memset(ret, JEMALLOC_ALLOC_JUNK,
+					    usize);
+				} else if (unlikely(opt_zero))
 					memset(ret, 0, usize);
 			}
 		} else
@@ -360,14 +399,15 @@ tcache_alloc_large(tsd_t *tsd, arena_t *arena, tcache_t *tcache, size_t size,
 }
 
 JEMALLOC_ALWAYS_INLINE void
-tcache_dalloc_small(tsd_t *tsd, tcache_t *tcache, void *ptr, szind_t binind)
+tcache_dalloc_small(tsd_t *tsd, tcache_t *tcache, void *ptr, szind_t binind,
+    bool slow_path)
 {
 	tcache_bin_t *tbin;
 	tcache_bin_info_t *tbin_info;
 
-	assert(tcache_salloc(ptr) <= SMALL_MAXCLASS);
+	assert(tcache_salloc(tsd_tsdn(tsd), ptr) <= SMALL_MAXCLASS);
 
-	if (config_fill && unlikely(opt_junk_free))
+	if (slow_path && config_fill && unlikely(opt_junk_free))
 		arena_dalloc_junk_small(ptr, &arena_bin_info[binind]);
 
 	tbin = &tcache->tbins[binind];
@@ -377,26 +417,27 @@ tcache_dalloc_small(tsd_t *tsd, tcache_t *tcache, void *ptr, szind_t binind)
 		    (tbin_info->ncached_max >> 1));
 	}
 	assert(tbin->ncached < tbin_info->ncached_max);
-	tbin->avail[tbin->ncached] = ptr;
 	tbin->ncached++;
+	*(tbin->avail - tbin->ncached) = ptr;
 
 	tcache_event(tsd, tcache);
 }
 
 JEMALLOC_ALWAYS_INLINE void
-tcache_dalloc_large(tsd_t *tsd, tcache_t *tcache, void *ptr, size_t size)
+tcache_dalloc_large(tsd_t *tsd, tcache_t *tcache, void *ptr, size_t size,
+    bool slow_path)
 {
 	szind_t binind;
 	tcache_bin_t *tbin;
 	tcache_bin_info_t *tbin_info;
 
 	assert((size & PAGE_MASK) == 0);
-	assert(tcache_salloc(ptr) > SMALL_MAXCLASS);
-	assert(tcache_salloc(ptr) <= tcache_maxclass);
+	assert(tcache_salloc(tsd_tsdn(tsd), ptr) > SMALL_MAXCLASS);
+	assert(tcache_salloc(tsd_tsdn(tsd), ptr) <= tcache_maxclass);
 
 	binind = size2index(size);
 
-	if (config_fill && unlikely(opt_junk_free))
+	if (slow_path && config_fill && unlikely(opt_junk_free))
 		arena_dalloc_junk_large(ptr, size);
 
 	tbin = &tcache->tbins[binind];
@@ -406,8 +447,8 @@ tcache_dalloc_large(tsd_t *tsd, tcache_t *tcache, void *ptr, size_t size)
 		    (tbin_info->ncached_max >> 1), tcache);
 	}
 	assert(tbin->ncached < tbin_info->ncached_max);
-	tbin->avail[tbin->ncached] = ptr;
 	tbin->ncached++;
+	*(tbin->avail - tbin->ncached) = ptr;
 
 	tcache_event(tsd, tcache);
 }
@@ -416,8 +457,10 @@ JEMALLOC_ALWAYS_INLINE tcache_t *
 tcaches_get(tsd_t *tsd, unsigned ind)
 {
 	tcaches_t *elm = &tcaches[ind];
-	if (unlikely(elm->tcache == NULL))
-		elm->tcache = tcache_create(tsd, arena_choose(tsd, NULL));
+	if (unlikely(elm->tcache == NULL)) {
+		elm->tcache = tcache_create(tsd_tsdn(tsd), arena_choose(tsd,
+		    NULL));
+	}
 	return (elm->tcache);
 }
 #endif
diff --git a/deps/jemalloc/include/jemalloc/internal/ticker.h b/deps/jemalloc/include/jemalloc/internal/ticker.h
new file mode 100644
index 000000000..4696e56d2
--- /dev/null
+++ b/deps/jemalloc/include/jemalloc/internal/ticker.h
@@ -0,0 +1,75 @@
+/******************************************************************************/
+#ifdef JEMALLOC_H_TYPES
+
+typedef struct ticker_s ticker_t;
+
+#endif /* JEMALLOC_H_TYPES */
+/******************************************************************************/
+#ifdef JEMALLOC_H_STRUCTS
+
+struct ticker_s {
+	int32_t	tick;
+	int32_t	nticks;
+};
+
+#endif /* JEMALLOC_H_STRUCTS */
+/******************************************************************************/
+#ifdef JEMALLOC_H_EXTERNS
+
+#endif /* JEMALLOC_H_EXTERNS */
+/******************************************************************************/
+#ifdef JEMALLOC_H_INLINES
+
+#ifndef JEMALLOC_ENABLE_INLINE
+void	ticker_init(ticker_t *ticker, int32_t nticks);
+void	ticker_copy(ticker_t *ticker, const ticker_t *other);
+int32_t	ticker_read(const ticker_t *ticker);
+bool	ticker_ticks(ticker_t *ticker, int32_t nticks);
+bool	ticker_tick(ticker_t *ticker);
+#endif
+
+#if (defined(JEMALLOC_ENABLE_INLINE) || defined(JEMALLOC_TICKER_C_))
+JEMALLOC_INLINE void
+ticker_init(ticker_t *ticker, int32_t nticks)
+{
+
+	ticker->tick = nticks;
+	ticker->nticks = nticks;
+}
+
+JEMALLOC_INLINE void
+ticker_copy(ticker_t *ticker, const ticker_t *other)
+{
+
+	*ticker = *other;
+}
+
+JEMALLOC_INLINE int32_t
+ticker_read(const ticker_t *ticker)
+{
+
+	return (ticker->tick);
+}
+
+JEMALLOC_INLINE bool
+ticker_ticks(ticker_t *ticker, int32_t nticks)
+{
+
+	if (unlikely(ticker->tick < nticks)) {
+		ticker->tick = ticker->nticks;
+		return (true);
+	}
+	ticker->tick -= nticks;
+	return(false);
+}
+
+JEMALLOC_INLINE bool
+ticker_tick(ticker_t *ticker)
+{
+
+	return (ticker_ticks(ticker, 1));
+}
+#endif
+
+#endif /* JEMALLOC_H_INLINES */
+/******************************************************************************/
diff --git a/deps/jemalloc/include/jemalloc/internal/tsd.h b/deps/jemalloc/include/jemalloc/internal/tsd.h
index eed7aa013..9055acafd 100644
--- a/deps/jemalloc/include/jemalloc/internal/tsd.h
+++ b/deps/jemalloc/include/jemalloc/internal/tsd.h
@@ -13,6 +13,9 @@ typedef struct tsd_init_head_s tsd_init_head_t;
 #endif
 
 typedef struct tsd_s tsd_t;
+typedef struct tsdn_s tsdn_t;
+
+#define	TSDN_NULL	((tsdn_t *)0)
 
 typedef enum {
 	tsd_state_uninitialized,
@@ -44,7 +47,8 @@ typedef enum {
  * The result is a set of generated functions, e.g.:
  *
  *   bool example_tsd_boot(void) {...}
- *   example_t *example_tsd_get() {...}
+ *   bool example_tsd_booted_get(void) {...}
+ *   example_t *example_tsd_get(bool init) {...}
  *   void example_tsd_set(example_t *val) {...}
  *
  * Note that all of the functions deal in terms of (a_type *) rather than
@@ -98,8 +102,10 @@ a_attr void								\
 a_name##tsd_boot1(void);						\
 a_attr bool								\
 a_name##tsd_boot(void);							\
+a_attr bool								\
+a_name##tsd_booted_get(void);						\
 a_attr a_type *								\
-a_name##tsd_get(void);							\
+a_name##tsd_get(bool init);						\
 a_attr void								\
 a_name##tsd_set(a_type *val);
 
@@ -201,9 +207,21 @@ a_name##tsd_boot(void)							\
 									\
 	return (a_name##tsd_boot0());					\
 }									\
+a_attr bool								\
+a_name##tsd_booted_get(void)						\
+{									\
+									\
+	return (a_name##tsd_booted);					\
+}									\
+a_attr bool								\
+a_name##tsd_get_allocates(void)						\
+{									\
+									\
+	return (false);							\
+}									\
 /* Get/set. */								\
 a_attr a_type *								\
-a_name##tsd_get(void)							\
+a_name##tsd_get(bool init)						\
 {									\
 									\
 	assert(a_name##tsd_booted);					\
@@ -246,9 +264,21 @@ a_name##tsd_boot(void)							\
 									\
 	return (a_name##tsd_boot0());					\
 }									\
+a_attr bool								\
+a_name##tsd_booted_get(void)						\
+{									\
+									\
+	return (a_name##tsd_booted);					\
+}									\
+a_attr bool								\
+a_name##tsd_get_allocates(void)						\
+{									\
+									\
+	return (false);							\
+}									\
 /* Get/set. */								\
 a_attr a_type *								\
-a_name##tsd_get(void)							\
+a_name##tsd_get(bool init)						\
 {									\
 									\
 	assert(a_name##tsd_booted);					\
@@ -307,14 +337,14 @@ a_name##tsd_wrapper_set(a_name##tsd_wrapper_t *wrapper)			\
 	}								\
 }									\
 a_attr a_name##tsd_wrapper_t *						\
-a_name##tsd_wrapper_get(void)						\
+a_name##tsd_wrapper_get(bool init)					\
 {									\
 	DWORD error = GetLastError();					\
 	a_name##tsd_wrapper_t *wrapper = (a_name##tsd_wrapper_t *)	\
 	    TlsGetValue(a_name##tsd_tsd);				\
 	SetLastError(error);						\
 									\
-	if (unlikely(wrapper == NULL)) {				\
+	if (init && unlikely(wrapper == NULL)) {			\
 		wrapper = (a_name##tsd_wrapper_t *)			\
 		    malloc_tsd_malloc(sizeof(a_name##tsd_wrapper_t));	\
 		if (wrapper == NULL) {					\
@@ -368,14 +398,28 @@ a_name##tsd_boot(void)							\
 	a_name##tsd_boot1();						\
 	return (false);							\
 }									\
+a_attr bool								\
+a_name##tsd_booted_get(void)						\
+{									\
+									\
+	return (a_name##tsd_booted);					\
+}									\
+a_attr bool								\
+a_name##tsd_get_allocates(void)						\
+{									\
+									\
+	return (true);							\
+}									\
 /* Get/set. */								\
 a_attr a_type *								\
-a_name##tsd_get(void)							\
+a_name##tsd_get(bool init)						\
 {									\
 	a_name##tsd_wrapper_t *wrapper;					\
 									\
 	assert(a_name##tsd_booted);					\
-	wrapper = a_name##tsd_wrapper_get();				\
+	wrapper = a_name##tsd_wrapper_get(init);			\
+	if (a_name##tsd_get_allocates() && !init && wrapper == NULL)	\
+		return (NULL);						\
 	return (&wrapper->val);						\
 }									\
 a_attr void								\
@@ -384,7 +428,7 @@ a_name##tsd_set(a_type *val)						\
 	a_name##tsd_wrapper_t *wrapper;					\
 									\
 	assert(a_name##tsd_booted);					\
-	wrapper = a_name##tsd_wrapper_get();				\
+	wrapper = a_name##tsd_wrapper_get(true);			\
 	wrapper->val = *(val);						\
 	if (a_cleanup != malloc_tsd_no_cleanup)				\
 		wrapper->initialized = true;				\
@@ -428,12 +472,12 @@ a_name##tsd_wrapper_set(a_name##tsd_wrapper_t *wrapper)			\
 	}								\
 }									\
 a_attr a_name##tsd_wrapper_t *						\
-a_name##tsd_wrapper_get(void)						\
+a_name##tsd_wrapper_get(bool init)					\
 {									\
 	a_name##tsd_wrapper_t *wrapper = (a_name##tsd_wrapper_t *)	\
 	    pthread_getspecific(a_name##tsd_tsd);			\
 									\
-	if (unlikely(wrapper == NULL)) {				\
+	if (init && unlikely(wrapper == NULL)) {			\
 		tsd_init_block_t block;					\
 		wrapper = tsd_init_check_recursion(			\
 		    &a_name##tsd_init_head, &block);			\
@@ -490,14 +534,28 @@ a_name##tsd_boot(void)							\
 	a_name##tsd_boot1();						\
 	return (false);							\
 }									\
+a_attr bool								\
+a_name##tsd_booted_get(void)						\
+{									\
+									\
+	return (a_name##tsd_booted);					\
+}									\
+a_attr bool								\
+a_name##tsd_get_allocates(void)						\
+{									\
+									\
+	return (true);							\
+}									\
 /* Get/set. */								\
 a_attr a_type *								\
-a_name##tsd_get(void)							\
+a_name##tsd_get(bool init)						\
 {									\
 	a_name##tsd_wrapper_t *wrapper;					\
 									\
 	assert(a_name##tsd_booted);					\
-	wrapper = a_name##tsd_wrapper_get();				\
+	wrapper = a_name##tsd_wrapper_get(init);			\
+	if (a_name##tsd_get_allocates() && !init && wrapper == NULL)	\
+		return (NULL);						\
 	return (&wrapper->val);						\
 }									\
 a_attr void								\
@@ -506,7 +564,7 @@ a_name##tsd_set(a_type *val)						\
 	a_name##tsd_wrapper_t *wrapper;					\
 									\
 	assert(a_name##tsd_booted);					\
-	wrapper = a_name##tsd_wrapper_get();				\
+	wrapper = a_name##tsd_wrapper_get(true);			\
 	wrapper->val = *(val);						\
 	if (a_cleanup != malloc_tsd_no_cleanup)				\
 		wrapper->initialized = true;				\
@@ -536,12 +594,15 @@ struct tsd_init_head_s {
     O(thread_allocated,		uint64_t)				\
     O(thread_deallocated,	uint64_t)				\
     O(prof_tdata,		prof_tdata_t *)				\
+    O(iarena,			arena_t *)				\
     O(arena,			arena_t *)				\
-    O(arenas_cache,		arena_t **)				\
-    O(narenas_cache,		unsigned)				\
-    O(arenas_cache_bypass,	bool)					\
+    O(arenas_tdata,		arena_tdata_t *)			\
+    O(narenas_tdata,		unsigned)				\
+    O(arenas_tdata_bypass,	bool)					\
     O(tcache_enabled,		tcache_enabled_t)			\
     O(quarantine,		quarantine_t *)				\
+    O(witnesses,		witness_list_t)				\
+    O(witness_fork,		bool)					\
 
 #define	TSD_INITIALIZER {						\
     tsd_state_uninitialized,						\
@@ -551,10 +612,13 @@ struct tsd_init_head_s {
     NULL,								\
     NULL,								\
     NULL,								\
+    NULL,								\
     0,									\
     false,								\
     tcache_enabled_default,						\
-    NULL								\
+    NULL,								\
+    ql_head_initializer(witnesses),					\
+    false								\
 }
 
 struct tsd_s {
@@ -565,6 +629,15 @@ MALLOC_TSD
 #undef O
 };
 
+/*
+ * Wrapper around tsd_t that makes it possible to avoid implicit conversion
+ * between tsd_t and tsdn_t, where tsdn_t is "nullable" and has to be
+ * explicitly converted to tsd_t, which is non-nullable.
+ */
+struct tsdn_s {
+	tsd_t	tsd;
+};
+
 static const tsd_t tsd_initializer = TSD_INITIALIZER;
 
 malloc_tsd_types(, tsd_t)
@@ -577,7 +650,7 @@ void	*malloc_tsd_malloc(size_t size);
 void	malloc_tsd_dalloc(void *wrapper);
 void	malloc_tsd_no_cleanup(void *arg);
 void	malloc_tsd_cleanup_register(bool (*f)(void));
-bool	malloc_tsd_boot0(void);
+tsd_t	*malloc_tsd_boot0(void);
 void	malloc_tsd_boot1(void);
 #if (!defined(JEMALLOC_MALLOC_THREAD_CLEANUP) && !defined(JEMALLOC_TLS) && \
     !defined(_WIN32))
@@ -594,7 +667,9 @@ void	tsd_cleanup(void *arg);
 #ifndef JEMALLOC_ENABLE_INLINE
 malloc_tsd_protos(JEMALLOC_ATTR(unused), , tsd_t)
 
+tsd_t	*tsd_fetch_impl(bool init);
 tsd_t	*tsd_fetch(void);
+tsdn_t	*tsd_tsdn(tsd_t *tsd);
 bool	tsd_nominal(tsd_t *tsd);
 #define	O(n, t)								\
 t	*tsd_##n##p_get(tsd_t *tsd);					\
@@ -602,6 +677,9 @@ t	tsd_##n##_get(tsd_t *tsd);					\
 void	tsd_##n##_set(tsd_t *tsd, t n);
 MALLOC_TSD
 #undef O
+tsdn_t	*tsdn_fetch(void);
+bool	tsdn_null(const tsdn_t *tsdn);
+tsd_t	*tsdn_tsd(tsdn_t *tsdn);
 #endif
 
 #if (defined(JEMALLOC_ENABLE_INLINE) || defined(JEMALLOC_TSD_C_))
@@ -609,9 +687,13 @@ malloc_tsd_externs(, tsd_t)
 malloc_tsd_funcs(JEMALLOC_ALWAYS_INLINE, , tsd_t, tsd_initializer, tsd_cleanup)
 
 JEMALLOC_ALWAYS_INLINE tsd_t *
-tsd_fetch(void)
+tsd_fetch_impl(bool init)
 {
-	tsd_t *tsd = tsd_get();
+	tsd_t *tsd = tsd_get(init);
+
+	if (!init && tsd_get_allocates() && tsd == NULL)
+		return (NULL);
+	assert(tsd != NULL);
 
 	if (unlikely(tsd->state != tsd_state_nominal)) {
 		if (tsd->state == tsd_state_uninitialized) {
@@ -628,6 +710,20 @@ tsd_fetch(void)
 	return (tsd);
 }
 
+JEMALLOC_ALWAYS_INLINE tsd_t *
+tsd_fetch(void)
+{
+
+	return (tsd_fetch_impl(true));
+}
+
+JEMALLOC_ALWAYS_INLINE tsdn_t *
+tsd_tsdn(tsd_t *tsd)
+{
+
+	return ((tsdn_t *)tsd);
+}
+
 JEMALLOC_INLINE bool
 tsd_nominal(tsd_t *tsd)
 {
@@ -659,6 +755,32 @@ tsd_##n##_set(tsd_t *tsd, t n)						\
 }
 MALLOC_TSD
 #undef O
+
+JEMALLOC_ALWAYS_INLINE tsdn_t *
+tsdn_fetch(void)
+{
+
+	if (!tsd_booted_get())
+		return (NULL);
+
+	return (tsd_tsdn(tsd_fetch_impl(false)));
+}
+
+JEMALLOC_ALWAYS_INLINE bool
+tsdn_null(const tsdn_t *tsdn)
+{
+
+	return (tsdn == NULL);
+}
+
+JEMALLOC_ALWAYS_INLINE tsd_t *
+tsdn_tsd(tsdn_t *tsdn)
+{
+
+	assert(!tsdn_null(tsdn));
+
+	return (&tsdn->tsd);
+}
 #endif
 
 #endif /* JEMALLOC_H_INLINES */
diff --git a/deps/jemalloc/include/jemalloc/internal/util.h b/deps/jemalloc/include/jemalloc/internal/util.h
index b2ea740fd..4b56d652e 100644
--- a/deps/jemalloc/include/jemalloc/internal/util.h
+++ b/deps/jemalloc/include/jemalloc/internal/util.h
@@ -40,6 +40,14 @@
  */
 #define	MALLOC_PRINTF_BUFSIZE	4096
 
+/* Junk fill patterns. */
+#ifndef JEMALLOC_ALLOC_JUNK
+#  define JEMALLOC_ALLOC_JUNK	((uint8_t)0xa5)
+#endif
+#ifndef JEMALLOC_FREE_JUNK
+#  define JEMALLOC_FREE_JUNK	((uint8_t)0x5a)
+#endif
+
 /*
  * Wrap a cpp argument that contains commas such that it isn't broken up into
  * multiple arguments.
@@ -57,73 +65,21 @@
 #	define JEMALLOC_CC_SILENCE_INIT(v)
 #endif
 
-#define	JEMALLOC_GNUC_PREREQ(major, minor)				\
-    (!defined(__clang__) &&						\
-    (__GNUC__ > (major) || (__GNUC__ == (major) && __GNUC_MINOR__ >= (minor))))
-#ifndef __has_builtin
-#  define __has_builtin(builtin) (0)
-#endif
-#define	JEMALLOC_CLANG_HAS_BUILTIN(builtin)				\
-    (defined(__clang__) && __has_builtin(builtin))
-
 #ifdef __GNUC__
 #	define likely(x)   __builtin_expect(!!(x), 1)
 #	define unlikely(x) __builtin_expect(!!(x), 0)
-#  if JEMALLOC_GNUC_PREREQ(4, 6) ||					\
-      JEMALLOC_CLANG_HAS_BUILTIN(__builtin_unreachable)
-#	define unreachable() __builtin_unreachable()
-#  else
-#	define unreachable()
-#  endif
 #else
 #	define likely(x)   !!(x)
 #	define unlikely(x) !!(x)
-#	define unreachable()
 #endif
 
-/*
- * Define a custom assert() in order to reduce the chances of deadlock during
- * assertion failure.
- */
-#ifndef assert
-#define	assert(e) do {							\
-	if (unlikely(config_debug && !(e))) {				\
-		malloc_printf(						\
-		    "<jemalloc>: %s:%d: Failed assertion: \"%s\"\n",	\
-		    __FILE__, __LINE__, #e);				\
-		abort();						\
-	}								\
-} while (0)
+#if !defined(JEMALLOC_INTERNAL_UNREACHABLE)
+#  error JEMALLOC_INTERNAL_UNREACHABLE should have been defined by configure
 #endif
 
-#ifndef not_reached
-#define	not_reached() do {						\
-	if (config_debug) {						\
-		malloc_printf(						\
-		    "<jemalloc>: %s:%d: Unreachable code reached\n",	\
-		    __FILE__, __LINE__);				\
-		abort();						\
-	}								\
-	unreachable();							\
-} while (0)
-#endif
+#define unreachable() JEMALLOC_INTERNAL_UNREACHABLE()
 
-#ifndef not_implemented
-#define	not_implemented() do {						\
-	if (config_debug) {						\
-		malloc_printf("<jemalloc>: %s:%d: Not implemented\n",	\
-		    __FILE__, __LINE__);				\
-		abort();						\
-	}								\
-} while (0)
-#endif
-
-#ifndef assert_not_implemented
-#define	assert_not_implemented(e) do {					\
-	if (unlikely(config_debug && !(e)))				\
-		not_implemented();					\
-} while (0)
-#endif
+#include "jemalloc/internal/assert.h"
 
 /* Use to assert a particular configuration, e.g., cassert(config_debug). */
 #define	cassert(c) do {							\
@@ -148,9 +104,9 @@ void	malloc_write(const char *s);
  * malloc_vsnprintf() supports a subset of snprintf(3) that avoids floating
  * point math.
  */
-int	malloc_vsnprintf(char *str, size_t size, const char *format,
+size_t	malloc_vsnprintf(char *str, size_t size, const char *format,
     va_list ap);
-int	malloc_snprintf(char *str, size_t size, const char *format, ...)
+size_t	malloc_snprintf(char *str, size_t size, const char *format, ...)
     JEMALLOC_FORMAT_PRINTF(3, 4);
 void	malloc_vcprintf(void (*write_cb)(void *, const char *), void *cbopaque,
     const char *format, va_list ap);
@@ -163,10 +119,16 @@ void	malloc_printf(const char *format, ...) JEMALLOC_FORMAT_PRINTF(1, 2);
 #ifdef JEMALLOC_H_INLINES
 
 #ifndef JEMALLOC_ENABLE_INLINE
-int	jemalloc_ffsl(long bitmap);
-int	jemalloc_ffs(int bitmap);
-size_t	pow2_ceil(size_t x);
-size_t	lg_floor(size_t x);
+unsigned	ffs_llu(unsigned long long bitmap);
+unsigned	ffs_lu(unsigned long bitmap);
+unsigned	ffs_u(unsigned bitmap);
+unsigned	ffs_zu(size_t bitmap);
+unsigned	ffs_u64(uint64_t bitmap);
+unsigned	ffs_u32(uint32_t bitmap);
+uint64_t	pow2_ceil_u64(uint64_t x);
+uint32_t	pow2_ceil_u32(uint32_t x);
+size_t	pow2_ceil_zu(size_t x);
+unsigned	lg_floor(size_t x);
 void	set_errno(int errnum);
 int	get_errno(void);
 #endif
@@ -174,27 +136,74 @@ int	get_errno(void);
 #if (defined(JEMALLOC_ENABLE_INLINE) || defined(JEMALLOC_UTIL_C_))
 
 /* Sanity check. */
-#if !defined(JEMALLOC_INTERNAL_FFSL) || !defined(JEMALLOC_INTERNAL_FFS)
-#  error Both JEMALLOC_INTERNAL_FFSL && JEMALLOC_INTERNAL_FFS should have been defined by configure
+#if !defined(JEMALLOC_INTERNAL_FFSLL) || !defined(JEMALLOC_INTERNAL_FFSL) \
+    || !defined(JEMALLOC_INTERNAL_FFS)
+#  error JEMALLOC_INTERNAL_FFS{,L,LL} should have been defined by configure
 #endif
 
-JEMALLOC_ALWAYS_INLINE int
-jemalloc_ffsl(long bitmap)
+JEMALLOC_ALWAYS_INLINE unsigned
+ffs_llu(unsigned long long bitmap)
+{
+
+	return (JEMALLOC_INTERNAL_FFSLL(bitmap));
+}
+
+JEMALLOC_ALWAYS_INLINE unsigned
+ffs_lu(unsigned long bitmap)
 {
 
 	return (JEMALLOC_INTERNAL_FFSL(bitmap));
 }
 
-JEMALLOC_ALWAYS_INLINE int
-jemalloc_ffs(int bitmap)
+JEMALLOC_ALWAYS_INLINE unsigned
+ffs_u(unsigned bitmap)
 {
 
 	return (JEMALLOC_INTERNAL_FFS(bitmap));
 }
 
-/* Compute the smallest power of 2 that is >= x. */
-JEMALLOC_INLINE size_t
-pow2_ceil(size_t x)
+JEMALLOC_ALWAYS_INLINE unsigned
+ffs_zu(size_t bitmap)
+{
+
+#if LG_SIZEOF_PTR == LG_SIZEOF_INT
+	return (ffs_u(bitmap));
+#elif LG_SIZEOF_PTR == LG_SIZEOF_LONG
+	return (ffs_lu(bitmap));
+#elif LG_SIZEOF_PTR == LG_SIZEOF_LONG_LONG
+	return (ffs_llu(bitmap));
+#else
+#error No implementation for size_t ffs()
+#endif
+}
+
+JEMALLOC_ALWAYS_INLINE unsigned
+ffs_u64(uint64_t bitmap)
+{
+
+#if LG_SIZEOF_LONG == 3
+	return (ffs_lu(bitmap));
+#elif LG_SIZEOF_LONG_LONG == 3
+	return (ffs_llu(bitmap));
+#else
+#error No implementation for 64-bit ffs()
+#endif
+}
+
+JEMALLOC_ALWAYS_INLINE unsigned
+ffs_u32(uint32_t bitmap)
+{
+
+#if LG_SIZEOF_INT == 2
+	return (ffs_u(bitmap));
+#else
+#error No implementation for 32-bit ffs()
+#endif
+	return (ffs_u(bitmap));
+}
+
+JEMALLOC_INLINE uint64_t
+pow2_ceil_u64(uint64_t x)
 {
 
 	x--;
@@ -203,15 +212,39 @@ pow2_ceil(size_t x)
 	x |= x >> 4;
 	x |= x >> 8;
 	x |= x >> 16;
-#if (LG_SIZEOF_PTR == 3)
 	x |= x >> 32;
-#endif
 	x++;
 	return (x);
 }
 
-#if (defined(__i386__) || defined(__amd64__) || defined(__x86_64__))
+JEMALLOC_INLINE uint32_t
+pow2_ceil_u32(uint32_t x)
+{
+
+	x--;
+	x |= x >> 1;
+	x |= x >> 2;
+	x |= x >> 4;
+	x |= x >> 8;
+	x |= x >> 16;
+	x++;
+	return (x);
+}
+
+/* Compute the smallest power of 2 that is >= x. */
 JEMALLOC_INLINE size_t
+pow2_ceil_zu(size_t x)
+{
+
+#if (LG_SIZEOF_PTR == 3)
+	return (pow2_ceil_u64(x));
+#else
+	return (pow2_ceil_u32(x));
+#endif
+}
+
+#if (defined(__i386__) || defined(__amd64__) || defined(__x86_64__))
+JEMALLOC_INLINE unsigned
 lg_floor(size_t x)
 {
 	size_t ret;
@@ -222,10 +255,11 @@ lg_floor(size_t x)
 	    : "=r"(ret) // Outputs.
 	    : "r"(x)    // Inputs.
 	    );
-	return (ret);
+	assert(ret < UINT_MAX);
+	return ((unsigned)ret);
 }
 #elif (defined(_MSC_VER))
-JEMALLOC_INLINE size_t
+JEMALLOC_INLINE unsigned
 lg_floor(size_t x)
 {
 	unsigned long ret;
@@ -237,12 +271,13 @@ lg_floor(size_t x)
 #elif (LG_SIZEOF_PTR == 2)
 	_BitScanReverse(&ret, x);
 #else
-#  error "Unsupported type sizes for lg_floor()"
+#  error "Unsupported type size for lg_floor()"
 #endif
-	return (ret);
+	assert(ret < UINT_MAX);
+	return ((unsigned)ret);
 }
 #elif (defined(JEMALLOC_HAVE_BUILTIN_CLZ))
-JEMALLOC_INLINE size_t
+JEMALLOC_INLINE unsigned
 lg_floor(size_t x)
 {
 
@@ -253,11 +288,11 @@ lg_floor(size_t x)
 #elif (LG_SIZEOF_PTR == LG_SIZEOF_LONG)
 	return (((8 << LG_SIZEOF_PTR) - 1) - __builtin_clzl(x));
 #else
-#  error "Unsupported type sizes for lg_floor()"
+#  error "Unsupported type size for lg_floor()"
 #endif
 }
 #else
-JEMALLOC_INLINE size_t
+JEMALLOC_INLINE unsigned
 lg_floor(size_t x)
 {
 
@@ -268,20 +303,13 @@ lg_floor(size_t x)
 	x |= (x >> 4);
 	x |= (x >> 8);
 	x |= (x >> 16);
-#if (LG_SIZEOF_PTR == 3 && LG_SIZEOF_PTR == LG_SIZEOF_LONG)
+#if (LG_SIZEOF_PTR == 3)
 	x |= (x >> 32);
-	if (x == KZU(0xffffffffffffffff))
-		return (63);
-	x++;
-	return (jemalloc_ffsl(x) - 2);
-#elif (LG_SIZEOF_PTR == 2)
-	if (x == KZU(0xffffffff))
-		return (31);
-	x++;
-	return (jemalloc_ffs(x) - 2);
-#else
-#  error "Unsupported type sizes for lg_floor()"
 #endif
+	if (x == SIZE_T_MAX)
+		return ((8 << LG_SIZEOF_PTR) - 1);
+	x++;
+	return (ffs_zu(x) - 2);
 }
 #endif
 
diff --git a/deps/jemalloc/include/jemalloc/internal/valgrind.h b/deps/jemalloc/include/jemalloc/internal/valgrind.h
index a3380df92..877a142b6 100644
--- a/deps/jemalloc/include/jemalloc/internal/valgrind.h
+++ b/deps/jemalloc/include/jemalloc/internal/valgrind.h
@@ -30,17 +30,31 @@
  * calls must be embedded in macros rather than in functions so that when
  * Valgrind reports errors, there are no extra stack frames in the backtraces.
  */
-#define	JEMALLOC_VALGRIND_MALLOC(cond, ptr, usize, zero) do {		\
-	if (unlikely(in_valgrind && cond))				\
-		VALGRIND_MALLOCLIKE_BLOCK(ptr, usize, p2rz(ptr), zero);	\
+#define	JEMALLOC_VALGRIND_MALLOC(cond, tsdn, ptr, usize, zero) do {	\
+	if (unlikely(in_valgrind && cond)) {				\
+		VALGRIND_MALLOCLIKE_BLOCK(ptr, usize, p2rz(tsdn, ptr),	\
+		    zero);						\
+	}								\
 } while (0)
-#define	JEMALLOC_VALGRIND_REALLOC(maybe_moved, ptr, usize,		\
-    ptr_maybe_null, old_ptr, old_usize, old_rzsize, old_ptr_maybe_null,	\
-    zero) do {								\
+#define	JEMALLOC_VALGRIND_REALLOC_MOVED_no(ptr, old_ptr)		\
+    (false)
+#define	JEMALLOC_VALGRIND_REALLOC_MOVED_maybe(ptr, old_ptr)		\
+    ((ptr) != (old_ptr))
+#define	JEMALLOC_VALGRIND_REALLOC_PTR_NULL_no(ptr)			\
+    (false)
+#define	JEMALLOC_VALGRIND_REALLOC_PTR_NULL_maybe(ptr)			\
+    (ptr == NULL)
+#define	JEMALLOC_VALGRIND_REALLOC_OLD_PTR_NULL_no(old_ptr)		\
+    (false)
+#define	JEMALLOC_VALGRIND_REALLOC_OLD_PTR_NULL_maybe(old_ptr)		\
+    (old_ptr == NULL)
+#define	JEMALLOC_VALGRIND_REALLOC(moved, tsdn, ptr, usize, ptr_null,	\
+    old_ptr, old_usize, old_rzsize, old_ptr_null, zero) do {		\
 	if (unlikely(in_valgrind)) {					\
-		size_t rzsize = p2rz(ptr);				\
+		size_t rzsize = p2rz(tsdn, ptr);			\
 									\
-		if (!maybe_moved || ptr == old_ptr) {			\
+		if (!JEMALLOC_VALGRIND_REALLOC_MOVED_##moved(ptr,	\
+		    old_ptr)) {						\
 			VALGRIND_RESIZEINPLACE_BLOCK(ptr, old_usize,	\
 			    usize, rzsize);				\
 			if (zero && old_usize < usize) {		\
@@ -49,11 +63,13 @@
 				    old_usize), usize - old_usize);	\
 			}						\
 		} else {						\
-			if (!old_ptr_maybe_null || old_ptr != NULL) {	\
+			if (!JEMALLOC_VALGRIND_REALLOC_OLD_PTR_NULL_##	\
+			    old_ptr_null(old_ptr)) {			\
 				valgrind_freelike_block(old_ptr,	\
 				    old_rzsize);			\
 			}						\
-			if (!ptr_maybe_null || ptr != NULL) {		\
+			if (!JEMALLOC_VALGRIND_REALLOC_PTR_NULL_##	\
+			    ptr_null(ptr)) {				\
 				size_t copy_size = (old_usize < usize)	\
 				    ?  old_usize : usize;		\
 				size_t tail_size = usize - copy_size;	\
@@ -81,8 +97,8 @@
 #define	JEMALLOC_VALGRIND_MAKE_MEM_NOACCESS(ptr, usize) do {} while (0)
 #define	JEMALLOC_VALGRIND_MAKE_MEM_UNDEFINED(ptr, usize) do {} while (0)
 #define	JEMALLOC_VALGRIND_MAKE_MEM_DEFINED(ptr, usize) do {} while (0)
-#define	JEMALLOC_VALGRIND_MALLOC(cond, ptr, usize, zero) do {} while (0)
-#define	JEMALLOC_VALGRIND_REALLOC(maybe_moved, ptr, usize,		\
+#define	JEMALLOC_VALGRIND_MALLOC(cond, tsdn, ptr, usize, zero) do {} while (0)
+#define	JEMALLOC_VALGRIND_REALLOC(maybe_moved, tsdn, ptr, usize,	\
     ptr_maybe_null, old_ptr, old_usize, old_rzsize, old_ptr_maybe_null,	\
     zero) do {} while (0)
 #define	JEMALLOC_VALGRIND_FREE(ptr, rzsize) do {} while (0)
diff --git a/deps/jemalloc/include/jemalloc/internal/witness.h b/deps/jemalloc/include/jemalloc/internal/witness.h
new file mode 100644
index 000000000..cdf15d797
--- /dev/null
+++ b/deps/jemalloc/include/jemalloc/internal/witness.h
@@ -0,0 +1,266 @@
+/******************************************************************************/
+#ifdef JEMALLOC_H_TYPES
+
+typedef struct witness_s witness_t;
+typedef unsigned witness_rank_t;
+typedef ql_head(witness_t) witness_list_t;
+typedef int witness_comp_t (const witness_t *, const witness_t *);
+
+/*
+ * Lock ranks.  Witnesses with rank WITNESS_RANK_OMIT are completely ignored by
+ * the witness machinery.
+ */
+#define	WITNESS_RANK_OMIT		0U
+
+#define	WITNESS_RANK_INIT		1U
+#define	WITNESS_RANK_CTL		1U
+#define	WITNESS_RANK_ARENAS		2U
+
+#define	WITNESS_RANK_PROF_DUMP		3U
+#define	WITNESS_RANK_PROF_BT2GCTX	4U
+#define	WITNESS_RANK_PROF_TDATAS	5U
+#define	WITNESS_RANK_PROF_TDATA		6U
+#define	WITNESS_RANK_PROF_GCTX		7U
+
+#define	WITNESS_RANK_ARENA		8U
+#define	WITNESS_RANK_ARENA_CHUNKS	9U
+#define	WITNESS_RANK_ARENA_NODE_CACHE	10
+
+#define	WITNESS_RANK_BASE		11U
+
+#define	WITNESS_RANK_LEAF		0xffffffffU
+#define	WITNESS_RANK_ARENA_BIN		WITNESS_RANK_LEAF
+#define	WITNESS_RANK_ARENA_HUGE		WITNESS_RANK_LEAF
+#define	WITNESS_RANK_DSS		WITNESS_RANK_LEAF
+#define	WITNESS_RANK_PROF_ACTIVE	WITNESS_RANK_LEAF
+#define	WITNESS_RANK_PROF_DUMP_SEQ	WITNESS_RANK_LEAF
+#define	WITNESS_RANK_PROF_GDUMP		WITNESS_RANK_LEAF
+#define	WITNESS_RANK_PROF_NEXT_THR_UID	WITNESS_RANK_LEAF
+#define	WITNESS_RANK_PROF_THREAD_ACTIVE_INIT	WITNESS_RANK_LEAF
+
+#define	WITNESS_INITIALIZER(rank) {"initializer", rank, NULL, {NULL, NULL}}
+
+#endif /* JEMALLOC_H_TYPES */
+/******************************************************************************/
+#ifdef JEMALLOC_H_STRUCTS
+
+struct witness_s {
+	/* Name, used for printing lock order reversal messages. */
+	const char		*name;
+
+	/*
+	 * Witness rank, where 0 is lowest and UINT_MAX is highest.  Witnesses
+	 * must be acquired in order of increasing rank.
+	 */
+	witness_rank_t		rank;
+
+	/*
+	 * If two witnesses are of equal rank and they have the samp comp
+	 * function pointer, it is called as a last attempt to differentiate
+	 * between witnesses of equal rank.
+	 */
+	witness_comp_t		*comp;
+
+	/* Linkage for thread's currently owned locks. */
+	ql_elm(witness_t)	link;
+};
+
+#endif /* JEMALLOC_H_STRUCTS */
+/******************************************************************************/
+#ifdef JEMALLOC_H_EXTERNS
+
+void	witness_init(witness_t *witness, const char *name, witness_rank_t rank,
+    witness_comp_t *comp);
+#ifdef JEMALLOC_JET
+typedef void (witness_lock_error_t)(const witness_list_t *, const witness_t *);
+extern witness_lock_error_t *witness_lock_error;
+#else
+void	witness_lock_error(const witness_list_t *witnesses,
+    const witness_t *witness);
+#endif
+#ifdef JEMALLOC_JET
+typedef void (witness_owner_error_t)(const witness_t *);
+extern witness_owner_error_t *witness_owner_error;
+#else
+void	witness_owner_error(const witness_t *witness);
+#endif
+#ifdef JEMALLOC_JET
+typedef void (witness_not_owner_error_t)(const witness_t *);
+extern witness_not_owner_error_t *witness_not_owner_error;
+#else
+void	witness_not_owner_error(const witness_t *witness);
+#endif
+#ifdef JEMALLOC_JET
+typedef void (witness_lockless_error_t)(const witness_list_t *);
+extern witness_lockless_error_t *witness_lockless_error;
+#else
+void	witness_lockless_error(const witness_list_t *witnesses);
+#endif
+
+void	witnesses_cleanup(tsd_t *tsd);
+void	witness_fork_cleanup(tsd_t *tsd);
+void	witness_prefork(tsd_t *tsd);
+void	witness_postfork_parent(tsd_t *tsd);
+void	witness_postfork_child(tsd_t *tsd);
+
+#endif /* JEMALLOC_H_EXTERNS */
+/******************************************************************************/
+#ifdef JEMALLOC_H_INLINES
+
+#ifndef JEMALLOC_ENABLE_INLINE
+bool	witness_owner(tsd_t *tsd, const witness_t *witness);
+void	witness_assert_owner(tsdn_t *tsdn, const witness_t *witness);
+void	witness_assert_not_owner(tsdn_t *tsdn, const witness_t *witness);
+void	witness_assert_lockless(tsdn_t *tsdn);
+void	witness_lock(tsdn_t *tsdn, witness_t *witness);
+void	witness_unlock(tsdn_t *tsdn, witness_t *witness);
+#endif
+
+#if (defined(JEMALLOC_ENABLE_INLINE) || defined(JEMALLOC_MUTEX_C_))
+JEMALLOC_INLINE bool
+witness_owner(tsd_t *tsd, const witness_t *witness)
+{
+	witness_list_t *witnesses;
+	witness_t *w;
+
+	witnesses = tsd_witnessesp_get(tsd);
+	ql_foreach(w, witnesses, link) {
+		if (w == witness)
+			return (true);
+	}
+
+	return (false);
+}
+
+JEMALLOC_INLINE void
+witness_assert_owner(tsdn_t *tsdn, const witness_t *witness)
+{
+	tsd_t *tsd;
+
+	if (!config_debug)
+		return;
+
+	if (tsdn_null(tsdn))
+		return;
+	tsd = tsdn_tsd(tsdn);
+	if (witness->rank == WITNESS_RANK_OMIT)
+		return;
+
+	if (witness_owner(tsd, witness))
+		return;
+	witness_owner_error(witness);
+}
+
+JEMALLOC_INLINE void
+witness_assert_not_owner(tsdn_t *tsdn, const witness_t *witness)
+{
+	tsd_t *tsd;
+	witness_list_t *witnesses;
+	witness_t *w;
+
+	if (!config_debug)
+		return;
+
+	if (tsdn_null(tsdn))
+		return;
+	tsd = tsdn_tsd(tsdn);
+	if (witness->rank == WITNESS_RANK_OMIT)
+		return;
+
+	witnesses = tsd_witnessesp_get(tsd);
+	ql_foreach(w, witnesses, link) {
+		if (w == witness)
+			witness_not_owner_error(witness);
+	}
+}
+
+JEMALLOC_INLINE void
+witness_assert_lockless(tsdn_t *tsdn)
+{
+	tsd_t *tsd;
+	witness_list_t *witnesses;
+	witness_t *w;
+
+	if (!config_debug)
+		return;
+
+	if (tsdn_null(tsdn))
+		return;
+	tsd = tsdn_tsd(tsdn);
+
+	witnesses = tsd_witnessesp_get(tsd);
+	w = ql_last(witnesses, link);
+	if (w != NULL)
+		witness_lockless_error(witnesses);
+}
+
+JEMALLOC_INLINE void
+witness_lock(tsdn_t *tsdn, witness_t *witness)
+{
+	tsd_t *tsd;
+	witness_list_t *witnesses;
+	witness_t *w;
+
+	if (!config_debug)
+		return;
+
+	if (tsdn_null(tsdn))
+		return;
+	tsd = tsdn_tsd(tsdn);
+	if (witness->rank == WITNESS_RANK_OMIT)
+		return;
+
+	witness_assert_not_owner(tsdn, witness);
+
+	witnesses = tsd_witnessesp_get(tsd);
+	w = ql_last(witnesses, link);
+	if (w == NULL) {
+		/* No other locks; do nothing. */
+	} else if (tsd_witness_fork_get(tsd) && w->rank <= witness->rank) {
+		/* Forking, and relaxed ranking satisfied. */
+	} else if (w->rank > witness->rank) {
+		/* Not forking, rank order reversal. */
+		witness_lock_error(witnesses, witness);
+	} else if (w->rank == witness->rank && (w->comp == NULL || w->comp !=
+	    witness->comp || w->comp(w, witness) > 0)) {
+		/*
+		 * Missing/incompatible comparison function, or comparison
+		 * function indicates rank order reversal.
+		 */
+		witness_lock_error(witnesses, witness);
+	}
+
+	ql_elm_new(witness, link);
+	ql_tail_insert(witnesses, witness, link);
+}
+
+JEMALLOC_INLINE void
+witness_unlock(tsdn_t *tsdn, witness_t *witness)
+{
+	tsd_t *tsd;
+	witness_list_t *witnesses;
+
+	if (!config_debug)
+		return;
+
+	if (tsdn_null(tsdn))
+		return;
+	tsd = tsdn_tsd(tsdn);
+	if (witness->rank == WITNESS_RANK_OMIT)
+		return;
+
+	/*
+	 * Check whether owner before removal, rather than relying on
+	 * witness_assert_owner() to abort, so that unit tests can test this
+	 * function's failure mode without causing undefined behavior.
+	 */
+	if (witness_owner(tsd, witness)) {
+		witnesses = tsd_witnessesp_get(tsd);
+		ql_remove(witnesses, witness, link);
+	} else
+		witness_assert_owner(tsdn, witness);
+}
+#endif
+
+#endif /* JEMALLOC_H_INLINES */
+/******************************************************************************/
diff --git a/deps/jemalloc/include/jemalloc/jemalloc_defs.h.in b/deps/jemalloc/include/jemalloc/jemalloc_defs.h.in
index ab13c3758..6d89435c2 100644
--- a/deps/jemalloc/include/jemalloc/jemalloc_defs.h.in
+++ b/deps/jemalloc/include/jemalloc/jemalloc_defs.h.in
@@ -33,5 +33,13 @@
  */
 #undef JEMALLOC_USE_CXX_THROW
 
+#ifdef _MSC_VER
+#  ifdef _WIN64
+#    define LG_SIZEOF_PTR_WIN 3
+#  else
+#    define LG_SIZEOF_PTR_WIN 2
+#  endif
+#endif
+
 /* sizeof(void *) == 2^LG_SIZEOF_PTR. */
 #undef LG_SIZEOF_PTR
diff --git a/deps/jemalloc/include/jemalloc/jemalloc_macros.h.in b/deps/jemalloc/include/jemalloc/jemalloc_macros.h.in
index a7028db34..2956c7bb4 100644
--- a/deps/jemalloc/include/jemalloc/jemalloc_macros.h.in
+++ b/deps/jemalloc/include/jemalloc/jemalloc_macros.h.in
@@ -11,12 +11,13 @@
 #define	JEMALLOC_VERSION_NREV @jemalloc_version_nrev@
 #define	JEMALLOC_VERSION_GID "@jemalloc_version_gid@"
 
-#  define MALLOCX_LG_ALIGN(la)	(la)
+#  define MALLOCX_LG_ALIGN(la)	((int)(la))
 #  if LG_SIZEOF_PTR == 2
-#    define MALLOCX_ALIGN(a)	(ffs(a)-1)
+#    define MALLOCX_ALIGN(a)	((int)(ffs((int)(a))-1))
 #  else
 #    define MALLOCX_ALIGN(a)						\
-	 ((a < (size_t)INT_MAX) ? ffs(a)-1 : ffs(a>>32)+31)
+       ((int)(((size_t)(a) < (size_t)INT_MAX) ? ffs((int)(a))-1 :	\
+       ffs((int)(((size_t)(a))>>32))+31))
 #  endif
 #  define MALLOCX_ZERO	((int)0x40)
 /*
@@ -28,7 +29,7 @@
 /*
  * Bias arena index bits so that 0 encodes "use an automatically chosen arena".
  */
-#  define MALLOCX_ARENA(a)	((int)(((a)+1) << 20))
+#  define MALLOCX_ARENA(a)	((((int)(a))+1) << 20)
 
 #if defined(__cplusplus) && defined(JEMALLOC_USE_CXX_THROW)
 #  define JEMALLOC_CXX_THROW throw()
@@ -36,32 +37,7 @@
 #  define JEMALLOC_CXX_THROW
 #endif
 
-#ifdef JEMALLOC_HAVE_ATTR
-#  define JEMALLOC_ATTR(s) __attribute__((s))
-#  define JEMALLOC_ALIGNED(s) JEMALLOC_ATTR(aligned(s))
-#  ifdef JEMALLOC_HAVE_ATTR_ALLOC_SIZE
-#    define JEMALLOC_ALLOC_SIZE(s) JEMALLOC_ATTR(alloc_size(s))
-#    define JEMALLOC_ALLOC_SIZE2(s1, s2) JEMALLOC_ATTR(alloc_size(s1, s2))
-#  else
-#    define JEMALLOC_ALLOC_SIZE(s)
-#    define JEMALLOC_ALLOC_SIZE2(s1, s2)
-#  endif
-#  ifndef JEMALLOC_EXPORT
-#    define JEMALLOC_EXPORT JEMALLOC_ATTR(visibility("default"))
-#  endif
-#  ifdef JEMALLOC_HAVE_ATTR_FORMAT_GNU_PRINTF
-#    define JEMALLOC_FORMAT_PRINTF(s, i) JEMALLOC_ATTR(format(gnu_printf, s, i))
-#  elif defined(JEMALLOC_HAVE_ATTR_FORMAT_PRINTF)
-#    define JEMALLOC_FORMAT_PRINTF(s, i) JEMALLOC_ATTR(format(printf, s, i))
-#  else
-#    define JEMALLOC_FORMAT_PRINTF(s, i)
-#  endif
-#  define JEMALLOC_NOINLINE JEMALLOC_ATTR(noinline)
-#  define JEMALLOC_NOTHROW JEMALLOC_ATTR(nothrow)
-#  define JEMALLOC_SECTION(s) JEMALLOC_ATTR(section(s))
-#  define JEMALLOC_RESTRICT_RETURN
-#  define JEMALLOC_ALLOCATOR
-#elif _MSC_VER
+#if _MSC_VER
 #  define JEMALLOC_ATTR(s)
 #  define JEMALLOC_ALIGNED(s) __declspec(align(s))
 #  define JEMALLOC_ALLOC_SIZE(s)
@@ -87,6 +63,31 @@
 #  else
 #    define JEMALLOC_ALLOCATOR
 #  endif
+#elif defined(JEMALLOC_HAVE_ATTR)
+#  define JEMALLOC_ATTR(s) __attribute__((s))
+#  define JEMALLOC_ALIGNED(s) JEMALLOC_ATTR(aligned(s))
+#  ifdef JEMALLOC_HAVE_ATTR_ALLOC_SIZE
+#    define JEMALLOC_ALLOC_SIZE(s) JEMALLOC_ATTR(alloc_size(s))
+#    define JEMALLOC_ALLOC_SIZE2(s1, s2) JEMALLOC_ATTR(alloc_size(s1, s2))
+#  else
+#    define JEMALLOC_ALLOC_SIZE(s)
+#    define JEMALLOC_ALLOC_SIZE2(s1, s2)
+#  endif
+#  ifndef JEMALLOC_EXPORT
+#    define JEMALLOC_EXPORT JEMALLOC_ATTR(visibility("default"))
+#  endif
+#  ifdef JEMALLOC_HAVE_ATTR_FORMAT_GNU_PRINTF
+#    define JEMALLOC_FORMAT_PRINTF(s, i) JEMALLOC_ATTR(format(gnu_printf, s, i))
+#  elif defined(JEMALLOC_HAVE_ATTR_FORMAT_PRINTF)
+#    define JEMALLOC_FORMAT_PRINTF(s, i) JEMALLOC_ATTR(format(printf, s, i))
+#  else
+#    define JEMALLOC_FORMAT_PRINTF(s, i)
+#  endif
+#  define JEMALLOC_NOINLINE JEMALLOC_ATTR(noinline)
+#  define JEMALLOC_NOTHROW JEMALLOC_ATTR(nothrow)
+#  define JEMALLOC_SECTION(s) JEMALLOC_ATTR(section(s))
+#  define JEMALLOC_RESTRICT_RETURN
+#  define JEMALLOC_ALLOCATOR
 #else
 #  define JEMALLOC_ATTR(s)
 #  define JEMALLOC_ALIGNED(s)
diff --git a/deps/jemalloc/include/msvc_compat/strings.h b/deps/jemalloc/include/msvc_compat/strings.h
index f01ffdd18..a3ee25063 100644
--- a/deps/jemalloc/include/msvc_compat/strings.h
+++ b/deps/jemalloc/include/msvc_compat/strings.h
@@ -21,7 +21,37 @@ static __forceinline int ffs(int x)
 	return (ffsl(x));
 }
 
+#  ifdef  _M_X64
+#    pragma intrinsic(_BitScanForward64)
+#  endif
+
+static __forceinline int ffsll(unsigned __int64 x)
+{
+	unsigned long i;
+#ifdef  _M_X64
+	if (_BitScanForward64(&i, x))
+		return (i + 1);
+	return (0);
+#else
+// Fallback for 32-bit build where 64-bit version not available
+// assuming little endian
+	union {
+		unsigned __int64 ll;
+		unsigned   long l[2];
+	} s;
+
+	s.ll = x;
+
+	if (_BitScanForward(&i, s.l[0]))
+		return (i + 1);
+	else if(_BitScanForward(&i, s.l[1]))
+		return (i + 33);
+	return (0);
+#endif
+}
+
 #else
+#  define ffsll(x) __builtin_ffsll(x)
 #  define ffsl(x) __builtin_ffsl(x)
 #  define ffs(x) __builtin_ffs(x)
 #endif
diff --git a/deps/jemalloc/include/msvc_compat/windows_extra.h b/deps/jemalloc/include/msvc_compat/windows_extra.h
index 0c5e323ff..3008faa37 100644
--- a/deps/jemalloc/include/msvc_compat/windows_extra.h
+++ b/deps/jemalloc/include/msvc_compat/windows_extra.h
@@ -1,26 +1,6 @@
 #ifndef MSVC_COMPAT_WINDOWS_EXTRA_H
 #define	MSVC_COMPAT_WINDOWS_EXTRA_H
 
-#ifndef ENOENT
-#  define ENOENT ERROR_PATH_NOT_FOUND
-#endif
-#ifndef EINVAL
-#  define EINVAL ERROR_BAD_ARGUMENTS
-#endif
-#ifndef EAGAIN
-#  define EAGAIN ERROR_OUTOFMEMORY
-#endif
-#ifndef EPERM
-#  define EPERM  ERROR_WRITE_FAULT
-#endif
-#ifndef EFAULT
-#  define EFAULT ERROR_INVALID_ADDRESS
-#endif
-#ifndef ENOMEM
-#  define ENOMEM ERROR_NOT_ENOUGH_MEMORY
-#endif
-#ifndef ERANGE
-#  define ERANGE ERROR_INVALID_DATA
-#endif
+#include <errno.h>
 
 #endif /* MSVC_COMPAT_WINDOWS_EXTRA_H */
diff --git a/deps/jemalloc/install-sh b/deps/jemalloc/install-sh
deleted file mode 100755
index ebc66913e..000000000
--- a/deps/jemalloc/install-sh
+++ /dev/null
@@ -1,250 +0,0 @@
-#! /bin/sh
-#
-# install - install a program, script, or datafile
-# This comes from X11R5 (mit/util/scripts/install.sh).
-#
-# Copyright 1991 by the Massachusetts Institute of Technology
-#
-# Permission to use, copy, modify, distribute, and sell this software and its
-# documentation for any purpose is hereby granted without fee, provided that
-# the above copyright notice appear in all copies and that both that
-# copyright notice and this permission notice appear in supporting
-# documentation, and that the name of M.I.T. not be used in advertising or
-# publicity pertaining to distribution of the software without specific,
-# written prior permission.  M.I.T. makes no representations about the
-# suitability of this software for any purpose.  It is provided "as is"
-# without express or implied warranty.
-#
-# Calling this script install-sh is preferred over install.sh, to prevent
-# `make' implicit rules from creating a file called install from it
-# when there is no Makefile.
-#
-# This script is compatible with the BSD install script, but was written
-# from scratch.  It can only install one file at a time, a restriction
-# shared with many OS's install programs.
-
-
-# set DOITPROG to echo to test this script
-
-# Don't use :- since 4.3BSD and earlier shells don't like it.
-doit="${DOITPROG-}"
-
-
-# put in absolute paths if you don't have them in your path; or use env. vars.
-
-mvprog="${MVPROG-mv}"
-cpprog="${CPPROG-cp}"
-chmodprog="${CHMODPROG-chmod}"
-chownprog="${CHOWNPROG-chown}"
-chgrpprog="${CHGRPPROG-chgrp}"
-stripprog="${STRIPPROG-strip}"
-rmprog="${RMPROG-rm}"
-mkdirprog="${MKDIRPROG-mkdir}"
-
-transformbasename=""
-transform_arg=""
-instcmd="$mvprog"
-chmodcmd="$chmodprog 0755"
-chowncmd=""
-chgrpcmd=""
-stripcmd=""
-rmcmd="$rmprog -f"
-mvcmd="$mvprog"
-src=""
-dst=""
-dir_arg=""
-
-while [ x"$1" != x ]; do
-    case $1 in
-	-c) instcmd="$cpprog"
-	    shift
-	    continue;;
-
-	-d) dir_arg=true
-	    shift
-	    continue;;
-
-	-m) chmodcmd="$chmodprog $2"
-	    shift
-	    shift
-	    continue;;
-
-	-o) chowncmd="$chownprog $2"
-	    shift
-	    shift
-	    continue;;
-
-	-g) chgrpcmd="$chgrpprog $2"
-	    shift
-	    shift
-	    continue;;
-
-	-s) stripcmd="$stripprog"
-	    shift
-	    continue;;
-
-	-t=*) transformarg=`echo $1 | sed 's/-t=//'`
-	    shift
-	    continue;;
-
-	-b=*) transformbasename=`echo $1 | sed 's/-b=//'`
-	    shift
-	    continue;;
-
-	*)  if [ x"$src" = x ]
-	    then
-		src=$1
-	    else
-		# this colon is to work around a 386BSD /bin/sh bug
-		:
-		dst=$1
-	    fi
-	    shift
-	    continue;;
-    esac
-done
-
-if [ x"$src" = x ]
-then
-	echo "install:	no input file specified"
-	exit 1
-else
-	true
-fi
-
-if [ x"$dir_arg" != x ]; then
-	dst=$src
-	src=""
-	
-	if [ -d $dst ]; then
-		instcmd=:
-	else
-		instcmd=mkdir
-	fi
-else
-
-# Waiting for this to be detected by the "$instcmd $src $dsttmp" command
-# might cause directories to be created, which would be especially bad 
-# if $src (and thus $dsttmp) contains '*'.
-
-	if [ -f $src -o -d $src ]
-	then
-		true
-	else
-		echo "install:  $src does not exist"
-		exit 1
-	fi
-	
-	if [ x"$dst" = x ]
-	then
-		echo "install:	no destination specified"
-		exit 1
-	else
-		true
-	fi
-
-# If destination is a directory, append the input filename; if your system
-# does not like double slashes in filenames, you may need to add some logic
-
-	if [ -d $dst ]
-	then
-		dst="$dst"/`basename $src`
-	else
-		true
-	fi
-fi
-
-## this sed command emulates the dirname command
-dstdir=`echo $dst | sed -e 's,[^/]*$,,;s,/$,,;s,^$,.,'`
-
-# Make sure that the destination directory exists.
-#  this part is taken from Noah Friedman's mkinstalldirs script
-
-# Skip lots of stat calls in the usual case.
-if [ ! -d "$dstdir" ]; then
-defaultIFS='	
-'
-IFS="${IFS-${defaultIFS}}"
-
-oIFS="${IFS}"
-# Some sh's can't handle IFS=/ for some reason.
-IFS='%'
-set - `echo ${dstdir} | sed -e 's@/@%@g' -e 's@^%@/@'`
-IFS="${oIFS}"
-
-pathcomp=''
-
-while [ $# -ne 0 ] ; do
-	pathcomp="${pathcomp}${1}"
-	shift
-
-	if [ ! -d "${pathcomp}" ] ;
-        then
-		$mkdirprog "${pathcomp}"
-	else
-		true
-	fi
-
-	pathcomp="${pathcomp}/"
-done
-fi
-
-if [ x"$dir_arg" != x ]
-then
-	$doit $instcmd $dst &&
-
-	if [ x"$chowncmd" != x ]; then $doit $chowncmd $dst; else true ; fi &&
-	if [ x"$chgrpcmd" != x ]; then $doit $chgrpcmd $dst; else true ; fi &&
-	if [ x"$stripcmd" != x ]; then $doit $stripcmd $dst; else true ; fi &&
-	if [ x"$chmodcmd" != x ]; then $doit $chmodcmd $dst; else true ; fi
-else
-
-# If we're going to rename the final executable, determine the name now.
-
-	if [ x"$transformarg" = x ] 
-	then
-		dstfile=`basename $dst`
-	else
-		dstfile=`basename $dst $transformbasename | 
-			sed $transformarg`$transformbasename
-	fi
-
-# don't allow the sed command to completely eliminate the filename
-
-	if [ x"$dstfile" = x ] 
-	then
-		dstfile=`basename $dst`
-	else
-		true
-	fi
-
-# Make a temp file name in the proper directory.
-
-	dsttmp=$dstdir/#inst.$$#
-
-# Move or copy the file name to the temp name
-
-	$doit $instcmd $src $dsttmp &&
-
-	trap "rm -f ${dsttmp}" 0 &&
-
-# and set any options; do chmod last to preserve setuid bits
-
-# If any of these fail, we abort the whole thing.  If we want to
-# ignore errors from any of these, just make sure not to ignore
-# errors from the above "$doit $instcmd $src $dsttmp" command.
-
-	if [ x"$chowncmd" != x ]; then $doit $chowncmd $dsttmp; else true;fi &&
-	if [ x"$chgrpcmd" != x ]; then $doit $chgrpcmd $dsttmp; else true;fi &&
-	if [ x"$stripcmd" != x ]; then $doit $stripcmd $dsttmp; else true;fi &&
-	if [ x"$chmodcmd" != x ]; then $doit $chmodcmd $dsttmp; else true;fi &&
-
-# Now rename the file to the real destination.
-
-	$doit $rmcmd -f $dstdir/$dstfile &&
-	$doit $mvcmd $dsttmp $dstdir/$dstfile 
-
-fi &&
-
-
-exit 0
diff --git a/deps/jemalloc/jemalloc.pc.in b/deps/jemalloc/jemalloc.pc.in
index 1a3ad9b34..a318e8dd3 100644
--- a/deps/jemalloc/jemalloc.pc.in
+++ b/deps/jemalloc/jemalloc.pc.in
@@ -6,7 +6,7 @@ install_suffix=@install_suffix@
 
 Name: jemalloc
 Description: A general purpose malloc(3) implementation that emphasizes fragmentation avoidance and scalable concurrency support.
-URL: http://www.canonware.com/jemalloc
+URL: http://jemalloc.net/
 Version: @jemalloc_version@
 Cflags: -I${includedir}
 Libs: -L${libdir} -ljemalloc${install_suffix}
diff --git a/deps/jemalloc/msvc/ReadMe.txt b/deps/jemalloc/msvc/ReadMe.txt
new file mode 100644
index 000000000..77d567da0
--- /dev/null
+++ b/deps/jemalloc/msvc/ReadMe.txt
@@ -0,0 +1,24 @@
+
+How to build jemalloc for Windows
+=================================
+
+1. Install Cygwin with at least the following packages:
+   * autoconf
+   * autogen
+   * gawk
+   * grep
+   * sed
+
+2. Install Visual Studio 2015 with Visual C++
+
+3. Add Cygwin\bin to the PATH environment variable
+
+4. Open "VS2015 x86 Native Tools Command Prompt"
+   (note: x86/x64 doesn't matter at this point)
+
+5. Generate header files:
+   sh -c "CC=cl ./autogen.sh"
+
+6. Now the project can be opened and built in Visual Studio:
+   msvc\jemalloc_vc2015.sln
+
diff --git a/deps/jemalloc/msvc/jemalloc_vc2015.sln b/deps/jemalloc/msvc/jemalloc_vc2015.sln
new file mode 100644
index 000000000..aedd5e5ea
--- /dev/null
+++ b/deps/jemalloc/msvc/jemalloc_vc2015.sln
@@ -0,0 +1,63 @@
+
+Microsoft Visual Studio Solution File, Format Version 12.00
+# Visual Studio 14
+VisualStudioVersion = 14.0.24720.0
+MinimumVisualStudioVersion = 10.0.40219.1
+Project("{2150E333-8FDC-42A3-9474-1A3956D46DE8}") = "Solution Items", "Solution Items", "{70A99006-6DE9-472B-8F83-4CEE6C616DF3}"
+	ProjectSection(SolutionItems) = preProject
+		ReadMe.txt = ReadMe.txt
+	EndProjectSection
+EndProject
+Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "jemalloc", "projects\vc2015\jemalloc\jemalloc.vcxproj", "{8D6BB292-9E1C-413D-9F98-4864BDC1514A}"
+EndProject
+Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "test_threads", "projects\vc2015\test_threads\test_threads.vcxproj", "{09028CFD-4EB7-491D-869C-0708DB97ED44}"
+EndProject
+Global
+	GlobalSection(SolutionConfigurationPlatforms) = preSolution
+		Debug|x64 = Debug|x64
+		Debug|x86 = Debug|x86
+		Debug-static|x64 = Debug-static|x64
+		Debug-static|x86 = Debug-static|x86
+		Release|x64 = Release|x64
+		Release|x86 = Release|x86
+		Release-static|x64 = Release-static|x64
+		Release-static|x86 = Release-static|x86
+	EndGlobalSection
+	GlobalSection(ProjectConfigurationPlatforms) = postSolution
+		{8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug|x64.ActiveCfg = Debug|x64
+		{8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug|x64.Build.0 = Debug|x64
+		{8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug|x86.ActiveCfg = Debug|Win32
+		{8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug|x86.Build.0 = Debug|Win32
+		{8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug-static|x64.ActiveCfg = Debug-static|x64
+		{8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug-static|x64.Build.0 = Debug-static|x64
+		{8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug-static|x86.ActiveCfg = Debug-static|Win32
+		{8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Debug-static|x86.Build.0 = Debug-static|Win32
+		{8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release|x64.ActiveCfg = Release|x64
+		{8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release|x64.Build.0 = Release|x64
+		{8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release|x86.ActiveCfg = Release|Win32
+		{8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release|x86.Build.0 = Release|Win32
+		{8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release-static|x64.ActiveCfg = Release-static|x64
+		{8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release-static|x64.Build.0 = Release-static|x64
+		{8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release-static|x86.ActiveCfg = Release-static|Win32
+		{8D6BB292-9E1C-413D-9F98-4864BDC1514A}.Release-static|x86.Build.0 = Release-static|Win32
+		{09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug|x64.ActiveCfg = Debug|x64
+		{09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug|x64.Build.0 = Debug|x64
+		{09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug|x86.ActiveCfg = Debug|Win32
+		{09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug|x86.Build.0 = Debug|Win32
+		{09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug-static|x64.ActiveCfg = Debug-static|x64
+		{09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug-static|x64.Build.0 = Debug-static|x64
+		{09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug-static|x86.ActiveCfg = Debug-static|Win32
+		{09028CFD-4EB7-491D-869C-0708DB97ED44}.Debug-static|x86.Build.0 = Debug-static|Win32
+		{09028CFD-4EB7-491D-869C-0708DB97ED44}.Release|x64.ActiveCfg = Release|x64
+		{09028CFD-4EB7-491D-869C-0708DB97ED44}.Release|x64.Build.0 = Release|x64
+		{09028CFD-4EB7-491D-869C-0708DB97ED44}.Release|x86.ActiveCfg = Release|Win32
+		{09028CFD-4EB7-491D-869C-0708DB97ED44}.Release|x86.Build.0 = Release|Win32
+		{09028CFD-4EB7-491D-869C-0708DB97ED44}.Release-static|x64.ActiveCfg = Release-static|x64
+		{09028CFD-4EB7-491D-869C-0708DB97ED44}.Release-static|x64.Build.0 = Release-static|x64
+		{09028CFD-4EB7-491D-869C-0708DB97ED44}.Release-static|x86.ActiveCfg = Release-static|Win32
+		{09028CFD-4EB7-491D-869C-0708DB97ED44}.Release-static|x86.Build.0 = Release-static|Win32
+	EndGlobalSection
+	GlobalSection(SolutionProperties) = preSolution
+		HideSolutionNode = FALSE
+	EndGlobalSection
+EndGlobal
diff --git a/deps/jemalloc/msvc/projects/vc2015/jemalloc/jemalloc.vcxproj b/deps/jemalloc/msvc/projects/vc2015/jemalloc/jemalloc.vcxproj
new file mode 100644
index 000000000..8342ab3ab
--- /dev/null
+++ b/deps/jemalloc/msvc/projects/vc2015/jemalloc/jemalloc.vcxproj
@@ -0,0 +1,402 @@
+<?xml version="1.0" encoding="utf-8"?>
+<Project DefaultTargets="Build" ToolsVersion="14.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
+  <ItemGroup Label="ProjectConfigurations">
+    <ProjectConfiguration Include="Debug-static|Win32">
+      <Configuration>Debug-static</Configuration>
+      <Platform>Win32</Platform>
+    </ProjectConfiguration>
+    <ProjectConfiguration Include="Debug-static|x64">
+      <Configuration>Debug-static</Configuration>
+      <Platform>x64</Platform>
+    </ProjectConfiguration>
+    <ProjectConfiguration Include="Debug|Win32">
+      <Configuration>Debug</Configuration>
+      <Platform>Win32</Platform>
+    </ProjectConfiguration>
+    <ProjectConfiguration Include="Release-static|Win32">
+      <Configuration>Release-static</Configuration>
+      <Platform>Win32</Platform>
+    </ProjectConfiguration>
+    <ProjectConfiguration Include="Release-static|x64">
+      <Configuration>Release-static</Configuration>
+      <Platform>x64</Platform>
+    </ProjectConfiguration>
+    <ProjectConfiguration Include="Release|Win32">
+      <Configuration>Release</Configuration>
+      <Platform>Win32</Platform>
+    </ProjectConfiguration>
+    <ProjectConfiguration Include="Debug|x64">
+      <Configuration>Debug</Configuration>
+      <Platform>x64</Platform>
+    </ProjectConfiguration>
+    <ProjectConfiguration Include="Release|x64">
+      <Configuration>Release</Configuration>
+      <Platform>x64</Platform>
+    </ProjectConfiguration>
+  </ItemGroup>
+  <ItemGroup>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\arena.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\assert.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\atomic.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\base.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\bitmap.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\chunk.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\chunk_dss.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\chunk_mmap.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\ckh.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\ctl.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\extent.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\hash.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\huge.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\jemalloc_internal.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\jemalloc_internal_decls.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\jemalloc_internal_defs.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\jemalloc_internal_macros.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\mb.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\mutex.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\nstime.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\pages.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\ph.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\private_namespace.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\private_unnamespace.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\prng.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\prof.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\public_namespace.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\public_unnamespace.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\ql.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\qr.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\quarantine.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\rb.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\rtree.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\size_classes.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\smoothstep.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\spin.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\stats.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\tcache.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\ticker.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\tsd.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\util.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\witness.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\jemalloc.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\jemalloc_defs.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\jemalloc_macros.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\jemalloc_mangle.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\jemalloc_protos.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\jemalloc_protos_jet.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\jemalloc_rename.h" />
+    <ClInclude Include="..\..\..\..\include\jemalloc\jemalloc_typedefs.h" />
+    <ClInclude Include="..\..\..\..\include\msvc_compat\C99\stdbool.h" />
+    <ClInclude Include="..\..\..\..\include\msvc_compat\C99\stdint.h" />
+    <ClInclude Include="..\..\..\..\include\msvc_compat\strings.h" />
+    <ClInclude Include="..\..\..\..\include\msvc_compat\windows_extra.h" />
+  </ItemGroup>
+  <ItemGroup>
+    <ClCompile Include="..\..\..\..\src\arena.c" />
+    <ClCompile Include="..\..\..\..\src\atomic.c" />
+    <ClCompile Include="..\..\..\..\src\base.c" />
+    <ClCompile Include="..\..\..\..\src\bitmap.c" />
+    <ClCompile Include="..\..\..\..\src\chunk.c" />
+    <ClCompile Include="..\..\..\..\src\chunk_dss.c" />
+    <ClCompile Include="..\..\..\..\src\chunk_mmap.c" />
+    <ClCompile Include="..\..\..\..\src\ckh.c" />
+    <ClCompile Include="..\..\..\..\src\ctl.c" />
+    <ClCompile Include="..\..\..\..\src\extent.c" />
+    <ClCompile Include="..\..\..\..\src\hash.c" />
+    <ClCompile Include="..\..\..\..\src\huge.c" />
+    <ClCompile Include="..\..\..\..\src\jemalloc.c" />
+    <ClCompile Include="..\..\..\..\src\mb.c" />
+    <ClCompile Include="..\..\..\..\src\mutex.c" />
+    <ClCompile Include="..\..\..\..\src\nstime.c" />
+    <ClCompile Include="..\..\..\..\src\pages.c" />
+    <ClCompile Include="..\..\..\..\src\prng.c" />
+    <ClCompile Include="..\..\..\..\src\prof.c" />
+    <ClCompile Include="..\..\..\..\src\quarantine.c" />
+    <ClCompile Include="..\..\..\..\src\rtree.c" />
+    <ClCompile Include="..\..\..\..\src\spin.c" />
+    <ClCompile Include="..\..\..\..\src\stats.c" />
+    <ClCompile Include="..\..\..\..\src\tcache.c" />
+    <ClCompile Include="..\..\..\..\src\ticker.c" />
+    <ClCompile Include="..\..\..\..\src\tsd.c" />
+    <ClCompile Include="..\..\..\..\src\util.c" />
+    <ClCompile Include="..\..\..\..\src\witness.c" />
+  </ItemGroup>
+  <PropertyGroup Label="Globals">
+    <ProjectGuid>{8D6BB292-9E1C-413D-9F98-4864BDC1514A}</ProjectGuid>
+    <Keyword>Win32Proj</Keyword>
+    <RootNamespace>jemalloc</RootNamespace>
+    <WindowsTargetPlatformVersion>8.1</WindowsTargetPlatformVersion>
+  </PropertyGroup>
+  <Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" />
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'" Label="Configuration">
+    <ConfigurationType>DynamicLibrary</ConfigurationType>
+    <UseDebugLibraries>true</UseDebugLibraries>
+    <PlatformToolset>v140</PlatformToolset>
+    <CharacterSet>MultiByte</CharacterSet>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'" Label="Configuration">
+    <ConfigurationType>StaticLibrary</ConfigurationType>
+    <UseDebugLibraries>true</UseDebugLibraries>
+    <PlatformToolset>v140</PlatformToolset>
+    <CharacterSet>MultiByte</CharacterSet>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'" Label="Configuration">
+    <ConfigurationType>DynamicLibrary</ConfigurationType>
+    <UseDebugLibraries>false</UseDebugLibraries>
+    <PlatformToolset>v140</PlatformToolset>
+    <WholeProgramOptimization>true</WholeProgramOptimization>
+    <CharacterSet>MultiByte</CharacterSet>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'" Label="Configuration">
+    <ConfigurationType>StaticLibrary</ConfigurationType>
+    <UseDebugLibraries>false</UseDebugLibraries>
+    <PlatformToolset>v140</PlatformToolset>
+    <WholeProgramOptimization>true</WholeProgramOptimization>
+    <CharacterSet>MultiByte</CharacterSet>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'" Label="Configuration">
+    <ConfigurationType>DynamicLibrary</ConfigurationType>
+    <UseDebugLibraries>true</UseDebugLibraries>
+    <PlatformToolset>v140</PlatformToolset>
+    <CharacterSet>MultiByte</CharacterSet>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'" Label="Configuration">
+    <ConfigurationType>StaticLibrary</ConfigurationType>
+    <UseDebugLibraries>true</UseDebugLibraries>
+    <PlatformToolset>v140</PlatformToolset>
+    <CharacterSet>MultiByte</CharacterSet>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'" Label="Configuration">
+    <ConfigurationType>DynamicLibrary</ConfigurationType>
+    <UseDebugLibraries>false</UseDebugLibraries>
+    <PlatformToolset>v140</PlatformToolset>
+    <WholeProgramOptimization>true</WholeProgramOptimization>
+    <CharacterSet>MultiByte</CharacterSet>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'" Label="Configuration">
+    <ConfigurationType>StaticLibrary</ConfigurationType>
+    <UseDebugLibraries>false</UseDebugLibraries>
+    <PlatformToolset>v140</PlatformToolset>
+    <WholeProgramOptimization>true</WholeProgramOptimization>
+    <CharacterSet>MultiByte</CharacterSet>
+  </PropertyGroup>
+  <Import Project="$(VCTargetsPath)\Microsoft.Cpp.props" />
+  <ImportGroup Label="ExtensionSettings">
+  </ImportGroup>
+  <ImportGroup Label="Shared">
+  </ImportGroup>
+  <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'">
+    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
+  </ImportGroup>
+  <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'" Label="PropertySheets">
+    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
+  </ImportGroup>
+  <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Release|Win32'">
+    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
+  </ImportGroup>
+  <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'" Label="PropertySheets">
+    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
+  </ImportGroup>
+  <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Debug|x64'">
+    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
+  </ImportGroup>
+  <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'" Label="PropertySheets">
+    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
+  </ImportGroup>
+  <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Release|x64'">
+    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
+  </ImportGroup>
+  <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'" Label="PropertySheets">
+    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
+  </ImportGroup>
+  <PropertyGroup Label="UserMacros" />
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'">
+    <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir>
+    <IntDir>$(Platform)\$(Configuration)\</IntDir>
+    <TargetName>$(ProjectName)d</TargetName>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'">
+    <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir>
+    <IntDir>$(Platform)\$(Configuration)\</IntDir>
+    <TargetName>$(ProjectName)-$(PlatformToolset)-$(Configuration)</TargetName>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'">
+    <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir>
+    <IntDir>$(Platform)\$(Configuration)\</IntDir>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'">
+    <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir>
+    <IntDir>$(Platform)\$(Configuration)\</IntDir>
+    <TargetName>$(ProjectName)-$(PlatformToolset)-$(Configuration)</TargetName>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'">
+    <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir>
+    <IntDir>$(Platform)\$(Configuration)\</IntDir>
+    <TargetName>$(ProjectName)d</TargetName>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'">
+    <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir>
+    <IntDir>$(Platform)\$(Configuration)\</IntDir>
+    <TargetName>$(ProjectName)-vc$(PlatformToolsetVersion)-$(Configuration)</TargetName>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'">
+    <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir>
+    <IntDir>$(Platform)\$(Configuration)\</IntDir>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'">
+    <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir>
+    <IntDir>$(Platform)\$(Configuration)\</IntDir>
+    <TargetName>$(ProjectName)-vc$(PlatformToolsetVersion)-$(Configuration)</TargetName>
+  </PropertyGroup>
+  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'">
+    <ClCompile>
+      <PrecompiledHeader>
+      </PrecompiledHeader>
+      <WarningLevel>Level3</WarningLevel>
+      <Optimization>Disabled</Optimization>
+      <PreprocessorDefinitions>_REENTRANT;_WINDLL;DLLEXPORT;JEMALLOC_DEBUG;_DEBUG;%(PreprocessorDefinitions)</PreprocessorDefinitions>
+      <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
+      <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings>
+      <ProgramDataBaseFileName>$(OutputPath)$(TargetName).pdb</ProgramDataBaseFileName>
+    </ClCompile>
+    <Link>
+      <SubSystem>Windows</SubSystem>
+      <GenerateDebugInformation>true</GenerateDebugInformation>
+    </Link>
+  </ItemDefinitionGroup>
+  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'">
+    <ClCompile>
+      <PrecompiledHeader>
+      </PrecompiledHeader>
+      <WarningLevel>Level3</WarningLevel>
+      <Optimization>Disabled</Optimization>
+      <PreprocessorDefinitions>JEMALLOC_DEBUG;_REENTRANT;JEMALLOC_EXPORT=;_DEBUG;_LIB;%(PreprocessorDefinitions)</PreprocessorDefinitions>
+      <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
+      <RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary>
+      <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings>
+      <ProgramDataBaseFileName>$(OutputPath)$(TargetName).pdb</ProgramDataBaseFileName>
+    </ClCompile>
+    <Link>
+      <SubSystem>Windows</SubSystem>
+      <GenerateDebugInformation>true</GenerateDebugInformation>
+    </Link>
+  </ItemDefinitionGroup>
+  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'">
+    <ClCompile>
+      <PrecompiledHeader>
+      </PrecompiledHeader>
+      <WarningLevel>Level3</WarningLevel>
+      <Optimization>Disabled</Optimization>
+      <PreprocessorDefinitions>_REENTRANT;_WINDLL;DLLEXPORT;JEMALLOC_DEBUG;_DEBUG;%(PreprocessorDefinitions)</PreprocessorDefinitions>
+      <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
+      <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings>
+      <ProgramDataBaseFileName>$(OutputPath)$(TargetName).pdb</ProgramDataBaseFileName>
+    </ClCompile>
+    <Link>
+      <SubSystem>Windows</SubSystem>
+      <GenerateDebugInformation>true</GenerateDebugInformation>
+    </Link>
+  </ItemDefinitionGroup>
+  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'">
+    <ClCompile>
+      <PrecompiledHeader>
+      </PrecompiledHeader>
+      <WarningLevel>Level3</WarningLevel>
+      <Optimization>Disabled</Optimization>
+      <PreprocessorDefinitions>JEMALLOC_DEBUG;_REENTRANT;JEMALLOC_EXPORT=;_DEBUG;_LIB;%(PreprocessorDefinitions)</PreprocessorDefinitions>
+      <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
+      <RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary>
+      <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings>
+      <DebugInformationFormat>OldStyle</DebugInformationFormat>
+      <MinimalRebuild>false</MinimalRebuild>
+    </ClCompile>
+    <Link>
+      <SubSystem>Windows</SubSystem>
+      <GenerateDebugInformation>true</GenerateDebugInformation>
+    </Link>
+  </ItemDefinitionGroup>
+  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'">
+    <ClCompile>
+      <WarningLevel>Level3</WarningLevel>
+      <PrecompiledHeader>
+      </PrecompiledHeader>
+      <Optimization>MaxSpeed</Optimization>
+      <FunctionLevelLinking>true</FunctionLevelLinking>
+      <IntrinsicFunctions>true</IntrinsicFunctions>
+      <PreprocessorDefinitions>_REENTRANT;_WINDLL;DLLEXPORT;NDEBUG;%(PreprocessorDefinitions)</PreprocessorDefinitions>
+      <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
+      <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings>
+      <ProgramDataBaseFileName>$(OutputPath)$(TargetName).pdb</ProgramDataBaseFileName>
+    </ClCompile>
+    <Link>
+      <SubSystem>Windows</SubSystem>
+      <GenerateDebugInformation>true</GenerateDebugInformation>
+      <EnableCOMDATFolding>true</EnableCOMDATFolding>
+      <OptimizeReferences>true</OptimizeReferences>
+    </Link>
+  </ItemDefinitionGroup>
+  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'">
+    <ClCompile>
+      <WarningLevel>Level3</WarningLevel>
+      <PrecompiledHeader>
+      </PrecompiledHeader>
+      <Optimization>MaxSpeed</Optimization>
+      <FunctionLevelLinking>true</FunctionLevelLinking>
+      <IntrinsicFunctions>true</IntrinsicFunctions>
+      <PreprocessorDefinitions>_REENTRANT;JEMALLOC_EXPORT=;NDEBUG;_LIB;%(PreprocessorDefinitions)</PreprocessorDefinitions>
+      <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
+      <RuntimeLibrary>MultiThreaded</RuntimeLibrary>
+      <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings>
+      <ProgramDataBaseFileName>$(OutputPath)$(TargetName).pdb</ProgramDataBaseFileName>
+    </ClCompile>
+    <Link>
+      <SubSystem>Windows</SubSystem>
+      <GenerateDebugInformation>true</GenerateDebugInformation>
+      <EnableCOMDATFolding>true</EnableCOMDATFolding>
+      <OptimizeReferences>true</OptimizeReferences>
+    </Link>
+  </ItemDefinitionGroup>
+  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'">
+    <ClCompile>
+      <WarningLevel>Level3</WarningLevel>
+      <PrecompiledHeader>
+      </PrecompiledHeader>
+      <Optimization>MaxSpeed</Optimization>
+      <FunctionLevelLinking>true</FunctionLevelLinking>
+      <IntrinsicFunctions>true</IntrinsicFunctions>
+      <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
+      <PreprocessorDefinitions>_REENTRANT;_WINDLL;DLLEXPORT;NDEBUG;%(PreprocessorDefinitions)</PreprocessorDefinitions>
+      <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings>
+      <ProgramDataBaseFileName>$(OutputPath)$(TargetName).pdb</ProgramDataBaseFileName>
+    </ClCompile>
+    <Link>
+      <SubSystem>Windows</SubSystem>
+      <GenerateDebugInformation>true</GenerateDebugInformation>
+      <EnableCOMDATFolding>true</EnableCOMDATFolding>
+      <OptimizeReferences>true</OptimizeReferences>
+    </Link>
+  </ItemDefinitionGroup>
+  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'">
+    <ClCompile>
+      <WarningLevel>Level3</WarningLevel>
+      <PrecompiledHeader>
+      </PrecompiledHeader>
+      <Optimization>MaxSpeed</Optimization>
+      <FunctionLevelLinking>true</FunctionLevelLinking>
+      <IntrinsicFunctions>true</IntrinsicFunctions>
+      <PreprocessorDefinitions>_REENTRANT;JEMALLOC_EXPORT=;NDEBUG;_LIB;%(PreprocessorDefinitions)</PreprocessorDefinitions>
+      <AdditionalIncludeDirectories>..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
+      <RuntimeLibrary>MultiThreaded</RuntimeLibrary>
+      <DisableSpecificWarnings>4090;4146;4267;4334</DisableSpecificWarnings>
+      <DebugInformationFormat>OldStyle</DebugInformationFormat>
+    </ClCompile>
+    <Link>
+      <SubSystem>Windows</SubSystem>
+      <GenerateDebugInformation>true</GenerateDebugInformation>
+      <EnableCOMDATFolding>true</EnableCOMDATFolding>
+      <OptimizeReferences>true</OptimizeReferences>
+    </Link>
+  </ItemDefinitionGroup>
+  <Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" />
+  <ImportGroup Label="ExtensionTargets">
+  </ImportGroup>
+</Project>
\ No newline at end of file
diff --git a/deps/jemalloc/msvc/projects/vc2015/jemalloc/jemalloc.vcxproj.filters b/deps/jemalloc/msvc/projects/vc2015/jemalloc/jemalloc.vcxproj.filters
new file mode 100644
index 000000000..37f0f02ae
--- /dev/null
+++ b/deps/jemalloc/msvc/projects/vc2015/jemalloc/jemalloc.vcxproj.filters
@@ -0,0 +1,272 @@
+<?xml version="1.0" encoding="utf-8"?>
+<Project ToolsVersion="4.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
+  <ItemGroup>
+    <Filter Include="Source Files">
+      <UniqueIdentifier>{4FC737F1-C7A5-4376-A066-2A32D752A2FF}</UniqueIdentifier>
+      <Extensions>cpp;c;cc;cxx;def;odl;idl;hpj;bat;asm;asmx</Extensions>
+    </Filter>
+    <Filter Include="Header Files">
+      <UniqueIdentifier>{93995380-89BD-4b04-88EB-625FBE52EBFB}</UniqueIdentifier>
+      <Extensions>h;hh;hpp;hxx;hm;inl;inc;xsd</Extensions>
+    </Filter>
+    <Filter Include="Header Files\internal">
+      <UniqueIdentifier>{5697dfa3-16cf-4932-b428-6e0ec6e9f98e}</UniqueIdentifier>
+    </Filter>
+    <Filter Include="Header Files\msvc_compat">
+      <UniqueIdentifier>{0cbd2ca6-42a7-4f82-8517-d7e7a14fd986}</UniqueIdentifier>
+    </Filter>
+    <Filter Include="Header Files\msvc_compat\C99">
+      <UniqueIdentifier>{0abe6f30-49b5-46dd-8aca-6e33363fa52c}</UniqueIdentifier>
+    </Filter>
+  </ItemGroup>
+  <ItemGroup>
+    <ClInclude Include="..\..\..\..\include\jemalloc\jemalloc.h">
+      <Filter>Header Files</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\jemalloc_defs.h">
+      <Filter>Header Files</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\jemalloc_macros.h">
+      <Filter>Header Files</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\jemalloc_mangle.h">
+      <Filter>Header Files</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\jemalloc_protos.h">
+      <Filter>Header Files</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\jemalloc_protos_jet.h">
+      <Filter>Header Files</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\jemalloc_rename.h">
+      <Filter>Header Files</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\jemalloc_typedefs.h">
+      <Filter>Header Files</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\arena.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\assert.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\atomic.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\base.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\bitmap.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\chunk.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\chunk_dss.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\chunk_mmap.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\ckh.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\ctl.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\extent.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\hash.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\huge.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\jemalloc_internal.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\jemalloc_internal_decls.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\jemalloc_internal_defs.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\jemalloc_internal_macros.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\mb.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\mutex.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\nstime.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\pages.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\ph.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\private_namespace.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\private_unnamespace.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\prng.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\prof.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\public_namespace.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\public_unnamespace.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\ql.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\qr.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\quarantine.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\rb.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\rtree.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\size_classes.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\smoothstep.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\spin.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\stats.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\tcache.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\ticker.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\tsd.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\util.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\jemalloc\internal\witness.h">
+      <Filter>Header Files\internal</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\msvc_compat\strings.h">
+      <Filter>Header Files\msvc_compat</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\msvc_compat\windows_extra.h">
+      <Filter>Header Files\msvc_compat</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\msvc_compat\C99\stdbool.h">
+      <Filter>Header Files\msvc_compat\C99</Filter>
+    </ClInclude>
+    <ClInclude Include="..\..\..\..\include\msvc_compat\C99\stdint.h">
+      <Filter>Header Files\msvc_compat\C99</Filter>
+    </ClInclude>
+  </ItemGroup>
+  <ItemGroup>
+    <ClCompile Include="..\..\..\..\src\arena.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\atomic.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\base.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\bitmap.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\chunk.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\chunk_dss.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\chunk_mmap.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\ckh.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\ctl.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\extent.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\hash.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\huge.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\jemalloc.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\mb.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\mutex.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\nstime.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\pages.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\prng.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\prof.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\quarantine.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\rtree.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\spin.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\stats.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\tcache.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\ticker.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\tsd.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\util.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="..\..\..\..\src\witness.c">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+  </ItemGroup>
+</Project>
diff --git a/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.cpp b/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.cpp
new file mode 100755
index 000000000..a3d1a792a
--- /dev/null
+++ b/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.cpp
@@ -0,0 +1,89 @@
+// jemalloc C++ threaded test
+// Author: Rustam Abdullaev
+// Public Domain
+
+#include <atomic>
+#include <functional>
+#include <future>
+#include <random>
+#include <thread>
+#include <vector>
+#include <stdio.h>
+#include <jemalloc/jemalloc.h>
+
+using std::vector;
+using std::thread;
+using std::uniform_int_distribution;
+using std::minstd_rand;
+
+int test_threads()
+{
+  je_malloc_conf = "narenas:3";
+  int narenas = 0;
+  size_t sz = sizeof(narenas);
+  je_mallctl("opt.narenas", (void *)&narenas, &sz, NULL, 0);
+  if (narenas != 3) {
+    printf("Error: unexpected number of arenas: %d\n", narenas);
+    return 1;
+  }
+  static const int sizes[] = { 7, 16, 32, 60, 91, 100, 120, 144, 169, 199, 255, 400, 670, 900, 917, 1025, 3333, 5190, 13131, 49192, 99999, 123123, 255265, 2333111 };
+  static const int numSizes = (int)(sizeof(sizes) / sizeof(sizes[0]));
+  vector<thread> workers;
+  static const int numThreads = narenas + 1, numAllocsMax = 25, numIter1 = 50, numIter2 = 50;
+  je_malloc_stats_print(NULL, NULL, NULL);
+  size_t allocated1;
+  size_t sz1 = sizeof(allocated1);
+  je_mallctl("stats.active", (void *)&allocated1, &sz1, NULL, 0);
+  printf("\nPress Enter to start threads...\n");
+  getchar();
+  printf("Starting %d threads x %d x %d iterations...\n", numThreads, numIter1, numIter2);
+  for (int i = 0; i < numThreads; i++) {
+    workers.emplace_back([tid=i]() {
+      uniform_int_distribution<int> sizeDist(0, numSizes - 1);
+      minstd_rand rnd(tid * 17);
+      uint8_t* ptrs[numAllocsMax];
+      int ptrsz[numAllocsMax];
+      for (int i = 0; i < numIter1; ++i) {
+        thread t([&]() {
+          for (int i = 0; i < numIter2; ++i) {
+            const int numAllocs = numAllocsMax - sizeDist(rnd);
+            for (int j = 0; j < numAllocs; j += 64) {
+              const int x = sizeDist(rnd);
+              const int sz = sizes[x];
+              ptrsz[j] = sz;
+              ptrs[j] = (uint8_t*)je_malloc(sz);
+              if (!ptrs[j]) {
+                printf("Unable to allocate %d bytes in thread %d, iter %d, alloc %d. %d\n", sz, tid, i, j, x);
+                exit(1);
+              }
+              for (int k = 0; k < sz; k++)
+                ptrs[j][k] = tid + k;
+            }
+            for (int j = 0; j < numAllocs; j += 64) {
+              for (int k = 0, sz = ptrsz[j]; k < sz; k++)
+                if (ptrs[j][k] != (uint8_t)(tid + k)) {
+                  printf("Memory error in thread %d, iter %d, alloc %d @ %d : %02X!=%02X\n", tid, i, j, k, ptrs[j][k], (uint8_t)(tid + k));
+                  exit(1);
+                }
+              je_free(ptrs[j]);
+            }
+          }
+        });
+        t.join();
+      }
+    });
+  }
+  for (thread& t : workers) {
+    t.join();
+  }
+  je_malloc_stats_print(NULL, NULL, NULL);
+  size_t allocated2;
+  je_mallctl("stats.active", (void *)&allocated2, &sz1, NULL, 0);
+  size_t leaked = allocated2 - allocated1;
+  printf("\nDone. Leaked: %zd bytes\n", leaked);
+  bool failed = leaked > 65536; // in case C++ runtime allocated something (e.g. iostream locale or facet)
+  printf("\nTest %s!\n", (failed ? "FAILED" : "successful"));
+  printf("\nPress Enter to continue...\n");
+  getchar();
+  return failed ? 1 : 0;
+}
diff --git a/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.h b/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.h
new file mode 100644
index 000000000..64d0cdb33
--- /dev/null
+++ b/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.h
@@ -0,0 +1,3 @@
+#pragma once
+
+int test_threads();
diff --git a/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.vcxproj b/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.vcxproj
new file mode 100644
index 000000000..f5e9898f2
--- /dev/null
+++ b/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.vcxproj
@@ -0,0 +1,327 @@
+<?xml version="1.0" encoding="utf-8"?>
+<Project DefaultTargets="Build" ToolsVersion="14.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
+  <ItemGroup Label="ProjectConfigurations">
+    <ProjectConfiguration Include="Debug-static|Win32">
+      <Configuration>Debug-static</Configuration>
+      <Platform>Win32</Platform>
+    </ProjectConfiguration>
+    <ProjectConfiguration Include="Debug-static|x64">
+      <Configuration>Debug-static</Configuration>
+      <Platform>x64</Platform>
+    </ProjectConfiguration>
+    <ProjectConfiguration Include="Debug|Win32">
+      <Configuration>Debug</Configuration>
+      <Platform>Win32</Platform>
+    </ProjectConfiguration>
+    <ProjectConfiguration Include="Release-static|Win32">
+      <Configuration>Release-static</Configuration>
+      <Platform>Win32</Platform>
+    </ProjectConfiguration>
+    <ProjectConfiguration Include="Release-static|x64">
+      <Configuration>Release-static</Configuration>
+      <Platform>x64</Platform>
+    </ProjectConfiguration>
+    <ProjectConfiguration Include="Release|Win32">
+      <Configuration>Release</Configuration>
+      <Platform>Win32</Platform>
+    </ProjectConfiguration>
+    <ProjectConfiguration Include="Debug|x64">
+      <Configuration>Debug</Configuration>
+      <Platform>x64</Platform>
+    </ProjectConfiguration>
+    <ProjectConfiguration Include="Release|x64">
+      <Configuration>Release</Configuration>
+      <Platform>x64</Platform>
+    </ProjectConfiguration>
+  </ItemGroup>
+  <PropertyGroup Label="Globals">
+    <ProjectGuid>{09028CFD-4EB7-491D-869C-0708DB97ED44}</ProjectGuid>
+    <Keyword>Win32Proj</Keyword>
+    <RootNamespace>test_threads</RootNamespace>
+    <WindowsTargetPlatformVersion>8.1</WindowsTargetPlatformVersion>
+  </PropertyGroup>
+  <Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" />
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'" Label="Configuration">
+    <ConfigurationType>Application</ConfigurationType>
+    <UseDebugLibraries>true</UseDebugLibraries>
+    <PlatformToolset>v140</PlatformToolset>
+    <CharacterSet>MultiByte</CharacterSet>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'" Label="Configuration">
+    <ConfigurationType>Application</ConfigurationType>
+    <UseDebugLibraries>true</UseDebugLibraries>
+    <PlatformToolset>v140</PlatformToolset>
+    <CharacterSet>MultiByte</CharacterSet>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'" Label="Configuration">
+    <ConfigurationType>Application</ConfigurationType>
+    <UseDebugLibraries>false</UseDebugLibraries>
+    <PlatformToolset>v140</PlatformToolset>
+    <WholeProgramOptimization>true</WholeProgramOptimization>
+    <CharacterSet>MultiByte</CharacterSet>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'" Label="Configuration">
+    <ConfigurationType>Application</ConfigurationType>
+    <UseDebugLibraries>false</UseDebugLibraries>
+    <PlatformToolset>v140</PlatformToolset>
+    <WholeProgramOptimization>true</WholeProgramOptimization>
+    <CharacterSet>MultiByte</CharacterSet>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'" Label="Configuration">
+    <ConfigurationType>Application</ConfigurationType>
+    <UseDebugLibraries>true</UseDebugLibraries>
+    <PlatformToolset>v140</PlatformToolset>
+    <CharacterSet>MultiByte</CharacterSet>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'" Label="Configuration">
+    <ConfigurationType>Application</ConfigurationType>
+    <UseDebugLibraries>true</UseDebugLibraries>
+    <PlatformToolset>v140</PlatformToolset>
+    <CharacterSet>MultiByte</CharacterSet>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'" Label="Configuration">
+    <ConfigurationType>Application</ConfigurationType>
+    <UseDebugLibraries>false</UseDebugLibraries>
+    <PlatformToolset>v140</PlatformToolset>
+    <WholeProgramOptimization>true</WholeProgramOptimization>
+    <CharacterSet>MultiByte</CharacterSet>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'" Label="Configuration">
+    <ConfigurationType>Application</ConfigurationType>
+    <UseDebugLibraries>false</UseDebugLibraries>
+    <PlatformToolset>v140</PlatformToolset>
+    <WholeProgramOptimization>true</WholeProgramOptimization>
+    <CharacterSet>MultiByte</CharacterSet>
+  </PropertyGroup>
+  <Import Project="$(VCTargetsPath)\Microsoft.Cpp.props" />
+  <ImportGroup Label="ExtensionSettings">
+  </ImportGroup>
+  <ImportGroup Label="Shared">
+  </ImportGroup>
+  <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'">
+    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
+  </ImportGroup>
+  <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'" Label="PropertySheets">
+    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
+  </ImportGroup>
+  <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Release|Win32'">
+    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
+  </ImportGroup>
+  <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'" Label="PropertySheets">
+    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
+  </ImportGroup>
+  <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Debug|x64'">
+    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
+  </ImportGroup>
+  <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'" Label="PropertySheets">
+    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
+  </ImportGroup>
+  <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Release|x64'">
+    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
+  </ImportGroup>
+  <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'" Label="PropertySheets">
+    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
+  </ImportGroup>
+  <PropertyGroup Label="UserMacros" />
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'">
+    <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir>
+    <IntDir>$(Platform)\$(Configuration)\</IntDir>
+    <LinkIncremental>true</LinkIncremental>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'">
+    <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir>
+    <IntDir>$(Platform)\$(Configuration)\</IntDir>
+    <LinkIncremental>true</LinkIncremental>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'">
+    <LinkIncremental>true</LinkIncremental>
+    <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'">
+    <LinkIncremental>true</LinkIncremental>
+    <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'">
+    <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir>
+    <IntDir>$(Platform)\$(Configuration)\</IntDir>
+    <LinkIncremental>false</LinkIncremental>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'">
+    <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir>
+    <IntDir>$(Platform)\$(Configuration)\</IntDir>
+    <LinkIncremental>false</LinkIncremental>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'">
+    <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir>
+    <IntDir>$(Platform)\$(Configuration)\</IntDir>
+    <LinkIncremental>false</LinkIncremental>
+  </PropertyGroup>
+  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'">
+    <OutDir>$(SolutionDir)$(Platform)\$(Configuration)\</OutDir>
+    <IntDir>$(Platform)\$(Configuration)\</IntDir>
+    <LinkIncremental>false</LinkIncremental>
+  </PropertyGroup>
+  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'">
+    <ClCompile>
+      <PrecompiledHeader>
+      </PrecompiledHeader>
+      <WarningLevel>Level3</WarningLevel>
+      <Optimization>Disabled</Optimization>
+      <PreprocessorDefinitions>WIN32;_DEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions>
+      <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
+    </ClCompile>
+    <Link>
+      <SubSystem>Console</SubSystem>
+      <GenerateDebugInformation>true</GenerateDebugInformation>
+      <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories>
+      <AdditionalDependencies>jemallocd.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
+    </Link>
+  </ItemDefinitionGroup>
+  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|Win32'">
+    <ClCompile>
+      <PrecompiledHeader>
+      </PrecompiledHeader>
+      <WarningLevel>Level3</WarningLevel>
+      <Optimization>Disabled</Optimization>
+      <PreprocessorDefinitions>JEMALLOC_EXPORT=;JEMALLOC_STATIC;_DEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions>
+      <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
+      <RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary>
+    </ClCompile>
+    <Link>
+      <SubSystem>Console</SubSystem>
+      <GenerateDebugInformation>true</GenerateDebugInformation>
+      <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories>
+      <AdditionalDependencies>jemalloc-$(PlatformToolset)-$(Configuration).lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
+    </Link>
+  </ItemDefinitionGroup>
+  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'">
+    <ClCompile>
+      <PrecompiledHeader>
+      </PrecompiledHeader>
+      <WarningLevel>Level3</WarningLevel>
+      <Optimization>Disabled</Optimization>
+      <PreprocessorDefinitions>_DEBUG;%(PreprocessorDefinitions)</PreprocessorDefinitions>
+      <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
+    </ClCompile>
+    <Link>
+      <SubSystem>Console</SubSystem>
+      <GenerateDebugInformation>true</GenerateDebugInformation>
+      <AdditionalDependencies>jemallocd.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
+      <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories>
+    </Link>
+  </ItemDefinitionGroup>
+  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug-static|x64'">
+    <ClCompile>
+      <PrecompiledHeader>
+      </PrecompiledHeader>
+      <WarningLevel>Level3</WarningLevel>
+      <Optimization>Disabled</Optimization>
+      <PreprocessorDefinitions>JEMALLOC_EXPORT=;JEMALLOC_STATIC;_DEBUG;%(PreprocessorDefinitions)</PreprocessorDefinitions>
+      <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
+      <RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary>
+    </ClCompile>
+    <Link>
+      <SubSystem>Console</SubSystem>
+      <GenerateDebugInformation>true</GenerateDebugInformation>
+      <AdditionalDependencies>jemalloc-vc$(PlatformToolsetVersion)-$(Configuration).lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
+      <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories>
+    </Link>
+  </ItemDefinitionGroup>
+  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'">
+    <ClCompile>
+      <WarningLevel>Level3</WarningLevel>
+      <PrecompiledHeader>
+      </PrecompiledHeader>
+      <Optimization>MaxSpeed</Optimization>
+      <FunctionLevelLinking>true</FunctionLevelLinking>
+      <IntrinsicFunctions>true</IntrinsicFunctions>
+      <PreprocessorDefinitions>WIN32;NDEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions>
+      <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
+    </ClCompile>
+    <Link>
+      <SubSystem>Console</SubSystem>
+      <GenerateDebugInformation>true</GenerateDebugInformation>
+      <EnableCOMDATFolding>true</EnableCOMDATFolding>
+      <OptimizeReferences>true</OptimizeReferences>
+      <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories>
+      <AdditionalDependencies>jemalloc.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
+    </Link>
+  </ItemDefinitionGroup>
+  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|Win32'">
+    <ClCompile>
+      <WarningLevel>Level3</WarningLevel>
+      <PrecompiledHeader>
+      </PrecompiledHeader>
+      <Optimization>MaxSpeed</Optimization>
+      <FunctionLevelLinking>true</FunctionLevelLinking>
+      <IntrinsicFunctions>true</IntrinsicFunctions>
+      <PreprocessorDefinitions>JEMALLOC_EXPORT=;JEMALLOC_STATIC;NDEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions>
+      <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
+      <RuntimeLibrary>MultiThreaded</RuntimeLibrary>
+    </ClCompile>
+    <Link>
+      <SubSystem>Console</SubSystem>
+      <GenerateDebugInformation>true</GenerateDebugInformation>
+      <EnableCOMDATFolding>true</EnableCOMDATFolding>
+      <OptimizeReferences>true</OptimizeReferences>
+      <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories>
+      <AdditionalDependencies>jemalloc-$(PlatformToolset)-$(Configuration).lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
+    </Link>
+  </ItemDefinitionGroup>
+  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'">
+    <ClCompile>
+      <WarningLevel>Level3</WarningLevel>
+      <PrecompiledHeader>
+      </PrecompiledHeader>
+      <Optimization>MaxSpeed</Optimization>
+      <FunctionLevelLinking>true</FunctionLevelLinking>
+      <IntrinsicFunctions>true</IntrinsicFunctions>
+      <PreprocessorDefinitions>NDEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions>
+      <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
+    </ClCompile>
+    <Link>
+      <SubSystem>Console</SubSystem>
+      <GenerateDebugInformation>true</GenerateDebugInformation>
+      <EnableCOMDATFolding>true</EnableCOMDATFolding>
+      <OptimizeReferences>true</OptimizeReferences>
+      <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories>
+      <AdditionalDependencies>jemalloc.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
+    </Link>
+  </ItemDefinitionGroup>
+  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release-static|x64'">
+    <ClCompile>
+      <WarningLevel>Level3</WarningLevel>
+      <PrecompiledHeader>
+      </PrecompiledHeader>
+      <Optimization>MaxSpeed</Optimization>
+      <FunctionLevelLinking>true</FunctionLevelLinking>
+      <IntrinsicFunctions>true</IntrinsicFunctions>
+      <PreprocessorDefinitions>JEMALLOC_EXPORT=;JEMALLOC_STATIC;NDEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions>
+      <AdditionalIncludeDirectories>..\..\..\..\test\include;..\..\..\..\include;..\..\..\..\include\msvc_compat;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
+      <RuntimeLibrary>MultiThreaded</RuntimeLibrary>
+    </ClCompile>
+    <Link>
+      <SubSystem>Console</SubSystem>
+      <GenerateDebugInformation>true</GenerateDebugInformation>
+      <EnableCOMDATFolding>true</EnableCOMDATFolding>
+      <OptimizeReferences>true</OptimizeReferences>
+      <AdditionalLibraryDirectories>$(SolutionDir)$(Platform)\$(Configuration)</AdditionalLibraryDirectories>
+      <AdditionalDependencies>jemalloc-vc$(PlatformToolsetVersion)-$(Configuration).lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
+    </Link>
+  </ItemDefinitionGroup>
+  <ItemGroup>
+    <ClCompile Include="test_threads.cpp" />
+    <ClCompile Include="test_threads_main.cpp" />
+  </ItemGroup>
+  <ItemGroup>
+    <ProjectReference Include="..\jemalloc\jemalloc.vcxproj">
+      <Project>{8d6bb292-9e1c-413d-9f98-4864bdc1514a}</Project>
+    </ProjectReference>
+  </ItemGroup>
+  <ItemGroup>
+    <ClInclude Include="test_threads.h" />
+  </ItemGroup>
+  <Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" />
+  <ImportGroup Label="ExtensionTargets">
+  </ImportGroup>
+</Project>
\ No newline at end of file
diff --git a/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.vcxproj.filters b/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.vcxproj.filters
new file mode 100644
index 000000000..4c2334073
--- /dev/null
+++ b/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads.vcxproj.filters
@@ -0,0 +1,26 @@
+<?xml version="1.0" encoding="utf-8"?>
+<Project ToolsVersion="4.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
+  <ItemGroup>
+    <Filter Include="Source Files">
+      <UniqueIdentifier>{4FC737F1-C7A5-4376-A066-2A32D752A2FF}</UniqueIdentifier>
+      <Extensions>cpp;c;cc;cxx;def;odl;idl;hpj;bat;asm;asmx</Extensions>
+    </Filter>
+    <Filter Include="Header Files">
+      <UniqueIdentifier>{93995380-89BD-4b04-88EB-625FBE52EBFB}</UniqueIdentifier>
+      <Extensions>h;hh;hpp;hxx;hm;inl;inc;xsd</Extensions>
+    </Filter>
+  </ItemGroup>
+  <ItemGroup>
+    <ClCompile Include="test_threads.cpp">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+    <ClCompile Include="test_threads_main.cpp">
+      <Filter>Source Files</Filter>
+    </ClCompile>
+  </ItemGroup>
+  <ItemGroup>
+    <ClInclude Include="test_threads.h">
+      <Filter>Header Files</Filter>
+    </ClInclude>
+  </ItemGroup>
+</Project>
\ No newline at end of file
diff --git a/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads_main.cpp b/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads_main.cpp
new file mode 100644
index 000000000..ffd96e6ab
--- /dev/null
+++ b/deps/jemalloc/msvc/projects/vc2015/test_threads/test_threads_main.cpp
@@ -0,0 +1,12 @@
+#include "test_threads.h"
+#include <future>
+#include <functional>
+#include <chrono>
+
+using namespace std::chrono_literals;
+
+int main(int argc, char** argv)
+{
+  int rc = test_threads();
+  return rc;
+}
diff --git a/deps/jemalloc/src/arena.c b/deps/jemalloc/src/arena.c
index 3081519cc..648a8da3a 100644
--- a/deps/jemalloc/src/arena.c
+++ b/deps/jemalloc/src/arena.c
@@ -4,16 +4,23 @@
 /******************************************************************************/
 /* Data. */
 
+purge_mode_t	opt_purge = PURGE_DEFAULT;
+const char	*purge_mode_names[] = {
+	"ratio",
+	"decay",
+	"N/A"
+};
 ssize_t		opt_lg_dirty_mult = LG_DIRTY_MULT_DEFAULT;
 static ssize_t	lg_dirty_mult_default;
+ssize_t		opt_decay_time = DECAY_TIME_DEFAULT;
+static ssize_t	decay_time_default;
+
 arena_bin_info_t	arena_bin_info[NBINS];
 
 size_t		map_bias;
 size_t		map_misc_offset;
 size_t		arena_maxrun; /* Max run size for arenas. */
 size_t		large_maxclass; /* Max large size class. */
-static size_t	small_maxrun; /* Max run size used for small size classes. */
-static bool	*small_run_tab; /* Valid small run page multiples. */
 unsigned	nlclasses; /* Number of large size classes. */
 unsigned	nhclasses; /* Number of huge size classes. */
 
@@ -23,60 +30,57 @@ unsigned	nhclasses; /* Number of huge size classes. */
  * definition.
  */
 
-static void	arena_purge(arena_t *arena, bool all);
-static void	arena_run_dalloc(arena_t *arena, arena_run_t *run, bool dirty,
-    bool cleaned, bool decommitted);
-static void	arena_dalloc_bin_run(arena_t *arena, arena_chunk_t *chunk,
-    arena_run_t *run, arena_bin_t *bin);
-static void	arena_bin_lower_run(arena_t *arena, arena_chunk_t *chunk,
-    arena_run_t *run, arena_bin_t *bin);
+static void	arena_chunk_dalloc(tsdn_t *tsdn, arena_t *arena,
+    arena_chunk_t *chunk);
+static void	arena_purge_to_limit(tsdn_t *tsdn, arena_t *arena,
+    size_t ndirty_limit);
+static void	arena_run_dalloc(tsdn_t *tsdn, arena_t *arena, arena_run_t *run,
+    bool dirty, bool cleaned, bool decommitted);
+static void	arena_dalloc_bin_run(tsdn_t *tsdn, arena_t *arena,
+    arena_chunk_t *chunk, arena_run_t *run, arena_bin_t *bin);
+static void	arena_bin_lower_run(arena_t *arena, arena_run_t *run,
+    arena_bin_t *bin);
 
 /******************************************************************************/
 
-#define	CHUNK_MAP_KEY		((uintptr_t)0x1U)
-
-JEMALLOC_INLINE_C arena_chunk_map_misc_t *
-arena_miscelm_key_create(size_t size)
+JEMALLOC_INLINE_C size_t
+arena_miscelm_size_get(const arena_chunk_map_misc_t *miscelm)
 {
+	arena_chunk_t *chunk;
+	size_t pageind, mapbits;
 
-	return ((arena_chunk_map_misc_t *)(arena_mapbits_size_encode(size) |
-	    CHUNK_MAP_KEY));
+	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(miscelm);
+	pageind = arena_miscelm_to_pageind(miscelm);
+	mapbits = arena_mapbits_get(chunk, pageind);
+	return (arena_mapbits_size_decode(mapbits));
 }
 
-JEMALLOC_INLINE_C bool
-arena_miscelm_is_key(const arena_chunk_map_misc_t *miscelm)
+JEMALLOC_INLINE_C const extent_node_t *
+arena_miscelm_extent_get(const arena_chunk_map_misc_t *miscelm)
 {
+	arena_chunk_t *chunk;
 
-	return (((uintptr_t)miscelm & CHUNK_MAP_KEY) != 0);
+	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(miscelm);
+	return (&chunk->node);
 }
 
-#undef CHUNK_MAP_KEY
-
-JEMALLOC_INLINE_C size_t
-arena_miscelm_key_size_get(const arena_chunk_map_misc_t *miscelm)
+JEMALLOC_INLINE_C int
+arena_sn_comp(const arena_chunk_map_misc_t *a, const arena_chunk_map_misc_t *b)
 {
+	size_t a_sn, b_sn;
 
-	assert(arena_miscelm_is_key(miscelm));
-
-	return (arena_mapbits_size_decode((uintptr_t)miscelm));
-}
-
-JEMALLOC_INLINE_C size_t
-arena_miscelm_size_get(arena_chunk_map_misc_t *miscelm)
-{
-	arena_chunk_t *chunk;
-	size_t pageind, mapbits;
+	assert(a != NULL);
+	assert(b != NULL);
 
-	assert(!arena_miscelm_is_key(miscelm));
+	a_sn = extent_node_sn_get(arena_miscelm_extent_get(a));
+	b_sn = extent_node_sn_get(arena_miscelm_extent_get(b));
 
-	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(miscelm);
-	pageind = arena_miscelm_to_pageind(miscelm);
-	mapbits = arena_mapbits_get(chunk, pageind);
-	return (arena_mapbits_size_decode(mapbits));
+	return ((a_sn > b_sn) - (a_sn < b_sn));
 }
 
 JEMALLOC_INLINE_C int
-arena_run_comp(arena_chunk_map_misc_t *a, arena_chunk_map_misc_t *b)
+arena_ad_comp(const arena_chunk_map_misc_t *a,
+    const arena_chunk_map_misc_t *b)
 {
 	uintptr_t a_miscelm = (uintptr_t)a;
 	uintptr_t b_miscelm = (uintptr_t)b;
@@ -87,74 +91,79 @@ arena_run_comp(arena_chunk_map_misc_t *a, arena_chunk_map_misc_t *b)
 	return ((a_miscelm > b_miscelm) - (a_miscelm < b_miscelm));
 }
 
-/* Generate red-black tree functions. */
-rb_gen(static UNUSED, arena_run_tree_, arena_run_tree_t, arena_chunk_map_misc_t,
-    rb_link, arena_run_comp)
-
-static size_t
-run_quantize(size_t size)
+JEMALLOC_INLINE_C int
+arena_snad_comp(const arena_chunk_map_misc_t *a,
+    const arena_chunk_map_misc_t *b)
 {
-	size_t qsize;
+	int ret;
 
-	assert(size != 0);
-	assert(size == PAGE_CEILING(size));
+	assert(a != NULL);
+	assert(b != NULL);
 
-	/* Don't change sizes that are valid small run sizes. */
-	if (size <= small_maxrun && small_run_tab[size >> LG_PAGE])
-		return (size);
+	ret = arena_sn_comp(a, b);
+	if (ret != 0)
+		return (ret);
 
-	/*
-	 * Round down to the nearest run size that can actually be requested
-	 * during normal large allocation.  Add large_pad so that cache index
-	 * randomization can offset the allocation from the page boundary.
-	 */
-	qsize = index2size(size2index(size - large_pad + 1) - 1) + large_pad;
-	if (qsize <= SMALL_MAXCLASS + large_pad)
-		return (run_quantize(size - large_pad));
-	assert(qsize <= size);
-	return (qsize);
+	ret = arena_ad_comp(a, b);
+	return (ret);
 }
 
+/* Generate pairing heap functions. */
+ph_gen(static UNUSED, arena_run_heap_, arena_run_heap_t, arena_chunk_map_misc_t,
+    ph_link, arena_snad_comp)
+
+#ifdef JEMALLOC_JET
+#undef run_quantize_floor
+#define	run_quantize_floor JEMALLOC_N(n_run_quantize_floor)
+#endif
 static size_t
-run_quantize_next(size_t size)
+run_quantize_floor(size_t size)
 {
-	size_t large_run_size_next;
+	size_t ret;
+	pszind_t pind;
+
+	assert(size > 0);
+	assert(size <= HUGE_MAXCLASS);
+	assert((size & PAGE_MASK) == 0);
 
 	assert(size != 0);
 	assert(size == PAGE_CEILING(size));
 
-	/*
-	 * Return the next quantized size greater than the input size.
-	 * Quantized sizes comprise the union of run sizes that back small
-	 * region runs, and run sizes that back large regions with no explicit
-	 * alignment constraints.
-	 */
-
-	if (size > SMALL_MAXCLASS) {
-		large_run_size_next = PAGE_CEILING(index2size(size2index(size -
-		    large_pad) + 1) + large_pad);
-	} else
-		large_run_size_next = SIZE_T_MAX;
-	if (size >= small_maxrun)
-		return (large_run_size_next);
-
-	while (true) {
-		size += PAGE;
-		assert(size <= small_maxrun);
-		if (small_run_tab[size >> LG_PAGE]) {
-			if (large_run_size_next < size)
-				return (large_run_size_next);
-			return (size);
-		}
+	pind = psz2ind(size - large_pad + 1);
+	if (pind == 0) {
+		/*
+		 * Avoid underflow.  This short-circuit would also do the right
+		 * thing for all sizes in the range for which there are
+		 * PAGE-spaced size classes, but it's simplest to just handle
+		 * the one case that would cause erroneous results.
+		 */
+		return (size);
 	}
+	ret = pind2sz(pind - 1) + large_pad;
+	assert(ret <= size);
+	return (ret);
 }
+#ifdef JEMALLOC_JET
+#undef run_quantize_floor
+#define	run_quantize_floor JEMALLOC_N(run_quantize_floor)
+run_quantize_t *run_quantize_floor = JEMALLOC_N(n_run_quantize_floor);
+#endif
 
+#ifdef JEMALLOC_JET
+#undef run_quantize_ceil
+#define	run_quantize_ceil JEMALLOC_N(n_run_quantize_ceil)
+#endif
 static size_t
-run_quantize_first(size_t size)
+run_quantize_ceil(size_t size)
 {
-	size_t qsize = run_quantize(size);
+	size_t ret;
+
+	assert(size > 0);
+	assert(size <= HUGE_MAXCLASS);
+	assert((size & PAGE_MASK) == 0);
 
-	if (qsize < size) {
+	ret = run_quantize_floor(size);
+	if (ret < size) {
 		/*
 		 * Skip a quantization that may have an adequately large run,
 		 * because under-sized runs may be mixed in.  This only happens
@@ -163,72 +172,50 @@ run_quantize_first(size_t size)
 		 * search would potentially find sufficiently aligned available
 		 * memory somewhere lower.
 		 */
-		qsize = run_quantize_next(size);
+		ret = pind2sz(psz2ind(ret - large_pad + 1)) + large_pad;
 	}
-	return (qsize);
-}
-
-JEMALLOC_INLINE_C int
-arena_avail_comp(arena_chunk_map_misc_t *a, arena_chunk_map_misc_t *b)
-{
-	int ret;
-	uintptr_t a_miscelm = (uintptr_t)a;
-	size_t a_qsize = run_quantize(arena_miscelm_is_key(a) ?
-	    arena_miscelm_key_size_get(a) : arena_miscelm_size_get(a));
-	size_t b_qsize = run_quantize(arena_miscelm_size_get(b));
-
-	/*
-	 * Compare based on quantized size rather than size, in order to sort
-	 * equally useful runs only by address.
-	 */
-	ret = (a_qsize > b_qsize) - (a_qsize < b_qsize);
-	if (ret == 0) {
-		if (!arena_miscelm_is_key(a)) {
-			uintptr_t b_miscelm = (uintptr_t)b;
-
-			ret = (a_miscelm > b_miscelm) - (a_miscelm < b_miscelm);
-		} else {
-			/*
-			 * Treat keys as if they are lower than anything else.
-			 */
-			ret = -1;
-		}
-	}
-
 	return (ret);
 }
-
-/* Generate red-black tree functions. */
-rb_gen(static UNUSED, arena_avail_tree_, arena_avail_tree_t,
-    arena_chunk_map_misc_t, rb_link, arena_avail_comp)
+#ifdef JEMALLOC_JET
+#undef run_quantize_ceil
+#define	run_quantize_ceil JEMALLOC_N(run_quantize_ceil)
+run_quantize_t *run_quantize_ceil = JEMALLOC_N(n_run_quantize_ceil);
+#endif
 
 static void
 arena_avail_insert(arena_t *arena, arena_chunk_t *chunk, size_t pageind,
     size_t npages)
 {
-
+	pszind_t pind = psz2ind(run_quantize_floor(arena_miscelm_size_get(
+	    arena_miscelm_get_const(chunk, pageind))));
 	assert(npages == (arena_mapbits_unallocated_size_get(chunk, pageind) >>
 	    LG_PAGE));
-	arena_avail_tree_insert(&arena->runs_avail, arena_miscelm_get(chunk,
-	    pageind));
+	assert((npages << LG_PAGE) < chunksize);
+	assert(pind2sz(pind) <= chunksize);
+	arena_run_heap_insert(&arena->runs_avail[pind],
+	    arena_miscelm_get_mutable(chunk, pageind));
 }
 
 static void
 arena_avail_remove(arena_t *arena, arena_chunk_t *chunk, size_t pageind,
     size_t npages)
 {
-
+	pszind_t pind = psz2ind(run_quantize_floor(arena_miscelm_size_get(
+	    arena_miscelm_get_const(chunk, pageind))));
 	assert(npages == (arena_mapbits_unallocated_size_get(chunk, pageind) >>
 	    LG_PAGE));
-	arena_avail_tree_remove(&arena->runs_avail, arena_miscelm_get(chunk,
-	    pageind));
+	assert((npages << LG_PAGE) < chunksize);
+	assert(pind2sz(pind) <= chunksize);
+	arena_run_heap_remove(&arena->runs_avail[pind],
+	    arena_miscelm_get_mutable(chunk, pageind));
 }
 
 static void
 arena_run_dirty_insert(arena_t *arena, arena_chunk_t *chunk, size_t pageind,
     size_t npages)
 {
-	arena_chunk_map_misc_t *miscelm = arena_miscelm_get(chunk, pageind);
+	arena_chunk_map_misc_t *miscelm = arena_miscelm_get_mutable(chunk,
+	    pageind);
 
 	assert(npages == (arena_mapbits_unallocated_size_get(chunk, pageind) >>
 	    LG_PAGE));
@@ -245,7 +232,8 @@ static void
 arena_run_dirty_remove(arena_t *arena, arena_chunk_t *chunk, size_t pageind,
     size_t npages)
 {
-	arena_chunk_map_misc_t *miscelm = arena_miscelm_get(chunk, pageind);
+	arena_chunk_map_misc_t *miscelm = arena_miscelm_get_mutable(chunk,
+	    pageind);
 
 	assert(npages == (arena_mapbits_unallocated_size_get(chunk, pageind) >>
 	    LG_PAGE));
@@ -292,14 +280,14 @@ JEMALLOC_INLINE_C void *
 arena_run_reg_alloc(arena_run_t *run, arena_bin_info_t *bin_info)
 {
 	void *ret;
-	unsigned regind;
+	size_t regind;
 	arena_chunk_map_misc_t *miscelm;
 	void *rpages;
 
 	assert(run->nfree > 0);
 	assert(!bitmap_full(run->bitmap, &bin_info->bitmap_info));
 
-	regind = bitmap_sfu(run->bitmap, &bin_info->bitmap_info);
+	regind = (unsigned)bitmap_sfu(run->bitmap, &bin_info->bitmap_info);
 	miscelm = arena_run_to_miscelm(run);
 	rpages = arena_miscelm_to_rpages(miscelm);
 	ret = (void *)((uintptr_t)rpages + (uintptr_t)bin_info->reg0_offset +
@@ -316,7 +304,7 @@ arena_run_reg_dalloc(arena_run_t *run, void *ptr)
 	size_t mapbits = arena_mapbits_get(chunk, pageind);
 	szind_t binind = arena_ptr_small_binind_get(ptr, mapbits);
 	arena_bin_info_t *bin_info = &arena_bin_info[binind];
-	unsigned regind = arena_run_regind(run, bin_info, ptr);
+	size_t regind = arena_run_regind(run, bin_info, ptr);
 
 	assert(run->nfree < bin_info->nregs);
 	/* Freeing an interior pointer can cause assertion failure. */
@@ -364,16 +352,30 @@ arena_run_page_validate_zeroed(arena_chunk_t *chunk, size_t run_ind)
 }
 
 static void
-arena_cactive_update(arena_t *arena, size_t add_pages, size_t sub_pages)
+arena_nactive_add(arena_t *arena, size_t add_pages)
 {
 
 	if (config_stats) {
-		ssize_t cactive_diff = CHUNK_CEILING((arena->nactive + add_pages
-		    - sub_pages) << LG_PAGE) - CHUNK_CEILING(arena->nactive <<
+		size_t cactive_add = CHUNK_CEILING((arena->nactive +
+		    add_pages) << LG_PAGE) - CHUNK_CEILING(arena->nactive <<
 		    LG_PAGE);
-		if (cactive_diff != 0)
-			stats_cactive_add(cactive_diff);
+		if (cactive_add != 0)
+			stats_cactive_add(cactive_add);
 	}
+	arena->nactive += add_pages;
+}
+
+static void
+arena_nactive_sub(arena_t *arena, size_t sub_pages)
+{
+
+	if (config_stats) {
+		size_t cactive_sub = CHUNK_CEILING(arena->nactive << LG_PAGE) -
+		    CHUNK_CEILING((arena->nactive - sub_pages) << LG_PAGE);
+		if (cactive_sub != 0)
+			stats_cactive_sub(cactive_sub);
+	}
+	arena->nactive -= sub_pages;
 }
 
 static void
@@ -394,8 +396,7 @@ arena_run_split_remove(arena_t *arena, arena_chunk_t *chunk, size_t run_ind,
 	arena_avail_remove(arena, chunk, run_ind, total_pages);
 	if (flag_dirty != 0)
 		arena_run_dirty_remove(arena, chunk, run_ind, total_pages);
-	arena_cactive_update(arena, need_pages, 0);
-	arena->nactive += need_pages;
+	arena_nactive_add(arena, need_pages);
 
 	/* Keep track of trailing unused pages for later use. */
 	if (rem_pages > 0) {
@@ -567,7 +568,8 @@ arena_chunk_init_spare(arena_t *arena)
 }
 
 static bool
-arena_chunk_register(arena_t *arena, arena_chunk_t *chunk, bool zero)
+arena_chunk_register(tsdn_t *tsdn, arena_t *arena, arena_chunk_t *chunk,
+    size_t sn, bool zero)
 {
 
 	/*
@@ -576,64 +578,67 @@ arena_chunk_register(arena_t *arena, arena_chunk_t *chunk, bool zero)
 	 * of runs is tracked individually, and upon chunk deallocation the
 	 * entire chunk is in a consistent commit state.
 	 */
-	extent_node_init(&chunk->node, arena, chunk, chunksize, zero, true);
+	extent_node_init(&chunk->node, arena, chunk, chunksize, sn, zero, true);
 	extent_node_achunk_set(&chunk->node, true);
-	return (chunk_register(chunk, &chunk->node));
+	return (chunk_register(tsdn, chunk, &chunk->node));
 }
 
 static arena_chunk_t *
-arena_chunk_alloc_internal_hard(arena_t *arena, chunk_hooks_t *chunk_hooks,
-    bool *zero, bool *commit)
+arena_chunk_alloc_internal_hard(tsdn_t *tsdn, arena_t *arena,
+    chunk_hooks_t *chunk_hooks, bool *zero, bool *commit)
 {
 	arena_chunk_t *chunk;
+	size_t sn;
 
-	malloc_mutex_unlock(&arena->lock);
+	malloc_mutex_unlock(tsdn, &arena->lock);
 
-	chunk = (arena_chunk_t *)chunk_alloc_wrapper(arena, chunk_hooks, NULL,
-	    chunksize, chunksize, zero, commit);
+	chunk = (arena_chunk_t *)chunk_alloc_wrapper(tsdn, arena, chunk_hooks,
+	    NULL, chunksize, chunksize, &sn, zero, commit);
 	if (chunk != NULL && !*commit) {
 		/* Commit header. */
 		if (chunk_hooks->commit(chunk, chunksize, 0, map_bias <<
 		    LG_PAGE, arena->ind)) {
-			chunk_dalloc_wrapper(arena, chunk_hooks,
-			    (void *)chunk, chunksize, *commit);
+			chunk_dalloc_wrapper(tsdn, arena, chunk_hooks,
+			    (void *)chunk, chunksize, sn, *zero, *commit);
 			chunk = NULL;
 		}
 	}
-	if (chunk != NULL && arena_chunk_register(arena, chunk, *zero)) {
+	if (chunk != NULL && arena_chunk_register(tsdn, arena, chunk, sn,
+	    *zero)) {
 		if (!*commit) {
 			/* Undo commit of header. */
 			chunk_hooks->decommit(chunk, chunksize, 0, map_bias <<
 			    LG_PAGE, arena->ind);
 		}
-		chunk_dalloc_wrapper(arena, chunk_hooks, (void *)chunk,
-		    chunksize, *commit);
+		chunk_dalloc_wrapper(tsdn, arena, chunk_hooks, (void *)chunk,
+		    chunksize, sn, *zero, *commit);
 		chunk = NULL;
 	}
 
-	malloc_mutex_lock(&arena->lock);
+	malloc_mutex_lock(tsdn, &arena->lock);
 	return (chunk);
 }
 
 static arena_chunk_t *
-arena_chunk_alloc_internal(arena_t *arena, bool *zero, bool *commit)
+arena_chunk_alloc_internal(tsdn_t *tsdn, arena_t *arena, bool *zero,
+    bool *commit)
 {
 	arena_chunk_t *chunk;
 	chunk_hooks_t chunk_hooks = CHUNK_HOOKS_INITIALIZER;
+	size_t sn;
 
-	chunk = chunk_alloc_cache(arena, &chunk_hooks, NULL, chunksize,
-	    chunksize, zero, true);
+	chunk = chunk_alloc_cache(tsdn, arena, &chunk_hooks, NULL, chunksize,
+	    chunksize, &sn, zero, commit, true);
 	if (chunk != NULL) {
-		if (arena_chunk_register(arena, chunk, *zero)) {
-			chunk_dalloc_cache(arena, &chunk_hooks, chunk,
-			    chunksize, true);
+		if (arena_chunk_register(tsdn, arena, chunk, sn, *zero)) {
+			chunk_dalloc_cache(tsdn, arena, &chunk_hooks, chunk,
+			    chunksize, sn, true);
 			return (NULL);
 		}
-		*commit = true;
 	}
 	if (chunk == NULL) {
-		chunk = arena_chunk_alloc_internal_hard(arena, &chunk_hooks,
-		    zero, commit);
+		chunk = arena_chunk_alloc_internal_hard(tsdn, arena,
+		    &chunk_hooks, zero, commit);
 	}
 
 	if (config_stats && chunk != NULL) {
@@ -645,7 +650,7 @@ arena_chunk_alloc_internal(arena_t *arena, bool *zero, bool *commit)
 }
 
 static arena_chunk_t *
-arena_chunk_init_hard(arena_t *arena)
+arena_chunk_init_hard(tsdn_t *tsdn, arena_t *arena)
 {
 	arena_chunk_t *chunk;
 	bool zero, commit;
@@ -655,14 +660,16 @@ arena_chunk_init_hard(arena_t *arena)
 
 	zero = false;
 	commit = false;
-	chunk = arena_chunk_alloc_internal(arena, &zero, &commit);
+	chunk = arena_chunk_alloc_internal(tsdn, arena, &zero, &commit);
 	if (chunk == NULL)
 		return (NULL);
 
+	chunk->hugepage = true;
+
 	/*
 	 * Initialize the map to contain one maximal free untouched run.  Mark
-	 * the pages as zeroed if chunk_alloc() returned a zeroed or decommitted
-	 * chunk.
+	 * the pages as zeroed if arena_chunk_alloc_internal() returned a zeroed
+	 * or decommitted chunk.
 	 */
 	flag_unzeroed = (zero || !commit) ? 0 : CHUNK_MAP_UNZEROED;
 	flag_decommitted = commit ? 0 : CHUNK_MAP_DECOMMITTED;
@@ -674,17 +681,18 @@ arena_chunk_init_hard(arena_t *arena)
 	 */
 	if (!zero) {
 		JEMALLOC_VALGRIND_MAKE_MEM_UNDEFINED(
-		    (void *)arena_bitselm_get(chunk, map_bias+1),
-		    (size_t)((uintptr_t) arena_bitselm_get(chunk,
-		    chunk_npages-1) - (uintptr_t)arena_bitselm_get(chunk,
-		    map_bias+1)));
+		    (void *)arena_bitselm_get_const(chunk, map_bias+1),
+		    (size_t)((uintptr_t)arena_bitselm_get_const(chunk,
+		    chunk_npages-1) -
+		    (uintptr_t)arena_bitselm_get_const(chunk, map_bias+1)));
 		for (i = map_bias+1; i < chunk_npages-1; i++)
 			arena_mapbits_internal_set(chunk, i, flag_unzeroed);
 	} else {
 		JEMALLOC_VALGRIND_MAKE_MEM_DEFINED((void
-		    *)arena_bitselm_get(chunk, map_bias+1), (size_t)((uintptr_t)
-		    arena_bitselm_get(chunk, chunk_npages-1) -
-		    (uintptr_t)arena_bitselm_get(chunk, map_bias+1)));
+		    *)arena_bitselm_get_const(chunk, map_bias+1),
+		    (size_t)((uintptr_t)arena_bitselm_get_const(chunk,
+		    chunk_npages-1) -
+		    (uintptr_t)arena_bitselm_get_const(chunk, map_bias+1)));
 		if (config_debug) {
 			for (i = map_bias+1; i < chunk_npages-1; i++) {
 				assert(arena_mapbits_unzeroed_get(chunk, i) ==
@@ -699,28 +707,85 @@ arena_chunk_init_hard(arena_t *arena)
 }
 
 static arena_chunk_t *
-arena_chunk_alloc(arena_t *arena)
+arena_chunk_alloc(tsdn_t *tsdn, arena_t *arena)
 {
 	arena_chunk_t *chunk;
 
 	if (arena->spare != NULL)
 		chunk = arena_chunk_init_spare(arena);
 	else {
-		chunk = arena_chunk_init_hard(arena);
+		chunk = arena_chunk_init_hard(tsdn, arena);
 		if (chunk == NULL)
 			return (NULL);
 	}
 
-	/* Insert the run into the runs_avail tree. */
+	ql_elm_new(&chunk->node, ql_link);
+	ql_tail_insert(&arena->achunks, &chunk->node, ql_link);
 	arena_avail_insert(arena, chunk, map_bias, chunk_npages-map_bias);
 
 	return (chunk);
 }
 
 static void
-arena_chunk_dalloc(arena_t *arena, arena_chunk_t *chunk)
+arena_chunk_discard(tsdn_t *tsdn, arena_t *arena, arena_chunk_t *chunk)
+{
+	size_t sn, hugepage;
+	bool committed;
+	chunk_hooks_t chunk_hooks = CHUNK_HOOKS_INITIALIZER;
+
+	chunk_deregister(chunk, &chunk->node);
+
+	sn = extent_node_sn_get(&chunk->node);
+	hugepage = chunk->hugepage;
+	committed = (arena_mapbits_decommitted_get(chunk, map_bias) == 0);
+	if (!committed) {
+		/*
+		 * Decommit the header.  Mark the chunk as decommitted even if
+		 * header decommit fails, since treating a partially committed
+		 * chunk as committed has a high potential for causing later
+		 * access of decommitted memory.
+		 */
+		chunk_hooks = chunk_hooks_get(tsdn, arena);
+		chunk_hooks.decommit(chunk, chunksize, 0, map_bias << LG_PAGE,
+		    arena->ind);
+	}
+	if (!hugepage) {
+		/*
+		 * Convert chunk back to the default state, so that all
+		 * subsequent chunk allocations start out with chunks that can
+		 * be backed by transparent huge pages.
+		 */
+		pages_huge(chunk, chunksize);
+	}
+
+	chunk_dalloc_cache(tsdn, arena, &chunk_hooks, (void *)chunk, chunksize,
+	    sn, committed);
+
+	if (config_stats) {
+		arena->stats.mapped -= chunksize;
+		arena->stats.metadata_mapped -= (map_bias << LG_PAGE);
+	}
+}
+
+static void
+arena_spare_discard(tsdn_t *tsdn, arena_t *arena, arena_chunk_t *spare)
 {
 
+	assert(arena->spare != spare);
+
+	if (arena_mapbits_dirty_get(spare, map_bias) != 0) {
+		arena_run_dirty_remove(arena, spare, map_bias,
+		    chunk_npages-map_bias);
+	}
+
+	arena_chunk_discard(tsdn, arena, spare);
+}
+
+static void
+arena_chunk_dalloc(tsdn_t *tsdn, arena_t *arena, arena_chunk_t *chunk)
+{
+	arena_chunk_t *spare;
+
 	assert(arena_mapbits_allocated_get(chunk, map_bias) == 0);
 	assert(arena_mapbits_allocated_get(chunk, chunk_npages-1) == 0);
 	assert(arena_mapbits_unallocated_size_get(chunk, map_bias) ==
@@ -732,49 +797,14 @@ arena_chunk_dalloc(arena_t *arena, arena_chunk_t *chunk)
 	assert(arena_mapbits_decommitted_get(chunk, map_bias) ==
 	    arena_mapbits_decommitted_get(chunk, chunk_npages-1));
 
-	/*
-	 * Remove run from the runs_avail tree, so that the arena does not use
-	 * it.
-	 */
+	/* Remove run from runs_avail, so that the arena does not use it. */
 	arena_avail_remove(arena, chunk, map_bias, chunk_npages-map_bias);
 
-	if (arena->spare != NULL) {
-		arena_chunk_t *spare = arena->spare;
-		chunk_hooks_t chunk_hooks = CHUNK_HOOKS_INITIALIZER;
-		bool committed;
-
-		arena->spare = chunk;
-		if (arena_mapbits_dirty_get(spare, map_bias) != 0) {
-			arena_run_dirty_remove(arena, spare, map_bias,
-			    chunk_npages-map_bias);
-		}
-
-		chunk_deregister(spare, &spare->node);
-
-		committed = (arena_mapbits_decommitted_get(spare, map_bias) ==
-		    0);
-		if (!committed) {
-			/*
-			 * Decommit the header.  Mark the chunk as decommitted
-			 * even if header decommit fails, since treating a
-			 * partially committed chunk as committed has a high
-			 * potential for causing later access of decommitted
-			 * memory.
-			 */
-			chunk_hooks = chunk_hooks_get(arena);
-			chunk_hooks.decommit(spare, chunksize, 0, map_bias <<
-			    LG_PAGE, arena->ind);
-		}
-
-		chunk_dalloc_cache(arena, &chunk_hooks, (void *)spare,
-		    chunksize, committed);
-
-		if (config_stats) {
-			arena->stats.mapped -= chunksize;
-			arena->stats.metadata_mapped -= (map_bias << LG_PAGE);
-		}
-	} else
-		arena->spare = chunk;
+	ql_remove(&arena->achunks, &chunk->node, ql_link);
+	spare = arena->spare;
+	arena->spare = chunk;
+	if (spare != NULL)
+		arena_spare_discard(tsdn, arena, spare);
 }
 
 static void
@@ -816,6 +846,17 @@ arena_huge_dalloc_stats_update(arena_t *arena, size_t usize)
 	arena->stats.hstats[index].curhchunks--;
 }
 
+static void
+arena_huge_reset_stats_cancel(arena_t *arena, size_t usize)
+{
+	szind_t index = size2index(usize) - nlclasses - NBINS;
+
+	cassert(config_stats);
+
+	arena->stats.ndalloc_huge++;
+	arena->stats.hstats[index].ndalloc--;
+}
+
 static void
 arena_huge_dalloc_stats_update_undo(arena_t *arena, size_t usize)
 {
@@ -847,243 +888,240 @@ arena_huge_ralloc_stats_update_undo(arena_t *arena, size_t oldsize,
 }
 
 extent_node_t *
-arena_node_alloc(arena_t *arena)
+arena_node_alloc(tsdn_t *tsdn, arena_t *arena)
 {
 	extent_node_t *node;
 
-	malloc_mutex_lock(&arena->node_cache_mtx);
+	malloc_mutex_lock(tsdn, &arena->node_cache_mtx);
 	node = ql_last(&arena->node_cache, ql_link);
 	if (node == NULL) {
-		malloc_mutex_unlock(&arena->node_cache_mtx);
-		return (base_alloc(sizeof(extent_node_t)));
+		malloc_mutex_unlock(tsdn, &arena->node_cache_mtx);
+		return (base_alloc(tsdn, sizeof(extent_node_t)));
 	}
 	ql_tail_remove(&arena->node_cache, extent_node_t, ql_link);
-	malloc_mutex_unlock(&arena->node_cache_mtx);
+	malloc_mutex_unlock(tsdn, &arena->node_cache_mtx);
 	return (node);
 }
 
 void
-arena_node_dalloc(arena_t *arena, extent_node_t *node)
+arena_node_dalloc(tsdn_t *tsdn, arena_t *arena, extent_node_t *node)
 {
 
-	malloc_mutex_lock(&arena->node_cache_mtx);
+	malloc_mutex_lock(tsdn, &arena->node_cache_mtx);
 	ql_elm_new(node, ql_link);
 	ql_tail_insert(&arena->node_cache, node, ql_link);
-	malloc_mutex_unlock(&arena->node_cache_mtx);
+	malloc_mutex_unlock(tsdn, &arena->node_cache_mtx);
 }
 
 static void *
-arena_chunk_alloc_huge_hard(arena_t *arena, chunk_hooks_t *chunk_hooks,
-    size_t usize, size_t alignment, bool *zero, size_t csize)
+arena_chunk_alloc_huge_hard(tsdn_t *tsdn, arena_t *arena,
+    chunk_hooks_t *chunk_hooks, size_t usize, size_t alignment, size_t *sn,
+    bool *zero, size_t csize)
 {
 	void *ret;
 	bool commit = true;
 
-	ret = chunk_alloc_wrapper(arena, chunk_hooks, NULL, csize, alignment,
-	    zero, &commit);
+	ret = chunk_alloc_wrapper(tsdn, arena, chunk_hooks, NULL, csize,
+	    alignment, sn, zero, &commit);
 	if (ret == NULL) {
 		/* Revert optimistic stats updates. */
-		malloc_mutex_lock(&arena->lock);
+		malloc_mutex_lock(tsdn, &arena->lock);
 		if (config_stats) {
 			arena_huge_malloc_stats_update_undo(arena, usize);
 			arena->stats.mapped -= usize;
 		}
-		arena->nactive -= (usize >> LG_PAGE);
-		malloc_mutex_unlock(&arena->lock);
+		arena_nactive_sub(arena, usize >> LG_PAGE);
+		malloc_mutex_unlock(tsdn, &arena->lock);
 	}
 
 	return (ret);
 }
 
 void *
-arena_chunk_alloc_huge(arena_t *arena, size_t usize, size_t alignment,
-    bool *zero)
+arena_chunk_alloc_huge(tsdn_t *tsdn, arena_t *arena, size_t usize,
+    size_t alignment, size_t *sn, bool *zero)
 {
 	void *ret;
 	chunk_hooks_t chunk_hooks = CHUNK_HOOKS_INITIALIZER;
 	size_t csize = CHUNK_CEILING(usize);
+	bool commit = true;
 
-	malloc_mutex_lock(&arena->lock);
+	malloc_mutex_lock(tsdn, &arena->lock);
 
 	/* Optimistically update stats. */
 	if (config_stats) {
 		arena_huge_malloc_stats_update(arena, usize);
 		arena->stats.mapped += usize;
 	}
-	arena->nactive += (usize >> LG_PAGE);
+	arena_nactive_add(arena, usize >> LG_PAGE);
 
-	ret = chunk_alloc_cache(arena, &chunk_hooks, NULL, csize, alignment,
-	    zero, true);
-	malloc_mutex_unlock(&arena->lock);
+	ret = chunk_alloc_cache(tsdn, arena, &chunk_hooks, NULL, csize,
+	    alignment, sn, zero, &commit, true);
+	malloc_mutex_unlock(tsdn, &arena->lock);
 	if (ret == NULL) {
-		ret = arena_chunk_alloc_huge_hard(arena, &chunk_hooks, usize,
-		    alignment, zero, csize);
+		ret = arena_chunk_alloc_huge_hard(tsdn, arena, &chunk_hooks,
+		    usize, alignment, sn, zero, csize);
 	}
 
-	if (config_stats && ret != NULL)
-		stats_cactive_add(usize);
 	return (ret);
 }
 
 void
-arena_chunk_dalloc_huge(arena_t *arena, void *chunk, size_t usize)
+arena_chunk_dalloc_huge(tsdn_t *tsdn, arena_t *arena, void *chunk, size_t usize,
+    size_t sn)
 {
 	chunk_hooks_t chunk_hooks = CHUNK_HOOKS_INITIALIZER;
 	size_t csize;
 
 	csize = CHUNK_CEILING(usize);
-	malloc_mutex_lock(&arena->lock);
+	malloc_mutex_lock(tsdn, &arena->lock);
 	if (config_stats) {
 		arena_huge_dalloc_stats_update(arena, usize);
 		arena->stats.mapped -= usize;
-		stats_cactive_sub(usize);
 	}
-	arena->nactive -= (usize >> LG_PAGE);
+	arena_nactive_sub(arena, usize >> LG_PAGE);
 
-	chunk_dalloc_cache(arena, &chunk_hooks, chunk, csize, true);
-	malloc_mutex_unlock(&arena->lock);
+	chunk_dalloc_cache(tsdn, arena, &chunk_hooks, chunk, csize, sn, true);
+	malloc_mutex_unlock(tsdn, &arena->lock);
 }
 
 void
-arena_chunk_ralloc_huge_similar(arena_t *arena, void *chunk, size_t oldsize,
-    size_t usize)
+arena_chunk_ralloc_huge_similar(tsdn_t *tsdn, arena_t *arena, void *chunk,
+    size_t oldsize, size_t usize)
 {
 
 	assert(CHUNK_CEILING(oldsize) == CHUNK_CEILING(usize));
 	assert(oldsize != usize);
 
-	malloc_mutex_lock(&arena->lock);
+	malloc_mutex_lock(tsdn, &arena->lock);
 	if (config_stats)
 		arena_huge_ralloc_stats_update(arena, oldsize, usize);
-	if (oldsize < usize) {
-		size_t udiff = usize - oldsize;
-		arena->nactive += udiff >> LG_PAGE;
-		if (config_stats)
-			stats_cactive_add(udiff);
-	} else {
-		size_t udiff = oldsize - usize;
-		arena->nactive -= udiff >> LG_PAGE;
-		if (config_stats)
-			stats_cactive_sub(udiff);
-	}
-	malloc_mutex_unlock(&arena->lock);
+	if (oldsize < usize)
+		arena_nactive_add(arena, (usize - oldsize) >> LG_PAGE);
+	else
+		arena_nactive_sub(arena, (oldsize - usize) >> LG_PAGE);
+	malloc_mutex_unlock(tsdn, &arena->lock);
 }
 
 void
-arena_chunk_ralloc_huge_shrink(arena_t *arena, void *chunk, size_t oldsize,
-    size_t usize)
+arena_chunk_ralloc_huge_shrink(tsdn_t *tsdn, arena_t *arena, void *chunk,
+    size_t oldsize, size_t usize, size_t sn)
 {
 	size_t udiff = oldsize - usize;
 	size_t cdiff = CHUNK_CEILING(oldsize) - CHUNK_CEILING(usize);
 
-	malloc_mutex_lock(&arena->lock);
+	malloc_mutex_lock(tsdn, &arena->lock);
 	if (config_stats) {
 		arena_huge_ralloc_stats_update(arena, oldsize, usize);
-		if (cdiff != 0) {
+		if (cdiff != 0)
 			arena->stats.mapped -= cdiff;
-			stats_cactive_sub(udiff);
-		}
 	}
-	arena->nactive -= udiff >> LG_PAGE;
+	arena_nactive_sub(arena, udiff >> LG_PAGE);
 
 	if (cdiff != 0) {
 		chunk_hooks_t chunk_hooks = CHUNK_HOOKS_INITIALIZER;
 		void *nchunk = (void *)((uintptr_t)chunk +
 		    CHUNK_CEILING(usize));
 
-		chunk_dalloc_cache(arena, &chunk_hooks, nchunk, cdiff, true);
+		chunk_dalloc_cache(tsdn, arena, &chunk_hooks, nchunk, cdiff,
+		    sn, true);
 	}
-	malloc_mutex_unlock(&arena->lock);
+	malloc_mutex_unlock(tsdn, &arena->lock);
 }
 
 static bool
-arena_chunk_ralloc_huge_expand_hard(arena_t *arena, chunk_hooks_t *chunk_hooks,
-    void *chunk, size_t oldsize, size_t usize, bool *zero, void *nchunk,
-    size_t udiff, size_t cdiff)
+arena_chunk_ralloc_huge_expand_hard(tsdn_t *tsdn, arena_t *arena,
+    chunk_hooks_t *chunk_hooks, void *chunk, size_t oldsize, size_t usize,
+    size_t *sn, bool *zero, void *nchunk, size_t udiff, size_t cdiff)
 {
 	bool err;
 	bool commit = true;
 
-	err = (chunk_alloc_wrapper(arena, chunk_hooks, nchunk, cdiff, chunksize,
-	    zero, &commit) == NULL);
+	err = (chunk_alloc_wrapper(tsdn, arena, chunk_hooks, nchunk, cdiff,
+	    chunksize, sn, zero, &commit) == NULL);
 	if (err) {
 		/* Revert optimistic stats updates. */
-		malloc_mutex_lock(&arena->lock);
+		malloc_mutex_lock(tsdn, &arena->lock);
 		if (config_stats) {
 			arena_huge_ralloc_stats_update_undo(arena, oldsize,
 			    usize);
 			arena->stats.mapped -= cdiff;
 		}
-		arena->nactive -= (udiff >> LG_PAGE);
-		malloc_mutex_unlock(&arena->lock);
+		arena_nactive_sub(arena, udiff >> LG_PAGE);
+		malloc_mutex_unlock(tsdn, &arena->lock);
 	} else if (chunk_hooks->merge(chunk, CHUNK_CEILING(oldsize), nchunk,
 	    cdiff, true, arena->ind)) {
-		chunk_dalloc_arena(arena, chunk_hooks, nchunk, cdiff, *zero,
-		    true);
+		chunk_dalloc_wrapper(tsdn, arena, chunk_hooks, nchunk, cdiff,
+		    *sn, *zero, true);
 		err = true;
 	}
 	return (err);
 }
 
 bool
-arena_chunk_ralloc_huge_expand(arena_t *arena, void *chunk, size_t oldsize,
-    size_t usize, bool *zero)
+arena_chunk_ralloc_huge_expand(tsdn_t *tsdn, arena_t *arena, void *chunk,
+    size_t oldsize, size_t usize, bool *zero)
 {
 	bool err;
-	chunk_hooks_t chunk_hooks = chunk_hooks_get(arena);
+	chunk_hooks_t chunk_hooks = chunk_hooks_get(tsdn, arena);
 	void *nchunk = (void *)((uintptr_t)chunk + CHUNK_CEILING(oldsize));
 	size_t udiff = usize - oldsize;
 	size_t cdiff = CHUNK_CEILING(usize) - CHUNK_CEILING(oldsize);
+	size_t sn;
+	bool commit = true;
 
-	malloc_mutex_lock(&arena->lock);
+	malloc_mutex_lock(tsdn, &arena->lock);
 
 	/* Optimistically update stats. */
 	if (config_stats) {
 		arena_huge_ralloc_stats_update(arena, oldsize, usize);
 		arena->stats.mapped += cdiff;
 	}
-	arena->nactive += (udiff >> LG_PAGE);
+	arena_nactive_add(arena, udiff >> LG_PAGE);
 
-	err = (chunk_alloc_cache(arena, &arena->chunk_hooks, nchunk, cdiff,
-	    chunksize, zero, true) == NULL);
-	malloc_mutex_unlock(&arena->lock);
+	err = (chunk_alloc_cache(tsdn, arena, &chunk_hooks, nchunk, cdiff,
+	    chunksize, &sn, zero, &commit, true) == NULL);
+	malloc_mutex_unlock(tsdn, &arena->lock);
 	if (err) {
-		err = arena_chunk_ralloc_huge_expand_hard(arena, &chunk_hooks,
-		    chunk, oldsize, usize, zero, nchunk, udiff,
-		    cdiff);
+		err = arena_chunk_ralloc_huge_expand_hard(tsdn, arena,
+		    &chunk_hooks, chunk, oldsize, usize, &sn, zero, nchunk,
+		    udiff, cdiff);
 	} else if (chunk_hooks.merge(chunk, CHUNK_CEILING(oldsize), nchunk,
 	    cdiff, true, arena->ind)) {
-		chunk_dalloc_arena(arena, &chunk_hooks, nchunk, cdiff, *zero,
-		    true);
+		chunk_dalloc_wrapper(tsdn, arena, &chunk_hooks, nchunk, cdiff,
+		    sn, *zero, true);
 		err = true;
 	}
 
-	if (config_stats && !err)
-		stats_cactive_add(udiff);
 	return (err);
 }
 
 /*
  * Do first-best-fit run selection, i.e. select the lowest run that best fits.
- * Run sizes are quantized, so not all candidate runs are necessarily exactly
- * the same size.
+ * Run sizes are indexed, so not all candidate runs are necessarily exactly the
+ * same size.
  */
 static arena_run_t *
 arena_run_first_best_fit(arena_t *arena, size_t size)
 {
-	size_t search_size = run_quantize_first(size);
-	arena_chunk_map_misc_t *key = arena_miscelm_key_create(search_size);
-	arena_chunk_map_misc_t *miscelm =
-	    arena_avail_tree_nsearch(&arena->runs_avail, key);
-	if (miscelm == NULL)
-		return (NULL);
-	return (&miscelm->run);
+	pszind_t pind, i;
+
+	pind = psz2ind(run_quantize_ceil(size));
+
+	for (i = pind; pind2sz(i) <= chunksize; i++) {
+		arena_chunk_map_misc_t *miscelm = arena_run_heap_first(
+		    &arena->runs_avail[i]);
+		if (miscelm != NULL)
+			return (&miscelm->run);
+	}
+
+	return (NULL);
 }
 
 static arena_run_t *
 arena_run_alloc_large_helper(arena_t *arena, size_t size, bool zero)
 {
-	arena_run_t *run = arena_run_first_best_fit(arena, s2u(size));
+	arena_run_t *run = arena_run_first_best_fit(arena, size);
 	if (run != NULL) {
 		if (arena_run_split_large(arena, run, size, zero))
 			run = NULL;
@@ -1092,7 +1130,7 @@ arena_run_alloc_large_helper(arena_t *arena, size_t size, bool zero)
 }
 
 static arena_run_t *
-arena_run_alloc_large(arena_t *arena, size_t size, bool zero)
+arena_run_alloc_large(tsdn_t *tsdn, arena_t *arena, size_t size, bool zero)
 {
 	arena_chunk_t *chunk;
 	arena_run_t *run;
@@ -1108,9 +1146,9 @@ arena_run_alloc_large(arena_t *arena, size_t size, bool zero)
 	/*
 	 * No usable runs.  Create a new chunk from which to allocate the run.
 	 */
-	chunk = arena_chunk_alloc(arena);
+	chunk = arena_chunk_alloc(tsdn, arena);
 	if (chunk != NULL) {
-		run = &arena_miscelm_get(chunk, map_bias)->run;
+		run = &arena_miscelm_get_mutable(chunk, map_bias)->run;
 		if (arena_run_split_large(arena, run, size, zero))
 			run = NULL;
 		return (run);
@@ -1136,7 +1174,7 @@ arena_run_alloc_small_helper(arena_t *arena, size_t size, szind_t binind)
 }
 
 static arena_run_t *
-arena_run_alloc_small(arena_t *arena, size_t size, szind_t binind)
+arena_run_alloc_small(tsdn_t *tsdn, arena_t *arena, size_t size, szind_t binind)
 {
 	arena_chunk_t *chunk;
 	arena_run_t *run;
@@ -1153,9 +1191,9 @@ arena_run_alloc_small(arena_t *arena, size_t size, szind_t binind)
 	/*
 	 * No usable runs.  Create a new chunk from which to allocate the run.
 	 */
-	chunk = arena_chunk_alloc(arena);
+	chunk = arena_chunk_alloc(tsdn, arena);
 	if (chunk != NULL) {
-		run = &arena_miscelm_get(chunk, map_bias)->run;
+		run = &arena_miscelm_get_mutable(chunk, map_bias)->run;
 		if (arena_run_split_small(arena, run, size, binind))
 			run = NULL;
 		return (run);
@@ -1178,42 +1216,239 @@ arena_lg_dirty_mult_valid(ssize_t lg_dirty_mult)
 }
 
 ssize_t
-arena_lg_dirty_mult_get(arena_t *arena)
+arena_lg_dirty_mult_get(tsdn_t *tsdn, arena_t *arena)
 {
 	ssize_t lg_dirty_mult;
 
-	malloc_mutex_lock(&arena->lock);
+	malloc_mutex_lock(tsdn, &arena->lock);
 	lg_dirty_mult = arena->lg_dirty_mult;
-	malloc_mutex_unlock(&arena->lock);
+	malloc_mutex_unlock(tsdn, &arena->lock);
 
 	return (lg_dirty_mult);
 }
 
 bool
-arena_lg_dirty_mult_set(arena_t *arena, ssize_t lg_dirty_mult)
+arena_lg_dirty_mult_set(tsdn_t *tsdn, arena_t *arena, ssize_t lg_dirty_mult)
 {
 
 	if (!arena_lg_dirty_mult_valid(lg_dirty_mult))
 		return (true);
 
-	malloc_mutex_lock(&arena->lock);
+	malloc_mutex_lock(tsdn, &arena->lock);
 	arena->lg_dirty_mult = lg_dirty_mult;
-	arena_maybe_purge(arena);
-	malloc_mutex_unlock(&arena->lock);
+	arena_maybe_purge(tsdn, arena);
+	malloc_mutex_unlock(tsdn, &arena->lock);
 
 	return (false);
 }
 
-void
-arena_maybe_purge(arena_t *arena)
+static void
+arena_decay_deadline_init(arena_t *arena)
+{
+
+	assert(opt_purge == purge_mode_decay);
+
+	/*
+	 * Generate a new deadline that is uniformly random within the next
+	 * epoch after the current one.
+	 */
+	nstime_copy(&arena->decay.deadline, &arena->decay.epoch);
+	nstime_add(&arena->decay.deadline, &arena->decay.interval);
+	if (arena->decay.time > 0) {
+		nstime_t jitter;
+
+		nstime_init(&jitter, prng_range_u64(&arena->decay.jitter_state,
+		    nstime_ns(&arena->decay.interval)));
+		nstime_add(&arena->decay.deadline, &jitter);
+	}
+}
+
+static bool
+arena_decay_deadline_reached(const arena_t *arena, const nstime_t *time)
+{
+
+	assert(opt_purge == purge_mode_decay);
+
+	return (nstime_compare(&arena->decay.deadline, time) <= 0);
+}
+
+static size_t
+arena_decay_backlog_npages_limit(const arena_t *arena)
+{
+	static const uint64_t h_steps[] = {
+#define	STEP(step, h, x, y) \
+		h,
+		SMOOTHSTEP
+#undef STEP
+	};
+	uint64_t sum;
+	size_t npages_limit_backlog;
+	unsigned i;
+
+	assert(opt_purge == purge_mode_decay);
+
+	/*
+	 * For each element of decay_backlog, multiply by the corresponding
+	 * fixed-point smoothstep decay factor.  Sum the products, then divide
+	 * to round down to the nearest whole number of pages.
+	 */
+	sum = 0;
+	for (i = 0; i < SMOOTHSTEP_NSTEPS; i++)
+		sum += arena->decay.backlog[i] * h_steps[i];
+	npages_limit_backlog = (size_t)(sum >> SMOOTHSTEP_BFP);
+
+	return (npages_limit_backlog);
+}
+
+static void
+arena_decay_backlog_update_last(arena_t *arena)
+{
+	size_t ndirty_delta = (arena->ndirty > arena->decay.ndirty) ?
+	    arena->ndirty - arena->decay.ndirty : 0;
+	arena->decay.backlog[SMOOTHSTEP_NSTEPS-1] = ndirty_delta;
+}
+
+static void
+arena_decay_backlog_update(arena_t *arena, uint64_t nadvance_u64)
+{
+
+	if (nadvance_u64 >= SMOOTHSTEP_NSTEPS) {
+		memset(arena->decay.backlog, 0, (SMOOTHSTEP_NSTEPS-1) *
+		    sizeof(size_t));
+	} else {
+		size_t nadvance_z = (size_t)nadvance_u64;
+
+		assert((uint64_t)nadvance_z == nadvance_u64);
+
+		memmove(arena->decay.backlog, &arena->decay.backlog[nadvance_z],
+		    (SMOOTHSTEP_NSTEPS - nadvance_z) * sizeof(size_t));
+		if (nadvance_z > 1) {
+			memset(&arena->decay.backlog[SMOOTHSTEP_NSTEPS -
+			    nadvance_z], 0, (nadvance_z-1) * sizeof(size_t));
+		}
+	}
+
+	arena_decay_backlog_update_last(arena);
+}
+
+static void
+arena_decay_epoch_advance_helper(arena_t *arena, const nstime_t *time)
+{
+	uint64_t nadvance_u64;
+	nstime_t delta;
+
+	assert(opt_purge == purge_mode_decay);
+	assert(arena_decay_deadline_reached(arena, time));
+
+	nstime_copy(&delta, time);
+	nstime_subtract(&delta, &arena->decay.epoch);
+	nadvance_u64 = nstime_divide(&delta, &arena->decay.interval);
+	assert(nadvance_u64 > 0);
+
+	/* Add nadvance_u64 decay intervals to epoch. */
+	nstime_copy(&delta, &arena->decay.interval);
+	nstime_imultiply(&delta, nadvance_u64);
+	nstime_add(&arena->decay.epoch, &delta);
+
+	/* Set a new deadline. */
+	arena_decay_deadline_init(arena);
+
+	/* Update the backlog. */
+	arena_decay_backlog_update(arena, nadvance_u64);
+}
+
+static void
+arena_decay_epoch_advance_purge(tsdn_t *tsdn, arena_t *arena)
+{
+	size_t ndirty_limit = arena_decay_backlog_npages_limit(arena);
+
+	if (arena->ndirty > ndirty_limit)
+		arena_purge_to_limit(tsdn, arena, ndirty_limit);
+	arena->decay.ndirty = arena->ndirty;
+}
+
+static void
+arena_decay_epoch_advance(tsdn_t *tsdn, arena_t *arena, const nstime_t *time)
+{
+
+	arena_decay_epoch_advance_helper(arena, time);
+	arena_decay_epoch_advance_purge(tsdn, arena);
+}
+
+static void
+arena_decay_init(arena_t *arena, ssize_t decay_time)
+{
+
+	arena->decay.time = decay_time;
+	if (decay_time > 0) {
+		nstime_init2(&arena->decay.interval, decay_time, 0);
+		nstime_idivide(&arena->decay.interval, SMOOTHSTEP_NSTEPS);
+	}
+
+	nstime_init(&arena->decay.epoch, 0);
+	nstime_update(&arena->decay.epoch);
+	arena->decay.jitter_state = (uint64_t)(uintptr_t)arena;
+	arena_decay_deadline_init(arena);
+	arena->decay.ndirty = arena->ndirty;
+	memset(arena->decay.backlog, 0, SMOOTHSTEP_NSTEPS * sizeof(size_t));
+}
+
+static bool
+arena_decay_time_valid(ssize_t decay_time)
+{
+
+	if (decay_time < -1)
+		return (false);
+	if (decay_time == -1 || (uint64_t)decay_time <= NSTIME_SEC_MAX)
+		return (true);
+	return (false);
+}
+
+ssize_t
+arena_decay_time_get(tsdn_t *tsdn, arena_t *arena)
+{
+	ssize_t decay_time;
+
+	malloc_mutex_lock(tsdn, &arena->lock);
+	decay_time = arena->decay.time;
+	malloc_mutex_unlock(tsdn, &arena->lock);
+
+	return (decay_time);
+}
+
+bool
+arena_decay_time_set(tsdn_t *tsdn, arena_t *arena, ssize_t decay_time)
+{
+
+	if (!arena_decay_time_valid(decay_time))
+		return (true);
+
+	malloc_mutex_lock(tsdn, &arena->lock);
+	/*
+	 * Restart decay backlog from scratch, which may cause many dirty pages
+	 * to be immediately purged.  It would conceptually be possible to map
+	 * the old backlog onto the new backlog, but there is no justification
+	 * for such complexity since decay_time changes are intended to be
+	 * infrequent, either between the {-1, 0, >0} states, or a one-time
+	 * arbitrary change during initial arena configuration.
+	 */
+	arena_decay_init(arena, decay_time);
+	arena_maybe_purge(tsdn, arena);
+	malloc_mutex_unlock(tsdn, &arena->lock);
+
+	return (false);
+}
+
+static void
+arena_maybe_purge_ratio(tsdn_t *tsdn, arena_t *arena)
 {
 
+	assert(opt_purge == purge_mode_ratio);
+
 	/* Don't purge if the option is disabled. */
 	if (arena->lg_dirty_mult < 0)
 		return;
-	/* Don't recursively purge. */
-	if (arena->purging)
-		return;
+
 	/*
 	 * Iterate, since preventing recursive purging could otherwise leave too
 	 * many dirty pages.
@@ -1228,8 +1463,66 @@ arena_maybe_purge(arena_t *arena)
 		 */
 		if (arena->ndirty <= threshold)
 			return;
-		arena_purge(arena, false);
+		arena_purge_to_limit(tsdn, arena, threshold);
+	}
+}
+
+static void
+arena_maybe_purge_decay(tsdn_t *tsdn, arena_t *arena)
+{
+	nstime_t time;
+
+	assert(opt_purge == purge_mode_decay);
+
+	/* Purge all or nothing if the option is disabled. */
+	if (arena->decay.time <= 0) {
+		if (arena->decay.time == 0)
+			arena_purge_to_limit(tsdn, arena, 0);
+		return;
+	}
+
+	nstime_init(&time, 0);
+	nstime_update(&time);
+	if (unlikely(!nstime_monotonic() && nstime_compare(&arena->decay.epoch,
+	    &time) > 0)) {
+		/*
+		 * Time went backwards.  Move the epoch back in time and
+		 * generate a new deadline, with the expectation that time
+		 * typically flows forward for long enough periods of time that
+		 * epochs complete.  Unfortunately, this strategy is susceptible
+		 * to clock jitter triggering premature epoch advances, but
+		 * clock jitter estimation and compensation isn't feasible here
+		 * because calls into this code are event-driven.
+		 */
+		nstime_copy(&arena->decay.epoch, &time);
+		arena_decay_deadline_init(arena);
+	} else {
+		/* Verify that time does not go backwards. */
+		assert(nstime_compare(&arena->decay.epoch, &time) <= 0);
 	}
+
+	/*
+	 * If the deadline has been reached, advance to the current epoch and
+	 * purge to the new limit if necessary.  Note that dirty pages created
+	 * during the current epoch are not subject to purge until a future
+	 * epoch, so as a result purging only happens during epoch advances.
+	 */
+	if (arena_decay_deadline_reached(arena, &time))
+		arena_decay_epoch_advance(tsdn, arena, &time);
+}
+
+void
+arena_maybe_purge(tsdn_t *tsdn, arena_t *arena)
+{
+
+	/* Don't recursively purge. */
+	if (arena->purging)
+		return;
+
+	if (opt_purge == purge_mode_ratio)
+		arena_maybe_purge_ratio(tsdn, arena);
+	else
+		arena_maybe_purge_decay(tsdn, arena);
 }
 
 static size_t
@@ -1253,49 +1546,29 @@ arena_dirty_count(arena_t *arena)
 			arena_chunk_map_misc_t *miscelm =
 			    arena_rd_to_miscelm(rdelm);
 			size_t pageind = arena_miscelm_to_pageind(miscelm);
-			assert(arena_mapbits_allocated_get(chunk, pageind) ==
-			    0);
-			assert(arena_mapbits_large_get(chunk, pageind) == 0);
-			assert(arena_mapbits_dirty_get(chunk, pageind) != 0);
-			npages = arena_mapbits_unallocated_size_get(chunk,
-			    pageind) >> LG_PAGE;
-		}
-		ndirty += npages;
-	}
-
-	return (ndirty);
-}
-
-static size_t
-arena_compute_npurge(arena_t *arena, bool all)
-{
-	size_t npurge;
-
-	/*
-	 * Compute the minimum number of pages that this thread should try to
-	 * purge.
-	 */
-	if (!all) {
-		size_t threshold = (arena->nactive >> arena->lg_dirty_mult);
-		threshold = threshold < chunk_npages ? chunk_npages : threshold;
-
-		npurge = arena->ndirty - threshold;
-	} else
-		npurge = arena->ndirty;
+			assert(arena_mapbits_allocated_get(chunk, pageind) ==
+			    0);
+			assert(arena_mapbits_large_get(chunk, pageind) == 0);
+			assert(arena_mapbits_dirty_get(chunk, pageind) != 0);
+			npages = arena_mapbits_unallocated_size_get(chunk,
+			    pageind) >> LG_PAGE;
+		}
+		ndirty += npages;
+	}
 
-	return (npurge);
+	return (ndirty);
 }
 
 static size_t
-arena_stash_dirty(arena_t *arena, chunk_hooks_t *chunk_hooks, bool all,
-    size_t npurge, arena_runs_dirty_link_t *purge_runs_sentinel,
+arena_stash_dirty(tsdn_t *tsdn, arena_t *arena, chunk_hooks_t *chunk_hooks,
+    size_t ndirty_limit, arena_runs_dirty_link_t *purge_runs_sentinel,
     extent_node_t *purge_chunks_sentinel)
 {
 	arena_runs_dirty_link_t *rdelm, *rdelm_next;
 	extent_node_t *chunkselm;
 	size_t nstashed = 0;
 
-	/* Stash at least npurge pages. */
+	/* Stash runs/chunks according to ndirty_limit. */
 	for (rdelm = qr_next(&arena->runs_dirty, rd_link),
 	    chunkselm = qr_next(&arena->chunks_cache, cc_link);
 	    rdelm != &arena->runs_dirty; rdelm = rdelm_next) {
@@ -1304,24 +1577,32 @@ arena_stash_dirty(arena_t *arena, chunk_hooks_t *chunk_hooks, bool all,
 
 		if (rdelm == &chunkselm->rd) {
 			extent_node_t *chunkselm_next;
-			bool zero;
+			size_t sn;
+			bool zero, commit;
 			UNUSED void *chunk;
 
+			npages = extent_node_size_get(chunkselm) >> LG_PAGE;
+			if (opt_purge == purge_mode_decay && arena->ndirty -
+			    (nstashed + npages) < ndirty_limit)
+				break;
+
 			chunkselm_next = qr_next(chunkselm, cc_link);
 			/*
 			 * Allocate.  chunkselm remains valid due to the
 			 * dalloc_node=false argument to chunk_alloc_cache().
 			 */
 			zero = false;
-			chunk = chunk_alloc_cache(arena, chunk_hooks,
+			commit = false;
+			chunk = chunk_alloc_cache(tsdn, arena, chunk_hooks,
 			    extent_node_addr_get(chunkselm),
-			    extent_node_size_get(chunkselm), chunksize, &zero,
-			    false);
+			    extent_node_size_get(chunkselm), chunksize, &sn,
+			    &zero, &commit, false);
 			assert(chunk == extent_node_addr_get(chunkselm));
 			assert(zero == extent_node_zeroed_get(chunkselm));
 			extent_node_dirty_insert(chunkselm, purge_runs_sentinel,
 			    purge_chunks_sentinel);
-			npages = extent_node_size_get(chunkselm) >> LG_PAGE;
+			assert(npages == (extent_node_size_get(chunkselm) >>
+			    LG_PAGE));
 			chunkselm = chunkselm_next;
 		} else {
 			arena_chunk_t *chunk =
@@ -1334,6 +1615,9 @@ arena_stash_dirty(arena_t *arena, chunk_hooks_t *chunk_hooks, bool all,
 			    arena_mapbits_unallocated_size_get(chunk, pageind);
 
 			npages = run_size >> LG_PAGE;
+			if (opt_purge == purge_mode_decay && arena->ndirty -
+			    (nstashed + npages) < ndirty_limit)
+				break;
 
 			assert(pageind + npages <= chunk_npages);
 			assert(arena_mapbits_dirty_get(chunk, pageind) ==
@@ -1344,7 +1628,7 @@ arena_stash_dirty(arena_t *arena, chunk_hooks_t *chunk_hooks, bool all,
 			 * prior to allocation.
 			 */
 			if (chunk == arena->spare)
-				arena_chunk_alloc(arena);
+				arena_chunk_alloc(tsdn, arena);
 
 			/* Temporarily allocate the free dirty run. */
 			arena_run_split_large(arena, run, run_size, false);
@@ -1359,7 +1643,8 @@ arena_stash_dirty(arena_t *arena, chunk_hooks_t *chunk_hooks, bool all,
 		}
 
 		nstashed += npages;
-		if (!all && nstashed >= npurge)
+		if (opt_purge == purge_mode_ratio && arena->ndirty - nstashed <=
+		    ndirty_limit)
 			break;
 	}
 
@@ -1367,7 +1652,7 @@ arena_stash_dirty(arena_t *arena, chunk_hooks_t *chunk_hooks, bool all,
 }
 
 static size_t
-arena_purge_stashed(arena_t *arena, chunk_hooks_t *chunk_hooks,
+arena_purge_stashed(tsdn_t *tsdn, arena_t *arena, chunk_hooks_t *chunk_hooks,
     arena_runs_dirty_link_t *purge_runs_sentinel,
     extent_node_t *purge_chunks_sentinel)
 {
@@ -1379,7 +1664,7 @@ arena_purge_stashed(arena_t *arena, chunk_hooks_t *chunk_hooks,
 		nmadvise = 0;
 	npurged = 0;
 
-	malloc_mutex_unlock(&arena->lock);
+	malloc_mutex_unlock(tsdn, &arena->lock);
 	for (rdelm = qr_next(purge_runs_sentinel, rd_link),
 	    chunkselm = qr_next(purge_chunks_sentinel, cc_link);
 	    rdelm != purge_runs_sentinel; rdelm = qr_next(rdelm, rd_link)) {
@@ -1408,6 +1693,17 @@ arena_purge_stashed(arena_t *arena, chunk_hooks_t *chunk_hooks,
 			run_size = arena_mapbits_large_size_get(chunk, pageind);
 			npages = run_size >> LG_PAGE;
 
+			/*
+			 * If this is the first run purged within chunk, mark
+			 * the chunk as non-huge.  This will prevent all use of
+			 * transparent huge pages for this chunk until the chunk
+			 * as a whole is deallocated.
+			 */
+			if (chunk->hugepage) {
+				pages_nohuge(chunk, chunksize);
+				chunk->hugepage = false;
+			}
+
 			assert(pageind + npages <= chunk_npages);
 			assert(!arena_mapbits_decommitted_get(chunk, pageind));
 			assert(!arena_mapbits_decommitted_get(chunk,
@@ -1418,7 +1714,7 @@ arena_purge_stashed(arena_t *arena, chunk_hooks_t *chunk_hooks,
 				flag_unzeroed = 0;
 				flags = CHUNK_MAP_DECOMMITTED;
 			} else {
-				flag_unzeroed = chunk_purge_wrapper(arena,
+				flag_unzeroed = chunk_purge_wrapper(tsdn, arena,
 				    chunk_hooks, chunk, chunksize, pageind <<
 				    LG_PAGE, run_size) ? CHUNK_MAP_UNZEROED : 0;
 				flags = flag_unzeroed;
@@ -1449,7 +1745,7 @@ arena_purge_stashed(arena_t *arena, chunk_hooks_t *chunk_hooks,
 		if (config_stats)
 			nmadvise++;
 	}
-	malloc_mutex_lock(&arena->lock);
+	malloc_mutex_lock(tsdn, &arena->lock);
 
 	if (config_stats) {
 		arena->stats.nmadvise += nmadvise;
@@ -1460,7 +1756,7 @@ arena_purge_stashed(arena_t *arena, chunk_hooks_t *chunk_hooks,
 }
 
 static void
-arena_unstash_purged(arena_t *arena, chunk_hooks_t *chunk_hooks,
+arena_unstash_purged(tsdn_t *tsdn, arena_t *arena, chunk_hooks_t *chunk_hooks,
     arena_runs_dirty_link_t *purge_runs_sentinel,
     extent_node_t *purge_chunks_sentinel)
 {
@@ -1477,13 +1773,14 @@ arena_unstash_purged(arena_t *arena, chunk_hooks_t *chunk_hooks,
 			    cc_link);
 			void *addr = extent_node_addr_get(chunkselm);
 			size_t size = extent_node_size_get(chunkselm);
+			size_t sn = extent_node_sn_get(chunkselm);
 			bool zeroed = extent_node_zeroed_get(chunkselm);
 			bool committed = extent_node_committed_get(chunkselm);
 			extent_node_dirty_remove(chunkselm);
-			arena_node_dalloc(arena, chunkselm);
+			arena_node_dalloc(tsdn, arena, chunkselm);
 			chunkselm = chunkselm_next;
-			chunk_dalloc_arena(arena, chunk_hooks, addr, size,
-			    zeroed, committed);
+			chunk_dalloc_wrapper(tsdn, arena, chunk_hooks, addr,
+			    size, sn, zeroed, committed);
 		} else {
 			arena_chunk_t *chunk =
 			    (arena_chunk_t *)CHUNK_ADDR2BASE(rdelm);
@@ -1494,16 +1791,26 @@ arena_unstash_purged(arena_t *arena, chunk_hooks_t *chunk_hooks,
 			    pageind) != 0);
 			arena_run_t *run = &miscelm->run;
 			qr_remove(rdelm, rd_link);
-			arena_run_dalloc(arena, run, false, true, decommitted);
+			arena_run_dalloc(tsdn, arena, run, false, true,
+			    decommitted);
 		}
 	}
 }
 
+/*
+ * NB: ndirty_limit is interpreted differently depending on opt_purge:
+ *   - purge_mode_ratio: Purge as few dirty run/chunks as possible to reach the
+ *                       desired state:
+ *                       (arena->ndirty <= ndirty_limit)
+ *   - purge_mode_decay: Purge as many dirty runs/chunks as possible without
+ *                       violating the invariant:
+ *                       (arena->ndirty >= ndirty_limit)
+ */
 static void
-arena_purge(arena_t *arena, bool all)
+arena_purge_to_limit(tsdn_t *tsdn, arena_t *arena, size_t ndirty_limit)
 {
-	chunk_hooks_t chunk_hooks = chunk_hooks_get(arena);
-	size_t npurge, npurgeable, npurged;
+	chunk_hooks_t chunk_hooks = chunk_hooks_get(tsdn, arena);
+	size_t npurge, npurged;
 	arena_runs_dirty_link_t purge_runs_sentinel;
 	extent_node_t purge_chunks_sentinel;
 
@@ -1517,34 +1824,183 @@ arena_purge(arena_t *arena, bool all)
 		size_t ndirty = arena_dirty_count(arena);
 		assert(ndirty == arena->ndirty);
 	}
-	assert((arena->nactive >> arena->lg_dirty_mult) < arena->ndirty || all);
-
-	if (config_stats)
-		arena->stats.npurge++;
+	assert(opt_purge != purge_mode_ratio || (arena->nactive >>
+	    arena->lg_dirty_mult) < arena->ndirty || ndirty_limit == 0);
 
-	npurge = arena_compute_npurge(arena, all);
 	qr_new(&purge_runs_sentinel, rd_link);
 	extent_node_dirty_linkage_init(&purge_chunks_sentinel);
 
-	npurgeable = arena_stash_dirty(arena, &chunk_hooks, all, npurge,
+	npurge = arena_stash_dirty(tsdn, arena, &chunk_hooks, ndirty_limit,
 	    &purge_runs_sentinel, &purge_chunks_sentinel);
-	assert(npurgeable >= npurge);
-	npurged = arena_purge_stashed(arena, &chunk_hooks, &purge_runs_sentinel,
-	    &purge_chunks_sentinel);
-	assert(npurged == npurgeable);
-	arena_unstash_purged(arena, &chunk_hooks, &purge_runs_sentinel,
+	if (npurge == 0)
+		goto label_return;
+	npurged = arena_purge_stashed(tsdn, arena, &chunk_hooks,
+	    &purge_runs_sentinel, &purge_chunks_sentinel);
+	assert(npurged == npurge);
+	arena_unstash_purged(tsdn, arena, &chunk_hooks, &purge_runs_sentinel,
 	    &purge_chunks_sentinel);
 
+	if (config_stats)
+		arena->stats.npurge++;
+
+label_return:
 	arena->purging = false;
 }
 
 void
-arena_purge_all(arena_t *arena)
+arena_purge(tsdn_t *tsdn, arena_t *arena, bool all)
+{
+
+	malloc_mutex_lock(tsdn, &arena->lock);
+	if (all)
+		arena_purge_to_limit(tsdn, arena, 0);
+	else
+		arena_maybe_purge(tsdn, arena);
+	malloc_mutex_unlock(tsdn, &arena->lock);
+}
+
+static void
+arena_achunk_prof_reset(tsd_t *tsd, arena_t *arena, arena_chunk_t *chunk)
+{
+	size_t pageind, npages;
+
+	cassert(config_prof);
+	assert(opt_prof);
+
+	/*
+	 * Iterate over the allocated runs and remove profiled allocations from
+	 * the sample set.
+	 */
+	for (pageind = map_bias; pageind < chunk_npages; pageind += npages) {
+		if (arena_mapbits_allocated_get(chunk, pageind) != 0) {
+			if (arena_mapbits_large_get(chunk, pageind) != 0) {
+				void *ptr = (void *)((uintptr_t)chunk + (pageind
+				    << LG_PAGE));
+				size_t usize = isalloc(tsd_tsdn(tsd), ptr,
+				    config_prof);
+
+				prof_free(tsd, ptr, usize);
+				npages = arena_mapbits_large_size_get(chunk,
+				    pageind) >> LG_PAGE;
+			} else {
+				/* Skip small run. */
+				size_t binind = arena_mapbits_binind_get(chunk,
+				    pageind);
+				arena_bin_info_t *bin_info =
+				    &arena_bin_info[binind];
+				npages = bin_info->run_size >> LG_PAGE;
+			}
+		} else {
+			/* Skip unallocated run. */
+			npages = arena_mapbits_unallocated_size_get(chunk,
+			    pageind) >> LG_PAGE;
+		}
+		assert(pageind + npages <= chunk_npages);
+	}
+}
+
+void
+arena_reset(tsd_t *tsd, arena_t *arena)
 {
+	unsigned i;
+	extent_node_t *node;
+
+	/*
+	 * Locking in this function is unintuitive.  The caller guarantees that
+	 * no concurrent operations are happening in this arena, but there are
+	 * still reasons that some locking is necessary:
+	 *
+	 * - Some of the functions in the transitive closure of calls assume
+	 *   appropriate locks are held, and in some cases these locks are
+	 *   temporarily dropped to avoid lock order reversal or deadlock due to
+	 *   reentry.
+	 * - mallctl("epoch", ...) may concurrently refresh stats.  While
+	 *   strictly speaking this is a "concurrent operation", disallowing
+	 *   stats refreshes would impose an inconvenient burden.
+	 */
+
+	/* Remove large allocations from prof sample set. */
+	if (config_prof && opt_prof) {
+		ql_foreach(node, &arena->achunks, ql_link) {
+			arena_achunk_prof_reset(tsd, arena,
+			    extent_node_addr_get(node));
+		}
+	}
+
+	/* Reset curruns for large size classes. */
+	if (config_stats) {
+		for (i = 0; i < nlclasses; i++)
+			arena->stats.lstats[i].curruns = 0;
+	}
+
+	/* Huge allocations. */
+	malloc_mutex_lock(tsd_tsdn(tsd), &arena->huge_mtx);
+	for (node = ql_last(&arena->huge, ql_link); node != NULL; node =
+	    ql_last(&arena->huge, ql_link)) {
+		void *ptr = extent_node_addr_get(node);
+		size_t usize;
+
+		malloc_mutex_unlock(tsd_tsdn(tsd), &arena->huge_mtx);
+		if (config_stats || (config_prof && opt_prof))
+			usize = isalloc(tsd_tsdn(tsd), ptr, config_prof);
+		/* Remove huge allocation from prof sample set. */
+		if (config_prof && opt_prof)
+			prof_free(tsd, ptr, usize);
+		huge_dalloc(tsd_tsdn(tsd), ptr);
+		malloc_mutex_lock(tsd_tsdn(tsd), &arena->huge_mtx);
+		/* Cancel out unwanted effects on stats. */
+		if (config_stats)
+			arena_huge_reset_stats_cancel(arena, usize);
+	}
+	malloc_mutex_unlock(tsd_tsdn(tsd), &arena->huge_mtx);
+
+	malloc_mutex_lock(tsd_tsdn(tsd), &arena->lock);
+
+	/* Bins. */
+	for (i = 0; i < NBINS; i++) {
+		arena_bin_t *bin = &arena->bins[i];
+		malloc_mutex_lock(tsd_tsdn(tsd), &bin->lock);
+		bin->runcur = NULL;
+		arena_run_heap_new(&bin->runs);
+		if (config_stats) {
+			bin->stats.curregs = 0;
+			bin->stats.curruns = 0;
+		}
+		malloc_mutex_unlock(tsd_tsdn(tsd), &bin->lock);
+	}
+
+	/*
+	 * Re-initialize runs_dirty such that the chunks_cache and runs_dirty
+	 * chains directly correspond.
+	 */
+	qr_new(&arena->runs_dirty, rd_link);
+	for (node = qr_next(&arena->chunks_cache, cc_link);
+	    node != &arena->chunks_cache; node = qr_next(node, cc_link)) {
+		qr_new(&node->rd, rd_link);
+		qr_meld(&arena->runs_dirty, &node->rd, rd_link);
+	}
+
+	/* Arena chunks. */
+	for (node = ql_last(&arena->achunks, ql_link); node != NULL; node =
+	    ql_last(&arena->achunks, ql_link)) {
+		ql_remove(&arena->achunks, node, ql_link);
+		arena_chunk_discard(tsd_tsdn(tsd), arena,
+		    extent_node_addr_get(node));
+	}
+
+	/* Spare. */
+	if (arena->spare != NULL) {
+		arena_chunk_discard(tsd_tsdn(tsd), arena, arena->spare);
+		arena->spare = NULL;
+	}
 
-	malloc_mutex_lock(&arena->lock);
-	arena_purge(arena, true);
-	malloc_mutex_unlock(&arena->lock);
+	assert(!arena->purging);
+	arena->nactive = 0;
+
+	for (i = 0; i < NPSIZES; i++)
+		arena_run_heap_new(&arena->runs_avail[i]);
+
+	malloc_mutex_unlock(tsd_tsdn(tsd), &arena->lock);
 }
 
 static void
@@ -1660,21 +2116,9 @@ arena_run_size_get(arena_t *arena, arena_chunk_t *chunk, arena_run_t *run,
 	return (size);
 }
 
-static bool
-arena_run_decommit(arena_t *arena, arena_chunk_t *chunk, arena_run_t *run)
-{
-	arena_chunk_map_misc_t *miscelm = arena_run_to_miscelm(run);
-	size_t run_ind = arena_miscelm_to_pageind(miscelm);
-	size_t offset = run_ind << LG_PAGE;
-	size_t length = arena_run_size_get(arena, chunk, run, run_ind);
-
-	return (arena->chunk_hooks.decommit(chunk, chunksize, offset, length,
-	    arena->ind));
-}
-
 static void
-arena_run_dalloc(arena_t *arena, arena_run_t *run, bool dirty, bool cleaned,
-    bool decommitted)
+arena_run_dalloc(tsdn_t *tsdn, arena_t *arena, arena_run_t *run, bool dirty,
+    bool cleaned, bool decommitted)
 {
 	arena_chunk_t *chunk;
 	arena_chunk_map_misc_t *miscelm;
@@ -1687,8 +2131,7 @@ arena_run_dalloc(arena_t *arena, arena_run_t *run, bool dirty, bool cleaned,
 	assert(run_ind < chunk_npages);
 	size = arena_run_size_get(arena, chunk, run, run_ind);
 	run_pages = (size >> LG_PAGE);
-	arena_cactive_update(arena, 0, run_pages);
-	arena->nactive -= run_pages;
+	arena_nactive_sub(arena, run_pages);
 
 	/*
 	 * The run is dirty if the caller claims to have dirtied it, as well as
@@ -1735,7 +2178,7 @@ arena_run_dalloc(arena_t *arena, arena_run_t *run, bool dirty, bool cleaned,
 	if (size == arena_maxrun) {
 		assert(run_ind == map_bias);
 		assert(run_pages == (arena_maxrun >> LG_PAGE));
-		arena_chunk_dalloc(arena, chunk);
+		arena_chunk_dalloc(tsdn, arena, chunk);
 	}
 
 	/*
@@ -1746,21 +2189,12 @@ arena_run_dalloc(arena_t *arena, arena_run_t *run, bool dirty, bool cleaned,
 	 * chances of spuriously crossing the dirty page purging threshold.
 	 */
 	if (dirty)
-		arena_maybe_purge(arena);
-}
-
-static void
-arena_run_dalloc_decommit(arena_t *arena, arena_chunk_t *chunk,
-    arena_run_t *run)
-{
-	bool committed = arena_run_decommit(arena, chunk, run);
-
-	arena_run_dalloc(arena, run, committed, false, !committed);
+		arena_maybe_purge(tsdn, arena);
 }
 
 static void
-arena_run_trim_head(arena_t *arena, arena_chunk_t *chunk, arena_run_t *run,
-    size_t oldsize, size_t newsize)
+arena_run_trim_head(tsdn_t *tsdn, arena_t *arena, arena_chunk_t *chunk,
+    arena_run_t *run, size_t oldsize, size_t newsize)
 {
 	arena_chunk_map_misc_t *miscelm = arena_run_to_miscelm(run);
 	size_t pageind = arena_miscelm_to_pageind(miscelm);
@@ -1795,12 +2229,13 @@ arena_run_trim_head(arena_t *arena, arena_chunk_t *chunk, arena_run_t *run,
 	    flag_dirty | (flag_unzeroed_mask & arena_mapbits_unzeroed_get(chunk,
 	    pageind+head_npages)));
 
-	arena_run_dalloc(arena, run, false, false, (flag_decommitted != 0));
+	arena_run_dalloc(tsdn, arena, run, false, false, (flag_decommitted !=
+	    0));
 }
 
 static void
-arena_run_trim_tail(arena_t *arena, arena_chunk_t *chunk, arena_run_t *run,
-    size_t oldsize, size_t newsize, bool dirty)
+arena_run_trim_tail(tsdn_t *tsdn, arena_t *arena, arena_chunk_t *chunk,
+    arena_run_t *run, size_t oldsize, size_t newsize, bool dirty)
 {
 	arena_chunk_map_misc_t *miscelm = arena_run_to_miscelm(run);
 	size_t pageind = arena_miscelm_to_pageind(miscelm);
@@ -1837,20 +2272,10 @@ arena_run_trim_tail(arena_t *arena, arena_chunk_t *chunk, arena_run_t *run,
 	    flag_dirty | (flag_unzeroed_mask & arena_mapbits_unzeroed_get(chunk,
 	    pageind+head_npages)));
 
-	tail_miscelm = arena_miscelm_get(chunk, pageind + head_npages);
+	tail_miscelm = arena_miscelm_get_mutable(chunk, pageind + head_npages);
 	tail_run = &tail_miscelm->run;
-	arena_run_dalloc(arena, tail_run, dirty, false, (flag_decommitted !=
-	    0));
-}
-
-static arena_run_t *
-arena_bin_runs_first(arena_bin_t *bin)
-{
-	arena_chunk_map_misc_t *miscelm = arena_run_tree_first(&bin->runs);
-	if (miscelm != NULL)
-		return (&miscelm->run);
-
-	return (NULL);
+	arena_run_dalloc(tsdn, arena, tail_run, dirty, false, (flag_decommitted
+	    != 0));
 }
 
 static void
@@ -1858,35 +2283,25 @@ arena_bin_runs_insert(arena_bin_t *bin, arena_run_t *run)
 {
 	arena_chunk_map_misc_t *miscelm = arena_run_to_miscelm(run);
 
-	assert(arena_run_tree_search(&bin->runs, miscelm) == NULL);
-
-	arena_run_tree_insert(&bin->runs, miscelm);
-}
-
-static void
-arena_bin_runs_remove(arena_bin_t *bin, arena_run_t *run)
-{
-	arena_chunk_map_misc_t *miscelm = arena_run_to_miscelm(run);
-
-	assert(arena_run_tree_search(&bin->runs, miscelm) != NULL);
-
-	arena_run_tree_remove(&bin->runs, miscelm);
+	arena_run_heap_insert(&bin->runs, miscelm);
 }
 
 static arena_run_t *
 arena_bin_nonfull_run_tryget(arena_bin_t *bin)
 {
-	arena_run_t *run = arena_bin_runs_first(bin);
-	if (run != NULL) {
-		arena_bin_runs_remove(bin, run);
-		if (config_stats)
-			bin->stats.reruns++;
-	}
-	return (run);
+	arena_chunk_map_misc_t *miscelm;
+
+	miscelm = arena_run_heap_remove_first(&bin->runs);
+	if (miscelm == NULL)
+		return (NULL);
+	if (config_stats)
+		bin->stats.reruns++;
+
+	return (&miscelm->run);
 }
 
 static arena_run_t *
-arena_bin_nonfull_run_get(arena_t *arena, arena_bin_t *bin)
+arena_bin_nonfull_run_get(tsdn_t *tsdn, arena_t *arena, arena_bin_t *bin)
 {
 	arena_run_t *run;
 	szind_t binind;
@@ -1902,19 +2317,19 @@ arena_bin_nonfull_run_get(arena_t *arena, arena_bin_t *bin)
 	bin_info = &arena_bin_info[binind];
 
 	/* Allocate a new run. */
-	malloc_mutex_unlock(&bin->lock);
+	malloc_mutex_unlock(tsdn, &bin->lock);
 	/******************************/
-	malloc_mutex_lock(&arena->lock);
-	run = arena_run_alloc_small(arena, bin_info->run_size, binind);
+	malloc_mutex_lock(tsdn, &arena->lock);
+	run = arena_run_alloc_small(tsdn, arena, bin_info->run_size, binind);
 	if (run != NULL) {
 		/* Initialize run internals. */
 		run->binind = binind;
 		run->nfree = bin_info->nregs;
 		bitmap_init(run->bitmap, &bin_info->bitmap_info);
 	}
-	malloc_mutex_unlock(&arena->lock);
+	malloc_mutex_unlock(tsdn, &arena->lock);
 	/********************************/
-	malloc_mutex_lock(&bin->lock);
+	malloc_mutex_lock(tsdn, &bin->lock);
 	if (run != NULL) {
 		if (config_stats) {
 			bin->stats.nruns++;
@@ -1937,7 +2352,7 @@ arena_bin_nonfull_run_get(arena_t *arena, arena_bin_t *bin)
 
 /* Re-fill bin->runcur, then call arena_run_reg_alloc(). */
 static void *
-arena_bin_malloc_hard(arena_t *arena, arena_bin_t *bin)
+arena_bin_malloc_hard(tsdn_t *tsdn, arena_t *arena, arena_bin_t *bin)
 {
 	szind_t binind;
 	arena_bin_info_t *bin_info;
@@ -1946,7 +2361,7 @@ arena_bin_malloc_hard(arena_t *arena, arena_bin_t *bin)
 	binind = arena_bin_index(arena, bin);
 	bin_info = &arena_bin_info[binind];
 	bin->runcur = NULL;
-	run = arena_bin_nonfull_run_get(arena, bin);
+	run = arena_bin_nonfull_run_get(tsdn, arena, bin);
 	if (bin->runcur != NULL && bin->runcur->nfree > 0) {
 		/*
 		 * Another thread updated runcur while this one ran without the
@@ -1967,10 +2382,11 @@ arena_bin_malloc_hard(arena_t *arena, arena_bin_t *bin)
 			 * were just deallocated from the run.
 			 */
 			chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(run);
-			if (run->nfree == bin_info->nregs)
-				arena_dalloc_bin_run(arena, chunk, run, bin);
-			else
-				arena_bin_lower_run(arena, chunk, run, bin);
+			if (run->nfree == bin_info->nregs) {
+				arena_dalloc_bin_run(tsdn, arena, chunk, run,
+				    bin);
+			} else
+				arena_bin_lower_run(arena, run, bin);
 		}
 		return (ret);
 	}
@@ -1986,18 +2402,18 @@ arena_bin_malloc_hard(arena_t *arena, arena_bin_t *bin)
 }
 
 void
-arena_tcache_fill_small(arena_t *arena, tcache_bin_t *tbin, szind_t binind,
-    uint64_t prof_accumbytes)
+arena_tcache_fill_small(tsdn_t *tsdn, arena_t *arena, tcache_bin_t *tbin,
+    szind_t binind, uint64_t prof_accumbytes)
 {
 	unsigned i, nfill;
 	arena_bin_t *bin;
 
 	assert(tbin->ncached == 0);
 
-	if (config_prof && arena_prof_accum(arena, prof_accumbytes))
-		prof_idump();
+	if (config_prof && arena_prof_accum(tsdn, arena, prof_accumbytes))
+		prof_idump(tsdn);
 	bin = &arena->bins[binind];
-	malloc_mutex_lock(&bin->lock);
+	malloc_mutex_lock(tsdn, &bin->lock);
 	for (i = 0, nfill = (tcache_bin_info[binind].ncached_max >>
 	    tbin->lg_fill_div); i < nfill; i++) {
 		arena_run_t *run;
@@ -2005,16 +2421,15 @@ arena_tcache_fill_small(arena_t *arena, tcache_bin_t *tbin, szind_t binind,
 		if ((run = bin->runcur) != NULL && run->nfree > 0)
 			ptr = arena_run_reg_alloc(run, &arena_bin_info[binind]);
 		else
-			ptr = arena_bin_malloc_hard(arena, bin);
+			ptr = arena_bin_malloc_hard(tsdn, arena, bin);
 		if (ptr == NULL) {
 			/*
 			 * OOM.  tbin->avail isn't yet filled down to its first
 			 * element, so the successful allocations (if any) must
-			 * be moved to the base of tbin->avail before bailing
-			 * out.
+			 * be moved just before tbin->avail before bailing out.
 			 */
 			if (i > 0) {
-				memmove(tbin->avail, &tbin->avail[nfill - i],
+				memmove(tbin->avail - i, tbin->avail - nfill,
 				    i * sizeof(void *));
 			}
 			break;
@@ -2024,7 +2439,7 @@ arena_tcache_fill_small(arena_t *arena, tcache_bin_t *tbin, szind_t binind,
 			    true);
 		}
 		/* Insert such that low regions get used first. */
-		tbin->avail[nfill - 1 - i] = ptr;
+		*(tbin->avail - nfill + i) = ptr;
 	}
 	if (config_stats) {
 		bin->stats.nmalloc += i;
@@ -2033,29 +2448,31 @@ arena_tcache_fill_small(arena_t *arena, tcache_bin_t *tbin, szind_t binind,
 		bin->stats.nfills++;
 		tbin->tstats.nrequests = 0;
 	}
-	malloc_mutex_unlock(&bin->lock);
+	malloc_mutex_unlock(tsdn, &bin->lock);
 	tbin->ncached = i;
+	arena_decay_tick(tsdn, arena);
 }
 
 void
 arena_alloc_junk_small(void *ptr, arena_bin_info_t *bin_info, bool zero)
 {
 
+	size_t redzone_size = bin_info->redzone_size;
+
 	if (zero) {
-		size_t redzone_size = bin_info->redzone_size;
-		memset((void *)((uintptr_t)ptr - redzone_size), 0xa5,
-		    redzone_size);
-		memset((void *)((uintptr_t)ptr + bin_info->reg_size), 0xa5,
-		    redzone_size);
+		memset((void *)((uintptr_t)ptr - redzone_size),
+		    JEMALLOC_ALLOC_JUNK, redzone_size);
+		memset((void *)((uintptr_t)ptr + bin_info->reg_size),
+		    JEMALLOC_ALLOC_JUNK, redzone_size);
 	} else {
-		memset((void *)((uintptr_t)ptr - bin_info->redzone_size), 0xa5,
-		    bin_info->reg_interval);
+		memset((void *)((uintptr_t)ptr - redzone_size),
+		    JEMALLOC_ALLOC_JUNK, bin_info->reg_interval);
 	}
 }
 
 #ifdef JEMALLOC_JET
 #undef arena_redzone_corruption
-#define	arena_redzone_corruption JEMALLOC_N(arena_redzone_corruption_impl)
+#define	arena_redzone_corruption JEMALLOC_N(n_arena_redzone_corruption)
 #endif
 static void
 arena_redzone_corruption(void *ptr, size_t usize, bool after,
@@ -2070,7 +2487,7 @@ arena_redzone_corruption(void *ptr, size_t usize, bool after,
 #undef arena_redzone_corruption
 #define	arena_redzone_corruption JEMALLOC_N(arena_redzone_corruption)
 arena_redzone_corruption_t *arena_redzone_corruption =
-    JEMALLOC_N(arena_redzone_corruption_impl);
+    JEMALLOC_N(n_arena_redzone_corruption);
 #endif
 
 static void
@@ -2085,22 +2502,22 @@ arena_redzones_validate(void *ptr, arena_bin_info_t *bin_info, bool reset)
 
 		for (i = 1; i <= redzone_size; i++) {
 			uint8_t *byte = (uint8_t *)((uintptr_t)ptr - i);
-			if (*byte != 0xa5) {
+			if (*byte != JEMALLOC_ALLOC_JUNK) {
 				error = true;
 				arena_redzone_corruption(ptr, size, false, i,
 				    *byte);
 				if (reset)
-					*byte = 0xa5;
+					*byte = JEMALLOC_ALLOC_JUNK;
 			}
 		}
 		for (i = 0; i < redzone_size; i++) {
 			uint8_t *byte = (uint8_t *)((uintptr_t)ptr + size + i);
-			if (*byte != 0xa5) {
+			if (*byte != JEMALLOC_ALLOC_JUNK) {
 				error = true;
 				arena_redzone_corruption(ptr, size, true, i,
 				    *byte);
 				if (reset)
-					*byte = 0xa5;
+					*byte = JEMALLOC_ALLOC_JUNK;
 			}
 		}
 	}
@@ -2111,7 +2528,7 @@ arena_redzones_validate(void *ptr, arena_bin_info_t *bin_info, bool reset)
 
 #ifdef JEMALLOC_JET
 #undef arena_dalloc_junk_small
-#define	arena_dalloc_junk_small JEMALLOC_N(arena_dalloc_junk_small_impl)
+#define	arena_dalloc_junk_small JEMALLOC_N(n_arena_dalloc_junk_small)
 #endif
 void
 arena_dalloc_junk_small(void *ptr, arena_bin_info_t *bin_info)
@@ -2119,14 +2536,14 @@ arena_dalloc_junk_small(void *ptr, arena_bin_info_t *bin_info)
 	size_t redzone_size = bin_info->redzone_size;
 
 	arena_redzones_validate(ptr, bin_info, false);
-	memset((void *)((uintptr_t)ptr - redzone_size), 0x5a,
+	memset((void *)((uintptr_t)ptr - redzone_size), JEMALLOC_FREE_JUNK,
 	    bin_info->reg_interval);
 }
 #ifdef JEMALLOC_JET
 #undef arena_dalloc_junk_small
 #define	arena_dalloc_junk_small JEMALLOC_N(arena_dalloc_junk_small)
 arena_dalloc_junk_small_t *arena_dalloc_junk_small =
-    JEMALLOC_N(arena_dalloc_junk_small_impl);
+    JEMALLOC_N(n_arena_dalloc_junk_small);
 #endif
 
 void
@@ -2144,27 +2561,26 @@ arena_quarantine_junk_small(void *ptr, size_t usize)
 	arena_redzones_validate(ptr, bin_info, true);
 }
 
-void *
-arena_malloc_small(arena_t *arena, size_t size, bool zero)
+static void *
+arena_malloc_small(tsdn_t *tsdn, arena_t *arena, szind_t binind, bool zero)
 {
 	void *ret;
 	arena_bin_t *bin;
+	size_t usize;
 	arena_run_t *run;
-	szind_t binind;
 
-	binind = size2index(size);
 	assert(binind < NBINS);
 	bin = &arena->bins[binind];
-	size = index2size(binind);
+	usize = index2size(binind);
 
-	malloc_mutex_lock(&bin->lock);
+	malloc_mutex_lock(tsdn, &bin->lock);
 	if ((run = bin->runcur) != NULL && run->nfree > 0)
 		ret = arena_run_reg_alloc(run, &arena_bin_info[binind]);
 	else
-		ret = arena_bin_malloc_hard(arena, bin);
+		ret = arena_bin_malloc_hard(tsdn, arena, bin);
 
 	if (ret == NULL) {
-		malloc_mutex_unlock(&bin->lock);
+		malloc_mutex_unlock(tsdn, &bin->lock);
 		return (NULL);
 	}
 
@@ -2173,9 +2589,9 @@ arena_malloc_small(arena_t *arena, size_t size, bool zero)
 		bin->stats.nrequests++;
 		bin->stats.curregs++;
 	}
-	malloc_mutex_unlock(&bin->lock);
-	if (config_prof && !isthreaded && arena_prof_accum(arena, size))
-		prof_idump();
+	malloc_mutex_unlock(tsdn, &bin->lock);
+	if (config_prof && !isthreaded && arena_prof_accum(tsdn, arena, usize))
+		prof_idump(tsdn);
 
 	if (!zero) {
 		if (config_fill) {
@@ -2183,34 +2599,35 @@ arena_malloc_small(arena_t *arena, size_t size, bool zero)
 				arena_alloc_junk_small(ret,
 				    &arena_bin_info[binind], false);
 			} else if (unlikely(opt_zero))
-				memset(ret, 0, size);
+				memset(ret, 0, usize);
 		}
-		JEMALLOC_VALGRIND_MAKE_MEM_UNDEFINED(ret, size);
+		JEMALLOC_VALGRIND_MAKE_MEM_UNDEFINED(ret, usize);
 	} else {
 		if (config_fill && unlikely(opt_junk_alloc)) {
 			arena_alloc_junk_small(ret, &arena_bin_info[binind],
 			    true);
 		}
-		JEMALLOC_VALGRIND_MAKE_MEM_UNDEFINED(ret, size);
-		memset(ret, 0, size);
+		JEMALLOC_VALGRIND_MAKE_MEM_UNDEFINED(ret, usize);
+		memset(ret, 0, usize);
 	}
 
+	arena_decay_tick(tsdn, arena);
 	return (ret);
 }
 
 void *
-arena_malloc_large(arena_t *arena, size_t size, bool zero)
+arena_malloc_large(tsdn_t *tsdn, arena_t *arena, szind_t binind, bool zero)
 {
 	void *ret;
 	size_t usize;
 	uintptr_t random_offset;
 	arena_run_t *run;
 	arena_chunk_map_misc_t *miscelm;
-	UNUSED bool idump;
+	UNUSED bool idump JEMALLOC_CC_SILENCE_INIT(false);
 
 	/* Large allocation. */
-	usize = s2u(size);
-	malloc_mutex_lock(&arena->lock);
+	usize = index2size(binind);
+	malloc_mutex_lock(tsdn, &arena->lock);
 	if (config_cache_oblivious) {
 		uint64_t r;
 
@@ -2219,22 +2636,21 @@ arena_malloc_large(arena_t *arena, size_t size, bool zero)
 		 * that is a multiple of the cacheline size, e.g. [0 .. 63) * 64
 		 * for 4 KiB pages and 64-byte cachelines.
 		 */
-		prng64(r, LG_PAGE - LG_CACHELINE, arena->offset_state,
-		    UINT64_C(6364136223846793009),
-		    UINT64_C(1442695040888963409));
+		r = prng_lg_range_zu(&arena->offset_state, LG_PAGE -
+		    LG_CACHELINE, false);
 		random_offset = ((uintptr_t)r) << LG_CACHELINE;
 	} else
 		random_offset = 0;
-	run = arena_run_alloc_large(arena, usize + large_pad, zero);
+	run = arena_run_alloc_large(tsdn, arena, usize + large_pad, zero);
 	if (run == NULL) {
-		malloc_mutex_unlock(&arena->lock);
+		malloc_mutex_unlock(tsdn, &arena->lock);
 		return (NULL);
 	}
 	miscelm = arena_run_to_miscelm(run);
 	ret = (void *)((uintptr_t)arena_miscelm_to_rpages(miscelm) +
 	    random_offset);
 	if (config_stats) {
-		szind_t index = size2index(usize) - NBINS;
+		szind_t index = binind - NBINS;
 
 		arena->stats.nmalloc_large++;
 		arena->stats.nrequests_large++;
@@ -2245,25 +2661,45 @@ arena_malloc_large(arena_t *arena, size_t size, bool zero)
 	}
 	if (config_prof)
 		idump = arena_prof_accum_locked(arena, usize);
-	malloc_mutex_unlock(&arena->lock);
+	malloc_mutex_unlock(tsdn, &arena->lock);
 	if (config_prof && idump)
-		prof_idump();
+		prof_idump(tsdn);
 
 	if (!zero) {
 		if (config_fill) {
 			if (unlikely(opt_junk_alloc))
-				memset(ret, 0xa5, usize);
+				memset(ret, JEMALLOC_ALLOC_JUNK, usize);
 			else if (unlikely(opt_zero))
 				memset(ret, 0, usize);
 		}
 	}
 
+	arena_decay_tick(tsdn, arena);
 	return (ret);
 }
 
+void *
+arena_malloc_hard(tsdn_t *tsdn, arena_t *arena, size_t size, szind_t ind,
+    bool zero)
+{
+
+	assert(!tsdn_null(tsdn) || arena != NULL);
+
+	if (likely(!tsdn_null(tsdn)))
+		arena = arena_choose(tsdn_tsd(tsdn), arena);
+	if (unlikely(arena == NULL))
+		return (NULL);
+
+	if (likely(size <= SMALL_MAXCLASS))
+		return (arena_malloc_small(tsdn, arena, ind, zero));
+	if (likely(size <= large_maxclass))
+		return (arena_malloc_large(tsdn, arena, ind, zero));
+	return (huge_malloc(tsdn, arena, index2size(ind), zero));
+}
+
 /* Only handles large allocations that require more than page alignment. */
 static void *
-arena_palloc_large(tsd_t *tsd, arena_t *arena, size_t usize, size_t alignment,
+arena_palloc_large(tsdn_t *tsdn, arena_t *arena, size_t usize, size_t alignment,
     bool zero)
 {
 	void *ret;
@@ -2273,19 +2709,21 @@ arena_palloc_large(tsd_t *tsd, arena_t *arena, size_t usize, size_t alignment,
 	arena_chunk_map_misc_t *miscelm;
 	void *rpages;
 
+	assert(!tsdn_null(tsdn) || arena != NULL);
 	assert(usize == PAGE_CEILING(usize));
 
-	arena = arena_choose(tsd, arena);
+	if (likely(!tsdn_null(tsdn)))
+		arena = arena_choose(tsdn_tsd(tsdn), arena);
 	if (unlikely(arena == NULL))
 		return (NULL);
 
 	alignment = PAGE_CEILING(alignment);
 	alloc_size = usize + large_pad + alignment - PAGE;
 
-	malloc_mutex_lock(&arena->lock);
-	run = arena_run_alloc_large(arena, alloc_size, false);
+	malloc_mutex_lock(tsdn, &arena->lock);
+	run = arena_run_alloc_large(tsdn, arena, alloc_size, false);
 	if (run == NULL) {
-		malloc_mutex_unlock(&arena->lock);
+		malloc_mutex_unlock(tsdn, &arena->lock);
 		return (NULL);
 	}
 	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(run);
@@ -2300,16 +2738,16 @@ arena_palloc_large(tsd_t *tsd, arena_t *arena, size_t usize, size_t alignment,
 		arena_chunk_map_misc_t *head_miscelm = miscelm;
 		arena_run_t *head_run = run;
 
-		miscelm = arena_miscelm_get(chunk,
+		miscelm = arena_miscelm_get_mutable(chunk,
 		    arena_miscelm_to_pageind(head_miscelm) + (leadsize >>
 		    LG_PAGE));
 		run = &miscelm->run;
 
-		arena_run_trim_head(arena, chunk, head_run, alloc_size,
+		arena_run_trim_head(tsdn, arena, chunk, head_run, alloc_size,
 		    alloc_size - leadsize);
 	}
 	if (trailsize != 0) {
-		arena_run_trim_tail(arena, chunk, run, usize + large_pad +
+		arena_run_trim_tail(tsdn, arena, chunk, run, usize + large_pad +
 		    trailsize, usize + large_pad, false);
 	}
 	if (arena_run_init_large(arena, run, usize + large_pad, zero)) {
@@ -2320,8 +2758,8 @@ arena_palloc_large(tsd_t *tsd, arena_t *arena, size_t usize, size_t alignment,
 		    run_ind) != 0);
 
 		assert(decommitted); /* Cause of OOM. */
-		arena_run_dalloc(arena, run, dirty, false, decommitted);
-		malloc_mutex_unlock(&arena->lock);
+		arena_run_dalloc(tsdn, arena, run, dirty, false, decommitted);
+		malloc_mutex_unlock(tsdn, &arena->lock);
 		return (NULL);
 	}
 	ret = arena_miscelm_to_rpages(miscelm);
@@ -2336,19 +2774,20 @@ arena_palloc_large(tsd_t *tsd, arena_t *arena, size_t usize, size_t alignment,
 		arena->stats.lstats[index].nrequests++;
 		arena->stats.lstats[index].curruns++;
 	}
-	malloc_mutex_unlock(&arena->lock);
+	malloc_mutex_unlock(tsdn, &arena->lock);
 
 	if (config_fill && !zero) {
 		if (unlikely(opt_junk_alloc))
-			memset(ret, 0xa5, usize);
+			memset(ret, JEMALLOC_ALLOC_JUNK, usize);
 		else if (unlikely(opt_zero))
 			memset(ret, 0, usize);
 	}
+	arena_decay_tick(tsdn, arena);
 	return (ret);
 }
 
 void *
-arena_palloc(tsd_t *tsd, arena_t *arena, size_t usize, size_t alignment,
+arena_palloc(tsdn_t *tsdn, arena_t *arena, size_t usize, size_t alignment,
     bool zero, tcache_t *tcache)
 {
 	void *ret;
@@ -2356,7 +2795,8 @@ arena_palloc(tsd_t *tsd, arena_t *arena, size_t usize, size_t alignment,
 	if (usize <= SMALL_MAXCLASS && (alignment < PAGE || (alignment == PAGE
 	    && (usize & PAGE_MASK) == 0))) {
 		/* Small; alignment doesn't require special run placement. */
-		ret = arena_malloc(tsd, arena, usize, zero, tcache);
+		ret = arena_malloc(tsdn, arena, usize, size2index(usize), zero,
+		    tcache, true);
 	} else if (usize <= large_maxclass && alignment <= PAGE) {
 		/*
 		 * Large; alignment doesn't require special run placement.
@@ -2364,25 +2804,25 @@ arena_palloc(tsd_t *tsd, arena_t *arena, size_t usize, size_t alignment,
 		 * the base of the run, so do some bit manipulation to retrieve
 		 * the base.
 		 */
-		ret = arena_malloc(tsd, arena, usize, zero, tcache);
+		ret = arena_malloc(tsdn, arena, usize, size2index(usize), zero,
+		    tcache, true);
 		if (config_cache_oblivious)
 			ret = (void *)((uintptr_t)ret & ~PAGE_MASK);
 	} else {
 		if (likely(usize <= large_maxclass)) {
-			ret = arena_palloc_large(tsd, arena, usize, alignment,
+			ret = arena_palloc_large(tsdn, arena, usize, alignment,
 			    zero);
 		} else if (likely(alignment <= chunksize))
-			ret = huge_malloc(tsd, arena, usize, zero, tcache);
+			ret = huge_malloc(tsdn, arena, usize, zero);
 		else {
-			ret = huge_palloc(tsd, arena, usize, alignment, zero,
-			    tcache);
+			ret = huge_palloc(tsdn, arena, usize, alignment, zero);
 		}
 	}
 	return (ret);
 }
 
 void
-arena_prof_promoted(const void *ptr, size_t size)
+arena_prof_promoted(tsdn_t *tsdn, const void *ptr, size_t size)
 {
 	arena_chunk_t *chunk;
 	size_t pageind;
@@ -2391,8 +2831,8 @@ arena_prof_promoted(const void *ptr, size_t size)
 	cassert(config_prof);
 	assert(ptr != NULL);
 	assert(CHUNK_ADDR2BASE(ptr) != ptr);
-	assert(isalloc(ptr, false) == LARGE_MINCLASS);
-	assert(isalloc(ptr, true) == LARGE_MINCLASS);
+	assert(isalloc(tsdn, ptr, false) == LARGE_MINCLASS);
+	assert(isalloc(tsdn, ptr, true) == LARGE_MINCLASS);
 	assert(size <= SMALL_MAXCLASS);
 
 	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(ptr);
@@ -2401,8 +2841,8 @@ arena_prof_promoted(const void *ptr, size_t size)
 	assert(binind < NBINS);
 	arena_mapbits_large_binind_set(chunk, pageind, binind);
 
-	assert(isalloc(ptr, false) == LARGE_MINCLASS);
-	assert(isalloc(ptr, true) == size);
+	assert(isalloc(tsdn, ptr, false) == LARGE_MINCLASS);
+	assert(isalloc(tsdn, ptr, true) == size);
 }
 
 static void
@@ -2418,48 +2858,51 @@ arena_dissociate_bin_run(arena_chunk_t *chunk, arena_run_t *run,
 		    &chunk->node), bin);
 		arena_bin_info_t *bin_info = &arena_bin_info[binind];
 
+		/*
+		 * The following block's conditional is necessary because if the
+		 * run only contains one region, then it never gets inserted
+		 * into the non-full runs tree.
+		 */
 		if (bin_info->nregs != 1) {
-			/*
-			 * This block's conditional is necessary because if the
-			 * run only contains one region, then it never gets
-			 * inserted into the non-full runs tree.
-			 */
-			arena_bin_runs_remove(bin, run);
+			arena_chunk_map_misc_t *miscelm =
+			    arena_run_to_miscelm(run);
+
+			arena_run_heap_remove(&bin->runs, miscelm);
 		}
 	}
 }
 
 static void
-arena_dalloc_bin_run(arena_t *arena, arena_chunk_t *chunk, arena_run_t *run,
-    arena_bin_t *bin)
+arena_dalloc_bin_run(tsdn_t *tsdn, arena_t *arena, arena_chunk_t *chunk,
+    arena_run_t *run, arena_bin_t *bin)
 {
 
 	assert(run != bin->runcur);
-	assert(arena_run_tree_search(&bin->runs, arena_run_to_miscelm(run)) ==
-	    NULL);
 
-	malloc_mutex_unlock(&bin->lock);
+	malloc_mutex_unlock(tsdn, &bin->lock);
 	/******************************/
-	malloc_mutex_lock(&arena->lock);
-	arena_run_dalloc_decommit(arena, chunk, run);
-	malloc_mutex_unlock(&arena->lock);
+	malloc_mutex_lock(tsdn, &arena->lock);
+	arena_run_dalloc(tsdn, arena, run, true, false, false);
+	malloc_mutex_unlock(tsdn, &arena->lock);
 	/****************************/
-	malloc_mutex_lock(&bin->lock);
+	malloc_mutex_lock(tsdn, &bin->lock);
 	if (config_stats)
 		bin->stats.curruns--;
 }
 
 static void
-arena_bin_lower_run(arena_t *arena, arena_chunk_t *chunk, arena_run_t *run,
-    arena_bin_t *bin)
+arena_bin_lower_run(arena_t *arena, arena_run_t *run, arena_bin_t *bin)
 {
 
 	/*
-	 * Make sure that if bin->runcur is non-NULL, it refers to the lowest
-	 * non-full run.  It is okay to NULL runcur out rather than proactively
-	 * keeping it pointing at the lowest non-full run.
+	 * Make sure that if bin->runcur is non-NULL, it refers to the
+	 * oldest/lowest non-full run.  It is okay to NULL runcur out rather
+	 * than proactively keeping it pointing at the oldest/lowest non-full
+	 * run.
 	 */
-	if ((uintptr_t)run < (uintptr_t)bin->runcur) {
+	if (bin->runcur != NULL &&
+	    arena_snad_comp(arena_run_to_miscelm(bin->runcur),
+	    arena_run_to_miscelm(run)) > 0) {
 		/* Switch runcur. */
 		if (bin->runcur->nfree > 0)
 			arena_bin_runs_insert(bin, bin->runcur);
@@ -2471,8 +2914,8 @@ arena_bin_lower_run(arena_t *arena, arena_chunk_t *chunk, arena_run_t *run,
 }
 
 static void
-arena_dalloc_bin_locked_impl(arena_t *arena, arena_chunk_t *chunk, void *ptr,
-    arena_chunk_map_bits_t *bitselm, bool junked)
+arena_dalloc_bin_locked_impl(tsdn_t *tsdn, arena_t *arena, arena_chunk_t *chunk,
+    void *ptr, arena_chunk_map_bits_t *bitselm, bool junked)
 {
 	size_t pageind, rpages_ind;
 	arena_run_t *run;
@@ -2482,7 +2925,7 @@ arena_dalloc_bin_locked_impl(arena_t *arena, arena_chunk_t *chunk, void *ptr,
 
 	pageind = ((uintptr_t)ptr - (uintptr_t)chunk) >> LG_PAGE;
 	rpages_ind = pageind - arena_mapbits_small_runind_get(chunk, pageind);
-	run = &arena_miscelm_get(chunk, rpages_ind)->run;
+	run = &arena_miscelm_get_mutable(chunk, rpages_ind)->run;
 	binind = run->binind;
 	bin = &arena->bins[binind];
 	bin_info = &arena_bin_info[binind];
@@ -2493,9 +2936,9 @@ arena_dalloc_bin_locked_impl(arena_t *arena, arena_chunk_t *chunk, void *ptr,
 	arena_run_reg_dalloc(run, ptr);
 	if (run->nfree == bin_info->nregs) {
 		arena_dissociate_bin_run(chunk, run, bin);
-		arena_dalloc_bin_run(arena, chunk, run, bin);
+		arena_dalloc_bin_run(tsdn, arena, chunk, run, bin);
 	} else if (run->nfree == 1 && run != bin->runcur)
-		arena_bin_lower_run(arena, chunk, run, bin);
+		arena_bin_lower_run(arena, run, bin);
 
 	if (config_stats) {
 		bin->stats.ndalloc++;
@@ -2504,15 +2947,15 @@ arena_dalloc_bin_locked_impl(arena_t *arena, arena_chunk_t *chunk, void *ptr,
 }
 
 void
-arena_dalloc_bin_junked_locked(arena_t *arena, arena_chunk_t *chunk, void *ptr,
-    arena_chunk_map_bits_t *bitselm)
+arena_dalloc_bin_junked_locked(tsdn_t *tsdn, arena_t *arena,
+    arena_chunk_t *chunk, void *ptr, arena_chunk_map_bits_t *bitselm)
 {
 
-	arena_dalloc_bin_locked_impl(arena, chunk, ptr, bitselm, true);
+	arena_dalloc_bin_locked_impl(tsdn, arena, chunk, ptr, bitselm, true);
 }
 
 void
-arena_dalloc_bin(arena_t *arena, arena_chunk_t *chunk, void *ptr,
+arena_dalloc_bin(tsdn_t *tsdn, arena_t *arena, arena_chunk_t *chunk, void *ptr,
     size_t pageind, arena_chunk_map_bits_t *bitselm)
 {
 	arena_run_t *run;
@@ -2520,16 +2963,16 @@ arena_dalloc_bin(arena_t *arena, arena_chunk_t *chunk, void *ptr,
 	size_t rpages_ind;
 
 	rpages_ind = pageind - arena_mapbits_small_runind_get(chunk, pageind);
-	run = &arena_miscelm_get(chunk, rpages_ind)->run;
+	run = &arena_miscelm_get_mutable(chunk, rpages_ind)->run;
 	bin = &arena->bins[run->binind];
-	malloc_mutex_lock(&bin->lock);
-	arena_dalloc_bin_locked_impl(arena, chunk, ptr, bitselm, false);
-	malloc_mutex_unlock(&bin->lock);
+	malloc_mutex_lock(tsdn, &bin->lock);
+	arena_dalloc_bin_locked_impl(tsdn, arena, chunk, ptr, bitselm, false);
+	malloc_mutex_unlock(tsdn, &bin->lock);
 }
 
 void
-arena_dalloc_small(arena_t *arena, arena_chunk_t *chunk, void *ptr,
-    size_t pageind)
+arena_dalloc_small(tsdn_t *tsdn, arena_t *arena, arena_chunk_t *chunk,
+    void *ptr, size_t pageind)
 {
 	arena_chunk_map_bits_t *bitselm;
 
@@ -2538,34 +2981,36 @@ arena_dalloc_small(arena_t *arena, arena_chunk_t *chunk, void *ptr,
 		assert(arena_ptr_small_binind_get(ptr, arena_mapbits_get(chunk,
 		    pageind)) != BININD_INVALID);
 	}
-	bitselm = arena_bitselm_get(chunk, pageind);
-	arena_dalloc_bin(arena, chunk, ptr, pageind, bitselm);
+	bitselm = arena_bitselm_get_mutable(chunk, pageind);
+	arena_dalloc_bin(tsdn, arena, chunk, ptr, pageind, bitselm);
+	arena_decay_tick(tsdn, arena);
 }
 
 #ifdef JEMALLOC_JET
 #undef arena_dalloc_junk_large
-#define	arena_dalloc_junk_large JEMALLOC_N(arena_dalloc_junk_large_impl)
+#define	arena_dalloc_junk_large JEMALLOC_N(n_arena_dalloc_junk_large)
 #endif
 void
 arena_dalloc_junk_large(void *ptr, size_t usize)
 {
 
 	if (config_fill && unlikely(opt_junk_free))
-		memset(ptr, 0x5a, usize);
+		memset(ptr, JEMALLOC_FREE_JUNK, usize);
 }
 #ifdef JEMALLOC_JET
 #undef arena_dalloc_junk_large
 #define	arena_dalloc_junk_large JEMALLOC_N(arena_dalloc_junk_large)
 arena_dalloc_junk_large_t *arena_dalloc_junk_large =
-    JEMALLOC_N(arena_dalloc_junk_large_impl);
+    JEMALLOC_N(n_arena_dalloc_junk_large);
 #endif
 
 static void
-arena_dalloc_large_locked_impl(arena_t *arena, arena_chunk_t *chunk,
-    void *ptr, bool junked)
+arena_dalloc_large_locked_impl(tsdn_t *tsdn, arena_t *arena,
+    arena_chunk_t *chunk, void *ptr, bool junked)
 {
 	size_t pageind = ((uintptr_t)ptr - (uintptr_t)chunk) >> LG_PAGE;
-	arena_chunk_map_misc_t *miscelm = arena_miscelm_get(chunk, pageind);
+	arena_chunk_map_misc_t *miscelm = arena_miscelm_get_mutable(chunk,
+	    pageind);
 	arena_run_t *run = &miscelm->run;
 
 	if (config_fill || config_stats) {
@@ -2584,32 +3029,35 @@ arena_dalloc_large_locked_impl(arena_t *arena, arena_chunk_t *chunk,
 		}
 	}
 
-	arena_run_dalloc_decommit(arena, chunk, run);
+	arena_run_dalloc(tsdn, arena, run, true, false, false);
 }
 
 void
-arena_dalloc_large_junked_locked(arena_t *arena, arena_chunk_t *chunk,
-    void *ptr)
+arena_dalloc_large_junked_locked(tsdn_t *tsdn, arena_t *arena,
+    arena_chunk_t *chunk, void *ptr)
 {
 
-	arena_dalloc_large_locked_impl(arena, chunk, ptr, true);
+	arena_dalloc_large_locked_impl(tsdn, arena, chunk, ptr, true);
 }
 
 void
-arena_dalloc_large(arena_t *arena, arena_chunk_t *chunk, void *ptr)
+arena_dalloc_large(tsdn_t *tsdn, arena_t *arena, arena_chunk_t *chunk,
+    void *ptr)
 {
 
-	malloc_mutex_lock(&arena->lock);
-	arena_dalloc_large_locked_impl(arena, chunk, ptr, false);
-	malloc_mutex_unlock(&arena->lock);
+	malloc_mutex_lock(tsdn, &arena->lock);
+	arena_dalloc_large_locked_impl(tsdn, arena, chunk, ptr, false);
+	malloc_mutex_unlock(tsdn, &arena->lock);
+	arena_decay_tick(tsdn, arena);
 }
 
 static void
-arena_ralloc_large_shrink(arena_t *arena, arena_chunk_t *chunk, void *ptr,
-    size_t oldsize, size_t size)
+arena_ralloc_large_shrink(tsdn_t *tsdn, arena_t *arena, arena_chunk_t *chunk,
+    void *ptr, size_t oldsize, size_t size)
 {
 	size_t pageind = ((uintptr_t)ptr - (uintptr_t)chunk) >> LG_PAGE;
-	arena_chunk_map_misc_t *miscelm = arena_miscelm_get(chunk, pageind);
+	arena_chunk_map_misc_t *miscelm = arena_miscelm_get_mutable(chunk,
+	    pageind);
 	arena_run_t *run = &miscelm->run;
 
 	assert(size < oldsize);
@@ -2618,8 +3066,8 @@ arena_ralloc_large_shrink(arena_t *arena, arena_chunk_t *chunk, void *ptr,
 	 * Shrink the run, and make trailing pages available for other
 	 * allocations.
 	 */
-	malloc_mutex_lock(&arena->lock);
-	arena_run_trim_tail(arena, chunk, run, oldsize + large_pad, size +
+	malloc_mutex_lock(tsdn, &arena->lock);
+	arena_run_trim_tail(tsdn, arena, chunk, run, oldsize + large_pad, size +
 	    large_pad, true);
 	if (config_stats) {
 		szind_t oldindex = size2index(oldsize) - NBINS;
@@ -2637,12 +3085,12 @@ arena_ralloc_large_shrink(arena_t *arena, arena_chunk_t *chunk, void *ptr,
 		arena->stats.lstats[index].nrequests++;
 		arena->stats.lstats[index].curruns++;
 	}
-	malloc_mutex_unlock(&arena->lock);
+	malloc_mutex_unlock(tsdn, &arena->lock);
 }
 
 static bool
-arena_ralloc_large_grow(arena_t *arena, arena_chunk_t *chunk, void *ptr,
-    size_t oldsize, size_t usize_min, size_t usize_max, bool zero)
+arena_ralloc_large_grow(tsdn_t *tsdn, arena_t *arena, arena_chunk_t *chunk,
+    void *ptr, size_t oldsize, size_t usize_min, size_t usize_max, bool zero)
 {
 	size_t pageind = ((uintptr_t)ptr - (uintptr_t)chunk) >> LG_PAGE;
 	size_t npages = (oldsize + large_pad) >> LG_PAGE;
@@ -2652,7 +3100,7 @@ arena_ralloc_large_grow(arena_t *arena, arena_chunk_t *chunk, void *ptr,
 	    large_pad);
 
 	/* Try to extend the run. */
-	malloc_mutex_lock(&arena->lock);
+	malloc_mutex_lock(tsdn, &arena->lock);
 	if (pageind+npages >= chunk_npages || arena_mapbits_allocated_get(chunk,
 	    pageind+npages) != 0)
 		goto label_fail;
@@ -2675,7 +3123,7 @@ arena_ralloc_large_grow(arena_t *arena, arena_chunk_t *chunk, void *ptr,
 		if (splitsize == 0)
 			goto label_fail;
 
-		run = &arena_miscelm_get(chunk, pageind+npages)->run;
+		run = &arena_miscelm_get_mutable(chunk, pageind+npages)->run;
 		if (arena_run_split_large(arena, run, splitsize, zero))
 			goto label_fail;
 
@@ -2683,10 +3131,16 @@ arena_ralloc_large_grow(arena_t *arena, arena_chunk_t *chunk, void *ptr,
 			/*
 			 * Zero the trailing bytes of the original allocation's
 			 * last page, since they are in an indeterminate state.
+			 * There will always be trailing bytes, because ptr's
+			 * offset from the beginning of the run is a multiple of
+			 * CACHELINE in [0 .. PAGE).
 			 */
-			assert(PAGE_CEILING(oldsize) == oldsize);
-			memset((void *)((uintptr_t)ptr + oldsize), 0,
-			    PAGE_CEILING((uintptr_t)ptr) - (uintptr_t)ptr);
+			void *zbase = (void *)((uintptr_t)ptr + oldsize);
+			void *zpast = PAGE_ADDR2BASE((void *)((uintptr_t)zbase +
+			    PAGE));
+			size_t nzero = (uintptr_t)zpast - (uintptr_t)zbase;
+			assert(nzero > 0);
+			memset(zbase, 0, nzero);
 		}
 
 		size = oldsize + splitsize;
@@ -2726,24 +3180,24 @@ arena_ralloc_large_grow(arena_t *arena, arena_chunk_t *chunk, void *ptr,
 			arena->stats.lstats[index].nrequests++;
 			arena->stats.lstats[index].curruns++;
 		}
-		malloc_mutex_unlock(&arena->lock);
+		malloc_mutex_unlock(tsdn, &arena->lock);
 		return (false);
 	}
 label_fail:
-	malloc_mutex_unlock(&arena->lock);
+	malloc_mutex_unlock(tsdn, &arena->lock);
 	return (true);
 }
 
 #ifdef JEMALLOC_JET
 #undef arena_ralloc_junk_large
-#define	arena_ralloc_junk_large JEMALLOC_N(arena_ralloc_junk_large_impl)
+#define	arena_ralloc_junk_large JEMALLOC_N(n_arena_ralloc_junk_large)
 #endif
 static void
 arena_ralloc_junk_large(void *ptr, size_t old_usize, size_t usize)
 {
 
 	if (config_fill && unlikely(opt_junk_free)) {
-		memset((void *)((uintptr_t)ptr + usize), 0x5a,
+		memset((void *)((uintptr_t)ptr + usize), JEMALLOC_FREE_JUNK,
 		    old_usize - usize);
 	}
 }
@@ -2751,7 +3205,7 @@ arena_ralloc_junk_large(void *ptr, size_t old_usize, size_t usize)
 #undef arena_ralloc_junk_large
 #define	arena_ralloc_junk_large JEMALLOC_N(arena_ralloc_junk_large)
 arena_ralloc_junk_large_t *arena_ralloc_junk_large =
-    JEMALLOC_N(arena_ralloc_junk_large_impl);
+    JEMALLOC_N(n_arena_ralloc_junk_large);
 #endif
 
 /*
@@ -2759,7 +3213,7 @@ arena_ralloc_junk_large_t *arena_ralloc_junk_large =
  * always fail if growing an object, and the following run is already in use.
  */
 static bool
-arena_ralloc_large(void *ptr, size_t oldsize, size_t usize_min,
+arena_ralloc_large(tsdn_t *tsdn, void *ptr, size_t oldsize, size_t usize_min,
     size_t usize_max, bool zero)
 {
 	arena_chunk_t *chunk;
@@ -2774,15 +3228,16 @@ arena_ralloc_large(void *ptr, size_t oldsize, size_t usize_min,
 	arena = extent_node_arena_get(&chunk->node);
 
 	if (oldsize < usize_max) {
-		bool ret = arena_ralloc_large_grow(arena, chunk, ptr, oldsize,
-		    usize_min, usize_max, zero);
+		bool ret = arena_ralloc_large_grow(tsdn, arena, chunk, ptr,
+		    oldsize, usize_min, usize_max, zero);
 		if (config_fill && !ret && !zero) {
 			if (unlikely(opt_junk_alloc)) {
-				memset((void *)((uintptr_t)ptr + oldsize), 0xa5,
-				    isalloc(ptr, config_prof) - oldsize);
+				memset((void *)((uintptr_t)ptr + oldsize),
+				    JEMALLOC_ALLOC_JUNK,
+				    isalloc(tsdn, ptr, config_prof) - oldsize);
 			} else if (unlikely(opt_zero)) {
 				memset((void *)((uintptr_t)ptr + oldsize), 0,
-				    isalloc(ptr, config_prof) - oldsize);
+				    isalloc(tsdn, ptr, config_prof) - oldsize);
 			}
 		}
 		return (ret);
@@ -2791,19 +3246,27 @@ arena_ralloc_large(void *ptr, size_t oldsize, size_t usize_min,
 	assert(oldsize > usize_max);
 	/* Fill before shrinking in order avoid a race. */
 	arena_ralloc_junk_large(ptr, oldsize, usize_max);
-	arena_ralloc_large_shrink(arena, chunk, ptr, oldsize, usize_max);
+	arena_ralloc_large_shrink(tsdn, arena, chunk, ptr, oldsize, usize_max);
 	return (false);
 }
 
 bool
-arena_ralloc_no_move(void *ptr, size_t oldsize, size_t size, size_t extra,
-    bool zero)
+arena_ralloc_no_move(tsdn_t *tsdn, void *ptr, size_t oldsize, size_t size,
+    size_t extra, bool zero)
 {
 	size_t usize_min, usize_max;
 
+	/* Calls with non-zero extra had to clamp extra. */
+	assert(extra == 0 || size + extra <= HUGE_MAXCLASS);
+
+	if (unlikely(size > HUGE_MAXCLASS))
+		return (true);
+
 	usize_min = s2u(size);
 	usize_max = s2u(size + extra);
 	if (likely(oldsize <= large_maxclass && usize_min <= large_maxclass)) {
+		arena_chunk_t *chunk;
+
 		/*
 		 * Avoid moving the allocation if the size class can be left the
 		 * same.
@@ -2811,37 +3274,39 @@ arena_ralloc_no_move(void *ptr, size_t oldsize, size_t size, size_t extra,
 		if (oldsize <= SMALL_MAXCLASS) {
 			assert(arena_bin_info[size2index(oldsize)].reg_size ==
 			    oldsize);
-			if ((usize_max <= SMALL_MAXCLASS &&
-			    size2index(usize_max) == size2index(oldsize)) ||
-			    (size <= oldsize && usize_max >= oldsize))
-				return (false);
+			if ((usize_max > SMALL_MAXCLASS ||
+			    size2index(usize_max) != size2index(oldsize)) &&
+			    (size > oldsize || usize_max < oldsize))
+				return (true);
 		} else {
-			if (usize_max > SMALL_MAXCLASS) {
-				if (!arena_ralloc_large(ptr, oldsize, usize_min,
-				    usize_max, zero))
-					return (false);
-			}
+			if (usize_max <= SMALL_MAXCLASS)
+				return (true);
+			if (arena_ralloc_large(tsdn, ptr, oldsize, usize_min,
+			    usize_max, zero))
+				return (true);
 		}
 
-		/* Reallocation would require a move. */
-		return (true);
+		chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(ptr);
+		arena_decay_tick(tsdn, extent_node_arena_get(&chunk->node));
+		return (false);
 	} else {
-		return (huge_ralloc_no_move(ptr, oldsize, usize_min, usize_max,
-		    zero));
+		return (huge_ralloc_no_move(tsdn, ptr, oldsize, usize_min,
+		    usize_max, zero));
 	}
 }
 
 static void *
-arena_ralloc_move_helper(tsd_t *tsd, arena_t *arena, size_t usize,
+arena_ralloc_move_helper(tsdn_t *tsdn, arena_t *arena, size_t usize,
     size_t alignment, bool zero, tcache_t *tcache)
 {
 
 	if (alignment == 0)
-		return (arena_malloc(tsd, arena, usize, zero, tcache));
+		return (arena_malloc(tsdn, arena, usize, size2index(usize),
+		    zero, tcache, true));
 	usize = sa2u(usize, alignment);
-	if (usize == 0)
+	if (unlikely(usize == 0 || usize > HUGE_MAXCLASS))
 		return (NULL);
-	return (ipalloct(tsd, usize, alignment, zero, tcache, arena));
+	return (ipalloct(tsdn, usize, alignment, zero, tcache, arena));
 }
 
 void *
@@ -2852,14 +3317,15 @@ arena_ralloc(tsd_t *tsd, arena_t *arena, void *ptr, size_t oldsize, size_t size,
 	size_t usize;
 
 	usize = s2u(size);
-	if (usize == 0)
+	if (unlikely(usize == 0 || size > HUGE_MAXCLASS))
 		return (NULL);
 
 	if (likely(usize <= large_maxclass)) {
 		size_t copysize;
 
 		/* Try to avoid moving the allocation. */
-		if (!arena_ralloc_no_move(ptr, oldsize, usize, 0, zero))
+		if (!arena_ralloc_no_move(tsd_tsdn(tsd), ptr, oldsize, usize, 0,
+		    zero))
 			return (ptr);
 
 		/*
@@ -2867,8 +3333,8 @@ arena_ralloc(tsd_t *tsd, arena_t *arena, void *ptr, size_t oldsize, size_t size,
 		 * the object.  In that case, fall back to allocating new space
 		 * and copying.
 		 */
-		ret = arena_ralloc_move_helper(tsd, arena, usize, alignment,
-		    zero, tcache);
+		ret = arena_ralloc_move_helper(tsd_tsdn(tsd), arena, usize,
+		    alignment, zero, tcache);
 		if (ret == NULL)
 			return (NULL);
 
@@ -2880,7 +3346,7 @@ arena_ralloc(tsd_t *tsd, arena_t *arena, void *ptr, size_t oldsize, size_t size,
 		copysize = (usize < oldsize) ? usize : oldsize;
 		JEMALLOC_VALGRIND_MAKE_MEM_UNDEFINED(ret, copysize);
 		memcpy(ret, ptr, copysize);
-		isqalloc(tsd, ptr, oldsize, tcache);
+		isqalloc(tsd, ptr, oldsize, tcache, true);
 	} else {
 		ret = huge_ralloc(tsd, arena, ptr, oldsize, usize, alignment,
 		    zero, tcache);
@@ -2889,25 +3355,25 @@ arena_ralloc(tsd_t *tsd, arena_t *arena, void *ptr, size_t oldsize, size_t size,
 }
 
 dss_prec_t
-arena_dss_prec_get(arena_t *arena)
+arena_dss_prec_get(tsdn_t *tsdn, arena_t *arena)
 {
 	dss_prec_t ret;
 
-	malloc_mutex_lock(&arena->lock);
+	malloc_mutex_lock(tsdn, &arena->lock);
 	ret = arena->dss_prec;
-	malloc_mutex_unlock(&arena->lock);
+	malloc_mutex_unlock(tsdn, &arena->lock);
 	return (ret);
 }
 
 bool
-arena_dss_prec_set(arena_t *arena, dss_prec_t dss_prec)
+arena_dss_prec_set(tsdn_t *tsdn, arena_t *arena, dss_prec_t dss_prec)
 {
 
 	if (!have_dss)
 		return (dss_prec != dss_prec_disabled);
-	malloc_mutex_lock(&arena->lock);
+	malloc_mutex_lock(tsdn, &arena->lock);
 	arena->dss_prec = dss_prec;
-	malloc_mutex_unlock(&arena->lock);
+	malloc_mutex_unlock(tsdn, &arena->lock);
 	return (false);
 }
 
@@ -2922,27 +3388,76 @@ bool
 arena_lg_dirty_mult_default_set(ssize_t lg_dirty_mult)
 {
 
+	if (opt_purge != purge_mode_ratio)
+		return (true);
 	if (!arena_lg_dirty_mult_valid(lg_dirty_mult))
 		return (true);
 	atomic_write_z((size_t *)&lg_dirty_mult_default, (size_t)lg_dirty_mult);
 	return (false);
 }
 
+ssize_t
+arena_decay_time_default_get(void)
+{
+
+	return ((ssize_t)atomic_read_z((size_t *)&decay_time_default));
+}
+
+bool
+arena_decay_time_default_set(ssize_t decay_time)
+{
+
+	if (opt_purge != purge_mode_decay)
+		return (true);
+	if (!arena_decay_time_valid(decay_time))
+		return (true);
+	atomic_write_z((size_t *)&decay_time_default, (size_t)decay_time);
+	return (false);
+}
+
+static void
+arena_basic_stats_merge_locked(arena_t *arena, unsigned *nthreads,
+    const char **dss, ssize_t *lg_dirty_mult, ssize_t *decay_time,
+    size_t *nactive, size_t *ndirty)
+{
+
+	*nthreads += arena_nthreads_get(arena, false);
+	*dss = dss_prec_names[arena->dss_prec];
+	*lg_dirty_mult = arena->lg_dirty_mult;
+	*decay_time = arena->decay.time;
+	*nactive += arena->nactive;
+	*ndirty += arena->ndirty;
+}
+
+void
+arena_basic_stats_merge(tsdn_t *tsdn, arena_t *arena, unsigned *nthreads,
+    const char **dss, ssize_t *lg_dirty_mult, ssize_t *decay_time,
+    size_t *nactive, size_t *ndirty)
+{
+
+	malloc_mutex_lock(tsdn, &arena->lock);
+	arena_basic_stats_merge_locked(arena, nthreads, dss, lg_dirty_mult,
+	    decay_time, nactive, ndirty);
+	malloc_mutex_unlock(tsdn, &arena->lock);
+}
+
 void
-arena_stats_merge(arena_t *arena, const char **dss, ssize_t *lg_dirty_mult,
+arena_stats_merge(tsdn_t *tsdn, arena_t *arena, unsigned *nthreads,
+    const char **dss, ssize_t *lg_dirty_mult, ssize_t *decay_time,
     size_t *nactive, size_t *ndirty, arena_stats_t *astats,
     malloc_bin_stats_t *bstats, malloc_large_stats_t *lstats,
     malloc_huge_stats_t *hstats)
 {
 	unsigned i;
 
-	malloc_mutex_lock(&arena->lock);
-	*dss = dss_prec_names[arena->dss_prec];
-	*lg_dirty_mult = arena->lg_dirty_mult;
-	*nactive += arena->nactive;
-	*ndirty += arena->ndirty;
+	cassert(config_stats);
+
+	malloc_mutex_lock(tsdn, &arena->lock);
+	arena_basic_stats_merge_locked(arena, nthreads, dss, lg_dirty_mult,
+	    decay_time, nactive, ndirty);
 
 	astats->mapped += arena->stats.mapped;
+	astats->retained += arena->stats.retained;
 	astats->npurge += arena->stats.npurge;
 	astats->nmadvise += arena->stats.nmadvise;
 	astats->purged += arena->stats.purged;
@@ -2968,12 +3483,12 @@ arena_stats_merge(arena_t *arena, const char **dss, ssize_t *lg_dirty_mult,
 		hstats[i].ndalloc += arena->stats.hstats[i].ndalloc;
 		hstats[i].curhchunks += arena->stats.hstats[i].curhchunks;
 	}
-	malloc_mutex_unlock(&arena->lock);
+	malloc_mutex_unlock(tsdn, &arena->lock);
 
 	for (i = 0; i < NBINS; i++) {
 		arena_bin_t *bin = &arena->bins[i];
 
-		malloc_mutex_lock(&bin->lock);
+		malloc_mutex_lock(tsdn, &bin->lock);
 		bstats[i].nmalloc += bin->stats.nmalloc;
 		bstats[i].ndalloc += bin->stats.ndalloc;
 		bstats[i].nrequests += bin->stats.nrequests;
@@ -2985,33 +3500,61 @@ arena_stats_merge(arena_t *arena, const char **dss, ssize_t *lg_dirty_mult,
 		bstats[i].nruns += bin->stats.nruns;
 		bstats[i].reruns += bin->stats.reruns;
 		bstats[i].curruns += bin->stats.curruns;
-		malloc_mutex_unlock(&bin->lock);
+		malloc_mutex_unlock(tsdn, &bin->lock);
 	}
 }
 
+unsigned
+arena_nthreads_get(arena_t *arena, bool internal)
+{
+
+	return (atomic_read_u(&arena->nthreads[internal]));
+}
+
+void
+arena_nthreads_inc(arena_t *arena, bool internal)
+{
+
+	atomic_add_u(&arena->nthreads[internal], 1);
+}
+
+void
+arena_nthreads_dec(arena_t *arena, bool internal)
+{
+
+	atomic_sub_u(&arena->nthreads[internal], 1);
+}
+
+size_t
+arena_extent_sn_next(arena_t *arena)
+{
+
+	return (atomic_add_z(&arena->extent_sn_next, 1) - 1);
+}
+
 arena_t *
-arena_new(unsigned ind)
+arena_new(tsdn_t *tsdn, unsigned ind)
 {
 	arena_t *arena;
 	unsigned i;
-	arena_bin_t *bin;
 
 	/*
 	 * Allocate arena, arena->lstats, and arena->hstats contiguously, mainly
 	 * because there is no way to clean up if base_alloc() OOMs.
 	 */
 	if (config_stats) {
-		arena = (arena_t *)base_alloc(CACHELINE_CEILING(sizeof(arena_t))
-		    + QUANTUM_CEILING(nlclasses * sizeof(malloc_large_stats_t) +
-		    nhclasses) * sizeof(malloc_huge_stats_t));
+		arena = (arena_t *)base_alloc(tsdn,
+		    CACHELINE_CEILING(sizeof(arena_t)) +
+		    QUANTUM_CEILING((nlclasses * sizeof(malloc_large_stats_t)))
+		    + (nhclasses * sizeof(malloc_huge_stats_t)));
 	} else
-		arena = (arena_t *)base_alloc(sizeof(arena_t));
+		arena = (arena_t *)base_alloc(tsdn, sizeof(arena_t));
 	if (arena == NULL)
 		return (NULL);
 
 	arena->ind = ind;
-	arena->nthreads = 0;
-	if (malloc_mutex_init(&arena->lock))
+	arena->nthreads[0] = arena->nthreads[1] = 0;
+	if (malloc_mutex_init(&arena->lock, "arena", WITNESS_RANK_ARENA))
 		return (NULL);
 
 	if (config_stats) {
@@ -3041,11 +3584,15 @@ arena_new(unsigned ind)
 		 * deterministic seed.
 		 */
 		arena->offset_state = config_debug ? ind :
-		    (uint64_t)(uintptr_t)arena;
+		    (size_t)(uintptr_t)arena;
 	}
 
 	arena->dss_prec = chunk_dss_prec_get();
 
+	ql_new(&arena->achunks);
+
+	arena->extent_sn_next = 0;
+
 	arena->spare = NULL;
 
 	arena->lg_dirty_mult = arena_lg_dirty_mult_default_get();
@@ -3053,33 +3600,42 @@ arena_new(unsigned ind)
 	arena->nactive = 0;
 	arena->ndirty = 0;
 
-	arena_avail_tree_new(&arena->runs_avail);
+	for (i = 0; i < NPSIZES; i++)
+		arena_run_heap_new(&arena->runs_avail[i]);
+
 	qr_new(&arena->runs_dirty, rd_link);
 	qr_new(&arena->chunks_cache, cc_link);
 
+	if (opt_purge == purge_mode_decay)
+		arena_decay_init(arena, arena_decay_time_default_get());
+
 	ql_new(&arena->huge);
-	if (malloc_mutex_init(&arena->huge_mtx))
+	if (malloc_mutex_init(&arena->huge_mtx, "arena_huge",
+	    WITNESS_RANK_ARENA_HUGE))
 		return (NULL);
 
-	extent_tree_szad_new(&arena->chunks_szad_cached);
+	extent_tree_szsnad_new(&arena->chunks_szsnad_cached);
 	extent_tree_ad_new(&arena->chunks_ad_cached);
-	extent_tree_szad_new(&arena->chunks_szad_retained);
+	extent_tree_szsnad_new(&arena->chunks_szsnad_retained);
 	extent_tree_ad_new(&arena->chunks_ad_retained);
-	if (malloc_mutex_init(&arena->chunks_mtx))
+	if (malloc_mutex_init(&arena->chunks_mtx, "arena_chunks",
+	    WITNESS_RANK_ARENA_CHUNKS))
 		return (NULL);
 	ql_new(&arena->node_cache);
-	if (malloc_mutex_init(&arena->node_cache_mtx))
+	if (malloc_mutex_init(&arena->node_cache_mtx, "arena_node_cache",
+	    WITNESS_RANK_ARENA_NODE_CACHE))
 		return (NULL);
 
 	arena->chunk_hooks = chunk_hooks_default;
 
 	/* Initialize bins. */
 	for (i = 0; i < NBINS; i++) {
-		bin = &arena->bins[i];
-		if (malloc_mutex_init(&bin->lock))
+		arena_bin_t *bin = &arena->bins[i];
+		if (malloc_mutex_init(&bin->lock, "arena_bin",
+		    WITNESS_RANK_ARENA_BIN))
 			return (NULL);
 		bin->runcur = NULL;
-		arena_run_tree_new(&bin->runs);
+		arena_run_heap_new(&bin->runs);
 		if (config_stats)
 			memset(&bin->stats, 0, sizeof(malloc_bin_stats_t));
 	}
@@ -3111,8 +3667,7 @@ bin_info_run_size_calc(arena_bin_info_t *bin_info)
 	 * be twice as large in order to maintain alignment.
 	 */
 	if (config_fill && unlikely(opt_redzone)) {
-		size_t align_min = ZU(1) << (jemalloc_ffs(bin_info->reg_size) -
-		    1);
+		size_t align_min = ZU(1) << (ffs_zu(bin_info->reg_size) - 1);
 		if (align_min <= REDZONE_MINSIZE) {
 			bin_info->redzone_size = REDZONE_MINSIZE;
 			pad_size = 0;
@@ -3132,18 +3687,19 @@ bin_info_run_size_calc(arena_bin_info_t *bin_info)
 	 * size).
 	 */
 	try_run_size = PAGE;
-	try_nregs = try_run_size / bin_info->reg_size;
+	try_nregs = (uint32_t)(try_run_size / bin_info->reg_size);
 	do {
 		perfect_run_size = try_run_size;
 		perfect_nregs = try_nregs;
 
 		try_run_size += PAGE;
-		try_nregs = try_run_size / bin_info->reg_size;
+		try_nregs = (uint32_t)(try_run_size / bin_info->reg_size);
 	} while (perfect_run_size != perfect_nregs * bin_info->reg_size);
 	assert(perfect_nregs <= RUN_MAXREGS);
 
 	actual_run_size = perfect_run_size;
-	actual_nregs = (actual_run_size - pad_size) / bin_info->reg_interval;
+	actual_nregs = (uint32_t)((actual_run_size - pad_size) /
+	    bin_info->reg_interval);
 
 	/*
 	 * Redzones can require enough padding that not even a single region can
@@ -3155,8 +3711,8 @@ bin_info_run_size_calc(arena_bin_info_t *bin_info)
 		assert(config_fill && unlikely(opt_redzone));
 
 		actual_run_size += PAGE;
-		actual_nregs = (actual_run_size - pad_size) /
-		    bin_info->reg_interval;
+		actual_nregs = (uint32_t)((actual_run_size - pad_size) /
+		    bin_info->reg_interval);
 	}
 
 	/*
@@ -3164,8 +3720,8 @@ bin_info_run_size_calc(arena_bin_info_t *bin_info)
 	 */
 	while (actual_run_size > arena_maxrun) {
 		actual_run_size -= PAGE;
-		actual_nregs = (actual_run_size - pad_size) /
-		    bin_info->reg_interval;
+		actual_nregs = (uint32_t)((actual_run_size - pad_size) /
+		    bin_info->reg_interval);
 	}
 	assert(actual_nregs > 0);
 	assert(actual_run_size == s2u(actual_run_size));
@@ -3173,11 +3729,8 @@ bin_info_run_size_calc(arena_bin_info_t *bin_info)
 	/* Copy final settings. */
 	bin_info->run_size = actual_run_size;
 	bin_info->nregs = actual_nregs;
-	bin_info->reg0_offset = actual_run_size - (actual_nregs *
-	    bin_info->reg_interval) - pad_size + bin_info->redzone_size;
-
-	if (actual_run_size > small_maxrun)
-		small_maxrun = actual_run_size;
+	bin_info->reg0_offset = (uint32_t)(actual_run_size - (actual_nregs *
+	    bin_info->reg_interval) - pad_size + bin_info->redzone_size);
 
 	assert(bin_info->reg0_offset - bin_info->redzone_size + (bin_info->nregs
 	    * bin_info->reg_interval) + pad_size == bin_info->run_size);
@@ -3194,7 +3747,7 @@ bin_info_init(void)
 	bin_info_run_size_calc(bin_info);				\
 	bitmap_info_init(&bin_info->bitmap_info, bin_info->nregs);
 #define	BIN_INFO_INIT_bin_no(index, size)
-#define	SC(index, lg_grp, lg_delta, ndelta, bin, lg_delta_lookup)	\
+#define	SC(index, lg_grp, lg_delta, ndelta, psz, bin, lg_delta_lookup)	\
 	BIN_INFO_INIT_bin_##bin(index, (ZU(1)<<lg_grp) + (ZU(ndelta)<<lg_delta))
 	SIZE_CLASSES
 #undef BIN_INFO_INIT_bin_yes
@@ -3202,38 +3755,13 @@ bin_info_init(void)
 #undef SC
 }
 
-static bool
-small_run_size_init(void)
-{
-
-	assert(small_maxrun != 0);
-
-	small_run_tab = (bool *)base_alloc(sizeof(bool) * (small_maxrun >>
-	    LG_PAGE));
-	if (small_run_tab == NULL)
-		return (true);
-
-#define	TAB_INIT_bin_yes(index, size) {					\
-		arena_bin_info_t *bin_info = &arena_bin_info[index];	\
-		small_run_tab[bin_info->run_size >> LG_PAGE] = true;	\
-	}
-#define	TAB_INIT_bin_no(index, size)
-#define	SC(index, lg_grp, lg_delta, ndelta, bin, lg_delta_lookup)	\
-	TAB_INIT_bin_##bin(index, (ZU(1)<<lg_grp) + (ZU(ndelta)<<lg_delta))
-	SIZE_CLASSES
-#undef TAB_INIT_bin_yes
-#undef TAB_INIT_bin_no
-#undef SC
-
-	return (false);
-}
-
-bool
+void
 arena_boot(void)
 {
 	unsigned i;
 
 	arena_lg_dirty_mult_default_set(opt_lg_dirty_mult);
+	arena_decay_time_default_set(opt_decay_time);
 
 	/*
 	 * Compute the header size such that it is large enough to contain the
@@ -3275,44 +3803,61 @@ arena_boot(void)
 	nhclasses = NSIZES - nlclasses - NBINS;
 
 	bin_info_init();
-	return (small_run_size_init());
 }
 
 void
-arena_prefork(arena_t *arena)
+arena_prefork0(tsdn_t *tsdn, arena_t *arena)
+{
+
+	malloc_mutex_prefork(tsdn, &arena->lock);
+}
+
+void
+arena_prefork1(tsdn_t *tsdn, arena_t *arena)
+{
+
+	malloc_mutex_prefork(tsdn, &arena->chunks_mtx);
+}
+
+void
+arena_prefork2(tsdn_t *tsdn, arena_t *arena)
+{
+
+	malloc_mutex_prefork(tsdn, &arena->node_cache_mtx);
+}
+
+void
+arena_prefork3(tsdn_t *tsdn, arena_t *arena)
 {
 	unsigned i;
 
-	malloc_mutex_prefork(&arena->lock);
-	malloc_mutex_prefork(&arena->huge_mtx);
-	malloc_mutex_prefork(&arena->chunks_mtx);
-	malloc_mutex_prefork(&arena->node_cache_mtx);
 	for (i = 0; i < NBINS; i++)
-		malloc_mutex_prefork(&arena->bins[i].lock);
+		malloc_mutex_prefork(tsdn, &arena->bins[i].lock);
+	malloc_mutex_prefork(tsdn, &arena->huge_mtx);
 }
 
 void
-arena_postfork_parent(arena_t *arena)
+arena_postfork_parent(tsdn_t *tsdn, arena_t *arena)
 {
 	unsigned i;
 
+	malloc_mutex_postfork_parent(tsdn, &arena->huge_mtx);
 	for (i = 0; i < NBINS; i++)
-		malloc_mutex_postfork_parent(&arena->bins[i].lock);
-	malloc_mutex_postfork_parent(&arena->node_cache_mtx);
-	malloc_mutex_postfork_parent(&arena->chunks_mtx);
-	malloc_mutex_postfork_parent(&arena->huge_mtx);
-	malloc_mutex_postfork_parent(&arena->lock);
+		malloc_mutex_postfork_parent(tsdn, &arena->bins[i].lock);
+	malloc_mutex_postfork_parent(tsdn, &arena->node_cache_mtx);
+	malloc_mutex_postfork_parent(tsdn, &arena->chunks_mtx);
+	malloc_mutex_postfork_parent(tsdn, &arena->lock);
 }
 
 void
-arena_postfork_child(arena_t *arena)
+arena_postfork_child(tsdn_t *tsdn, arena_t *arena)
 {
 	unsigned i;
 
+	malloc_mutex_postfork_child(tsdn, &arena->huge_mtx);
 	for (i = 0; i < NBINS; i++)
-		malloc_mutex_postfork_child(&arena->bins[i].lock);
-	malloc_mutex_postfork_child(&arena->node_cache_mtx);
-	malloc_mutex_postfork_child(&arena->chunks_mtx);
-	malloc_mutex_postfork_child(&arena->huge_mtx);
-	malloc_mutex_postfork_child(&arena->lock);
+		malloc_mutex_postfork_child(tsdn, &arena->bins[i].lock);
+	malloc_mutex_postfork_child(tsdn, &arena->node_cache_mtx);
+	malloc_mutex_postfork_child(tsdn, &arena->chunks_mtx);
+	malloc_mutex_postfork_child(tsdn, &arena->lock);
 }
diff --git a/deps/jemalloc/src/base.c b/deps/jemalloc/src/base.c
index 7cdcfed86..5681a3f36 100644
--- a/deps/jemalloc/src/base.c
+++ b/deps/jemalloc/src/base.c
@@ -5,7 +5,8 @@
 /* Data. */
 
 static malloc_mutex_t	base_mtx;
-static extent_tree_t	base_avail_szad;
+static size_t		base_extent_sn_next;
+static extent_tree_t	base_avail_szsnad;
 static extent_node_t	*base_nodes;
 static size_t		base_allocated;
 static size_t		base_resident;
@@ -13,12 +14,13 @@ static size_t		base_mapped;
 
 /******************************************************************************/
 
-/* base_mtx must be held. */
 static extent_node_t *
-base_node_try_alloc(void)
+base_node_try_alloc(tsdn_t *tsdn)
 {
 	extent_node_t *node;
 
+	malloc_mutex_assert_owner(tsdn, &base_mtx);
+
 	if (base_nodes == NULL)
 		return (NULL);
 	node = base_nodes;
@@ -27,33 +29,42 @@ base_node_try_alloc(void)
 	return (node);
 }
 
-/* base_mtx must be held. */
 static void
-base_node_dalloc(extent_node_t *node)
+base_node_dalloc(tsdn_t *tsdn, extent_node_t *node)
 {
 
+	malloc_mutex_assert_owner(tsdn, &base_mtx);
+
 	JEMALLOC_VALGRIND_MAKE_MEM_UNDEFINED(node, sizeof(extent_node_t));
 	*(extent_node_t **)node = base_nodes;
 	base_nodes = node;
 }
 
-/* base_mtx must be held. */
+static void
+base_extent_node_init(extent_node_t *node, void *addr, size_t size)
+{
+	size_t sn = atomic_add_z(&base_extent_sn_next, 1) - 1;
+
+	extent_node_init(node, NULL, addr, size, sn, true, true);
+}
+
 static extent_node_t *
-base_chunk_alloc(size_t minsize)
+base_chunk_alloc(tsdn_t *tsdn, size_t minsize)
 {
 	extent_node_t *node;
 	size_t csize, nsize;
 	void *addr;
 
+	malloc_mutex_assert_owner(tsdn, &base_mtx);
 	assert(minsize != 0);
-	node = base_node_try_alloc();
+	node = base_node_try_alloc(tsdn);
 	/* Allocate enough space to also carve a node out if necessary. */
 	nsize = (node == NULL) ? CACHELINE_CEILING(sizeof(extent_node_t)) : 0;
 	csize = CHUNK_CEILING(minsize + nsize);
 	addr = chunk_alloc_base(csize);
 	if (addr == NULL) {
 		if (node != NULL)
-			base_node_dalloc(node);
+			base_node_dalloc(tsdn, node);
 		return (NULL);
 	}
 	base_mapped += csize;
@@ -66,7 +77,7 @@ base_chunk_alloc(size_t minsize)
 			base_resident += PAGE_CEILING(nsize);
 		}
 	}
-	extent_node_init(node, NULL, addr, csize, true, true);
+	base_extent_node_init(node, addr, csize);
 	return (node);
 }
 
@@ -76,7 +87,7 @@ base_chunk_alloc(size_t minsize)
  * physical memory usage.
  */
 void *
-base_alloc(size_t size)
+base_alloc(tsdn_t *tsdn, size_t size)
 {
 	void *ret;
 	size_t csize, usize;
@@ -90,15 +101,15 @@ base_alloc(size_t size)
 	csize = CACHELINE_CEILING(size);
 
 	usize = s2u(csize);
-	extent_node_init(&key, NULL, NULL, usize, false, false);
-	malloc_mutex_lock(&base_mtx);
-	node = extent_tree_szad_nsearch(&base_avail_szad, &key);
+	extent_node_init(&key, NULL, NULL, usize, 0, false, false);
+	malloc_mutex_lock(tsdn, &base_mtx);
+	node = extent_tree_szsnad_nsearch(&base_avail_szsnad, &key);
 	if (node != NULL) {
 		/* Use existing space. */
-		extent_tree_szad_remove(&base_avail_szad, node);
+		extent_tree_szsnad_remove(&base_avail_szsnad, node);
 	} else {
 		/* Try to allocate more space. */
-		node = base_chunk_alloc(csize);
+		node = base_chunk_alloc(tsdn, csize);
 	}
 	if (node == NULL) {
 		ret = NULL;
@@ -109,9 +120,9 @@ base_alloc(size_t size)
 	if (extent_node_size_get(node) > csize) {
 		extent_node_addr_set(node, (void *)((uintptr_t)ret + csize));
 		extent_node_size_set(node, extent_node_size_get(node) - csize);
-		extent_tree_szad_insert(&base_avail_szad, node);
+		extent_tree_szsnad_insert(&base_avail_szsnad, node);
 	} else
-		base_node_dalloc(node);
+		base_node_dalloc(tsdn, node);
 	if (config_stats) {
 		base_allocated += csize;
 		/*
@@ -123,52 +134,54 @@ base_alloc(size_t size)
 	}
 	JEMALLOC_VALGRIND_MAKE_MEM_DEFINED(ret, csize);
 label_return:
-	malloc_mutex_unlock(&base_mtx);
+	malloc_mutex_unlock(tsdn, &base_mtx);
 	return (ret);
 }
 
 void
-base_stats_get(size_t *allocated, size_t *resident, size_t *mapped)
+base_stats_get(tsdn_t *tsdn, size_t *allocated, size_t *resident,
+    size_t *mapped)
 {
 
-	malloc_mutex_lock(&base_mtx);
+	malloc_mutex_lock(tsdn, &base_mtx);
 	assert(base_allocated <= base_resident);
 	assert(base_resident <= base_mapped);
 	*allocated = base_allocated;
 	*resident = base_resident;
 	*mapped = base_mapped;
-	malloc_mutex_unlock(&base_mtx);
+	malloc_mutex_unlock(tsdn, &base_mtx);
 }
 
 bool
 base_boot(void)
 {
 
-	if (malloc_mutex_init(&base_mtx))
+	if (malloc_mutex_init(&base_mtx, "base", WITNESS_RANK_BASE))
 		return (true);
-	extent_tree_szad_new(&base_avail_szad);
+	base_extent_sn_next = 0;
+	extent_tree_szsnad_new(&base_avail_szsnad);
 	base_nodes = NULL;
 
 	return (false);
 }
 
 void
-base_prefork(void)
+base_prefork(tsdn_t *tsdn)
 {
 
-	malloc_mutex_prefork(&base_mtx);
+	malloc_mutex_prefork(tsdn, &base_mtx);
 }
 
 void
-base_postfork_parent(void)
+base_postfork_parent(tsdn_t *tsdn)
 {
 
-	malloc_mutex_postfork_parent(&base_mtx);
+	malloc_mutex_postfork_parent(tsdn, &base_mtx);
 }
 
 void
-base_postfork_child(void)
+base_postfork_child(tsdn_t *tsdn)
 {
 
-	malloc_mutex_postfork_child(&base_mtx);
+	malloc_mutex_postfork_child(tsdn, &base_mtx);
 }
diff --git a/deps/jemalloc/src/bitmap.c b/deps/jemalloc/src/bitmap.c
index c733372b4..ac0f3b381 100644
--- a/deps/jemalloc/src/bitmap.c
+++ b/deps/jemalloc/src/bitmap.c
@@ -3,6 +3,8 @@
 
 /******************************************************************************/
 
+#ifdef USE_TREE
+
 void
 bitmap_info_init(bitmap_info_t *binfo, size_t nbits)
 {
@@ -32,20 +34,11 @@ bitmap_info_init(bitmap_info_t *binfo, size_t nbits)
 	binfo->nbits = nbits;
 }
 
-size_t
+static size_t
 bitmap_info_ngroups(const bitmap_info_t *binfo)
 {
 
-	return (binfo->levels[binfo->nlevels].group_offset << LG_SIZEOF_BITMAP);
-}
-
-size_t
-bitmap_size(size_t nbits)
-{
-	bitmap_info_t binfo;
-
-	bitmap_info_init(&binfo, nbits);
-	return (bitmap_info_ngroups(&binfo));
+	return (binfo->levels[binfo->nlevels].group_offset);
 }
 
 void
@@ -61,8 +54,7 @@ bitmap_init(bitmap_t *bitmap, const bitmap_info_t *binfo)
 	 * correspond to the first logical bit in the group, so extra bits
 	 * are the most significant bits of the last group.
 	 */
-	memset(bitmap, 0xffU, binfo->levels[binfo->nlevels].group_offset <<
-	    LG_SIZEOF_BITMAP);
+	memset(bitmap, 0xffU, bitmap_size(binfo));
 	extra = (BITMAP_GROUP_NBITS - (binfo->nbits & BITMAP_GROUP_NBITS_MASK))
 	    & BITMAP_GROUP_NBITS_MASK;
 	if (extra != 0)
@@ -76,3 +68,44 @@ bitmap_init(bitmap_t *bitmap, const bitmap_info_t *binfo)
 			bitmap[binfo->levels[i+1].group_offset - 1] >>= extra;
 	}
 }
+
+#else /* USE_TREE */
+
+void
+bitmap_info_init(bitmap_info_t *binfo, size_t nbits)
+{
+
+	assert(nbits > 0);
+	assert(nbits <= (ZU(1) << LG_BITMAP_MAXBITS));
+
+	binfo->ngroups = BITMAP_BITS2GROUPS(nbits);
+	binfo->nbits = nbits;
+}
+
+static size_t
+bitmap_info_ngroups(const bitmap_info_t *binfo)
+{
+
+	return (binfo->ngroups);
+}
+
+void
+bitmap_init(bitmap_t *bitmap, const bitmap_info_t *binfo)
+{
+	size_t extra;
+
+	memset(bitmap, 0xffU, bitmap_size(binfo));
+	extra = (BITMAP_GROUP_NBITS - (binfo->nbits & BITMAP_GROUP_NBITS_MASK))
+	    & BITMAP_GROUP_NBITS_MASK;
+	if (extra != 0)
+		bitmap[binfo->ngroups - 1] >>= extra;
+}
+
+#endif /* USE_TREE */
+
+size_t
+bitmap_size(const bitmap_info_t *binfo)
+{
+
+	return (bitmap_info_ngroups(binfo) << LG_SIZEOF_BITMAP);
+}
diff --git a/deps/jemalloc/src/chunk.c b/deps/jemalloc/src/chunk.c
index 6ba1ca7a5..c1c514a86 100644
--- a/deps/jemalloc/src/chunk.c
+++ b/deps/jemalloc/src/chunk.c
@@ -49,9 +49,10 @@ const chunk_hooks_t	chunk_hooks_default = {
  * definition.
  */
 
-static void	chunk_record(arena_t *arena, chunk_hooks_t *chunk_hooks,
-    extent_tree_t *chunks_szad, extent_tree_t *chunks_ad, bool cache,
-    void *chunk, size_t size, bool zeroed, bool committed);
+static void	chunk_record(tsdn_t *tsdn, arena_t *arena,
+    chunk_hooks_t *chunk_hooks, extent_tree_t *chunks_szsnad,
+    extent_tree_t *chunks_ad, bool cache, void *chunk, size_t size, size_t sn,
+    bool zeroed, bool committed);
 
 /******************************************************************************/
 
@@ -63,23 +64,23 @@ chunk_hooks_get_locked(arena_t *arena)
 }
 
 chunk_hooks_t
-chunk_hooks_get(arena_t *arena)
+chunk_hooks_get(tsdn_t *tsdn, arena_t *arena)
 {
 	chunk_hooks_t chunk_hooks;
 
-	malloc_mutex_lock(&arena->chunks_mtx);
+	malloc_mutex_lock(tsdn, &arena->chunks_mtx);
 	chunk_hooks = chunk_hooks_get_locked(arena);
-	malloc_mutex_unlock(&arena->chunks_mtx);
+	malloc_mutex_unlock(tsdn, &arena->chunks_mtx);
 
 	return (chunk_hooks);
 }
 
 chunk_hooks_t
-chunk_hooks_set(arena_t *arena, const chunk_hooks_t *chunk_hooks)
+chunk_hooks_set(tsdn_t *tsdn, arena_t *arena, const chunk_hooks_t *chunk_hooks)
 {
 	chunk_hooks_t old_chunk_hooks;
 
-	malloc_mutex_lock(&arena->chunks_mtx);
+	malloc_mutex_lock(tsdn, &arena->chunks_mtx);
 	old_chunk_hooks = arena->chunk_hooks;
 	/*
 	 * Copy each field atomically so that it is impossible for readers to
@@ -104,14 +105,14 @@ chunk_hooks_set(arena_t *arena, const chunk_hooks_t *chunk_hooks)
 	ATOMIC_COPY_HOOK(split);
 	ATOMIC_COPY_HOOK(merge);
 #undef ATOMIC_COPY_HOOK
-	malloc_mutex_unlock(&arena->chunks_mtx);
+	malloc_mutex_unlock(tsdn, &arena->chunks_mtx);
 
 	return (old_chunk_hooks);
 }
 
 static void
-chunk_hooks_assure_initialized_impl(arena_t *arena, chunk_hooks_t *chunk_hooks,
-    bool locked)
+chunk_hooks_assure_initialized_impl(tsdn_t *tsdn, arena_t *arena,
+    chunk_hooks_t *chunk_hooks, bool locked)
 {
 	static const chunk_hooks_t uninitialized_hooks =
 	    CHUNK_HOOKS_INITIALIZER;
@@ -119,27 +120,28 @@ chunk_hooks_assure_initialized_impl(arena_t *arena, chunk_hooks_t *chunk_hooks,
 	if (memcmp(chunk_hooks, &uninitialized_hooks, sizeof(chunk_hooks_t)) ==
 	    0) {
 		*chunk_hooks = locked ? chunk_hooks_get_locked(arena) :
-		    chunk_hooks_get(arena);
+		    chunk_hooks_get(tsdn, arena);
 	}
 }
 
 static void
-chunk_hooks_assure_initialized_locked(arena_t *arena,
+chunk_hooks_assure_initialized_locked(tsdn_t *tsdn, arena_t *arena,
     chunk_hooks_t *chunk_hooks)
 {
 
-	chunk_hooks_assure_initialized_impl(arena, chunk_hooks, true);
+	chunk_hooks_assure_initialized_impl(tsdn, arena, chunk_hooks, true);
 }
 
 static void
-chunk_hooks_assure_initialized(arena_t *arena, chunk_hooks_t *chunk_hooks)
+chunk_hooks_assure_initialized(tsdn_t *tsdn, arena_t *arena,
+    chunk_hooks_t *chunk_hooks)
 {
 
-	chunk_hooks_assure_initialized_impl(arena, chunk_hooks, false);
+	chunk_hooks_assure_initialized_impl(tsdn, arena, chunk_hooks, false);
 }
 
 bool
-chunk_register(const void *chunk, const extent_node_t *node)
+chunk_register(tsdn_t *tsdn, const void *chunk, const extent_node_t *node)
 {
 
 	assert(extent_node_addr_get(node) == chunk);
@@ -159,7 +161,7 @@ chunk_register(const void *chunk, const extent_node_t *node)
 			high = atomic_read_z(&highchunks);
 		}
 		if (cur > high && prof_gdump_get_unlocked())
-			prof_gdump();
+			prof_gdump(tsdn);
 	}
 
 	return (false);
@@ -181,33 +183,35 @@ chunk_deregister(const void *chunk, const extent_node_t *node)
 }
 
 /*
- * Do first-best-fit chunk selection, i.e. select the lowest chunk that best
- * fits.
+ * Do first-best-fit chunk selection, i.e. select the oldest/lowest chunk that
+ * best fits.
  */
 static extent_node_t *
-chunk_first_best_fit(arena_t *arena, extent_tree_t *chunks_szad,
-    extent_tree_t *chunks_ad, size_t size)
+chunk_first_best_fit(arena_t *arena, extent_tree_t *chunks_szsnad, size_t size)
 {
 	extent_node_t key;
 
 	assert(size == CHUNK_CEILING(size));
 
-	extent_node_init(&key, arena, NULL, size, false, false);
-	return (extent_tree_szad_nsearch(chunks_szad, &key));
+	extent_node_init(&key, arena, NULL, size, 0, false, false);
+	return (extent_tree_szsnad_nsearch(chunks_szsnad, &key));
 }
 
 static void *
-chunk_recycle(arena_t *arena, chunk_hooks_t *chunk_hooks,
-    extent_tree_t *chunks_szad, extent_tree_t *chunks_ad, bool cache,
-    void *new_addr, size_t size, size_t alignment, bool *zero, bool *commit,
-    bool dalloc_node)
+chunk_recycle(tsdn_t *tsdn, arena_t *arena, chunk_hooks_t *chunk_hooks,
+    extent_tree_t *chunks_szsnad, extent_tree_t *chunks_ad, bool cache,
+    void *new_addr, size_t size, size_t alignment, size_t *sn, bool *zero,
+    bool *commit, bool dalloc_node)
 {
 	void *ret;
 	extent_node_t *node;
 	size_t alloc_size, leadsize, trailsize;
 	bool zeroed, committed;
 
+	assert(CHUNK_CEILING(size) == size);
+	assert(alignment > 0);
 	assert(new_addr == NULL || alignment == chunksize);
+	assert(CHUNK_ADDR2BASE(new_addr) == new_addr);
 	/*
 	 * Cached chunks use the node linkage embedded in their headers, in
 	 * which case dalloc_node is true, and new_addr is non-NULL because
@@ -215,24 +219,23 @@ chunk_recycle(arena_t *arena, chunk_hooks_t *chunk_hooks,
 	 */
 	assert(dalloc_node || new_addr != NULL);
 
-	alloc_size = CHUNK_CEILING(s2u(size + alignment - chunksize));
+	alloc_size = size + CHUNK_CEILING(alignment) - chunksize;
 	/* Beware size_t wrap-around. */
 	if (alloc_size < size)
 		return (NULL);
-	malloc_mutex_lock(&arena->chunks_mtx);
-	chunk_hooks_assure_initialized_locked(arena, chunk_hooks);
+	malloc_mutex_lock(tsdn, &arena->chunks_mtx);
+	chunk_hooks_assure_initialized_locked(tsdn, arena, chunk_hooks);
 	if (new_addr != NULL) {
 		extent_node_t key;
-		extent_node_init(&key, arena, new_addr, alloc_size, false,
+		extent_node_init(&key, arena, new_addr, alloc_size, 0, false,
 		    false);
 		node = extent_tree_ad_search(chunks_ad, &key);
 	} else {
-		node = chunk_first_best_fit(arena, chunks_szad, chunks_ad,
-		    alloc_size);
+		node = chunk_first_best_fit(arena, chunks_szsnad, alloc_size);
 	}
 	if (node == NULL || (new_addr != NULL && extent_node_size_get(node) <
 	    size)) {
-		malloc_mutex_unlock(&arena->chunks_mtx);
+		malloc_mutex_unlock(tsdn, &arena->chunks_mtx);
 		return (NULL);
 	}
 	leadsize = ALIGNMENT_CEILING((uintptr_t)extent_node_addr_get(node),
@@ -241,6 +244,7 @@ chunk_recycle(arena_t *arena, chunk_hooks_t *chunk_hooks,
 	assert(extent_node_size_get(node) >= leadsize + size);
 	trailsize = extent_node_size_get(node) - leadsize - size;
 	ret = (void *)((uintptr_t)extent_node_addr_get(node) + leadsize);
+	*sn = extent_node_sn_get(node);
 	zeroed = extent_node_zeroed_get(node);
 	if (zeroed)
 		*zero = true;
@@ -251,17 +255,17 @@ chunk_recycle(arena_t *arena, chunk_hooks_t *chunk_hooks,
 	if (leadsize != 0 &&
 	    chunk_hooks->split(extent_node_addr_get(node),
 	    extent_node_size_get(node), leadsize, size, false, arena->ind)) {
-		malloc_mutex_unlock(&arena->chunks_mtx);
+		malloc_mutex_unlock(tsdn, &arena->chunks_mtx);
 		return (NULL);
 	}
 	/* Remove node from the tree. */
-	extent_tree_szad_remove(chunks_szad, node);
+	extent_tree_szsnad_remove(chunks_szsnad, node);
 	extent_tree_ad_remove(chunks_ad, node);
 	arena_chunk_cache_maybe_remove(arena, node, cache);
 	if (leadsize != 0) {
 		/* Insert the leading space as a smaller chunk. */
 		extent_node_size_set(node, leadsize);
-		extent_tree_szad_insert(chunks_szad, node);
+		extent_tree_szsnad_insert(chunks_szsnad, node);
 		extent_tree_ad_insert(chunks_ad, node);
 		arena_chunk_cache_maybe_insert(arena, node, cache);
 		node = NULL;
@@ -271,41 +275,42 @@ chunk_recycle(arena_t *arena, chunk_hooks_t *chunk_hooks,
 		if (chunk_hooks->split(ret, size + trailsize, size,
 		    trailsize, false, arena->ind)) {
 			if (dalloc_node && node != NULL)
-				arena_node_dalloc(arena, node);
-			malloc_mutex_unlock(&arena->chunks_mtx);
-			chunk_record(arena, chunk_hooks, chunks_szad, chunks_ad,
-			    cache, ret, size + trailsize, zeroed, committed);
+				arena_node_dalloc(tsdn, arena, node);
+			malloc_mutex_unlock(tsdn, &arena->chunks_mtx);
+			chunk_record(tsdn, arena, chunk_hooks, chunks_szsnad,
+			    chunks_ad, cache, ret, size + trailsize, *sn,
+			    zeroed, committed);
 			return (NULL);
 		}
 		/* Insert the trailing space as a smaller chunk. */
 		if (node == NULL) {
-			node = arena_node_alloc(arena);
+			node = arena_node_alloc(tsdn, arena);
 			if (node == NULL) {
-				malloc_mutex_unlock(&arena->chunks_mtx);
-				chunk_record(arena, chunk_hooks, chunks_szad,
-				    chunks_ad, cache, ret, size + trailsize,
-				    zeroed, committed);
+				malloc_mutex_unlock(tsdn, &arena->chunks_mtx);
+				chunk_record(tsdn, arena, chunk_hooks,
+				    chunks_szsnad, chunks_ad, cache, ret, size
+				    + trailsize, *sn, zeroed, committed);
 				return (NULL);
 			}
 		}
 		extent_node_init(node, arena, (void *)((uintptr_t)(ret) + size),
-		    trailsize, zeroed, committed);
-		extent_tree_szad_insert(chunks_szad, node);
+		    trailsize, *sn, zeroed, committed);
+		extent_tree_szsnad_insert(chunks_szsnad, node);
 		extent_tree_ad_insert(chunks_ad, node);
 		arena_chunk_cache_maybe_insert(arena, node, cache);
 		node = NULL;
 	}
 	if (!committed && chunk_hooks->commit(ret, size, 0, size, arena->ind)) {
-		malloc_mutex_unlock(&arena->chunks_mtx);
-		chunk_record(arena, chunk_hooks, chunks_szad, chunks_ad, cache,
-		    ret, size, zeroed, committed);
+		malloc_mutex_unlock(tsdn, &arena->chunks_mtx);
+		chunk_record(tsdn, arena, chunk_hooks, chunks_szsnad, chunks_ad,
+		    cache, ret, size, *sn, zeroed, committed);
 		return (NULL);
 	}
-	malloc_mutex_unlock(&arena->chunks_mtx);
+	malloc_mutex_unlock(tsdn, &arena->chunks_mtx);
 
 	assert(dalloc_node || node != NULL);
 	if (dalloc_node && node != NULL)
-		arena_node_dalloc(arena, node);
+		arena_node_dalloc(tsdn, arena, node);
 	if (*zero) {
 		if (!zeroed)
 			memset(ret, 0, size);
@@ -313,10 +318,11 @@ chunk_recycle(arena_t *arena, chunk_hooks_t *chunk_hooks,
 			size_t i;
 			size_t *p = (size_t *)(uintptr_t)ret;
 
-			JEMALLOC_VALGRIND_MAKE_MEM_DEFINED(ret, size);
 			for (i = 0; i < size / sizeof(size_t); i++)
 				assert(p[i] == 0);
 		}
+		if (config_valgrind)
+			JEMALLOC_VALGRIND_MAKE_MEM_DEFINED(ret, size);
 	}
 	return (ret);
 }
@@ -328,39 +334,29 @@ chunk_recycle(arena_t *arena, chunk_hooks_t *chunk_hooks,
  * them if they are returned.
  */
 static void *
-chunk_alloc_core(arena_t *arena, void *new_addr, size_t size, size_t alignment,
-    bool *zero, bool *commit, dss_prec_t dss_prec)
+chunk_alloc_core(tsdn_t *tsdn, arena_t *arena, void *new_addr, size_t size,
+    size_t alignment, bool *zero, bool *commit, dss_prec_t dss_prec)
 {
 	void *ret;
-	chunk_hooks_t chunk_hooks = CHUNK_HOOKS_INITIALIZER;
 
 	assert(size != 0);
 	assert((size & chunksize_mask) == 0);
 	assert(alignment != 0);
 	assert((alignment & chunksize_mask) == 0);
 
-	/* Retained. */
-	if ((ret = chunk_recycle(arena, &chunk_hooks,
-	    &arena->chunks_szad_retained, &arena->chunks_ad_retained, false,
-	    new_addr, size, alignment, zero, commit, true)) != NULL)
-		return (ret);
-
 	/* "primary" dss. */
 	if (have_dss && dss_prec == dss_prec_primary && (ret =
-	    chunk_alloc_dss(arena, new_addr, size, alignment, zero, commit)) !=
-	    NULL)
-		return (ret);
-	/*
-	 * mmap.  Requesting an address is not implemented for
-	 * chunk_alloc_mmap(), so only call it if (new_addr == NULL).
-	 */
-	if (new_addr == NULL && (ret = chunk_alloc_mmap(size, alignment, zero,
+	    chunk_alloc_dss(tsdn, arena, new_addr, size, alignment, zero,
 	    commit)) != NULL)
 		return (ret);
+	/* mmap. */
+	if ((ret = chunk_alloc_mmap(new_addr, size, alignment, zero, commit)) !=
+	    NULL)
+		return (ret);
 	/* "secondary" dss. */
 	if (have_dss && dss_prec == dss_prec_secondary && (ret =
-	    chunk_alloc_dss(arena, new_addr, size, alignment, zero, commit)) !=
-	    NULL)
+	    chunk_alloc_dss(tsdn, arena, new_addr, size, alignment, zero,
+	    commit)) != NULL)
 		return (ret);
 
 	/* All strategies for allocation failed. */
@@ -380,7 +376,7 @@ chunk_alloc_base(size_t size)
 	 */
 	zero = true;
 	commit = true;
-	ret = chunk_alloc_mmap(size, chunksize, &zero, &commit);
+	ret = chunk_alloc_mmap(NULL, size, chunksize, &zero, &commit);
 	if (ret == NULL)
 		return (NULL);
 	if (config_valgrind)
@@ -390,37 +386,33 @@ chunk_alloc_base(size_t size)
 }
 
 void *
-chunk_alloc_cache(arena_t *arena, chunk_hooks_t *chunk_hooks, void *new_addr,
-    size_t size, size_t alignment, bool *zero, bool dalloc_node)
+chunk_alloc_cache(tsdn_t *tsdn, arena_t *arena, chunk_hooks_t *chunk_hooks,
+    void *new_addr, size_t size, size_t alignment, size_t *sn, bool *zero,
+    bool *commit, bool dalloc_node)
 {
 	void *ret;
-	bool commit;
 
 	assert(size != 0);
 	assert((size & chunksize_mask) == 0);
 	assert(alignment != 0);
 	assert((alignment & chunksize_mask) == 0);
 
-	commit = true;
-	ret = chunk_recycle(arena, chunk_hooks, &arena->chunks_szad_cached,
-	    &arena->chunks_ad_cached, true, new_addr, size, alignment, zero,
-	    &commit, dalloc_node);
+	ret = chunk_recycle(tsdn, arena, chunk_hooks,
+	    &arena->chunks_szsnad_cached, &arena->chunks_ad_cached, true,
+	    new_addr, size, alignment, sn, zero, commit, dalloc_node);
 	if (ret == NULL)
 		return (NULL);
-	assert(commit);
 	if (config_valgrind)
 		JEMALLOC_VALGRIND_MAKE_MEM_UNDEFINED(ret, size);
 	return (ret);
 }
 
 static arena_t *
-chunk_arena_get(unsigned arena_ind)
+chunk_arena_get(tsdn_t *tsdn, unsigned arena_ind)
 {
 	arena_t *arena;
 
-	/* Dodge tsd for a0 in order to avoid bootstrapping issues. */
-	arena = (arena_ind == 0) ? a0get() : arena_get(tsd_fetch(), arena_ind,
-	     false, true);
+	arena = arena_get(tsdn, arena_ind, false);
 	/*
 	 * The arena we're allocating on behalf of must have been initialized
 	 * already.
@@ -430,14 +422,12 @@ chunk_arena_get(unsigned arena_ind)
 }
 
 static void *
-chunk_alloc_default(void *new_addr, size_t size, size_t alignment, bool *zero,
-    bool *commit, unsigned arena_ind)
+chunk_alloc_default_impl(tsdn_t *tsdn, arena_t *arena, void *new_addr,
+    size_t size, size_t alignment, bool *zero, bool *commit)
 {
 	void *ret;
-	arena_t *arena;
 
-	arena = chunk_arena_get(arena_ind);
-	ret = chunk_alloc_core(arena, new_addr, size, alignment, zero,
+	ret = chunk_alloc_core(tsdn, arena, new_addr, size, alignment, zero,
 	    commit, arena->dss_prec);
 	if (ret == NULL)
 		return (NULL);
@@ -447,26 +437,80 @@ chunk_alloc_default(void *new_addr, size_t size, size_t alignment, bool *zero,
 	return (ret);
 }
 
+static void *
+chunk_alloc_default(void *new_addr, size_t size, size_t alignment, bool *zero,
+    bool *commit, unsigned arena_ind)
+{
+	tsdn_t *tsdn;
+	arena_t *arena;
+
+	tsdn = tsdn_fetch();
+	arena = chunk_arena_get(tsdn, arena_ind);
+
+	return (chunk_alloc_default_impl(tsdn, arena, new_addr, size, alignment,
+	    zero, commit));
+}
+
+static void *
+chunk_alloc_retained(tsdn_t *tsdn, arena_t *arena, chunk_hooks_t *chunk_hooks,
+    void *new_addr, size_t size, size_t alignment, size_t *sn, bool *zero,
+    bool *commit)
+{
+	void *ret;
+
+	assert(size != 0);
+	assert((size & chunksize_mask) == 0);
+	assert(alignment != 0);
+	assert((alignment & chunksize_mask) == 0);
+
+	ret = chunk_recycle(tsdn, arena, chunk_hooks,
+	    &arena->chunks_szsnad_retained, &arena->chunks_ad_retained, false,
+	    new_addr, size, alignment, sn, zero, commit, true);
+
+	if (config_stats && ret != NULL)
+		arena->stats.retained -= size;
+
+	return (ret);
+}
+
 void *
-chunk_alloc_wrapper(arena_t *arena, chunk_hooks_t *chunk_hooks, void *new_addr,
-    size_t size, size_t alignment, bool *zero, bool *commit)
+chunk_alloc_wrapper(tsdn_t *tsdn, arena_t *arena, chunk_hooks_t *chunk_hooks,
+    void *new_addr, size_t size, size_t alignment, size_t *sn, bool *zero,
+    bool *commit)
 {
 	void *ret;
 
-	chunk_hooks_assure_initialized(arena, chunk_hooks);
-	ret = chunk_hooks->alloc(new_addr, size, alignment, zero, commit,
-	    arena->ind);
-	if (ret == NULL)
-		return (NULL);
-	if (config_valgrind && chunk_hooks->alloc != chunk_alloc_default)
-		JEMALLOC_VALGRIND_MAKE_MEM_UNDEFINED(ret, chunksize);
+	chunk_hooks_assure_initialized(tsdn, arena, chunk_hooks);
+
+	ret = chunk_alloc_retained(tsdn, arena, chunk_hooks, new_addr, size,
+	    alignment, sn, zero, commit);
+	if (ret == NULL) {
+		if (chunk_hooks->alloc == chunk_alloc_default) {
+			/* Call directly to propagate tsdn. */
+			ret = chunk_alloc_default_impl(tsdn, arena, new_addr,
+			    size, alignment, zero, commit);
+		} else {
+			ret = chunk_hooks->alloc(new_addr, size, alignment,
+			    zero, commit, arena->ind);
+		}
+
+		if (ret == NULL)
+			return (NULL);
+
+		*sn = arena_extent_sn_next(arena);
+
+		if (config_valgrind && chunk_hooks->alloc !=
+		    chunk_alloc_default)
+			JEMALLOC_VALGRIND_MAKE_MEM_UNDEFINED(ret, chunksize);
+	}
+
 	return (ret);
 }
 
 static void
-chunk_record(arena_t *arena, chunk_hooks_t *chunk_hooks,
-    extent_tree_t *chunks_szad, extent_tree_t *chunks_ad, bool cache,
-    void *chunk, size_t size, bool zeroed, bool committed)
+chunk_record(tsdn_t *tsdn, arena_t *arena, chunk_hooks_t *chunk_hooks,
+    extent_tree_t *chunks_szsnad, extent_tree_t *chunks_ad, bool cache,
+    void *chunk, size_t size, size_t sn, bool zeroed, bool committed)
 {
 	bool unzeroed;
 	extent_node_t *node, *prev;
@@ -476,9 +520,9 @@ chunk_record(arena_t *arena, chunk_hooks_t *chunk_hooks,
 	unzeroed = cache || !zeroed;
 	JEMALLOC_VALGRIND_MAKE_MEM_NOACCESS(chunk, size);
 
-	malloc_mutex_lock(&arena->chunks_mtx);
-	chunk_hooks_assure_initialized_locked(arena, chunk_hooks);
-	extent_node_init(&key, arena, (void *)((uintptr_t)chunk + size), 0,
+	malloc_mutex_lock(tsdn, &arena->chunks_mtx);
+	chunk_hooks_assure_initialized_locked(tsdn, arena, chunk_hooks);
+	extent_node_init(&key, arena, (void *)((uintptr_t)chunk + size), 0, 0,
 	    false, false);
 	node = extent_tree_ad_nsearch(chunks_ad, &key);
 	/* Try to coalesce forward. */
@@ -490,19 +534,21 @@ chunk_record(arena_t *arena, chunk_hooks_t *chunk_hooks,
 		/*
 		 * Coalesce chunk with the following address range.  This does
 		 * not change the position within chunks_ad, so only
-		 * remove/insert from/into chunks_szad.
+		 * remove/insert from/into chunks_szsnad.
 		 */
-		extent_tree_szad_remove(chunks_szad, node);
+		extent_tree_szsnad_remove(chunks_szsnad, node);
 		arena_chunk_cache_maybe_remove(arena, node, cache);
 		extent_node_addr_set(node, chunk);
 		extent_node_size_set(node, size + extent_node_size_get(node));
+		if (sn < extent_node_sn_get(node))
+			extent_node_sn_set(node, sn);
 		extent_node_zeroed_set(node, extent_node_zeroed_get(node) &&
 		    !unzeroed);
-		extent_tree_szad_insert(chunks_szad, node);
+		extent_tree_szsnad_insert(chunks_szsnad, node);
 		arena_chunk_cache_maybe_insert(arena, node, cache);
 	} else {
 		/* Coalescing forward failed, so insert a new node. */
-		node = arena_node_alloc(arena);
+		node = arena_node_alloc(tsdn, arena);
 		if (node == NULL) {
 			/*
 			 * Node allocation failed, which is an exceedingly
@@ -511,15 +557,15 @@ chunk_record(arena_t *arena, chunk_hooks_t *chunk_hooks,
 			 * a virtual memory leak.
 			 */
 			if (cache) {
-				chunk_purge_wrapper(arena, chunk_hooks, chunk,
-				    size, 0, size);
+				chunk_purge_wrapper(tsdn, arena, chunk_hooks,
+				    chunk, size, 0, size);
 			}
 			goto label_return;
 		}
-		extent_node_init(node, arena, chunk, size, !unzeroed,
+		extent_node_init(node, arena, chunk, size, sn, !unzeroed,
 		    committed);
 		extent_tree_ad_insert(chunks_ad, node);
-		extent_tree_szad_insert(chunks_szad, node);
+		extent_tree_szsnad_insert(chunks_szsnad, node);
 		arena_chunk_cache_maybe_insert(arena, node, cache);
 	}
 
@@ -533,31 +579,33 @@ chunk_record(arena_t *arena, chunk_hooks_t *chunk_hooks,
 		/*
 		 * Coalesce chunk with the previous address range.  This does
 		 * not change the position within chunks_ad, so only
-		 * remove/insert node from/into chunks_szad.
+		 * remove/insert node from/into chunks_szsnad.
 		 */
-		extent_tree_szad_remove(chunks_szad, prev);
+		extent_tree_szsnad_remove(chunks_szsnad, prev);
 		extent_tree_ad_remove(chunks_ad, prev);
 		arena_chunk_cache_maybe_remove(arena, prev, cache);
-		extent_tree_szad_remove(chunks_szad, node);
+		extent_tree_szsnad_remove(chunks_szsnad, node);
 		arena_chunk_cache_maybe_remove(arena, node, cache);
 		extent_node_addr_set(node, extent_node_addr_get(prev));
 		extent_node_size_set(node, extent_node_size_get(prev) +
 		    extent_node_size_get(node));
+		if (extent_node_sn_get(prev) < extent_node_sn_get(node))
+			extent_node_sn_set(node, extent_node_sn_get(prev));
 		extent_node_zeroed_set(node, extent_node_zeroed_get(prev) &&
 		    extent_node_zeroed_get(node));
-		extent_tree_szad_insert(chunks_szad, node);
+		extent_tree_szsnad_insert(chunks_szsnad, node);
 		arena_chunk_cache_maybe_insert(arena, node, cache);
 
-		arena_node_dalloc(arena, prev);
+		arena_node_dalloc(tsdn, arena, prev);
 	}
 
 label_return:
-	malloc_mutex_unlock(&arena->chunks_mtx);
+	malloc_mutex_unlock(tsdn, &arena->chunks_mtx);
 }
 
 void
-chunk_dalloc_cache(arena_t *arena, chunk_hooks_t *chunk_hooks, void *chunk,
-    size_t size, bool committed)
+chunk_dalloc_cache(tsdn_t *tsdn, arena_t *arena, chunk_hooks_t *chunk_hooks,
+    void *chunk, size_t size, size_t sn, bool committed)
 {
 
 	assert(chunk != NULL);
@@ -565,24 +613,49 @@ chunk_dalloc_cache(arena_t *arena, chunk_hooks_t *chunk_hooks, void *chunk,
 	assert(size != 0);
 	assert((size & chunksize_mask) == 0);
 
-	chunk_record(arena, chunk_hooks, &arena->chunks_szad_cached,
-	    &arena->chunks_ad_cached, true, chunk, size, false, committed);
-	arena_maybe_purge(arena);
+	chunk_record(tsdn, arena, chunk_hooks, &arena->chunks_szsnad_cached,
+	    &arena->chunks_ad_cached, true, chunk, size, sn, false,
+	    committed);
+	arena_maybe_purge(tsdn, arena);
+}
+
+static bool
+chunk_dalloc_default_impl(void *chunk, size_t size)
+{
+
+	if (!have_dss || !chunk_in_dss(chunk))
+		return (chunk_dalloc_mmap(chunk, size));
+	return (true);
+}
+
+static bool
+chunk_dalloc_default(void *chunk, size_t size, bool committed,
+    unsigned arena_ind)
+{
+
+	return (chunk_dalloc_default_impl(chunk, size));
 }
 
 void
-chunk_dalloc_arena(arena_t *arena, chunk_hooks_t *chunk_hooks, void *chunk,
-    size_t size, bool zeroed, bool committed)
+chunk_dalloc_wrapper(tsdn_t *tsdn, arena_t *arena, chunk_hooks_t *chunk_hooks,
+    void *chunk, size_t size, size_t sn, bool zeroed, bool committed)
 {
+	bool err;
 
 	assert(chunk != NULL);
 	assert(CHUNK_ADDR2BASE(chunk) == chunk);
 	assert(size != 0);
 	assert((size & chunksize_mask) == 0);
 
-	chunk_hooks_assure_initialized(arena, chunk_hooks);
+	chunk_hooks_assure_initialized(tsdn, arena, chunk_hooks);
 	/* Try to deallocate. */
-	if (!chunk_hooks->dalloc(chunk, size, committed, arena->ind))
+	if (chunk_hooks->dalloc == chunk_dalloc_default) {
+		/* Call directly to propagate tsdn. */
+		err = chunk_dalloc_default_impl(chunk, size);
+	} else
+		err = chunk_hooks->dalloc(chunk, size, committed, arena->ind);
+
+	if (!err)
 		return;
 	/* Try to decommit; purge if that fails. */
 	if (committed) {
@@ -591,29 +664,12 @@ chunk_dalloc_arena(arena_t *arena, chunk_hooks_t *chunk_hooks, void *chunk,
 	}
 	zeroed = !committed || !chunk_hooks->purge(chunk, size, 0, size,
 	    arena->ind);
-	chunk_record(arena, chunk_hooks, &arena->chunks_szad_retained,
-	    &arena->chunks_ad_retained, false, chunk, size, zeroed, committed);
-}
-
-static bool
-chunk_dalloc_default(void *chunk, size_t size, bool committed,
-    unsigned arena_ind)
-{
+	chunk_record(tsdn, arena, chunk_hooks, &arena->chunks_szsnad_retained,
+	    &arena->chunks_ad_retained, false, chunk, size, sn, zeroed,
+	    committed);
 
-	if (!have_dss || !chunk_in_dss(chunk))
-		return (chunk_dalloc_mmap(chunk, size));
-	return (true);
-}
-
-void
-chunk_dalloc_wrapper(arena_t *arena, chunk_hooks_t *chunk_hooks, void *chunk,
-    size_t size, bool committed)
-{
-
-	chunk_hooks_assure_initialized(arena, chunk_hooks);
-	chunk_hooks->dalloc(chunk, size, committed, arena->ind);
-	if (config_valgrind && chunk_hooks->dalloc != chunk_dalloc_default)
-		JEMALLOC_VALGRIND_MAKE_MEM_NOACCESS(chunk, size);
+	if (config_stats)
+		arena->stats.retained += size;
 }
 
 static bool
@@ -634,8 +690,9 @@ chunk_decommit_default(void *chunk, size_t size, size_t offset, size_t length,
 	    length));
 }
 
-bool
-chunk_purge_arena(arena_t *arena, void *chunk, size_t offset, size_t length)
+static bool
+chunk_purge_default(void *chunk, size_t size, size_t offset, size_t length,
+    unsigned arena_ind)
 {
 
 	assert(chunk != NULL);
@@ -648,21 +705,12 @@ chunk_purge_arena(arena_t *arena, void *chunk, size_t offset, size_t length)
 	    length));
 }
 
-static bool
-chunk_purge_default(void *chunk, size_t size, size_t offset, size_t length,
-    unsigned arena_ind)
-{
-
-	return (chunk_purge_arena(chunk_arena_get(arena_ind), chunk, offset,
-	    length));
-}
-
 bool
-chunk_purge_wrapper(arena_t *arena, chunk_hooks_t *chunk_hooks, void *chunk,
-    size_t size, size_t offset, size_t length)
+chunk_purge_wrapper(tsdn_t *tsdn, arena_t *arena, chunk_hooks_t *chunk_hooks,
+    void *chunk, size_t size, size_t offset, size_t length)
 {
 
-	chunk_hooks_assure_initialized(arena, chunk_hooks);
+	chunk_hooks_assure_initialized(tsdn, arena, chunk_hooks);
 	return (chunk_hooks->purge(chunk, size, offset, length, arena->ind));
 }
 
@@ -677,23 +725,30 @@ chunk_split_default(void *chunk, size_t size, size_t size_a, size_t size_b,
 }
 
 static bool
-chunk_merge_default(void *chunk_a, size_t size_a, void *chunk_b, size_t size_b,
-    bool committed, unsigned arena_ind)
+chunk_merge_default_impl(void *chunk_a, void *chunk_b)
 {
 
 	if (!maps_coalesce)
 		return (true);
-	if (have_dss && chunk_in_dss(chunk_a) != chunk_in_dss(chunk_b))
+	if (have_dss && !chunk_dss_mergeable(chunk_a, chunk_b))
 		return (true);
 
 	return (false);
 }
 
+static bool
+chunk_merge_default(void *chunk_a, size_t size_a, void *chunk_b, size_t size_b,
+    bool committed, unsigned arena_ind)
+{
+
+	return (chunk_merge_default_impl(chunk_a, chunk_b));
+}
+
 static rtree_node_elm_t *
 chunks_rtree_node_alloc(size_t nelms)
 {
 
-	return ((rtree_node_elm_t *)base_alloc(nelms *
+	return ((rtree_node_elm_t *)base_alloc(TSDN_NULL, nelms *
 	    sizeof(rtree_node_elm_t)));
 }
 
@@ -716,7 +771,7 @@ chunk_boot(void)
 	 * so pages_map will always take fast path.
 	 */
 	if (!opt_lg_chunk) {
-		opt_lg_chunk = jemalloc_ffs((int)info.dwAllocationGranularity)
+		opt_lg_chunk = ffs_u((unsigned)info.dwAllocationGranularity)
 		    - 1;
 	}
 #else
@@ -730,32 +785,11 @@ chunk_boot(void)
 	chunksize_mask = chunksize - 1;
 	chunk_npages = (chunksize >> LG_PAGE);
 
-	if (have_dss && chunk_dss_boot())
-		return (true);
-	if (rtree_new(&chunks_rtree, (ZU(1) << (LG_SIZEOF_PTR+3)) -
-	    opt_lg_chunk, chunks_rtree_node_alloc, NULL))
+	if (have_dss)
+		chunk_dss_boot();
+	if (rtree_new(&chunks_rtree, (unsigned)((ZU(1) << (LG_SIZEOF_PTR+3)) -
+	    opt_lg_chunk), chunks_rtree_node_alloc, NULL))
 		return (true);
 
 	return (false);
 }
-
-void
-chunk_prefork(void)
-{
-
-	chunk_dss_prefork();
-}
-
-void
-chunk_postfork_parent(void)
-{
-
-	chunk_dss_postfork_parent();
-}
-
-void
-chunk_postfork_child(void)
-{
-
-	chunk_dss_postfork_child();
-}
diff --git a/deps/jemalloc/src/chunk_dss.c b/deps/jemalloc/src/chunk_dss.c
index 61fc91696..ee3f83888 100644
--- a/deps/jemalloc/src/chunk_dss.c
+++ b/deps/jemalloc/src/chunk_dss.c
@@ -10,20 +10,19 @@ const char	*dss_prec_names[] = {
 	"N/A"
 };
 
-/* Current dss precedence default, used when creating new arenas. */
-static dss_prec_t	dss_prec_default = DSS_PREC_DEFAULT;
-
 /*
- * Protects sbrk() calls.  This avoids malloc races among threads, though it
- * does not protect against races with threads that call sbrk() directly.
+ * Current dss precedence default, used when creating new arenas.  NB: This is
+ * stored as unsigned rather than dss_prec_t because in principle there's no
+ * guarantee that sizeof(dss_prec_t) is the same as sizeof(unsigned), and we use
+ * atomic operations to synchronize the setting.
  */
-static malloc_mutex_t	dss_mtx;
+static unsigned		dss_prec_default = (unsigned)DSS_PREC_DEFAULT;
 
 /* Base address of the DSS. */
 static void		*dss_base;
-/* Current end of the DSS, or ((void *)-1) if the DSS is exhausted. */
-static void		*dss_prev;
-/* Current upper limit on DSS addresses. */
+/* Atomic boolean indicating whether the DSS is exhausted. */
+static unsigned		dss_exhausted;
+/* Atomic current upper limit on DSS addresses. */
 static void		*dss_max;
 
 /******************************************************************************/
@@ -47,9 +46,7 @@ chunk_dss_prec_get(void)
 
 	if (!have_dss)
 		return (dss_prec_disabled);
-	malloc_mutex_lock(&dss_mtx);
-	ret = dss_prec_default;
-	malloc_mutex_unlock(&dss_mtx);
+	ret = (dss_prec_t)atomic_read_u(&dss_prec_default);
 	return (ret);
 }
 
@@ -59,15 +56,46 @@ chunk_dss_prec_set(dss_prec_t dss_prec)
 
 	if (!have_dss)
 		return (dss_prec != dss_prec_disabled);
-	malloc_mutex_lock(&dss_mtx);
-	dss_prec_default = dss_prec;
-	malloc_mutex_unlock(&dss_mtx);
+	atomic_write_u(&dss_prec_default, (unsigned)dss_prec);
 	return (false);
 }
 
+static void *
+chunk_dss_max_update(void *new_addr)
+{
+	void *max_cur;
+	spin_t spinner;
+
+	/*
+	 * Get the current end of the DSS as max_cur and assure that dss_max is
+	 * up to date.
+	 */
+	spin_init(&spinner);
+	while (true) {
+		void *max_prev = atomic_read_p(&dss_max);
+
+		max_cur = chunk_dss_sbrk(0);
+		if ((uintptr_t)max_prev > (uintptr_t)max_cur) {
+			/*
+			 * Another thread optimistically updated dss_max.  Wait
+			 * for it to finish.
+			 */
+			spin_adaptive(&spinner);
+			continue;
+		}
+		if (!atomic_cas_p(&dss_max, max_prev, max_cur))
+			break;
+	}
+	/* Fixed new_addr can only be supported if it is at the edge of DSS. */
+	if (new_addr != NULL && max_cur != new_addr)
+		return (NULL);
+
+	return (max_cur);
+}
+
 void *
-chunk_alloc_dss(arena_t *arena, void *new_addr, size_t size, size_t alignment,
-    bool *zero, bool *commit)
+chunk_alloc_dss(tsdn_t *tsdn, arena_t *arena, void *new_addr, size_t size,
+    size_t alignment, bool *zero, bool *commit)
 {
 	cassert(have_dss);
 	assert(size > 0 && (size & chunksize_mask) == 0);
@@ -80,28 +108,20 @@ chunk_alloc_dss(arena_t *arena, void *new_addr, size_t size, size_t alignment,
 	if ((intptr_t)size < 0)
 		return (NULL);
 
-	malloc_mutex_lock(&dss_mtx);
-	if (dss_prev != (void *)-1) {
-
+	if (!atomic_read_u(&dss_exhausted)) {
 		/*
 		 * The loop is necessary to recover from races with other
 		 * threads that are using the DSS for something other than
 		 * malloc.
 		 */
-		do {
-			void *ret, *cpad, *dss_next;
+		while (true) {
+			void *ret, *cpad, *max_cur, *dss_next, *dss_prev;
 			size_t gap_size, cpad_size;
 			intptr_t incr;
-			/* Avoid an unnecessary system call. */
-			if (new_addr != NULL && dss_max != new_addr)
-				break;
-
-			/* Get the current end of the DSS. */
-			dss_max = chunk_dss_sbrk(0);
 
-			/* Make sure the earlier condition still holds. */
-			if (new_addr != NULL && dss_max != new_addr)
-				break;
+			max_cur = chunk_dss_max_update(new_addr);
+			if (max_cur == NULL)
+				goto label_oom;
 
 			/*
 			 * Calculate how much padding is necessary to
@@ -120,22 +140,29 @@ chunk_alloc_dss(arena_t *arena, void *new_addr, size_t size, size_t alignment,
 			cpad_size = (uintptr_t)ret - (uintptr_t)cpad;
 			dss_next = (void *)((uintptr_t)ret + size);
 			if ((uintptr_t)ret < (uintptr_t)dss_max ||
-			    (uintptr_t)dss_next < (uintptr_t)dss_max) {
-				/* Wrap-around. */
-				malloc_mutex_unlock(&dss_mtx);
-				return (NULL);
-			}
+			    (uintptr_t)dss_next < (uintptr_t)dss_max)
+				goto label_oom; /* Wrap-around. */
 			incr = gap_size + cpad_size + size;
+
+			/*
+			 * Optimistically update dss_max, and roll back below if
+			 * sbrk() fails.  No other thread will try to extend the
+			 * DSS while dss_max is greater than the current DSS
+			 * max reported by sbrk(0).
+			 */
+			if (atomic_cas_p(&dss_max, max_cur, dss_next))
+				continue;
+
+			/* Try to allocate. */
 			dss_prev = chunk_dss_sbrk(incr);
-			if (dss_prev == dss_max) {
+			if (dss_prev == max_cur) {
 				/* Success. */
-				dss_max = dss_next;
-				malloc_mutex_unlock(&dss_mtx);
 				if (cpad_size != 0) {
 					chunk_hooks_t chunk_hooks =
 					    CHUNK_HOOKS_INITIALIZER;
-					chunk_dalloc_wrapper(arena,
+					chunk_dalloc_wrapper(tsdn, arena,
 					    &chunk_hooks, cpad, cpad_size,
+					    arena_extent_sn_next(arena), false,
 					    true);
 				}
 				if (*zero) {
@@ -147,68 +174,65 @@ chunk_alloc_dss(arena_t *arena, void *new_addr, size_t size, size_t alignment,
 					*commit = pages_decommit(ret, size);
 				return (ret);
 			}
-		} while (dss_prev != (void *)-1);
-	}
-	malloc_mutex_unlock(&dss_mtx);
 
+			/*
+			 * Failure, whether due to OOM or a race with a raw
+			 * sbrk() call from outside the allocator.  Try to roll
+			 * back optimistic dss_max update; if rollback fails,
+			 * it's due to another caller of this function having
+			 * succeeded since this invocation started, in which
+			 * case rollback is not necessary.
+			 */
+			atomic_cas_p(&dss_max, dss_next, max_cur);
+			if (dss_prev == (void *)-1) {
+				/* OOM. */
+				atomic_write_u(&dss_exhausted, (unsigned)true);
+				goto label_oom;
+			}
+		}
+	}
+label_oom:
 	return (NULL);
 }
 
-bool
-chunk_in_dss(void *chunk)
+static bool
+chunk_in_dss_helper(void *chunk, void *max)
 {
-	bool ret;
 
-	cassert(have_dss);
-
-	malloc_mutex_lock(&dss_mtx);
-	if ((uintptr_t)chunk >= (uintptr_t)dss_base
-	    && (uintptr_t)chunk < (uintptr_t)dss_max)
-		ret = true;
-	else
-		ret = false;
-	malloc_mutex_unlock(&dss_mtx);
-
-	return (ret);
+	return ((uintptr_t)chunk >= (uintptr_t)dss_base && (uintptr_t)chunk <
+	    (uintptr_t)max);
 }
 
 bool
-chunk_dss_boot(void)
+chunk_in_dss(void *chunk)
 {
 
 	cassert(have_dss);
 
-	if (malloc_mutex_init(&dss_mtx))
-		return (true);
-	dss_base = chunk_dss_sbrk(0);
-	dss_prev = dss_base;
-	dss_max = dss_base;
-
-	return (false);
+	return (chunk_in_dss_helper(chunk, atomic_read_p(&dss_max)));
 }
 
-void
-chunk_dss_prefork(void)
+bool
+chunk_dss_mergeable(void *chunk_a, void *chunk_b)
 {
+	void *max;
 
-	if (have_dss)
-		malloc_mutex_prefork(&dss_mtx);
-}
-
-void
-chunk_dss_postfork_parent(void)
-{
+	cassert(have_dss);
 
-	if (have_dss)
-		malloc_mutex_postfork_parent(&dss_mtx);
+	max = atomic_read_p(&dss_max);
+	return (chunk_in_dss_helper(chunk_a, max) ==
+	    chunk_in_dss_helper(chunk_b, max));
 }
 
 void
-chunk_dss_postfork_child(void)
+chunk_dss_boot(void)
 {
 
-	if (have_dss)
-		malloc_mutex_postfork_child(&dss_mtx);
+	cassert(have_dss);
+
+	dss_base = chunk_dss_sbrk(0);
+	dss_exhausted = (unsigned)(dss_base == (void *)-1);
+	dss_max = dss_base;
 }
 
 /******************************************************************************/
diff --git a/deps/jemalloc/src/chunk_mmap.c b/deps/jemalloc/src/chunk_mmap.c
index b9ba74191..73fc497af 100644
--- a/deps/jemalloc/src/chunk_mmap.c
+++ b/deps/jemalloc/src/chunk_mmap.c
@@ -16,23 +16,22 @@ chunk_alloc_mmap_slow(size_t size, size_t alignment, bool *zero, bool *commit)
 	do {
 		void *pages;
 		size_t leadsize;
-		pages = pages_map(NULL, alloc_size);
+		pages = pages_map(NULL, alloc_size, commit);
 		if (pages == NULL)
 			return (NULL);
 		leadsize = ALIGNMENT_CEILING((uintptr_t)pages, alignment) -
 		    (uintptr_t)pages;
-		ret = pages_trim(pages, alloc_size, leadsize, size);
+		ret = pages_trim(pages, alloc_size, leadsize, size, commit);
 	} while (ret == NULL);
 
 	assert(ret != NULL);
 	*zero = true;
-	if (!*commit)
-		*commit = pages_decommit(ret, size);
 	return (ret);
 }
 
 void *
-chunk_alloc_mmap(size_t size, size_t alignment, bool *zero, bool *commit)
+chunk_alloc_mmap(void *new_addr, size_t size, size_t alignment, bool *zero,
+    bool *commit)
 {
 	void *ret;
 	size_t offset;
@@ -53,9 +52,10 @@ chunk_alloc_mmap(size_t size, size_t alignment, bool *zero, bool *commit)
 	assert(alignment != 0);
 	assert((alignment & chunksize_mask) == 0);
 
-	ret = pages_map(NULL, size);
-	if (ret == NULL)
-		return (NULL);
+	ret = pages_map(new_addr, size, commit);
+	if (ret == NULL || ret == new_addr)
+		return (ret);
+	assert(new_addr == NULL);
 	offset = ALIGNMENT_ADDR2OFFSET(ret, alignment);
 	if (offset != 0) {
 		pages_unmap(ret, size);
@@ -64,8 +64,6 @@ chunk_alloc_mmap(size_t size, size_t alignment, bool *zero, bool *commit)
 
 	assert(ret != NULL);
 	*zero = true;
-	if (!*commit)
-		*commit = pages_decommit(ret, size);
 	return (ret);
 }
 
diff --git a/deps/jemalloc/src/ckh.c b/deps/jemalloc/src/ckh.c
index 53a1c1ef1..159bd8ae1 100644
--- a/deps/jemalloc/src/ckh.c
+++ b/deps/jemalloc/src/ckh.c
@@ -99,7 +99,8 @@ ckh_try_bucket_insert(ckh_t *ckh, size_t bucket, const void *key,
 	 * Cycle through the cells in the bucket, starting at a random position.
 	 * The randomness avoids worst-case search overhead as buckets fill up.
 	 */
-	prng32(offset, LG_CKH_BUCKET_CELLS, ckh->prng_state, CKH_A, CKH_C);
+	offset = (unsigned)prng_lg_range_u64(&ckh->prng_state,
+	    LG_CKH_BUCKET_CELLS);
 	for (i = 0; i < (ZU(1) << LG_CKH_BUCKET_CELLS); i++) {
 		cell = &ckh->tab[(bucket << LG_CKH_BUCKET_CELLS) +
 		    ((i + offset) & ((ZU(1) << LG_CKH_BUCKET_CELLS) - 1))];
@@ -141,7 +142,8 @@ ckh_evict_reloc_insert(ckh_t *ckh, size_t argbucket, void const **argkey,
 		 * were an item for which both hashes indicated the same
 		 * bucket.
 		 */
-		prng32(i, LG_CKH_BUCKET_CELLS, ckh->prng_state, CKH_A, CKH_C);
+		i = (unsigned)prng_lg_range_u64(&ckh->prng_state,
+		    LG_CKH_BUCKET_CELLS);
 		cell = &ckh->tab[(bucket << LG_CKH_BUCKET_CELLS) + i];
 		assert(cell->key != NULL);
 
@@ -247,8 +249,7 @@ ckh_grow(tsd_t *tsd, ckh_t *ckh)
 {
 	bool ret;
 	ckhc_t *tab, *ttab;
-	size_t lg_curcells;
-	unsigned lg_prevbuckets;
+	unsigned lg_prevbuckets, lg_curcells;
 
 #ifdef CKH_COUNT
 	ckh->ngrows++;
@@ -266,12 +267,12 @@ ckh_grow(tsd_t *tsd, ckh_t *ckh)
 
 		lg_curcells++;
 		usize = sa2u(sizeof(ckhc_t) << lg_curcells, CACHELINE);
-		if (usize == 0) {
+		if (unlikely(usize == 0 || usize > HUGE_MAXCLASS)) {
 			ret = true;
 			goto label_return;
 		}
-		tab = (ckhc_t *)ipallocztm(tsd, usize, CACHELINE, true, NULL,
-		    true, NULL);
+		tab = (ckhc_t *)ipallocztm(tsd_tsdn(tsd), usize, CACHELINE,
+		    true, NULL, true, arena_ichoose(tsd, NULL));
 		if (tab == NULL) {
 			ret = true;
 			goto label_return;
@@ -283,12 +284,12 @@ ckh_grow(tsd_t *tsd, ckh_t *ckh)
 		ckh->lg_curbuckets = lg_curcells - LG_CKH_BUCKET_CELLS;
 
 		if (!ckh_rebuild(ckh, tab)) {
-			idalloctm(tsd, tab, tcache_get(tsd, false), true);
+			idalloctm(tsd_tsdn(tsd), tab, NULL, true, true);
 			break;
 		}
 
 		/* Rebuilding failed, so back out partially rebuilt table. */
-		idalloctm(tsd, ckh->tab, tcache_get(tsd, false), true);
+		idalloctm(tsd_tsdn(tsd), ckh->tab, NULL, true, true);
 		ckh->tab = tab;
 		ckh->lg_curbuckets = lg_prevbuckets;
 	}
@@ -302,8 +303,8 @@ static void
 ckh_shrink(tsd_t *tsd, ckh_t *ckh)
 {
 	ckhc_t *tab, *ttab;
-	size_t lg_curcells, usize;
-	unsigned lg_prevbuckets;
+	size_t usize;
+	unsigned lg_prevbuckets, lg_curcells;
 
 	/*
 	 * It is possible (though unlikely, given well behaved hashes) that the
@@ -312,10 +313,10 @@ ckh_shrink(tsd_t *tsd, ckh_t *ckh)
 	lg_prevbuckets = ckh->lg_curbuckets;
 	lg_curcells = ckh->lg_curbuckets + LG_CKH_BUCKET_CELLS - 1;
 	usize = sa2u(sizeof(ckhc_t) << lg_curcells, CACHELINE);
-	if (usize == 0)
+	if (unlikely(usize == 0 || usize > HUGE_MAXCLASS))
 		return;
-	tab = (ckhc_t *)ipallocztm(tsd, usize, CACHELINE, true, NULL, true,
-	    NULL);
+	tab = (ckhc_t *)ipallocztm(tsd_tsdn(tsd), usize, CACHELINE, true, NULL,
+	    true, arena_ichoose(tsd, NULL));
 	if (tab == NULL) {
 		/*
 		 * An OOM error isn't worth propagating, since it doesn't
@@ -330,7 +331,7 @@ ckh_shrink(tsd_t *tsd, ckh_t *ckh)
 	ckh->lg_curbuckets = lg_curcells - LG_CKH_BUCKET_CELLS;
 
 	if (!ckh_rebuild(ckh, tab)) {
-		idalloctm(tsd, tab, tcache_get(tsd, false), true);
+		idalloctm(tsd_tsdn(tsd), tab, NULL, true, true);
 #ifdef CKH_COUNT
 		ckh->nshrinks++;
 #endif
@@ -338,7 +339,7 @@ ckh_shrink(tsd_t *tsd, ckh_t *ckh)
 	}
 
 	/* Rebuilding failed, so back out partially rebuilt table. */
-	idalloctm(tsd, ckh->tab, tcache_get(tsd, false), true);
+	idalloctm(tsd_tsdn(tsd), ckh->tab, NULL, true, true);
 	ckh->tab = tab;
 	ckh->lg_curbuckets = lg_prevbuckets;
 #ifdef CKH_COUNT
@@ -387,12 +388,12 @@ ckh_new(tsd_t *tsd, ckh_t *ckh, size_t minitems, ckh_hash_t *hash,
 	ckh->keycomp = keycomp;
 
 	usize = sa2u(sizeof(ckhc_t) << lg_mincells, CACHELINE);
-	if (usize == 0) {
+	if (unlikely(usize == 0 || usize > HUGE_MAXCLASS)) {
 		ret = true;
 		goto label_return;
 	}
-	ckh->tab = (ckhc_t *)ipallocztm(tsd, usize, CACHELINE, true, NULL, true,
-	    NULL);
+	ckh->tab = (ckhc_t *)ipallocztm(tsd_tsdn(tsd), usize, CACHELINE, true,
+	    NULL, true, arena_ichoose(tsd, NULL));
 	if (ckh->tab == NULL) {
 		ret = true;
 		goto label_return;
@@ -421,9 +422,9 @@ ckh_delete(tsd_t *tsd, ckh_t *ckh)
 	    (unsigned long long)ckh->nrelocs);
 #endif
 
-	idalloctm(tsd, ckh->tab, tcache_get(tsd, false), true);
+	idalloctm(tsd_tsdn(tsd), ckh->tab, NULL, true, true);
 	if (config_debug)
-		memset(ckh, 0x5a, sizeof(ckh_t));
+		memset(ckh, JEMALLOC_FREE_JUNK, sizeof(ckh_t));
 }
 
 size_t
diff --git a/deps/jemalloc/src/ctl.c b/deps/jemalloc/src/ctl.c
index 3de8e602d..bc78b2055 100644
--- a/deps/jemalloc/src/ctl.c
+++ b/deps/jemalloc/src/ctl.c
@@ -24,7 +24,7 @@ ctl_named_node(const ctl_node_t *node)
 }
 
 JEMALLOC_INLINE_C const ctl_named_node_t *
-ctl_named_children(const ctl_named_node_t *node, int index)
+ctl_named_children(const ctl_named_node_t *node, size_t index)
 {
 	const ctl_named_node_t *children = ctl_named_node(node->children);
 
@@ -42,25 +42,25 @@ ctl_indexed_node(const ctl_node_t *node)
 /* Function prototypes for non-inline static functions. */
 
 #define	CTL_PROTO(n)							\
-static int	n##_ctl(const size_t *mib, size_t miblen, void *oldp,	\
-    size_t *oldlenp, void *newp, size_t newlen);
+static int	n##_ctl(tsd_t *tsd, const size_t *mib, size_t miblen,	\
+    void *oldp, size_t *oldlenp, void *newp, size_t newlen);
 
 #define	INDEX_PROTO(n)							\
-static const ctl_named_node_t	*n##_index(const size_t *mib,		\
-    size_t miblen, size_t i);
+static const ctl_named_node_t	*n##_index(tsdn_t *tsdn,		\
+    const size_t *mib, size_t miblen, size_t i);
 
 static bool	ctl_arena_init(ctl_arena_stats_t *astats);
 static void	ctl_arena_clear(ctl_arena_stats_t *astats);
-static void	ctl_arena_stats_amerge(ctl_arena_stats_t *cstats,
+static void	ctl_arena_stats_amerge(tsdn_t *tsdn, ctl_arena_stats_t *cstats,
     arena_t *arena);
 static void	ctl_arena_stats_smerge(ctl_arena_stats_t *sstats,
     ctl_arena_stats_t *astats);
-static void	ctl_arena_refresh(arena_t *arena, unsigned i);
-static bool	ctl_grow(void);
-static void	ctl_refresh(void);
-static bool	ctl_init(void);
-static int	ctl_lookup(const char *name, ctl_node_t const **nodesp,
-    size_t *mibp, size_t *depthp);
+static void	ctl_arena_refresh(tsdn_t *tsdn, arena_t *arena, unsigned i);
+static bool	ctl_grow(tsdn_t *tsdn);
+static void	ctl_refresh(tsdn_t *tsdn);
+static bool	ctl_init(tsdn_t *tsdn);
+static int	ctl_lookup(tsdn_t *tsdn, const char *name,
+    ctl_node_t const **nodesp, size_t *mibp, size_t *depthp);
 
 CTL_PROTO(version)
 CTL_PROTO(epoch)
@@ -77,6 +77,7 @@ CTL_PROTO(config_cache_oblivious)
 CTL_PROTO(config_debug)
 CTL_PROTO(config_fill)
 CTL_PROTO(config_lazy_lock)
+CTL_PROTO(config_malloc_conf)
 CTL_PROTO(config_munmap)
 CTL_PROTO(config_prof)
 CTL_PROTO(config_prof_libgcc)
@@ -91,7 +92,9 @@ CTL_PROTO(opt_abort)
 CTL_PROTO(opt_dss)
 CTL_PROTO(opt_lg_chunk)
 CTL_PROTO(opt_narenas)
+CTL_PROTO(opt_purge)
 CTL_PROTO(opt_lg_dirty_mult)
+CTL_PROTO(opt_decay_time)
 CTL_PROTO(opt_stats_print)
 CTL_PROTO(opt_junk)
 CTL_PROTO(opt_zero)
@@ -114,10 +117,13 @@ CTL_PROTO(opt_prof_accum)
 CTL_PROTO(tcache_create)
 CTL_PROTO(tcache_flush)
 CTL_PROTO(tcache_destroy)
+static void	arena_i_purge(tsdn_t *tsdn, unsigned arena_ind, bool all);
 CTL_PROTO(arena_i_purge)
-static void	arena_purge(unsigned arena_ind);
+CTL_PROTO(arena_i_decay)
+CTL_PROTO(arena_i_reset)
 CTL_PROTO(arena_i_dss)
 CTL_PROTO(arena_i_lg_dirty_mult)
+CTL_PROTO(arena_i_decay_time)
 CTL_PROTO(arena_i_chunk_hooks)
 INDEX_PROTO(arena_i)
 CTL_PROTO(arenas_bin_i_size)
@@ -131,6 +137,7 @@ INDEX_PROTO(arenas_hchunk_i)
 CTL_PROTO(arenas_narenas)
 CTL_PROTO(arenas_initialized)
 CTL_PROTO(arenas_lg_dirty_mult)
+CTL_PROTO(arenas_decay_time)
 CTL_PROTO(arenas_quantum)
 CTL_PROTO(arenas_page)
 CTL_PROTO(arenas_tcache_max)
@@ -181,9 +188,11 @@ INDEX_PROTO(stats_arenas_i_hchunks_j)
 CTL_PROTO(stats_arenas_i_nthreads)
 CTL_PROTO(stats_arenas_i_dss)
 CTL_PROTO(stats_arenas_i_lg_dirty_mult)
+CTL_PROTO(stats_arenas_i_decay_time)
 CTL_PROTO(stats_arenas_i_pactive)
 CTL_PROTO(stats_arenas_i_pdirty)
 CTL_PROTO(stats_arenas_i_mapped)
+CTL_PROTO(stats_arenas_i_retained)
 CTL_PROTO(stats_arenas_i_npurge)
 CTL_PROTO(stats_arenas_i_nmadvise)
 CTL_PROTO(stats_arenas_i_purged)
@@ -196,6 +205,7 @@ CTL_PROTO(stats_active)
 CTL_PROTO(stats_metadata)
 CTL_PROTO(stats_resident)
 CTL_PROTO(stats_mapped)
+CTL_PROTO(stats_retained)
 
 /******************************************************************************/
 /* mallctl tree. */
@@ -241,6 +251,7 @@ static const ctl_named_node_t	config_node[] = {
 	{NAME("debug"),		CTL(config_debug)},
 	{NAME("fill"),		CTL(config_fill)},
 	{NAME("lazy_lock"),	CTL(config_lazy_lock)},
+	{NAME("malloc_conf"),	CTL(config_malloc_conf)},
 	{NAME("munmap"),	CTL(config_munmap)},
 	{NAME("prof"),		CTL(config_prof)},
 	{NAME("prof_libgcc"),	CTL(config_prof_libgcc)},
@@ -258,7 +269,9 @@ static const ctl_named_node_t opt_node[] = {
 	{NAME("dss"),		CTL(opt_dss)},
 	{NAME("lg_chunk"),	CTL(opt_lg_chunk)},
 	{NAME("narenas"),	CTL(opt_narenas)},
+	{NAME("purge"),		CTL(opt_purge)},
 	{NAME("lg_dirty_mult"),	CTL(opt_lg_dirty_mult)},
+	{NAME("decay_time"),	CTL(opt_decay_time)},
 	{NAME("stats_print"),	CTL(opt_stats_print)},
 	{NAME("junk"),		CTL(opt_junk)},
 	{NAME("zero"),		CTL(opt_zero)},
@@ -288,8 +301,11 @@ static const ctl_named_node_t	tcache_node[] = {
 
 static const ctl_named_node_t arena_i_node[] = {
 	{NAME("purge"),		CTL(arena_i_purge)},
+	{NAME("decay"),		CTL(arena_i_decay)},
+	{NAME("reset"),		CTL(arena_i_reset)},
 	{NAME("dss"),		CTL(arena_i_dss)},
 	{NAME("lg_dirty_mult"),	CTL(arena_i_lg_dirty_mult)},
+	{NAME("decay_time"),	CTL(arena_i_decay_time)},
 	{NAME("chunk_hooks"),	CTL(arena_i_chunk_hooks)}
 };
 static const ctl_named_node_t super_arena_i_node[] = {
@@ -339,6 +355,7 @@ static const ctl_named_node_t arenas_node[] = {
 	{NAME("narenas"),	CTL(arenas_narenas)},
 	{NAME("initialized"),	CTL(arenas_initialized)},
 	{NAME("lg_dirty_mult"),	CTL(arenas_lg_dirty_mult)},
+	{NAME("decay_time"),	CTL(arenas_decay_time)},
 	{NAME("quantum"),	CTL(arenas_quantum)},
 	{NAME("page"),		CTL(arenas_page)},
 	{NAME("tcache_max"),	CTL(arenas_tcache_max)},
@@ -439,9 +456,11 @@ static const ctl_named_node_t stats_arenas_i_node[] = {
 	{NAME("nthreads"),	CTL(stats_arenas_i_nthreads)},
 	{NAME("dss"),		CTL(stats_arenas_i_dss)},
 	{NAME("lg_dirty_mult"),	CTL(stats_arenas_i_lg_dirty_mult)},
+	{NAME("decay_time"),	CTL(stats_arenas_i_decay_time)},
 	{NAME("pactive"),	CTL(stats_arenas_i_pactive)},
 	{NAME("pdirty"),	CTL(stats_arenas_i_pdirty)},
 	{NAME("mapped"),	CTL(stats_arenas_i_mapped)},
+	{NAME("retained"),	CTL(stats_arenas_i_retained)},
 	{NAME("npurge"),	CTL(stats_arenas_i_npurge)},
 	{NAME("nmadvise"),	CTL(stats_arenas_i_nmadvise)},
 	{NAME("purged"),	CTL(stats_arenas_i_purged)},
@@ -468,6 +487,7 @@ static const ctl_named_node_t stats_node[] = {
 	{NAME("metadata"),	CTL(stats_metadata)},
 	{NAME("resident"),	CTL(stats_resident)},
 	{NAME("mapped"),	CTL(stats_mapped)},
+	{NAME("retained"),	CTL(stats_retained)},
 	{NAME("arenas"),	CHILD(indexed, stats_arenas)}
 };
 
@@ -519,8 +539,10 @@ static void
 ctl_arena_clear(ctl_arena_stats_t *astats)
 {
 
+	astats->nthreads = 0;
 	astats->dss = dss_prec_names[dss_prec_limit];
 	astats->lg_dirty_mult = -1;
+	astats->decay_time = -1;
 	astats->pactive = 0;
 	astats->pdirty = 0;
 	if (config_stats) {
@@ -538,20 +560,27 @@ ctl_arena_clear(ctl_arena_stats_t *astats)
 }
 
 static void
-ctl_arena_stats_amerge(ctl_arena_stats_t *cstats, arena_t *arena)
+ctl_arena_stats_amerge(tsdn_t *tsdn, ctl_arena_stats_t *cstats, arena_t *arena)
 {
 	unsigned i;
 
-	arena_stats_merge(arena, &cstats->dss, &cstats->lg_dirty_mult,
-	    &cstats->pactive, &cstats->pdirty, &cstats->astats, cstats->bstats,
-	    cstats->lstats, cstats->hstats);
-
-	for (i = 0; i < NBINS; i++) {
-		cstats->allocated_small += cstats->bstats[i].curregs *
-		    index2size(i);
-		cstats->nmalloc_small += cstats->bstats[i].nmalloc;
-		cstats->ndalloc_small += cstats->bstats[i].ndalloc;
-		cstats->nrequests_small += cstats->bstats[i].nrequests;
+	if (config_stats) {
+		arena_stats_merge(tsdn, arena, &cstats->nthreads, &cstats->dss,
+		    &cstats->lg_dirty_mult, &cstats->decay_time,
+		    &cstats->pactive, &cstats->pdirty, &cstats->astats,
+		    cstats->bstats, cstats->lstats, cstats->hstats);
+
+		for (i = 0; i < NBINS; i++) {
+			cstats->allocated_small += cstats->bstats[i].curregs *
+			    index2size(i);
+			cstats->nmalloc_small += cstats->bstats[i].nmalloc;
+			cstats->ndalloc_small += cstats->bstats[i].ndalloc;
+			cstats->nrequests_small += cstats->bstats[i].nrequests;
+		}
+	} else {
+		arena_basic_stats_merge(tsdn, arena, &cstats->nthreads,
+		    &cstats->dss, &cstats->lg_dirty_mult, &cstats->decay_time,
+		    &cstats->pactive, &cstats->pdirty);
 	}
 }
 
@@ -560,89 +589,91 @@ ctl_arena_stats_smerge(ctl_arena_stats_t *sstats, ctl_arena_stats_t *astats)
 {
 	unsigned i;
 
+	sstats->nthreads += astats->nthreads;
 	sstats->pactive += astats->pactive;
 	sstats->pdirty += astats->pdirty;
 
-	sstats->astats.mapped += astats->astats.mapped;
-	sstats->astats.npurge += astats->astats.npurge;
-	sstats->astats.nmadvise += astats->astats.nmadvise;
-	sstats->astats.purged += astats->astats.purged;
-
-	sstats->astats.metadata_mapped += astats->astats.metadata_mapped;
-	sstats->astats.metadata_allocated += astats->astats.metadata_allocated;
-
-	sstats->allocated_small += astats->allocated_small;
-	sstats->nmalloc_small += astats->nmalloc_small;
-	sstats->ndalloc_small += astats->ndalloc_small;
-	sstats->nrequests_small += astats->nrequests_small;
-
-	sstats->astats.allocated_large += astats->astats.allocated_large;
-	sstats->astats.nmalloc_large += astats->astats.nmalloc_large;
-	sstats->astats.ndalloc_large += astats->astats.ndalloc_large;
-	sstats->astats.nrequests_large += astats->astats.nrequests_large;
-
-	sstats->astats.allocated_huge += astats->astats.allocated_huge;
-	sstats->astats.nmalloc_huge += astats->astats.nmalloc_huge;
-	sstats->astats.ndalloc_huge += astats->astats.ndalloc_huge;
-
-	for (i = 0; i < NBINS; i++) {
-		sstats->bstats[i].nmalloc += astats->bstats[i].nmalloc;
-		sstats->bstats[i].ndalloc += astats->bstats[i].ndalloc;
-		sstats->bstats[i].nrequests += astats->bstats[i].nrequests;
-		sstats->bstats[i].curregs += astats->bstats[i].curregs;
-		if (config_tcache) {
-			sstats->bstats[i].nfills += astats->bstats[i].nfills;
-			sstats->bstats[i].nflushes +=
-			    astats->bstats[i].nflushes;
+	if (config_stats) {
+		sstats->astats.mapped += astats->astats.mapped;
+		sstats->astats.retained += astats->astats.retained;
+		sstats->astats.npurge += astats->astats.npurge;
+		sstats->astats.nmadvise += astats->astats.nmadvise;
+		sstats->astats.purged += astats->astats.purged;
+
+		sstats->astats.metadata_mapped +=
+		    astats->astats.metadata_mapped;
+		sstats->astats.metadata_allocated +=
+		    astats->astats.metadata_allocated;
+
+		sstats->allocated_small += astats->allocated_small;
+		sstats->nmalloc_small += astats->nmalloc_small;
+		sstats->ndalloc_small += astats->ndalloc_small;
+		sstats->nrequests_small += astats->nrequests_small;
+
+		sstats->astats.allocated_large +=
+		    astats->astats.allocated_large;
+		sstats->astats.nmalloc_large += astats->astats.nmalloc_large;
+		sstats->astats.ndalloc_large += astats->astats.ndalloc_large;
+		sstats->astats.nrequests_large +=
+		    astats->astats.nrequests_large;
+
+		sstats->astats.allocated_huge += astats->astats.allocated_huge;
+		sstats->astats.nmalloc_huge += astats->astats.nmalloc_huge;
+		sstats->astats.ndalloc_huge += astats->astats.ndalloc_huge;
+
+		for (i = 0; i < NBINS; i++) {
+			sstats->bstats[i].nmalloc += astats->bstats[i].nmalloc;
+			sstats->bstats[i].ndalloc += astats->bstats[i].ndalloc;
+			sstats->bstats[i].nrequests +=
+			    astats->bstats[i].nrequests;
+			sstats->bstats[i].curregs += astats->bstats[i].curregs;
+			if (config_tcache) {
+				sstats->bstats[i].nfills +=
+				    astats->bstats[i].nfills;
+				sstats->bstats[i].nflushes +=
+				    astats->bstats[i].nflushes;
+			}
+			sstats->bstats[i].nruns += astats->bstats[i].nruns;
+			sstats->bstats[i].reruns += astats->bstats[i].reruns;
+			sstats->bstats[i].curruns += astats->bstats[i].curruns;
 		}
-		sstats->bstats[i].nruns += astats->bstats[i].nruns;
-		sstats->bstats[i].reruns += astats->bstats[i].reruns;
-		sstats->bstats[i].curruns += astats->bstats[i].curruns;
-	}
 
-	for (i = 0; i < nlclasses; i++) {
-		sstats->lstats[i].nmalloc += astats->lstats[i].nmalloc;
-		sstats->lstats[i].ndalloc += astats->lstats[i].ndalloc;
-		sstats->lstats[i].nrequests += astats->lstats[i].nrequests;
-		sstats->lstats[i].curruns += astats->lstats[i].curruns;
-	}
+		for (i = 0; i < nlclasses; i++) {
+			sstats->lstats[i].nmalloc += astats->lstats[i].nmalloc;
+			sstats->lstats[i].ndalloc += astats->lstats[i].ndalloc;
+			sstats->lstats[i].nrequests +=
+			    astats->lstats[i].nrequests;
+			sstats->lstats[i].curruns += astats->lstats[i].curruns;
+		}
 
-	for (i = 0; i < nhclasses; i++) {
-		sstats->hstats[i].nmalloc += astats->hstats[i].nmalloc;
-		sstats->hstats[i].ndalloc += astats->hstats[i].ndalloc;
-		sstats->hstats[i].curhchunks += astats->hstats[i].curhchunks;
+		for (i = 0; i < nhclasses; i++) {
+			sstats->hstats[i].nmalloc += astats->hstats[i].nmalloc;
+			sstats->hstats[i].ndalloc += astats->hstats[i].ndalloc;
+			sstats->hstats[i].curhchunks +=
+			    astats->hstats[i].curhchunks;
+		}
 	}
 }
 
 static void
-ctl_arena_refresh(arena_t *arena, unsigned i)
+ctl_arena_refresh(tsdn_t *tsdn, arena_t *arena, unsigned i)
 {
 	ctl_arena_stats_t *astats = &ctl_stats.arenas[i];
 	ctl_arena_stats_t *sstats = &ctl_stats.arenas[ctl_stats.narenas];
 
 	ctl_arena_clear(astats);
-
-	sstats->nthreads += astats->nthreads;
-	if (config_stats) {
-		ctl_arena_stats_amerge(astats, arena);
-		/* Merge into sum stats as well. */
-		ctl_arena_stats_smerge(sstats, astats);
-	} else {
-		astats->pactive += arena->nactive;
-		astats->pdirty += arena->ndirty;
-		/* Merge into sum stats as well. */
-		sstats->pactive += arena->nactive;
-		sstats->pdirty += arena->ndirty;
-	}
+	ctl_arena_stats_amerge(tsdn, astats, arena);
+	/* Merge into sum stats as well. */
+	ctl_arena_stats_smerge(sstats, astats);
 }
 
 static bool
-ctl_grow(void)
+ctl_grow(tsdn_t *tsdn)
 {
 	ctl_arena_stats_t *astats;
 
 	/* Initialize new arena. */
-	if (arena_init(ctl_stats.narenas) == NULL)
+	if (arena_init(tsdn, ctl_stats.narenas) == NULL)
 		return (true);
 
 	/* Allocate extended arena stats. */
@@ -677,47 +708,32 @@ ctl_grow(void)
 }
 
 static void
-ctl_refresh(void)
+ctl_refresh(tsdn_t *tsdn)
 {
-	tsd_t *tsd;
 	unsigned i;
-	bool refreshed;
 	VARIABLE_ARRAY(arena_t *, tarenas, ctl_stats.narenas);
 
 	/*
 	 * Clear sum stats, since they will be merged into by
 	 * ctl_arena_refresh().
 	 */
-	ctl_stats.arenas[ctl_stats.narenas].nthreads = 0;
 	ctl_arena_clear(&ctl_stats.arenas[ctl_stats.narenas]);
 
-	tsd = tsd_fetch();
-	for (i = 0, refreshed = false; i < ctl_stats.narenas; i++) {
-		tarenas[i] = arena_get(tsd, i, false, false);
-		if (tarenas[i] == NULL && !refreshed) {
-			tarenas[i] = arena_get(tsd, i, false, true);
-			refreshed = true;
-		}
-	}
-
-	for (i = 0; i < ctl_stats.narenas; i++) {
-		if (tarenas[i] != NULL)
-			ctl_stats.arenas[i].nthreads = arena_nbound(i);
-		else
-			ctl_stats.arenas[i].nthreads = 0;
-	}
+	for (i = 0; i < ctl_stats.narenas; i++)
+		tarenas[i] = arena_get(tsdn, i, false);
 
 	for (i = 0; i < ctl_stats.narenas; i++) {
 		bool initialized = (tarenas[i] != NULL);
 
 		ctl_stats.arenas[i].initialized = initialized;
 		if (initialized)
-			ctl_arena_refresh(tarenas[i], i);
+			ctl_arena_refresh(tsdn, tarenas[i], i);
 	}
 
 	if (config_stats) {
 		size_t base_allocated, base_resident, base_mapped;
-		base_stats_get(&base_allocated, &base_resident, &base_mapped);
+		base_stats_get(tsdn, &base_allocated, &base_resident,
+		    &base_mapped);
 		ctl_stats.allocated =
 		    ctl_stats.arenas[ctl_stats.narenas].allocated_small +
 		    ctl_stats.arenas[ctl_stats.narenas].astats.allocated_large +
@@ -734,17 +750,19 @@ ctl_refresh(void)
 		    ctl_stats.arenas[ctl_stats.narenas].pdirty) << LG_PAGE);
 		ctl_stats.mapped = base_mapped +
 		    ctl_stats.arenas[ctl_stats.narenas].astats.mapped;
+		ctl_stats.retained =
+		    ctl_stats.arenas[ctl_stats.narenas].astats.retained;
 	}
 
 	ctl_epoch++;
 }
 
 static bool
-ctl_init(void)
+ctl_init(tsdn_t *tsdn)
 {
 	bool ret;
 
-	malloc_mutex_lock(&ctl_mtx);
+	malloc_mutex_lock(tsdn, &ctl_mtx);
 	if (!ctl_initialized) {
 		/*
 		 * Allocate space for one extra arena stats element, which
@@ -786,19 +804,19 @@ ctl_init(void)
 		ctl_stats.arenas[ctl_stats.narenas].initialized = true;
 
 		ctl_epoch = 0;
-		ctl_refresh();
+		ctl_refresh(tsdn);
 		ctl_initialized = true;
 	}
 
 	ret = false;
 label_return:
-	malloc_mutex_unlock(&ctl_mtx);
+	malloc_mutex_unlock(tsdn, &ctl_mtx);
 	return (ret);
 }
 
 static int
-ctl_lookup(const char *name, ctl_node_t const **nodesp, size_t *mibp,
-    size_t *depthp)
+ctl_lookup(tsdn_t *tsdn, const char *name, ctl_node_t const **nodesp,
+    size_t *mibp, size_t *depthp)
 {
 	int ret;
 	const char *elm, *tdot, *dot;
@@ -850,7 +868,7 @@ ctl_lookup(const char *name, ctl_node_t const **nodesp, size_t *mibp,
 			}
 
 			inode = ctl_indexed_node(node->children);
-			node = inode->index(mibp, *depthp, (size_t)index);
+			node = inode->index(tsdn, mibp, *depthp, (size_t)index);
 			if (node == NULL) {
 				ret = ENOENT;
 				goto label_return;
@@ -894,8 +912,8 @@ label_return:
 }
 
 int
-ctl_byname(const char *name, void *oldp, size_t *oldlenp, void *newp,
-    size_t newlen)
+ctl_byname(tsd_t *tsd, const char *name, void *oldp, size_t *oldlenp,
+    void *newp, size_t newlen)
 {
 	int ret;
 	size_t depth;
@@ -903,19 +921,19 @@ ctl_byname(const char *name, void *oldp, size_t *oldlenp, void *newp,
 	size_t mib[CTL_MAX_DEPTH];
 	const ctl_named_node_t *node;
 
-	if (!ctl_initialized && ctl_init()) {
+	if (!ctl_initialized && ctl_init(tsd_tsdn(tsd))) {
 		ret = EAGAIN;
 		goto label_return;
 	}
 
 	depth = CTL_MAX_DEPTH;
-	ret = ctl_lookup(name, nodes, mib, &depth);
+	ret = ctl_lookup(tsd_tsdn(tsd), name, nodes, mib, &depth);
 	if (ret != 0)
 		goto label_return;
 
 	node = ctl_named_node(nodes[depth-1]);
 	if (node != NULL && node->ctl)
-		ret = node->ctl(mib, depth, oldp, oldlenp, newp, newlen);
+		ret = node->ctl(tsd, mib, depth, oldp, oldlenp, newp, newlen);
 	else {
 		/* The name refers to a partial path through the ctl tree. */
 		ret = ENOENT;
@@ -926,29 +944,29 @@ label_return:
 }
 
 int
-ctl_nametomib(const char *name, size_t *mibp, size_t *miblenp)
+ctl_nametomib(tsdn_t *tsdn, const char *name, size_t *mibp, size_t *miblenp)
 {
 	int ret;
 
-	if (!ctl_initialized && ctl_init()) {
+	if (!ctl_initialized && ctl_init(tsdn)) {
 		ret = EAGAIN;
 		goto label_return;
 	}
 
-	ret = ctl_lookup(name, NULL, mibp, miblenp);
+	ret = ctl_lookup(tsdn, name, NULL, mibp, miblenp);
 label_return:
 	return(ret);
 }
 
 int
-ctl_bymib(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
-    void *newp, size_t newlen)
+ctl_bymib(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
 	const ctl_named_node_t *node;
 	size_t i;
 
-	if (!ctl_initialized && ctl_init()) {
+	if (!ctl_initialized && ctl_init(tsd_tsdn(tsd))) {
 		ret = EAGAIN;
 		goto label_return;
 	}
@@ -960,7 +978,7 @@ ctl_bymib(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
 		assert(node->nchildren > 0);
 		if (ctl_named_node(node->children) != NULL) {
 			/* Children are named. */
-			if (node->nchildren <= mib[i]) {
+			if (node->nchildren <= (unsigned)mib[i]) {
 				ret = ENOENT;
 				goto label_return;
 			}
@@ -970,7 +988,7 @@ ctl_bymib(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
 
 			/* Indexed element. */
 			inode = ctl_indexed_node(node->children);
-			node = inode->index(mib, miblen, mib[i]);
+			node = inode->index(tsd_tsdn(tsd), mib, miblen, mib[i]);
 			if (node == NULL) {
 				ret = ENOENT;
 				goto label_return;
@@ -980,7 +998,7 @@ ctl_bymib(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
 
 	/* Call the ctl function. */
 	if (node && node->ctl)
-		ret = node->ctl(mib, miblen, oldp, oldlenp, newp, newlen);
+		ret = node->ctl(tsd, mib, miblen, oldp, oldlenp, newp, newlen);
 	else {
 		/* Partial MIB. */
 		ret = ENOENT;
@@ -994,7 +1012,7 @@ bool
 ctl_boot(void)
 {
 
-	if (malloc_mutex_init(&ctl_mtx))
+	if (malloc_mutex_init(&ctl_mtx, "ctl", WITNESS_RANK_CTL))
 		return (true);
 
 	ctl_initialized = false;
@@ -1003,24 +1021,24 @@ ctl_boot(void)
 }
 
 void
-ctl_prefork(void)
+ctl_prefork(tsdn_t *tsdn)
 {
 
-	malloc_mutex_prefork(&ctl_mtx);
+	malloc_mutex_prefork(tsdn, &ctl_mtx);
 }
 
 void
-ctl_postfork_parent(void)
+ctl_postfork_parent(tsdn_t *tsdn)
 {
 
-	malloc_mutex_postfork_parent(&ctl_mtx);
+	malloc_mutex_postfork_parent(tsdn, &ctl_mtx);
 }
 
 void
-ctl_postfork_child(void)
+ctl_postfork_child(tsdn_t *tsdn)
 {
 
-	malloc_mutex_postfork_child(&ctl_mtx);
+	malloc_mutex_postfork_child(tsdn, &ctl_mtx);
 }
 
 /******************************************************************************/
@@ -1077,8 +1095,8 @@ ctl_postfork_child(void)
  */
 #define	CTL_RO_CLGEN(c, l, n, v, t)					\
 static int								\
-n##_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,	\
-    void *newp, size_t newlen)						\
+n##_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,	\
+    size_t *oldlenp, void *newp, size_t newlen)				\
 {									\
 	int ret;							\
 	t oldval;							\
@@ -1086,7 +1104,7 @@ n##_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,	\
 	if (!(c))							\
 		return (ENOENT);					\
 	if (l)								\
-		malloc_mutex_lock(&ctl_mtx);				\
+		malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx);		\
 	READONLY();							\
 	oldval = (v);							\
 	READ(oldval, t);						\
@@ -1094,47 +1112,47 @@ n##_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,	\
 	ret = 0;							\
 label_return:								\
 	if (l)								\
-		malloc_mutex_unlock(&ctl_mtx);				\
+		malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx);		\
 	return (ret);							\
 }
 
 #define	CTL_RO_CGEN(c, n, v, t)						\
 static int								\
-n##_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,	\
-    void *newp, size_t newlen)						\
+n##_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,	\
+    size_t *oldlenp, void *newp, size_t newlen)				\
 {									\
 	int ret;							\
 	t oldval;							\
 									\
 	if (!(c))							\
 		return (ENOENT);					\
-	malloc_mutex_lock(&ctl_mtx);					\
+	malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx);			\
 	READONLY();							\
 	oldval = (v);							\
 	READ(oldval, t);						\
 									\
 	ret = 0;							\
 label_return:								\
-	malloc_mutex_unlock(&ctl_mtx);					\
+	malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx);			\
 	return (ret);							\
 }
 
 #define	CTL_RO_GEN(n, v, t)						\
 static int								\
-n##_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,	\
-    void *newp, size_t newlen)						\
+n##_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,	\
+    size_t *oldlenp, void *newp, size_t newlen)				\
 {									\
 	int ret;							\
 	t oldval;							\
 									\
-	malloc_mutex_lock(&ctl_mtx);					\
+	malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx);			\
 	READONLY();							\
 	oldval = (v);							\
 	READ(oldval, t);						\
 									\
 	ret = 0;							\
 label_return:								\
-	malloc_mutex_unlock(&ctl_mtx);					\
+	malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx);			\
 	return (ret);							\
 }
 
@@ -1144,8 +1162,8 @@ label_return:								\
  */
 #define	CTL_RO_NL_CGEN(c, n, v, t)					\
 static int								\
-n##_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,	\
-    void *newp, size_t newlen)						\
+n##_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,	\
+    size_t *oldlenp, void *newp, size_t newlen)				\
 {									\
 	int ret;							\
 	t oldval;							\
@@ -1163,8 +1181,8 @@ label_return:								\
 
 #define	CTL_RO_NL_GEN(n, v, t)						\
 static int								\
-n##_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,	\
-    void *newp, size_t newlen)						\
+n##_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,	\
+    size_t *oldlenp, void *newp, size_t newlen)				\
 {									\
 	int ret;							\
 	t oldval;							\
@@ -1180,17 +1198,15 @@ label_return:								\
 
 #define	CTL_TSD_RO_NL_CGEN(c, n, m, t)					\
 static int								\
-n##_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,	\
-    void *newp, size_t newlen)						\
+n##_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,	\
+    size_t *oldlenp, void *newp, size_t newlen)				\
 {									\
 	int ret;							\
 	t oldval;							\
-	tsd_t *tsd;							\
 									\
 	if (!(c))							\
 		return (ENOENT);					\
 	READONLY();							\
-	tsd = tsd_fetch();						\
 	oldval = (m(tsd));						\
 	READ(oldval, t);						\
 									\
@@ -1199,17 +1215,17 @@ label_return:								\
 	return (ret);							\
 }
 
-#define	CTL_RO_BOOL_CONFIG_GEN(n)					\
+#define	CTL_RO_CONFIG_GEN(n, t)						\
 static int								\
-n##_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,	\
-    void *newp, size_t newlen)						\
+n##_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,	\
+    size_t *oldlenp, void *newp, size_t newlen)				\
 {									\
 	int ret;							\
-	bool oldval;							\
+	t oldval;							\
 									\
 	READONLY();							\
 	oldval = n;							\
-	READ(oldval, bool);						\
+	READ(oldval, t);						\
 									\
 	ret = 0;							\
 label_return:								\
@@ -1221,48 +1237,51 @@ label_return:								\
 CTL_RO_NL_GEN(version, JEMALLOC_VERSION, const char *)
 
 static int
-epoch_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
-    void *newp, size_t newlen)
+epoch_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
 	UNUSED uint64_t newval;
 
-	malloc_mutex_lock(&ctl_mtx);
+	malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx);
 	WRITE(newval, uint64_t);
 	if (newp != NULL)
-		ctl_refresh();
+		ctl_refresh(tsd_tsdn(tsd));
 	READ(ctl_epoch, uint64_t);
 
 	ret = 0;
 label_return:
-	malloc_mutex_unlock(&ctl_mtx);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx);
 	return (ret);
 }
 
 /******************************************************************************/
 
-CTL_RO_BOOL_CONFIG_GEN(config_cache_oblivious)
-CTL_RO_BOOL_CONFIG_GEN(config_debug)
-CTL_RO_BOOL_CONFIG_GEN(config_fill)
-CTL_RO_BOOL_CONFIG_GEN(config_lazy_lock)
-CTL_RO_BOOL_CONFIG_GEN(config_munmap)
-CTL_RO_BOOL_CONFIG_GEN(config_prof)
-CTL_RO_BOOL_CONFIG_GEN(config_prof_libgcc)
-CTL_RO_BOOL_CONFIG_GEN(config_prof_libunwind)
-CTL_RO_BOOL_CONFIG_GEN(config_stats)
-CTL_RO_BOOL_CONFIG_GEN(config_tcache)
-CTL_RO_BOOL_CONFIG_GEN(config_tls)
-CTL_RO_BOOL_CONFIG_GEN(config_utrace)
-CTL_RO_BOOL_CONFIG_GEN(config_valgrind)
-CTL_RO_BOOL_CONFIG_GEN(config_xmalloc)
+CTL_RO_CONFIG_GEN(config_cache_oblivious, bool)
+CTL_RO_CONFIG_GEN(config_debug, bool)
+CTL_RO_CONFIG_GEN(config_fill, bool)
+CTL_RO_CONFIG_GEN(config_lazy_lock, bool)
+CTL_RO_CONFIG_GEN(config_malloc_conf, const char *)
+CTL_RO_CONFIG_GEN(config_munmap, bool)
+CTL_RO_CONFIG_GEN(config_prof, bool)
+CTL_RO_CONFIG_GEN(config_prof_libgcc, bool)
+CTL_RO_CONFIG_GEN(config_prof_libunwind, bool)
+CTL_RO_CONFIG_GEN(config_stats, bool)
+CTL_RO_CONFIG_GEN(config_tcache, bool)
+CTL_RO_CONFIG_GEN(config_tls, bool)
+CTL_RO_CONFIG_GEN(config_utrace, bool)
+CTL_RO_CONFIG_GEN(config_valgrind, bool)
+CTL_RO_CONFIG_GEN(config_xmalloc, bool)
 
 /******************************************************************************/
 
 CTL_RO_NL_GEN(opt_abort, opt_abort, bool)
 CTL_RO_NL_GEN(opt_dss, opt_dss, const char *)
 CTL_RO_NL_GEN(opt_lg_chunk, opt_lg_chunk, size_t)
-CTL_RO_NL_GEN(opt_narenas, opt_narenas, size_t)
+CTL_RO_NL_GEN(opt_narenas, opt_narenas, unsigned)
+CTL_RO_NL_GEN(opt_purge, purge_mode_names[opt_purge], const char *)
 CTL_RO_NL_GEN(opt_lg_dirty_mult, opt_lg_dirty_mult, ssize_t)
+CTL_RO_NL_GEN(opt_decay_time, opt_decay_time, ssize_t)
 CTL_RO_NL_GEN(opt_stats_print, opt_stats_print, bool)
 CTL_RO_NL_CGEN(config_fill, opt_junk, opt_junk, const char *)
 CTL_RO_NL_CGEN(config_fill, opt_quarantine, opt_quarantine, size_t)
@@ -1287,20 +1306,18 @@ CTL_RO_NL_CGEN(config_prof, opt_prof_leak, opt_prof_leak, bool)
 /******************************************************************************/
 
 static int
-thread_arena_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
-    void *newp, size_t newlen)
+thread_arena_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
-	tsd_t *tsd;
 	arena_t *oldarena;
 	unsigned newind, oldind;
 
-	tsd = tsd_fetch();
 	oldarena = arena_choose(tsd, NULL);
 	if (oldarena == NULL)
 		return (EAGAIN);
 
-	malloc_mutex_lock(&ctl_mtx);
+	malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx);
 	newind = oldind = oldarena->ind;
 	WRITE(newind, unsigned);
 	READ(oldind, unsigned);
@@ -1314,7 +1331,7 @@ thread_arena_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
 		}
 
 		/* Initialize arena if necessary. */
-		newarena = arena_get(tsd, newind, true, true);
+		newarena = arena_get(tsd_tsdn(tsd), newind, true);
 		if (newarena == NULL) {
 			ret = EAGAIN;
 			goto label_return;
@@ -1324,15 +1341,15 @@ thread_arena_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
 		if (config_tcache) {
 			tcache_t *tcache = tsd_tcache_get(tsd);
 			if (tcache != NULL) {
-				tcache_arena_reassociate(tcache, oldarena,
-				    newarena);
+				tcache_arena_reassociate(tsd_tsdn(tsd), tcache,
+				    oldarena, newarena);
 			}
 		}
 	}
 
 	ret = 0;
 label_return:
-	malloc_mutex_unlock(&ctl_mtx);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx);
 	return (ret);
 }
 
@@ -1346,8 +1363,8 @@ CTL_TSD_RO_NL_CGEN(config_stats, thread_deallocatedp,
     tsd_thread_deallocatedp_get, uint64_t *)
 
 static int
-thread_tcache_enabled_ctl(const size_t *mib, size_t miblen, void *oldp,
-    size_t *oldlenp, void *newp, size_t newlen)
+thread_tcache_enabled_ctl(tsd_t *tsd, const size_t *mib, size_t miblen,
+    void *oldp, size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
 	bool oldval;
@@ -1371,8 +1388,8 @@ label_return:
 }
 
 static int
-thread_tcache_flush_ctl(const size_t *mib, size_t miblen, void *oldp,
-    size_t *oldlenp, void *newp, size_t newlen)
+thread_tcache_flush_ctl(tsd_t *tsd, const size_t *mib, size_t miblen,
+    void *oldp, size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
 
@@ -1390,7 +1407,7 @@ label_return:
 }
 
 static int
-thread_prof_name_ctl(const size_t *mib, size_t miblen, void *oldp,
+thread_prof_name_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
     size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
@@ -1401,20 +1418,16 @@ thread_prof_name_ctl(const size_t *mib, size_t miblen, void *oldp,
 	READ_XOR_WRITE();
 
 	if (newp != NULL) {
-		tsd_t *tsd;
-
 		if (newlen != sizeof(const char *)) {
 			ret = EINVAL;
 			goto label_return;
 		}
 
-		tsd = tsd_fetch();
-
 		if ((ret = prof_thread_name_set(tsd, *(const char **)newp)) !=
 		    0)
 			goto label_return;
 	} else {
-		const char *oldname = prof_thread_name_get();
+		const char *oldname = prof_thread_name_get(tsd);
 		READ(oldname, const char *);
 	}
 
@@ -1424,7 +1437,7 @@ label_return:
 }
 
 static int
-thread_prof_active_ctl(const size_t *mib, size_t miblen, void *oldp,
+thread_prof_active_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
     size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
@@ -1433,13 +1446,13 @@ thread_prof_active_ctl(const size_t *mib, size_t miblen, void *oldp,
 	if (!config_prof)
 		return (ENOENT);
 
-	oldval = prof_thread_active_get();
+	oldval = prof_thread_active_get(tsd);
 	if (newp != NULL) {
 		if (newlen != sizeof(bool)) {
 			ret = EINVAL;
 			goto label_return;
 		}
-		if (prof_thread_active_set(*(bool *)newp)) {
+		if (prof_thread_active_set(tsd, *(bool *)newp)) {
 			ret = EAGAIN;
 			goto label_return;
 		}
@@ -1454,19 +1467,16 @@ label_return:
 /******************************************************************************/
 
 static int
-tcache_create_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
-    void *newp, size_t newlen)
+tcache_create_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
-	tsd_t *tsd;
 	unsigned tcache_ind;
 
 	if (!config_tcache)
 		return (ENOENT);
 
-	tsd = tsd_fetch();
-
-	malloc_mutex_lock(&ctl_mtx);
+	malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx);
 	READONLY();
 	if (tcaches_create(tsd, &tcache_ind)) {
 		ret = EFAULT;
@@ -1476,23 +1486,20 @@ tcache_create_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
 
 	ret = 0;
 label_return:
-	malloc_mutex_unlock(&ctl_mtx);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx);
 	return (ret);
 }
 
 static int
-tcache_flush_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
-    void *newp, size_t newlen)
+tcache_flush_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
-	tsd_t *tsd;
 	unsigned tcache_ind;
 
 	if (!config_tcache)
 		return (ENOENT);
 
-	tsd = tsd_fetch();
-
 	WRITEONLY();
 	tcache_ind = UINT_MAX;
 	WRITE(tcache_ind, unsigned);
@@ -1508,18 +1515,15 @@ label_return:
 }
 
 static int
-tcache_destroy_ctl(const size_t *mib, size_t miblen, void *oldp,
+tcache_destroy_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
     size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
-	tsd_t *tsd;
 	unsigned tcache_ind;
 
 	if (!config_tcache)
 		return (ENOENT);
 
-	tsd = tsd_fetch();
-
 	WRITEONLY();
 	tcache_ind = UINT_MAX;
 	WRITE(tcache_ind, unsigned);
@@ -1536,48 +1540,56 @@ label_return:
 
 /******************************************************************************/
 
-/* ctl_mutex must be held during execution of this function. */
 static void
-arena_purge(unsigned arena_ind)
+arena_i_purge(tsdn_t *tsdn, unsigned arena_ind, bool all)
 {
-	tsd_t *tsd;
-	unsigned i;
-	bool refreshed;
-	VARIABLE_ARRAY(arena_t *, tarenas, ctl_stats.narenas);
 
-	tsd = tsd_fetch();
-	for (i = 0, refreshed = false; i < ctl_stats.narenas; i++) {
-		tarenas[i] = arena_get(tsd, i, false, false);
-		if (tarenas[i] == NULL && !refreshed) {
-			tarenas[i] = arena_get(tsd, i, false, true);
-			refreshed = true;
-		}
-	}
+	malloc_mutex_lock(tsdn, &ctl_mtx);
+	{
+		unsigned narenas = ctl_stats.narenas;
 
-	if (arena_ind == ctl_stats.narenas) {
-		unsigned i;
-		for (i = 0; i < ctl_stats.narenas; i++) {
-			if (tarenas[i] != NULL)
-				arena_purge_all(tarenas[i]);
+		if (arena_ind == narenas) {
+			unsigned i;
+			VARIABLE_ARRAY(arena_t *, tarenas, narenas);
+
+			for (i = 0; i < narenas; i++)
+				tarenas[i] = arena_get(tsdn, i, false);
+
+			/*
+			 * No further need to hold ctl_mtx, since narenas and
+			 * tarenas contain everything needed below.
+			 */
+			malloc_mutex_unlock(tsdn, &ctl_mtx);
+
+			for (i = 0; i < narenas; i++) {
+				if (tarenas[i] != NULL)
+					arena_purge(tsdn, tarenas[i], all);
+			}
+		} else {
+			arena_t *tarena;
+
+			assert(arena_ind < narenas);
+
+			tarena = arena_get(tsdn, arena_ind, false);
+
+			/* No further need to hold ctl_mtx. */
+			malloc_mutex_unlock(tsdn, &ctl_mtx);
+
+			if (tarena != NULL)
+				arena_purge(tsdn, tarena, all);
 		}
-	} else {
-		assert(arena_ind < ctl_stats.narenas);
-		if (tarenas[arena_ind] != NULL)
-			arena_purge_all(tarenas[arena_ind]);
 	}
 }
 
 static int
-arena_i_purge_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
-    void *newp, size_t newlen)
+arena_i_purge_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
 
 	READONLY();
 	WRITEONLY();
-	malloc_mutex_lock(&ctl_mtx);
-	arena_purge(mib[1]);
-	malloc_mutex_unlock(&ctl_mtx);
+	arena_i_purge(tsd_tsdn(tsd), (unsigned)mib[1], true);
 
 	ret = 0;
 label_return:
@@ -1585,16 +1597,65 @@ label_return:
 }
 
 static int
-arena_i_dss_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
-    void *newp, size_t newlen)
+arena_i_decay_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen)
+{
+	int ret;
+
+	READONLY();
+	WRITEONLY();
+	arena_i_purge(tsd_tsdn(tsd), (unsigned)mib[1], false);
+
+	ret = 0;
+label_return:
+	return (ret);
+}
+
+static int
+arena_i_reset_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen)
+{
+	int ret;
+	unsigned arena_ind;
+	arena_t *arena;
+
+	READONLY();
+	WRITEONLY();
+
+	if ((config_valgrind && unlikely(in_valgrind)) || (config_fill &&
+	    unlikely(opt_quarantine))) {
+		ret = EFAULT;
+		goto label_return;
+	}
+
+	arena_ind = (unsigned)mib[1];
+	if (config_debug) {
+		malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx);
+		assert(arena_ind < ctl_stats.narenas);
+		malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx);
+	}
+	assert(arena_ind >= opt_narenas);
+
+	arena = arena_get(tsd_tsdn(tsd), arena_ind, false);
+
+	arena_reset(tsd, arena);
+
+	ret = 0;
+label_return:
+	return (ret);
+}
+
+static int
+arena_i_dss_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
 	const char *dss = NULL;
-	unsigned arena_ind = mib[1];
+	unsigned arena_ind = (unsigned)mib[1];
 	dss_prec_t dss_prec_old = dss_prec_limit;
 	dss_prec_t dss_prec = dss_prec_limit;
 
-	malloc_mutex_lock(&ctl_mtx);
+	malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx);
 	WRITE(dss, const char *);
 	if (dss != NULL) {
 		int i;
@@ -1615,13 +1676,13 @@ arena_i_dss_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
 	}
 
 	if (arena_ind < ctl_stats.narenas) {
-		arena_t *arena = arena_get(tsd_fetch(), arena_ind, false, true);
+		arena_t *arena = arena_get(tsd_tsdn(tsd), arena_ind, false);
 		if (arena == NULL || (dss_prec != dss_prec_limit &&
-		    arena_dss_prec_set(arena, dss_prec))) {
+		    arena_dss_prec_set(tsd_tsdn(tsd), arena, dss_prec))) {
 			ret = EFAULT;
 			goto label_return;
 		}
-		dss_prec_old = arena_dss_prec_get(arena);
+		dss_prec_old = arena_dss_prec_get(tsd_tsdn(tsd), arena);
 	} else {
 		if (dss_prec != dss_prec_limit &&
 		    chunk_dss_prec_set(dss_prec)) {
@@ -1636,26 +1697,26 @@ arena_i_dss_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
 
 	ret = 0;
 label_return:
-	malloc_mutex_unlock(&ctl_mtx);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx);
 	return (ret);
 }
 
 static int
-arena_i_lg_dirty_mult_ctl(const size_t *mib, size_t miblen, void *oldp,
-    size_t *oldlenp, void *newp, size_t newlen)
+arena_i_lg_dirty_mult_ctl(tsd_t *tsd, const size_t *mib, size_t miblen,
+    void *oldp, size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
-	unsigned arena_ind = mib[1];
+	unsigned arena_ind = (unsigned)mib[1];
 	arena_t *arena;
 
-	arena = arena_get(tsd_fetch(), arena_ind, false, true);
+	arena = arena_get(tsd_tsdn(tsd), arena_ind, false);
 	if (arena == NULL) {
 		ret = EFAULT;
 		goto label_return;
 	}
 
 	if (oldp != NULL && oldlenp != NULL) {
-		size_t oldval = arena_lg_dirty_mult_get(arena);
+		size_t oldval = arena_lg_dirty_mult_get(tsd_tsdn(tsd), arena);
 		READ(oldval, ssize_t);
 	}
 	if (newp != NULL) {
@@ -1663,7 +1724,8 @@ arena_i_lg_dirty_mult_ctl(const size_t *mib, size_t miblen, void *oldp,
 			ret = EINVAL;
 			goto label_return;
 		}
-		if (arena_lg_dirty_mult_set(arena, *(ssize_t *)newp)) {
+		if (arena_lg_dirty_mult_set(tsd_tsdn(tsd), arena,
+		    *(ssize_t *)newp)) {
 			ret = EFAULT;
 			goto label_return;
 		}
@@ -1675,24 +1737,60 @@ label_return:
 }
 
 static int
-arena_i_chunk_hooks_ctl(const size_t *mib, size_t miblen, void *oldp,
+arena_i_decay_time_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
     size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
-	unsigned arena_ind = mib[1];
+	unsigned arena_ind = (unsigned)mib[1];
 	arena_t *arena;
 
-	malloc_mutex_lock(&ctl_mtx);
+	arena = arena_get(tsd_tsdn(tsd), arena_ind, false);
+	if (arena == NULL) {
+		ret = EFAULT;
+		goto label_return;
+	}
+
+	if (oldp != NULL && oldlenp != NULL) {
+		size_t oldval = arena_decay_time_get(tsd_tsdn(tsd), arena);
+		READ(oldval, ssize_t);
+	}
+	if (newp != NULL) {
+		if (newlen != sizeof(ssize_t)) {
+			ret = EINVAL;
+			goto label_return;
+		}
+		if (arena_decay_time_set(tsd_tsdn(tsd), arena,
+		    *(ssize_t *)newp)) {
+			ret = EFAULT;
+			goto label_return;
+		}
+	}
+
+	ret = 0;
+label_return:
+	return (ret);
+}
+
+static int
+arena_i_chunk_hooks_ctl(tsd_t *tsd, const size_t *mib, size_t miblen,
+    void *oldp, size_t *oldlenp, void *newp, size_t newlen)
+{
+	int ret;
+	unsigned arena_ind = (unsigned)mib[1];
+	arena_t *arena;
+
+	malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx);
 	if (arena_ind < narenas_total_get() && (arena =
-	    arena_get(tsd_fetch(), arena_ind, false, true)) != NULL) {
+	    arena_get(tsd_tsdn(tsd), arena_ind, false)) != NULL) {
 		if (newp != NULL) {
 			chunk_hooks_t old_chunk_hooks, new_chunk_hooks;
 			WRITE(new_chunk_hooks, chunk_hooks_t);
-			old_chunk_hooks = chunk_hooks_set(arena,
+			old_chunk_hooks = chunk_hooks_set(tsd_tsdn(tsd), arena,
 			    &new_chunk_hooks);
 			READ(old_chunk_hooks, chunk_hooks_t);
 		} else {
-			chunk_hooks_t old_chunk_hooks = chunk_hooks_get(arena);
+			chunk_hooks_t old_chunk_hooks =
+			    chunk_hooks_get(tsd_tsdn(tsd), arena);
 			READ(old_chunk_hooks, chunk_hooks_t);
 		}
 	} else {
@@ -1701,16 +1799,16 @@ arena_i_chunk_hooks_ctl(const size_t *mib, size_t miblen, void *oldp,
 	}
 	ret = 0;
 label_return:
-	malloc_mutex_unlock(&ctl_mtx);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx);
 	return (ret);
 }
 
 static const ctl_named_node_t *
-arena_i_index(const size_t *mib, size_t miblen, size_t i)
+arena_i_index(tsdn_t *tsdn, const size_t *mib, size_t miblen, size_t i)
 {
-	const ctl_named_node_t * ret;
+	const ctl_named_node_t *ret;
 
-	malloc_mutex_lock(&ctl_mtx);
+	malloc_mutex_lock(tsdn, &ctl_mtx);
 	if (i > ctl_stats.narenas) {
 		ret = NULL;
 		goto label_return;
@@ -1718,20 +1816,20 @@ arena_i_index(const size_t *mib, size_t miblen, size_t i)
 
 	ret = super_arena_i_node;
 label_return:
-	malloc_mutex_unlock(&ctl_mtx);
+	malloc_mutex_unlock(tsdn, &ctl_mtx);
 	return (ret);
 }
 
 /******************************************************************************/
 
 static int
-arenas_narenas_ctl(const size_t *mib, size_t miblen, void *oldp,
+arenas_narenas_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
     size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
 	unsigned narenas;
 
-	malloc_mutex_lock(&ctl_mtx);
+	malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx);
 	READONLY();
 	if (*oldlenp != sizeof(unsigned)) {
 		ret = EINVAL;
@@ -1742,23 +1840,23 @@ arenas_narenas_ctl(const size_t *mib, size_t miblen, void *oldp,
 
 	ret = 0;
 label_return:
-	malloc_mutex_unlock(&ctl_mtx);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx);
 	return (ret);
 }
 
 static int
-arenas_initialized_ctl(const size_t *mib, size_t miblen, void *oldp,
+arenas_initialized_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
     size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
 	unsigned nread, i;
 
-	malloc_mutex_lock(&ctl_mtx);
+	malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx);
 	READONLY();
 	if (*oldlenp != ctl_stats.narenas * sizeof(bool)) {
 		ret = EINVAL;
 		nread = (*oldlenp < ctl_stats.narenas * sizeof(bool))
-		    ? (*oldlenp / sizeof(bool)) : ctl_stats.narenas;
+		    ? (unsigned)(*oldlenp / sizeof(bool)) : ctl_stats.narenas;
 	} else {
 		ret = 0;
 		nread = ctl_stats.narenas;
@@ -1768,13 +1866,13 @@ arenas_initialized_ctl(const size_t *mib, size_t miblen, void *oldp,
 		((bool *)oldp)[i] = ctl_stats.arenas[i].initialized;
 
 label_return:
-	malloc_mutex_unlock(&ctl_mtx);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx);
 	return (ret);
 }
 
 static int
-arenas_lg_dirty_mult_ctl(const size_t *mib, size_t miblen, void *oldp,
-    size_t *oldlenp, void *newp, size_t newlen)
+arenas_lg_dirty_mult_ctl(tsd_t *tsd, const size_t *mib, size_t miblen,
+    void *oldp, size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
 
@@ -1798,6 +1896,32 @@ label_return:
 	return (ret);
 }
 
+static int
+arenas_decay_time_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen)
+{
+	int ret;
+
+	if (oldp != NULL && oldlenp != NULL) {
+		size_t oldval = arena_decay_time_default_get();
+		READ(oldval, ssize_t);
+	}
+	if (newp != NULL) {
+		if (newlen != sizeof(ssize_t)) {
+			ret = EINVAL;
+			goto label_return;
+		}
+		if (arena_decay_time_default_set(*(ssize_t *)newp)) {
+			ret = EFAULT;
+			goto label_return;
+		}
+	}
+
+	ret = 0;
+label_return:
+	return (ret);
+}
+
 CTL_RO_NL_GEN(arenas_quantum, QUANTUM, size_t)
 CTL_RO_NL_GEN(arenas_page, PAGE, size_t)
 CTL_RO_NL_CGEN(config_tcache, arenas_tcache_max, tcache_maxclass, size_t)
@@ -1807,7 +1931,7 @@ CTL_RO_NL_GEN(arenas_bin_i_size, arena_bin_info[mib[2]].reg_size, size_t)
 CTL_RO_NL_GEN(arenas_bin_i_nregs, arena_bin_info[mib[2]].nregs, uint32_t)
 CTL_RO_NL_GEN(arenas_bin_i_run_size, arena_bin_info[mib[2]].run_size, size_t)
 static const ctl_named_node_t *
-arenas_bin_i_index(const size_t *mib, size_t miblen, size_t i)
+arenas_bin_i_index(tsdn_t *tsdn, const size_t *mib, size_t miblen, size_t i)
 {
 
 	if (i > NBINS)
@@ -1816,9 +1940,9 @@ arenas_bin_i_index(const size_t *mib, size_t miblen, size_t i)
 }
 
 CTL_RO_NL_GEN(arenas_nlruns, nlclasses, unsigned)
-CTL_RO_NL_GEN(arenas_lrun_i_size, index2size(NBINS+mib[2]), size_t)
+CTL_RO_NL_GEN(arenas_lrun_i_size, index2size(NBINS+(szind_t)mib[2]), size_t)
 static const ctl_named_node_t *
-arenas_lrun_i_index(const size_t *mib, size_t miblen, size_t i)
+arenas_lrun_i_index(tsdn_t *tsdn, const size_t *mib, size_t miblen, size_t i)
 {
 
 	if (i > nlclasses)
@@ -1827,9 +1951,10 @@ arenas_lrun_i_index(const size_t *mib, size_t miblen, size_t i)
 }
 
 CTL_RO_NL_GEN(arenas_nhchunks, nhclasses, unsigned)
-CTL_RO_NL_GEN(arenas_hchunk_i_size, index2size(NBINS+nlclasses+mib[2]), size_t)
+CTL_RO_NL_GEN(arenas_hchunk_i_size, index2size(NBINS+nlclasses+(szind_t)mib[2]),
+    size_t)
 static const ctl_named_node_t *
-arenas_hchunk_i_index(const size_t *mib, size_t miblen, size_t i)
+arenas_hchunk_i_index(tsdn_t *tsdn, const size_t *mib, size_t miblen, size_t i)
 {
 
 	if (i > nhclasses)
@@ -1838,15 +1963,15 @@ arenas_hchunk_i_index(const size_t *mib, size_t miblen, size_t i)
 }
 
 static int
-arenas_extend_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
-    void *newp, size_t newlen)
+arenas_extend_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
 	unsigned narenas;
 
-	malloc_mutex_lock(&ctl_mtx);
+	malloc_mutex_lock(tsd_tsdn(tsd), &ctl_mtx);
 	READONLY();
-	if (ctl_grow()) {
+	if (ctl_grow(tsd_tsdn(tsd))) {
 		ret = EAGAIN;
 		goto label_return;
 	}
@@ -1855,15 +1980,15 @@ arenas_extend_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
 
 	ret = 0;
 label_return:
-	malloc_mutex_unlock(&ctl_mtx);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &ctl_mtx);
 	return (ret);
 }
 
 /******************************************************************************/
 
 static int
-prof_thread_active_init_ctl(const size_t *mib, size_t miblen, void *oldp,
-    size_t *oldlenp, void *newp, size_t newlen)
+prof_thread_active_init_ctl(tsd_t *tsd, const size_t *mib, size_t miblen,
+    void *oldp, size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
 	bool oldval;
@@ -1876,9 +2001,10 @@ prof_thread_active_init_ctl(const size_t *mib, size_t miblen, void *oldp,
 			ret = EINVAL;
 			goto label_return;
 		}
-		oldval = prof_thread_active_init_set(*(bool *)newp);
+		oldval = prof_thread_active_init_set(tsd_tsdn(tsd),
+		    *(bool *)newp);
 	} else
-		oldval = prof_thread_active_init_get();
+		oldval = prof_thread_active_init_get(tsd_tsdn(tsd));
 	READ(oldval, bool);
 
 	ret = 0;
@@ -1887,8 +2013,8 @@ label_return:
 }
 
 static int
-prof_active_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
-    void *newp, size_t newlen)
+prof_active_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
 	bool oldval;
@@ -1901,9 +2027,9 @@ prof_active_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
 			ret = EINVAL;
 			goto label_return;
 		}
-		oldval = prof_active_set(*(bool *)newp);
+		oldval = prof_active_set(tsd_tsdn(tsd), *(bool *)newp);
 	} else
-		oldval = prof_active_get();
+		oldval = prof_active_get(tsd_tsdn(tsd));
 	READ(oldval, bool);
 
 	ret = 0;
@@ -1912,8 +2038,8 @@ label_return:
 }
 
 static int
-prof_dump_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
-    void *newp, size_t newlen)
+prof_dump_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
 	const char *filename = NULL;
@@ -1924,7 +2050,7 @@ prof_dump_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
 	WRITEONLY();
 	WRITE(filename, const char *);
 
-	if (prof_mdump(filename)) {
+	if (prof_mdump(tsd, filename)) {
 		ret = EFAULT;
 		goto label_return;
 	}
@@ -1935,8 +2061,8 @@ label_return:
 }
 
 static int
-prof_gdump_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
-    void *newp, size_t newlen)
+prof_gdump_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
 	bool oldval;
@@ -1949,9 +2075,9 @@ prof_gdump_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
 			ret = EINVAL;
 			goto label_return;
 		}
-		oldval = prof_gdump_set(*(bool *)newp);
+		oldval = prof_gdump_set(tsd_tsdn(tsd), *(bool *)newp);
 	} else
-		oldval = prof_gdump_get();
+		oldval = prof_gdump_get(tsd_tsdn(tsd));
 	READ(oldval, bool);
 
 	ret = 0;
@@ -1960,12 +2086,11 @@ label_return:
 }
 
 static int
-prof_reset_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
-    void *newp, size_t newlen)
+prof_reset_ctl(tsd_t *tsd, const size_t *mib, size_t miblen, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen)
 {
 	int ret;
 	size_t lg_sample = lg_prof_sample;
-	tsd_t *tsd;
 
 	if (!config_prof)
 		return (ENOENT);
@@ -1975,8 +2100,6 @@ prof_reset_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
 	if (lg_sample >= (sizeof(uint64_t) << 3))
 		lg_sample = (sizeof(uint64_t) << 3) - 1;
 
-	tsd = tsd_fetch();
-
 	prof_reset(tsd, lg_sample);
 
 	ret = 0;
@@ -1995,15 +2118,20 @@ CTL_RO_CGEN(config_stats, stats_active, ctl_stats.active, size_t)
 CTL_RO_CGEN(config_stats, stats_metadata, ctl_stats.metadata, size_t)
 CTL_RO_CGEN(config_stats, stats_resident, ctl_stats.resident, size_t)
 CTL_RO_CGEN(config_stats, stats_mapped, ctl_stats.mapped, size_t)
+CTL_RO_CGEN(config_stats, stats_retained, ctl_stats.retained, size_t)
 
 CTL_RO_GEN(stats_arenas_i_dss, ctl_stats.arenas[mib[2]].dss, const char *)
 CTL_RO_GEN(stats_arenas_i_lg_dirty_mult, ctl_stats.arenas[mib[2]].lg_dirty_mult,
     ssize_t)
+CTL_RO_GEN(stats_arenas_i_decay_time, ctl_stats.arenas[mib[2]].decay_time,
+    ssize_t)
 CTL_RO_GEN(stats_arenas_i_nthreads, ctl_stats.arenas[mib[2]].nthreads, unsigned)
 CTL_RO_GEN(stats_arenas_i_pactive, ctl_stats.arenas[mib[2]].pactive, size_t)
 CTL_RO_GEN(stats_arenas_i_pdirty, ctl_stats.arenas[mib[2]].pdirty, size_t)
 CTL_RO_CGEN(config_stats, stats_arenas_i_mapped,
     ctl_stats.arenas[mib[2]].astats.mapped, size_t)
+CTL_RO_CGEN(config_stats, stats_arenas_i_retained,
+    ctl_stats.arenas[mib[2]].astats.retained, size_t)
 CTL_RO_CGEN(config_stats, stats_arenas_i_npurge,
     ctl_stats.arenas[mib[2]].astats.npurge, uint64_t)
 CTL_RO_CGEN(config_stats, stats_arenas_i_nmadvise,
@@ -2060,7 +2188,8 @@ CTL_RO_CGEN(config_stats, stats_arenas_i_bins_j_curruns,
     ctl_stats.arenas[mib[2]].bstats[mib[4]].curruns, size_t)
 
 static const ctl_named_node_t *
-stats_arenas_i_bins_j_index(const size_t *mib, size_t miblen, size_t j)
+stats_arenas_i_bins_j_index(tsdn_t *tsdn, const size_t *mib, size_t miblen,
+    size_t j)
 {
 
 	if (j > NBINS)
@@ -2078,7 +2207,8 @@ CTL_RO_CGEN(config_stats, stats_arenas_i_lruns_j_curruns,
     ctl_stats.arenas[mib[2]].lstats[mib[4]].curruns, size_t)
 
 static const ctl_named_node_t *
-stats_arenas_i_lruns_j_index(const size_t *mib, size_t miblen, size_t j)
+stats_arenas_i_lruns_j_index(tsdn_t *tsdn, const size_t *mib, size_t miblen,
+    size_t j)
 {
 
 	if (j > nlclasses)
@@ -2097,7 +2227,8 @@ CTL_RO_CGEN(config_stats, stats_arenas_i_hchunks_j_curhchunks,
     ctl_stats.arenas[mib[2]].hstats[mib[4]].curhchunks, size_t)
 
 static const ctl_named_node_t *
-stats_arenas_i_hchunks_j_index(const size_t *mib, size_t miblen, size_t j)
+stats_arenas_i_hchunks_j_index(tsdn_t *tsdn, const size_t *mib, size_t miblen,
+    size_t j)
 {
 
 	if (j > nhclasses)
@@ -2106,11 +2237,11 @@ stats_arenas_i_hchunks_j_index(const size_t *mib, size_t miblen, size_t j)
 }
 
 static const ctl_named_node_t *
-stats_arenas_i_index(const size_t *mib, size_t miblen, size_t i)
+stats_arenas_i_index(tsdn_t *tsdn, const size_t *mib, size_t miblen, size_t i)
 {
 	const ctl_named_node_t * ret;
 
-	malloc_mutex_lock(&ctl_mtx);
+	malloc_mutex_lock(tsdn, &ctl_mtx);
 	if (i > ctl_stats.narenas || !ctl_stats.arenas[i].initialized) {
 		ret = NULL;
 		goto label_return;
@@ -2118,6 +2249,6 @@ stats_arenas_i_index(const size_t *mib, size_t miblen, size_t i)
 
 	ret = super_stats_arenas_i_node;
 label_return:
-	malloc_mutex_unlock(&ctl_mtx);
+	malloc_mutex_unlock(tsdn, &ctl_mtx);
 	return (ret);
 }
diff --git a/deps/jemalloc/src/extent.c b/deps/jemalloc/src/extent.c
index 13f94411c..218156c60 100644
--- a/deps/jemalloc/src/extent.c
+++ b/deps/jemalloc/src/extent.c
@@ -3,45 +3,48 @@
 
 /******************************************************************************/
 
+/*
+ * Round down to the nearest chunk size that can actually be requested during
+ * normal huge allocation.
+ */
 JEMALLOC_INLINE_C size_t
 extent_quantize(size_t size)
 {
+	size_t ret;
+	szind_t ind;
 
-	/*
-	 * Round down to the nearest chunk size that can actually be requested
-	 * during normal huge allocation.
-	 */
-	return (index2size(size2index(size + 1) - 1));
+	assert(size > 0);
+
+	ind = size2index(size + 1);
+	if (ind == 0) {
+		/* Avoid underflow. */
+		return (index2size(0));
+	}
+	ret = index2size(ind - 1);
+	assert(ret <= size);
+	return (ret);
 }
 
 JEMALLOC_INLINE_C int
-extent_szad_comp(extent_node_t *a, extent_node_t *b)
+extent_sz_comp(const extent_node_t *a, const extent_node_t *b)
 {
-	int ret;
 	size_t a_qsize = extent_quantize(extent_node_size_get(a));
 	size_t b_qsize = extent_quantize(extent_node_size_get(b));
 
-	/*
-	 * Compare based on quantized size rather than size, in order to sort
-	 * equally useful extents only by address.
-	 */
-	ret = (a_qsize > b_qsize) - (a_qsize < b_qsize);
-	if (ret == 0) {
-		uintptr_t a_addr = (uintptr_t)extent_node_addr_get(a);
-		uintptr_t b_addr = (uintptr_t)extent_node_addr_get(b);
+	return ((a_qsize > b_qsize) - (a_qsize < b_qsize));
+}
 
-		ret = (a_addr > b_addr) - (a_addr < b_addr);
-	}
+JEMALLOC_INLINE_C int
+extent_sn_comp(const extent_node_t *a, const extent_node_t *b)
+{
+	size_t a_sn = extent_node_sn_get(a);
+	size_t b_sn = extent_node_sn_get(b);
 
-	return (ret);
+	return ((a_sn > b_sn) - (a_sn < b_sn));
 }
 
-/* Generate red-black tree functions. */
-rb_gen(, extent_tree_szad_, extent_tree_t, extent_node_t, szad_link,
-    extent_szad_comp)
-
 JEMALLOC_INLINE_C int
-extent_ad_comp(extent_node_t *a, extent_node_t *b)
+extent_ad_comp(const extent_node_t *a, const extent_node_t *b)
 {
 	uintptr_t a_addr = (uintptr_t)extent_node_addr_get(a);
 	uintptr_t b_addr = (uintptr_t)extent_node_addr_get(b);
@@ -49,5 +52,26 @@ extent_ad_comp(extent_node_t *a, extent_node_t *b)
 	return ((a_addr > b_addr) - (a_addr < b_addr));
 }
 
+JEMALLOC_INLINE_C int
+extent_szsnad_comp(const extent_node_t *a, const extent_node_t *b)
+{
+	int ret;
+
+	ret = extent_sz_comp(a, b);
+	if (ret != 0)
+		return (ret);
+
+	ret = extent_sn_comp(a, b);
+	if (ret != 0)
+		return (ret);
+
+	ret = extent_ad_comp(a, b);
+	return (ret);
+}
+
+/* Generate red-black tree functions. */
+rb_gen(, extent_tree_szsnad_, extent_tree_t, extent_node_t, szsnad_link,
+    extent_szsnad_comp)
+
 /* Generate red-black tree functions. */
 rb_gen(, extent_tree_ad_, extent_tree_t, extent_node_t, ad_link, extent_ad_comp)
diff --git a/deps/jemalloc/src/huge.c b/deps/jemalloc/src/huge.c
index 1e9a66512..8abd8c00c 100644
--- a/deps/jemalloc/src/huge.c
+++ b/deps/jemalloc/src/huge.c
@@ -15,12 +15,21 @@ huge_node_get(const void *ptr)
 }
 
 static bool
-huge_node_set(const void *ptr, extent_node_t *node)
+huge_node_set(tsdn_t *tsdn, const void *ptr, extent_node_t *node)
 {
 
 	assert(extent_node_addr_get(node) == ptr);
 	assert(!extent_node_achunk_get(node));
-	return (chunk_register(ptr, node));
+	return (chunk_register(tsdn, ptr, node));
+}
+
+static void
+huge_node_reset(tsdn_t *tsdn, const void *ptr, extent_node_t *node)
+{
+	bool err;
+
+	err = huge_node_set(tsdn, ptr, node);
+	assert(!err);
 }
 
 static void
@@ -31,39 +40,39 @@ huge_node_unset(const void *ptr, const extent_node_t *node)
 }
 
 void *
-huge_malloc(tsd_t *tsd, arena_t *arena, size_t size, bool zero,
-    tcache_t *tcache)
+huge_malloc(tsdn_t *tsdn, arena_t *arena, size_t usize, bool zero)
 {
-	size_t usize;
 
-	usize = s2u(size);
-	if (usize == 0) {
-		/* size_t overflow. */
-		return (NULL);
-	}
+	assert(usize == s2u(usize));
 
-	return (huge_palloc(tsd, arena, usize, chunksize, zero, tcache));
+	return (huge_palloc(tsdn, arena, usize, chunksize, zero));
 }
 
 void *
-huge_palloc(tsd_t *tsd, arena_t *arena, size_t size, size_t alignment,
-    bool zero, tcache_t *tcache)
+huge_palloc(tsdn_t *tsdn, arena_t *arena, size_t usize, size_t alignment,
+    bool zero)
 {
 	void *ret;
-	size_t usize;
+	size_t ausize;
+	arena_t *iarena;
 	extent_node_t *node;
+	size_t sn;
 	bool is_zeroed;
 
 	/* Allocate one or more contiguous chunks for this request. */
 
-	usize = sa2u(size, alignment);
-	if (unlikely(usize == 0))
+	assert(!tsdn_null(tsdn) || arena != NULL);
+
+	ausize = sa2u(usize, alignment);
+	if (unlikely(ausize == 0 || ausize > HUGE_MAXCLASS))
 		return (NULL);
-	assert(usize >= chunksize);
+	assert(ausize >= chunksize);
 
 	/* Allocate an extent node with which to track the chunk. */
-	node = ipallocztm(tsd, CACHELINE_CEILING(sizeof(extent_node_t)),
-	    CACHELINE, false, tcache, true, arena);
+	iarena = (!tsdn_null(tsdn)) ? arena_ichoose(tsdn_tsd(tsdn), NULL) :
+	    a0get();
+	node = ipallocztm(tsdn, CACHELINE_CEILING(sizeof(extent_node_t)),
+	    CACHELINE, false, NULL, true, iarena);
 	if (node == NULL)
 		return (NULL);
 
@@ -72,33 +81,35 @@ huge_palloc(tsd_t *tsd, arena_t *arena, size_t size, size_t alignment,
 	 * it is possible to make correct junk/zero fill decisions below.
 	 */
 	is_zeroed = zero;
-	arena = arena_choose(tsd, arena);
-	if (unlikely(arena == NULL) || (ret = arena_chunk_alloc_huge(arena,
-	    size, alignment, &is_zeroed)) == NULL) {
-		idalloctm(tsd, node, tcache, true);
+	if (likely(!tsdn_null(tsdn)))
+		arena = arena_choose(tsdn_tsd(tsdn), arena);
+	if (unlikely(arena == NULL) || (ret = arena_chunk_alloc_huge(tsdn,
+	    arena, usize, alignment, &sn, &is_zeroed)) == NULL) {
+		idalloctm(tsdn, node, NULL, true, true);
 		return (NULL);
 	}
 
-	extent_node_init(node, arena, ret, size, is_zeroed, true);
+	extent_node_init(node, arena, ret, usize, sn, is_zeroed, true);
 
-	if (huge_node_set(ret, node)) {
-		arena_chunk_dalloc_huge(arena, ret, size);
-		idalloctm(tsd, node, tcache, true);
+	if (huge_node_set(tsdn, ret, node)) {
+		arena_chunk_dalloc_huge(tsdn, arena, ret, usize, sn);
+		idalloctm(tsdn, node, NULL, true, true);
 		return (NULL);
 	}
 
 	/* Insert node into huge. */
-	malloc_mutex_lock(&arena->huge_mtx);
+	malloc_mutex_lock(tsdn, &arena->huge_mtx);
 	ql_elm_new(node, ql_link);
 	ql_tail_insert(&arena->huge, node, ql_link);
-	malloc_mutex_unlock(&arena->huge_mtx);
+	malloc_mutex_unlock(tsdn, &arena->huge_mtx);
 
 	if (zero || (config_fill && unlikely(opt_zero))) {
 		if (!is_zeroed)
-			memset(ret, 0, size);
+			memset(ret, 0, usize);
 	} else if (config_fill && unlikely(opt_junk_alloc))
-		memset(ret, 0xa5, size);
+		memset(ret, JEMALLOC_ALLOC_JUNK, usize);
 
+	arena_decay_tick(tsdn, arena);
 	return (ret);
 }
 
@@ -116,7 +127,7 @@ huge_dalloc_junk(void *ptr, size_t usize)
 		 * unmapped.
 		 */
 		if (!config_munmap || (have_dss && chunk_in_dss(ptr)))
-			memset(ptr, 0x5a, usize);
+			memset(ptr, JEMALLOC_FREE_JUNK, usize);
 	}
 }
 #ifdef JEMALLOC_JET
@@ -126,8 +137,8 @@ huge_dalloc_junk_t *huge_dalloc_junk = JEMALLOC_N(huge_dalloc_junk_impl);
 #endif
 
 static void
-huge_ralloc_no_move_similar(void *ptr, size_t oldsize, size_t usize_min,
-    size_t usize_max, bool zero)
+huge_ralloc_no_move_similar(tsdn_t *tsdn, void *ptr, size_t oldsize,
+    size_t usize_min, size_t usize_max, bool zero)
 {
 	size_t usize, usize_next;
 	extent_node_t *node;
@@ -151,24 +162,28 @@ huge_ralloc_no_move_similar(void *ptr, size_t oldsize, size_t usize_min,
 	if (oldsize > usize) {
 		size_t sdiff = oldsize - usize;
 		if (config_fill && unlikely(opt_junk_free)) {
-			memset((void *)((uintptr_t)ptr + usize), 0x5a, sdiff);
+			memset((void *)((uintptr_t)ptr + usize),
+			    JEMALLOC_FREE_JUNK, sdiff);
 			post_zeroed = false;
 		} else {
-			post_zeroed = !chunk_purge_wrapper(arena, &chunk_hooks,
-			    ptr, CHUNK_CEILING(oldsize), usize, sdiff);
+			post_zeroed = !chunk_purge_wrapper(tsdn, arena,
+			    &chunk_hooks, ptr, CHUNK_CEILING(oldsize), usize,
+			    sdiff);
 		}
 	} else
 		post_zeroed = pre_zeroed;
 
-	malloc_mutex_lock(&arena->huge_mtx);
+	malloc_mutex_lock(tsdn, &arena->huge_mtx);
 	/* Update the size of the huge allocation. */
+	huge_node_unset(ptr, node);
 	assert(extent_node_size_get(node) != usize);
 	extent_node_size_set(node, usize);
+	huge_node_reset(tsdn, ptr, node);
 	/* Update zeroed. */
 	extent_node_zeroed_set(node, post_zeroed);
-	malloc_mutex_unlock(&arena->huge_mtx);
+	malloc_mutex_unlock(tsdn, &arena->huge_mtx);
 
-	arena_chunk_ralloc_huge_similar(arena, ptr, oldsize, usize);
+	arena_chunk_ralloc_huge_similar(tsdn, arena, ptr, oldsize, usize);
 
 	/* Fill if necessary (growing). */
 	if (oldsize < usize) {
@@ -178,14 +193,15 @@ huge_ralloc_no_move_similar(void *ptr, size_t oldsize, size_t usize_min,
 				    usize - oldsize);
 			}
 		} else if (config_fill && unlikely(opt_junk_alloc)) {
-			memset((void *)((uintptr_t)ptr + oldsize), 0xa5, usize -
-			    oldsize);
+			memset((void *)((uintptr_t)ptr + oldsize),
+			    JEMALLOC_ALLOC_JUNK, usize - oldsize);
 		}
 	}
 }
 
 static bool
-huge_ralloc_no_move_shrink(void *ptr, size_t oldsize, size_t usize)
+huge_ralloc_no_move_shrink(tsdn_t *tsdn, void *ptr, size_t oldsize,
+    size_t usize)
 {
 	extent_node_t *node;
 	arena_t *arena;
@@ -196,7 +212,7 @@ huge_ralloc_no_move_shrink(void *ptr, size_t oldsize, size_t usize)
 	node = huge_node_get(ptr);
 	arena = extent_node_arena_get(node);
 	pre_zeroed = extent_node_zeroed_get(node);
-	chunk_hooks = chunk_hooks_get(arena);
+	chunk_hooks = chunk_hooks_get(tsdn, arena);
 
 	assert(oldsize > usize);
 
@@ -213,53 +229,59 @@ huge_ralloc_no_move_shrink(void *ptr, size_t oldsize, size_t usize)
 			    sdiff);
 			post_zeroed = false;
 		} else {
-			post_zeroed = !chunk_purge_wrapper(arena, &chunk_hooks,
-			    CHUNK_ADDR2BASE((uintptr_t)ptr + usize),
-			    CHUNK_CEILING(oldsize),
+			post_zeroed = !chunk_purge_wrapper(tsdn, arena,
+			    &chunk_hooks, CHUNK_ADDR2BASE((uintptr_t)ptr +
+			    usize), CHUNK_CEILING(oldsize),
 			    CHUNK_ADDR2OFFSET((uintptr_t)ptr + usize), sdiff);
 		}
 	} else
 		post_zeroed = pre_zeroed;
 
-	malloc_mutex_lock(&arena->huge_mtx);
+	malloc_mutex_lock(tsdn, &arena->huge_mtx);
 	/* Update the size of the huge allocation. */
+	huge_node_unset(ptr, node);
 	extent_node_size_set(node, usize);
+	huge_node_reset(tsdn, ptr, node);
 	/* Update zeroed. */
 	extent_node_zeroed_set(node, post_zeroed);
-	malloc_mutex_unlock(&arena->huge_mtx);
+	malloc_mutex_unlock(tsdn, &arena->huge_mtx);
 
 	/* Zap the excess chunks. */
-	arena_chunk_ralloc_huge_shrink(arena, ptr, oldsize, usize);
+	arena_chunk_ralloc_huge_shrink(tsdn, arena, ptr, oldsize, usize,
+	    extent_node_sn_get(node));
 
 	return (false);
 }
 
 static bool
-huge_ralloc_no_move_expand(void *ptr, size_t oldsize, size_t usize, bool zero) {
+huge_ralloc_no_move_expand(tsdn_t *tsdn, void *ptr, size_t oldsize,
+    size_t usize, bool zero) {
 	extent_node_t *node;
 	arena_t *arena;
 	bool is_zeroed_subchunk, is_zeroed_chunk;
 
 	node = huge_node_get(ptr);
 	arena = extent_node_arena_get(node);
-	malloc_mutex_lock(&arena->huge_mtx);
+	malloc_mutex_lock(tsdn, &arena->huge_mtx);
 	is_zeroed_subchunk = extent_node_zeroed_get(node);
-	malloc_mutex_unlock(&arena->huge_mtx);
+	malloc_mutex_unlock(tsdn, &arena->huge_mtx);
 
 	/*
-	 * Copy zero into is_zeroed_chunk and pass the copy to chunk_alloc(), so
-	 * that it is possible to make correct junk/zero fill decisions below.
+	 * Use is_zeroed_chunk to detect whether the trailing memory is zeroed,
+	 * update extent's zeroed field, and zero as necessary.
 	 */
-	is_zeroed_chunk = zero;
-
-	if (arena_chunk_ralloc_huge_expand(arena, ptr, oldsize, usize,
+	is_zeroed_chunk = false;
+	if (arena_chunk_ralloc_huge_expand(tsdn, arena, ptr, oldsize, usize,
 	     &is_zeroed_chunk))
 		return (true);
 
-	malloc_mutex_lock(&arena->huge_mtx);
-	/* Update the size of the huge allocation. */
+	malloc_mutex_lock(tsdn, &arena->huge_mtx);
+	huge_node_unset(ptr, node);
 	extent_node_size_set(node, usize);
-	malloc_mutex_unlock(&arena->huge_mtx);
+	extent_node_zeroed_set(node, extent_node_zeroed_get(node) &&
+	    is_zeroed_chunk);
+	huge_node_reset(tsdn, ptr, node);
+	malloc_mutex_unlock(tsdn, &arena->huge_mtx);
 
 	if (zero || (config_fill && unlikely(opt_zero))) {
 		if (!is_zeroed_subchunk) {
@@ -272,19 +294,21 @@ huge_ralloc_no_move_expand(void *ptr, size_t oldsize, size_t usize, bool zero) {
 			    CHUNK_CEILING(oldsize));
 		}
 	} else if (config_fill && unlikely(opt_junk_alloc)) {
-		memset((void *)((uintptr_t)ptr + oldsize), 0xa5, usize -
-		    oldsize);
+		memset((void *)((uintptr_t)ptr + oldsize), JEMALLOC_ALLOC_JUNK,
+		    usize - oldsize);
 	}
 
 	return (false);
 }
 
 bool
-huge_ralloc_no_move(void *ptr, size_t oldsize, size_t usize_min,
+huge_ralloc_no_move(tsdn_t *tsdn, void *ptr, size_t oldsize, size_t usize_min,
     size_t usize_max, bool zero)
 {
 
 	assert(s2u(oldsize) == oldsize);
+	/* The following should have been caught by callers. */
+	assert(usize_min > 0 && usize_max <= HUGE_MAXCLASS);
 
 	/* Both allocations must be huge to avoid a move. */
 	if (oldsize < chunksize || usize_max < chunksize)
@@ -292,13 +316,18 @@ huge_ralloc_no_move(void *ptr, size_t oldsize, size_t usize_min,
 
 	if (CHUNK_CEILING(usize_max) > CHUNK_CEILING(oldsize)) {
 		/* Attempt to expand the allocation in-place. */
-		if (!huge_ralloc_no_move_expand(ptr, oldsize, usize_max, zero))
+		if (!huge_ralloc_no_move_expand(tsdn, ptr, oldsize, usize_max,
+		    zero)) {
+			arena_decay_tick(tsdn, huge_aalloc(ptr));
 			return (false);
+		}
 		/* Try again, this time with usize_min. */
 		if (usize_min < usize_max && CHUNK_CEILING(usize_min) >
-		    CHUNK_CEILING(oldsize) && huge_ralloc_no_move_expand(ptr,
-		    oldsize, usize_min, zero))
+		    CHUNK_CEILING(oldsize) && huge_ralloc_no_move_expand(tsdn,
+		    ptr, oldsize, usize_min, zero)) {
+			arena_decay_tick(tsdn, huge_aalloc(ptr));
 			return (false);
+		}
 	}
 
 	/*
@@ -307,36 +336,46 @@ huge_ralloc_no_move(void *ptr, size_t oldsize, size_t usize_min,
 	 */
 	if (CHUNK_CEILING(oldsize) >= CHUNK_CEILING(usize_min)
 	    && CHUNK_CEILING(oldsize) <= CHUNK_CEILING(usize_max)) {
-		huge_ralloc_no_move_similar(ptr, oldsize, usize_min, usize_max,
-		    zero);
+		huge_ralloc_no_move_similar(tsdn, ptr, oldsize, usize_min,
+		    usize_max, zero);
+		arena_decay_tick(tsdn, huge_aalloc(ptr));
 		return (false);
 	}
 
 	/* Attempt to shrink the allocation in-place. */
-	if (CHUNK_CEILING(oldsize) > CHUNK_CEILING(usize_max))
-		return (huge_ralloc_no_move_shrink(ptr, oldsize, usize_max));
+	if (CHUNK_CEILING(oldsize) > CHUNK_CEILING(usize_max)) {
+		if (!huge_ralloc_no_move_shrink(tsdn, ptr, oldsize,
+		    usize_max)) {
+			arena_decay_tick(tsdn, huge_aalloc(ptr));
+			return (false);
+		}
+	}
 	return (true);
 }
 
 static void *
-huge_ralloc_move_helper(tsd_t *tsd, arena_t *arena, size_t usize,
-    size_t alignment, bool zero, tcache_t *tcache)
+huge_ralloc_move_helper(tsdn_t *tsdn, arena_t *arena, size_t usize,
+    size_t alignment, bool zero)
 {
 
 	if (alignment <= chunksize)
-		return (huge_malloc(tsd, arena, usize, zero, tcache));
-	return (huge_palloc(tsd, arena, usize, alignment, zero, tcache));
+		return (huge_malloc(tsdn, arena, usize, zero));
+	return (huge_palloc(tsdn, arena, usize, alignment, zero));
 }
 
 void *
-huge_ralloc(tsd_t *tsd, arena_t *arena, void *ptr, size_t oldsize, size_t usize,
-    size_t alignment, bool zero, tcache_t *tcache)
+huge_ralloc(tsd_t *tsd, arena_t *arena, void *ptr, size_t oldsize,
+    size_t usize, size_t alignment, bool zero, tcache_t *tcache)
 {
 	void *ret;
 	size_t copysize;
 
+	/* The following should have been caught by callers. */
+	assert(usize > 0 && usize <= HUGE_MAXCLASS);
+
 	/* Try to avoid moving the allocation. */
-	if (!huge_ralloc_no_move(ptr, oldsize, usize, usize, zero))
+	if (!huge_ralloc_no_move(tsd_tsdn(tsd), ptr, oldsize, usize, usize,
+	    zero))
 		return (ptr);
 
 	/*
@@ -344,19 +383,19 @@ huge_ralloc(tsd_t *tsd, arena_t *arena, void *ptr, size_t oldsize, size_t usize,
 	 * different size class.  In that case, fall back to allocating new
 	 * space and copying.
 	 */
-	ret = huge_ralloc_move_helper(tsd, arena, usize, alignment, zero,
-	    tcache);
+	ret = huge_ralloc_move_helper(tsd_tsdn(tsd), arena, usize, alignment,
+	    zero);
 	if (ret == NULL)
 		return (NULL);
 
 	copysize = (usize < oldsize) ? usize : oldsize;
 	memcpy(ret, ptr, copysize);
-	isqalloc(tsd, ptr, oldsize, tcache);
+	isqalloc(tsd, ptr, oldsize, tcache, true);
 	return (ret);
 }
 
 void
-huge_dalloc(tsd_t *tsd, void *ptr, tcache_t *tcache)
+huge_dalloc(tsdn_t *tsdn, void *ptr)
 {
 	extent_node_t *node;
 	arena_t *arena;
@@ -364,15 +403,18 @@ huge_dalloc(tsd_t *tsd, void *ptr, tcache_t *tcache)
 	node = huge_node_get(ptr);
 	arena = extent_node_arena_get(node);
 	huge_node_unset(ptr, node);
-	malloc_mutex_lock(&arena->huge_mtx);
+	malloc_mutex_lock(tsdn, &arena->huge_mtx);
 	ql_remove(&arena->huge, node, ql_link);
-	malloc_mutex_unlock(&arena->huge_mtx);
+	malloc_mutex_unlock(tsdn, &arena->huge_mtx);
 
 	huge_dalloc_junk(extent_node_addr_get(node),
 	    extent_node_size_get(node));
-	arena_chunk_dalloc_huge(extent_node_arena_get(node),
-	    extent_node_addr_get(node), extent_node_size_get(node));
-	idalloctm(tsd, node, tcache, true);
+	arena_chunk_dalloc_huge(tsdn, extent_node_arena_get(node),
+	    extent_node_addr_get(node), extent_node_size_get(node),
+	    extent_node_sn_get(node));
+	idalloctm(tsdn, node, NULL, true, true);
+
+	arena_decay_tick(tsdn, arena);
 }
 
 arena_t *
@@ -383,7 +425,7 @@ huge_aalloc(const void *ptr)
 }
 
 size_t
-huge_salloc(const void *ptr)
+huge_salloc(tsdn_t *tsdn, const void *ptr)
 {
 	size_t size;
 	extent_node_t *node;
@@ -391,15 +433,15 @@ huge_salloc(const void *ptr)
 
 	node = huge_node_get(ptr);
 	arena = extent_node_arena_get(node);
-	malloc_mutex_lock(&arena->huge_mtx);
+	malloc_mutex_lock(tsdn, &arena->huge_mtx);
 	size = extent_node_size_get(node);
-	malloc_mutex_unlock(&arena->huge_mtx);
+	malloc_mutex_unlock(tsdn, &arena->huge_mtx);
 
 	return (size);
 }
 
 prof_tctx_t *
-huge_prof_tctx_get(const void *ptr)
+huge_prof_tctx_get(tsdn_t *tsdn, const void *ptr)
 {
 	prof_tctx_t *tctx;
 	extent_node_t *node;
@@ -407,29 +449,29 @@ huge_prof_tctx_get(const void *ptr)
 
 	node = huge_node_get(ptr);
 	arena = extent_node_arena_get(node);
-	malloc_mutex_lock(&arena->huge_mtx);
+	malloc_mutex_lock(tsdn, &arena->huge_mtx);
 	tctx = extent_node_prof_tctx_get(node);
-	malloc_mutex_unlock(&arena->huge_mtx);
+	malloc_mutex_unlock(tsdn, &arena->huge_mtx);
 
 	return (tctx);
 }
 
 void
-huge_prof_tctx_set(const void *ptr, prof_tctx_t *tctx)
+huge_prof_tctx_set(tsdn_t *tsdn, const void *ptr, prof_tctx_t *tctx)
 {
 	extent_node_t *node;
 	arena_t *arena;
 
 	node = huge_node_get(ptr);
 	arena = extent_node_arena_get(node);
-	malloc_mutex_lock(&arena->huge_mtx);
+	malloc_mutex_lock(tsdn, &arena->huge_mtx);
 	extent_node_prof_tctx_set(node, tctx);
-	malloc_mutex_unlock(&arena->huge_mtx);
+	malloc_mutex_unlock(tsdn, &arena->huge_mtx);
 }
 
 void
-huge_prof_tctx_reset(const void *ptr)
+huge_prof_tctx_reset(tsdn_t *tsdn, const void *ptr)
 {
 
-	huge_prof_tctx_set(ptr, (prof_tctx_t *)(uintptr_t)1U);
+	huge_prof_tctx_set(tsdn, ptr, (prof_tctx_t *)(uintptr_t)1U);
 }
diff --git a/deps/jemalloc/src/jemalloc.c b/deps/jemalloc/src/jemalloc.c
index fe77c2475..07389ca2f 100644
--- a/deps/jemalloc/src/jemalloc.c
+++ b/deps/jemalloc/src/jemalloc.c
@@ -5,7 +5,11 @@
 /* Data. */
 
 /* Runtime configuration options. */
-const char	*je_malloc_conf JEMALLOC_ATTR(weak);
+const char	*je_malloc_conf
+#ifndef _WIN32
+    JEMALLOC_ATTR(weak)
+#endif
+    ;
 bool	opt_abort =
 #ifdef JEMALLOC_DEBUG
     true
@@ -40,14 +44,14 @@ bool	opt_redzone = false;
 bool	opt_utrace = false;
 bool	opt_xmalloc = false;
 bool	opt_zero = false;
-size_t	opt_narenas = 0;
+unsigned	opt_narenas = 0;
 
 /* Initialized to true if the process is running inside Valgrind. */
 bool	in_valgrind;
 
 unsigned	ncpus;
 
-/* Protects arenas initialization (arenas, narenas_total). */
+/* Protects arenas initialization. */
 static malloc_mutex_t	arenas_lock;
 /*
  * Arenas that are used to service external requests.  Not all elements of the
@@ -57,10 +61,10 @@ static malloc_mutex_t	arenas_lock;
  * arenas.  arenas[narenas_auto..narenas_total) are only used if the application
  * takes some action to create them and allocate from them.
  */
-static arena_t		**arenas;
-static unsigned		narenas_total;
+arena_t			**arenas;
+static unsigned		narenas_total; /* Use narenas_total_*(). */
 static arena_t		*a0; /* arenas[0]; read-only after initialization. */
-static unsigned		narenas_auto; /* Read-only after initialization. */
+unsigned		narenas_auto; /* Read-only after initialization. */
 
 typedef enum {
 	malloc_init_uninitialized	= 3,
@@ -70,9 +74,37 @@ typedef enum {
 } malloc_init_t;
 static malloc_init_t	malloc_init_state = malloc_init_uninitialized;
 
+/* False should be the common case.  Set to true to trigger initialization. */
+static bool	malloc_slow = true;
+
+/* When malloc_slow is true, set the corresponding bits for sanity check. */
+enum {
+	flag_opt_junk_alloc	= (1U),
+	flag_opt_junk_free	= (1U << 1),
+	flag_opt_quarantine	= (1U << 2),
+	flag_opt_zero		= (1U << 3),
+	flag_opt_utrace		= (1U << 4),
+	flag_in_valgrind	= (1U << 5),
+	flag_opt_xmalloc	= (1U << 6)
+};
+static uint8_t	malloc_slow_flags;
+
+JEMALLOC_ALIGNED(CACHELINE)
+const size_t	pind2sz_tab[NPSIZES] = {
+#define	PSZ_yes(lg_grp, ndelta, lg_delta)				\
+	(((ZU(1)<<lg_grp) + (ZU(ndelta)<<lg_delta))),
+#define	PSZ_no(lg_grp, ndelta, lg_delta)
+#define	SC(index, lg_grp, lg_delta, ndelta, psz, bin, lg_delta_lookup)	\
+	PSZ_##psz(lg_grp, ndelta, lg_delta)
+	SIZE_CLASSES
+#undef PSZ_yes
+#undef PSZ_no
+#undef SC
+};
+
 JEMALLOC_ALIGNED(CACHELINE)
 const size_t	index2size_tab[NSIZES] = {
-#define	SC(index, lg_grp, lg_delta, ndelta, bin, lg_delta_lookup) \
+#define	SC(index, lg_grp, lg_delta, ndelta, psz, bin, lg_delta_lookup)	\
 	((ZU(1)<<lg_grp) + (ZU(ndelta)<<lg_delta)),
 	SIZE_CLASSES
 #undef SC
@@ -144,7 +176,7 @@ const uint8_t	size2index_tab[] = {
 #define	S2B_11(i)	S2B_10(i) S2B_10(i)
 #endif
 #define	S2B_no(i)
-#define	SC(index, lg_grp, lg_delta, ndelta, bin, lg_delta_lookup) \
+#define	SC(index, lg_grp, lg_delta, ndelta, psz, bin, lg_delta_lookup)	\
 	S2B_##lg_delta_lookup(index)
 	SIZE_CLASSES
 #undef S2B_3
@@ -195,7 +227,7 @@ _init_init_lock(void)
 	 * really only matters early in the process creation, before any
 	 * separate thread normally starts doing anything. */
 	if (!init_lock_initialized)
-		malloc_mutex_init(&init_lock);
+		malloc_mutex_init(&init_lock, "init", WITNESS_RANK_INIT);
 	init_lock_initialized = true;
 }
 
@@ -290,18 +322,10 @@ malloc_init(void)
 }
 
 /*
- * The a0*() functions are used instead of i[mcd]alloc() in situations that
+ * The a0*() functions are used instead of i{d,}alloc() in situations that
  * cannot tolerate TLS variable access.
  */
 
-arena_t *
-a0get(void)
-{
-
-	assert(a0 != NULL);
-	return (a0);
-}
-
 static void *
 a0ialloc(size_t size, bool zero, bool is_metadata)
 {
@@ -309,14 +333,22 @@ a0ialloc(size_t size, bool zero, bool is_metadata)
 	if (unlikely(malloc_init_a0()))
 		return (NULL);
 
-	return (iallocztm(NULL, size, zero, false, is_metadata, a0get()));
+	return (iallocztm(TSDN_NULL, size, size2index(size), zero, NULL,
+	    is_metadata, arena_get(TSDN_NULL, 0, true), true));
 }
 
 static void
 a0idalloc(void *ptr, bool is_metadata)
 {
 
-	idalloctm(NULL, ptr, false, is_metadata);
+	idalloctm(TSDN_NULL, ptr, false, is_metadata, true);
+}
+
+arena_t *
+a0get(void)
+{
+
+	return (a0);
 }
 
 void *
@@ -373,92 +405,88 @@ bootstrap_free(void *ptr)
 	a0idalloc(ptr, false);
 }
 
+static void
+arena_set(unsigned ind, arena_t *arena)
+{
+
+	atomic_write_p((void **)&arenas[ind], arena);
+}
+
+static void
+narenas_total_set(unsigned narenas)
+{
+
+	atomic_write_u(&narenas_total, narenas);
+}
+
+static void
+narenas_total_inc(void)
+{
+
+	atomic_add_u(&narenas_total, 1);
+}
+
+unsigned
+narenas_total_get(void)
+{
+
+	return (atomic_read_u(&narenas_total));
+}
+
 /* Create a new arena and insert it into the arenas array at index ind. */
 static arena_t *
-arena_init_locked(unsigned ind)
+arena_init_locked(tsdn_t *tsdn, unsigned ind)
 {
 	arena_t *arena;
 
-	/* Expand arenas if necessary. */
-	assert(ind <= narenas_total);
+	assert(ind <= narenas_total_get());
 	if (ind > MALLOCX_ARENA_MAX)
 		return (NULL);
-	if (ind == narenas_total) {
-		unsigned narenas_new = narenas_total + 1;
-		arena_t **arenas_new =
-		    (arena_t **)a0malloc(CACHELINE_CEILING(narenas_new *
-		    sizeof(arena_t *)));
-		if (arenas_new == NULL)
-			return (NULL);
-		memcpy(arenas_new, arenas, narenas_total * sizeof(arena_t *));
-		arenas_new[ind] = NULL;
-		/*
-		 * Deallocate only if arenas came from a0malloc() (not
-		 * base_alloc()).
-		 */
-		if (narenas_total != narenas_auto)
-			a0dalloc(arenas);
-		arenas = arenas_new;
-		narenas_total = narenas_new;
-	}
+	if (ind == narenas_total_get())
+		narenas_total_inc();
 
 	/*
 	 * Another thread may have already initialized arenas[ind] if it's an
 	 * auto arena.
 	 */
-	arena = arenas[ind];
+	arena = arena_get(tsdn, ind, false);
 	if (arena != NULL) {
 		assert(ind < narenas_auto);
 		return (arena);
 	}
 
 	/* Actually initialize the arena. */
-	arena = arenas[ind] = arena_new(ind);
+	arena = arena_new(tsdn, ind);
+	arena_set(ind, arena);
 	return (arena);
 }
 
 arena_t *
-arena_init(unsigned ind)
+arena_init(tsdn_t *tsdn, unsigned ind)
 {
 	arena_t *arena;
 
-	malloc_mutex_lock(&arenas_lock);
-	arena = arena_init_locked(ind);
-	malloc_mutex_unlock(&arenas_lock);
+	malloc_mutex_lock(tsdn, &arenas_lock);
+	arena = arena_init_locked(tsdn, ind);
+	malloc_mutex_unlock(tsdn, &arenas_lock);
 	return (arena);
 }
 
-unsigned
-narenas_total_get(void)
-{
-	unsigned narenas;
-
-	malloc_mutex_lock(&arenas_lock);
-	narenas = narenas_total;
-	malloc_mutex_unlock(&arenas_lock);
-
-	return (narenas);
-}
-
 static void
-arena_bind_locked(tsd_t *tsd, unsigned ind)
+arena_bind(tsd_t *tsd, unsigned ind, bool internal)
 {
 	arena_t *arena;
 
-	arena = arenas[ind];
-	arena->nthreads++;
-
-	if (tsd_nominal(tsd))
-		tsd_arena_set(tsd, arena);
-}
+	if (!tsd_nominal(tsd))
+		return;
 
-static void
-arena_bind(tsd_t *tsd, unsigned ind)
-{
+	arena = arena_get(tsd_tsdn(tsd), ind, false);
+	arena_nthreads_inc(arena, internal);
 
-	malloc_mutex_lock(&arenas_lock);
-	arena_bind_locked(tsd, ind);
-	malloc_mutex_unlock(&arenas_lock);
+	if (internal)
+		tsd_iarena_set(tsd, arena);
+	else
+		tsd_arena_set(tsd, arena);
 }
 
 void
@@ -466,131 +494,139 @@ arena_migrate(tsd_t *tsd, unsigned oldind, unsigned newind)
 {
 	arena_t *oldarena, *newarena;
 
-	malloc_mutex_lock(&arenas_lock);
-	oldarena = arenas[oldind];
-	newarena = arenas[newind];
-	oldarena->nthreads--;
-	newarena->nthreads++;
-	malloc_mutex_unlock(&arenas_lock);
+	oldarena = arena_get(tsd_tsdn(tsd), oldind, false);
+	newarena = arena_get(tsd_tsdn(tsd), newind, false);
+	arena_nthreads_dec(oldarena, false);
+	arena_nthreads_inc(newarena, false);
 	tsd_arena_set(tsd, newarena);
 }
 
-unsigned
-arena_nbound(unsigned ind)
-{
-	unsigned nthreads;
-
-	malloc_mutex_lock(&arenas_lock);
-	nthreads = arenas[ind]->nthreads;
-	malloc_mutex_unlock(&arenas_lock);
-	return (nthreads);
-}
-
 static void
-arena_unbind(tsd_t *tsd, unsigned ind)
+arena_unbind(tsd_t *tsd, unsigned ind, bool internal)
 {
 	arena_t *arena;
 
-	malloc_mutex_lock(&arenas_lock);
-	arena = arenas[ind];
-	arena->nthreads--;
-	malloc_mutex_unlock(&arenas_lock);
-	tsd_arena_set(tsd, NULL);
+	arena = arena_get(tsd_tsdn(tsd), ind, false);
+	arena_nthreads_dec(arena, internal);
+	if (internal)
+		tsd_iarena_set(tsd, NULL);
+	else
+		tsd_arena_set(tsd, NULL);
 }
 
-arena_t *
-arena_get_hard(tsd_t *tsd, unsigned ind, bool init_if_missing)
+arena_tdata_t *
+arena_tdata_get_hard(tsd_t *tsd, unsigned ind)
 {
-	arena_t *arena;
-	arena_t **arenas_cache = tsd_arenas_cache_get(tsd);
-	unsigned narenas_cache = tsd_narenas_cache_get(tsd);
+	arena_tdata_t *tdata, *arenas_tdata_old;
+	arena_tdata_t *arenas_tdata = tsd_arenas_tdata_get(tsd);
+	unsigned narenas_tdata_old, i;
+	unsigned narenas_tdata = tsd_narenas_tdata_get(tsd);
 	unsigned narenas_actual = narenas_total_get();
 
-	/* Deallocate old cache if it's too small. */
-	if (arenas_cache != NULL && narenas_cache < narenas_actual) {
-		a0dalloc(arenas_cache);
-		arenas_cache = NULL;
-		narenas_cache = 0;
-		tsd_arenas_cache_set(tsd, arenas_cache);
-		tsd_narenas_cache_set(tsd, narenas_cache);
-	}
-
-	/* Allocate cache if it's missing. */
-	if (arenas_cache == NULL) {
-		bool *arenas_cache_bypassp = tsd_arenas_cache_bypassp_get(tsd);
-		assert(ind < narenas_actual || !init_if_missing);
-		narenas_cache = (ind < narenas_actual) ? narenas_actual : ind+1;
-
-		if (tsd_nominal(tsd) && !*arenas_cache_bypassp) {
-			*arenas_cache_bypassp = true;
-			arenas_cache = (arena_t **)a0malloc(sizeof(arena_t *) *
-			    narenas_cache);
-			*arenas_cache_bypassp = false;
+	/*
+	 * Dissociate old tdata array (and set up for deallocation upon return)
+	 * if it's too small.
+	 */
+	if (arenas_tdata != NULL && narenas_tdata < narenas_actual) {
+		arenas_tdata_old = arenas_tdata;
+		narenas_tdata_old = narenas_tdata;
+		arenas_tdata = NULL;
+		narenas_tdata = 0;
+		tsd_arenas_tdata_set(tsd, arenas_tdata);
+		tsd_narenas_tdata_set(tsd, narenas_tdata);
+	} else {
+		arenas_tdata_old = NULL;
+		narenas_tdata_old = 0;
+	}
+
+	/* Allocate tdata array if it's missing. */
+	if (arenas_tdata == NULL) {
+		bool *arenas_tdata_bypassp = tsd_arenas_tdata_bypassp_get(tsd);
+		narenas_tdata = (ind < narenas_actual) ? narenas_actual : ind+1;
+
+		if (tsd_nominal(tsd) && !*arenas_tdata_bypassp) {
+			*arenas_tdata_bypassp = true;
+			arenas_tdata = (arena_tdata_t *)a0malloc(
+			    sizeof(arena_tdata_t) * narenas_tdata);
+			*arenas_tdata_bypassp = false;
 		}
-		if (arenas_cache == NULL) {
-			/*
-			 * This function must always tell the truth, even if
-			 * it's slow, so don't let OOM, thread cleanup (note
-			 * tsd_nominal check), nor recursive allocation
-			 * avoidance (note arenas_cache_bypass check) get in the
-			 * way.
-			 */
-			if (ind >= narenas_actual)
-				return (NULL);
-			malloc_mutex_lock(&arenas_lock);
-			arena = arenas[ind];
-			malloc_mutex_unlock(&arenas_lock);
-			return (arena);
+		if (arenas_tdata == NULL) {
+			tdata = NULL;
+			goto label_return;
 		}
-		assert(tsd_nominal(tsd) && !*arenas_cache_bypassp);
-		tsd_arenas_cache_set(tsd, arenas_cache);
-		tsd_narenas_cache_set(tsd, narenas_cache);
+		assert(tsd_nominal(tsd) && !*arenas_tdata_bypassp);
+		tsd_arenas_tdata_set(tsd, arenas_tdata);
+		tsd_narenas_tdata_set(tsd, narenas_tdata);
 	}
 
 	/*
-	 * Copy to cache.  It's possible that the actual number of arenas has
-	 * increased since narenas_total_get() was called above, but that causes
-	 * no correctness issues unless two threads concurrently execute the
-	 * arenas.extend mallctl, which we trust mallctl synchronization to
+	 * Copy to tdata array.  It's possible that the actual number of arenas
+	 * has increased since narenas_total_get() was called above, but that
+	 * causes no correctness issues unless two threads concurrently execute
+	 * the arenas.extend mallctl, which we trust mallctl synchronization to
 	 * prevent.
 	 */
-	malloc_mutex_lock(&arenas_lock);
-	memcpy(arenas_cache, arenas, sizeof(arena_t *) * narenas_actual);
-	malloc_mutex_unlock(&arenas_lock);
-	if (narenas_cache > narenas_actual) {
-		memset(&arenas_cache[narenas_actual], 0, sizeof(arena_t *) *
-		    (narenas_cache - narenas_actual));
-	}
-
-	/* Read the refreshed cache, and init the arena if necessary. */
-	arena = arenas_cache[ind];
-	if (init_if_missing && arena == NULL)
-		arena = arenas_cache[ind] = arena_init(ind);
-	return (arena);
+
+	/* Copy/initialize tickers. */
+	for (i = 0; i < narenas_actual; i++) {
+		if (i < narenas_tdata_old) {
+			ticker_copy(&arenas_tdata[i].decay_ticker,
+			    &arenas_tdata_old[i].decay_ticker);
+		} else {
+			ticker_init(&arenas_tdata[i].decay_ticker,
+			    DECAY_NTICKS_PER_UPDATE);
+		}
+	}
+	if (narenas_tdata > narenas_actual) {
+		memset(&arenas_tdata[narenas_actual], 0, sizeof(arena_tdata_t)
+		    * (narenas_tdata - narenas_actual));
+	}
+
+	/* Read the refreshed tdata array. */
+	tdata = &arenas_tdata[ind];
+label_return:
+	if (arenas_tdata_old != NULL)
+		a0dalloc(arenas_tdata_old);
+	return (tdata);
 }
 
 /* Slow path, called only by arena_choose(). */
 arena_t *
-arena_choose_hard(tsd_t *tsd)
+arena_choose_hard(tsd_t *tsd, bool internal)
 {
-	arena_t *ret;
+	arena_t *ret JEMALLOC_CC_SILENCE_INIT(NULL);
 
 	if (narenas_auto > 1) {
-		unsigned i, choose, first_null;
+		unsigned i, j, choose[2], first_null;
+
+		/*
+		 * Determine binding for both non-internal and internal
+		 * allocation.
+		 *
+		 *   choose[0]: For application allocation.
+		 *   choose[1]: For internal metadata allocation.
+		 */
+
+		for (j = 0; j < 2; j++)
+			choose[j] = 0;
 
-		choose = 0;
 		first_null = narenas_auto;
-		malloc_mutex_lock(&arenas_lock);
-		assert(a0get() != NULL);
+		malloc_mutex_lock(tsd_tsdn(tsd), &arenas_lock);
+		assert(arena_get(tsd_tsdn(tsd), 0, false) != NULL);
 		for (i = 1; i < narenas_auto; i++) {
-			if (arenas[i] != NULL) {
+			if (arena_get(tsd_tsdn(tsd), i, false) != NULL) {
 				/*
 				 * Choose the first arena that has the lowest
 				 * number of threads assigned to it.
 				 */
-				if (arenas[i]->nthreads <
-				    arenas[choose]->nthreads)
-					choose = i;
+				for (j = 0; j < 2; j++) {
+					if (arena_nthreads_get(arena_get(
+					    tsd_tsdn(tsd), i, false), !!j) <
+					    arena_nthreads_get(arena_get(
+					    tsd_tsdn(tsd), choose[j], false),
+					    !!j))
+						choose[j] = i;
+				}
 			} else if (first_null == narenas_auto) {
 				/*
 				 * Record the index of the first uninitialized
@@ -605,27 +641,40 @@ arena_choose_hard(tsd_t *tsd)
 			}
 		}
 
-		if (arenas[choose]->nthreads == 0
-		    || first_null == narenas_auto) {
-			/*
-			 * Use an unloaded arena, or the least loaded arena if
-			 * all arenas are already initialized.
-			 */
-			ret = arenas[choose];
-		} else {
-			/* Initialize a new arena. */
-			choose = first_null;
-			ret = arena_init_locked(choose);
-			if (ret == NULL) {
-				malloc_mutex_unlock(&arenas_lock);
-				return (NULL);
+		for (j = 0; j < 2; j++) {
+			if (arena_nthreads_get(arena_get(tsd_tsdn(tsd),
+			    choose[j], false), !!j) == 0 || first_null ==
+			    narenas_auto) {
+				/*
+				 * Use an unloaded arena, or the least loaded
+				 * arena if all arenas are already initialized.
+				 */
+				if (!!j == internal) {
+					ret = arena_get(tsd_tsdn(tsd),
+					    choose[j], false);
+				}
+			} else {
+				arena_t *arena;
+
+				/* Initialize a new arena. */
+				choose[j] = first_null;
+				arena = arena_init_locked(tsd_tsdn(tsd),
+				    choose[j]);
+				if (arena == NULL) {
+					malloc_mutex_unlock(tsd_tsdn(tsd),
+					    &arenas_lock);
+					return (NULL);
+				}
+				if (!!j == internal)
+					ret = arena;
 			}
+			arena_bind(tsd, choose[j], !!j);
 		}
-		arena_bind_locked(tsd, choose);
-		malloc_mutex_unlock(&arenas_lock);
+		malloc_mutex_unlock(tsd_tsdn(tsd), &arenas_lock);
 	} else {
-		ret = a0get();
-		arena_bind(tsd, 0);
+		ret = arena_get(tsd_tsdn(tsd), 0, false);
+		arena_bind(tsd, 0, false);
+		arena_bind(tsd, 0, true);
 	}
 
 	return (ret);
@@ -645,6 +694,16 @@ thread_deallocated_cleanup(tsd_t *tsd)
 	/* Do nothing. */
 }
 
+void
+iarena_cleanup(tsd_t *tsd)
+{
+	arena_t *iarena;
+
+	iarena = tsd_iarena_get(tsd);
+	if (iarena != NULL)
+		arena_unbind(tsd, iarena->ind, true);
+}
+
 void
 arena_cleanup(tsd_t *tsd)
 {
@@ -652,30 +711,33 @@ arena_cleanup(tsd_t *tsd)
 
 	arena = tsd_arena_get(tsd);
 	if (arena != NULL)
-		arena_unbind(tsd, arena->ind);
+		arena_unbind(tsd, arena->ind, false);
 }
 
 void
-arenas_cache_cleanup(tsd_t *tsd)
+arenas_tdata_cleanup(tsd_t *tsd)
 {
-	arena_t **arenas_cache;
+	arena_tdata_t *arenas_tdata;
 
-	arenas_cache = tsd_arenas_cache_get(tsd);
-	if (arenas_cache != NULL) {
-		tsd_arenas_cache_set(tsd, NULL);
-		a0dalloc(arenas_cache);
+	/* Prevent tsd->arenas_tdata from being (re)created. */
+	*tsd_arenas_tdata_bypassp_get(tsd) = true;
+
+	arenas_tdata = tsd_arenas_tdata_get(tsd);
+	if (arenas_tdata != NULL) {
+		tsd_arenas_tdata_set(tsd, NULL);
+		a0dalloc(arenas_tdata);
 	}
 }
 
 void
-narenas_cache_cleanup(tsd_t *tsd)
+narenas_tdata_cleanup(tsd_t *tsd)
 {
 
 	/* Do nothing. */
 }
 
 void
-arenas_cache_bypass_cleanup(tsd_t *tsd)
+arenas_tdata_bypass_cleanup(tsd_t *tsd)
 {
 
 	/* Do nothing. */
@@ -686,8 +748,11 @@ stats_print_atexit(void)
 {
 
 	if (config_tcache && config_stats) {
+		tsdn_t *tsdn;
 		unsigned narenas, i;
 
+		tsdn = tsdn_fetch();
+
 		/*
 		 * Merge stats from extant threads.  This is racy, since
 		 * individual threads do not lock when recording tcache stats
@@ -696,7 +761,7 @@ stats_print_atexit(void)
 		 * continue to allocate.
 		 */
 		for (i = 0, narenas = narenas_total_get(); i < narenas; i++) {
-			arena_t *arena = arenas[i];
+			arena_t *arena = arena_get(tsdn, i, false);
 			if (arena != NULL) {
 				tcache_t *tcache;
 
@@ -706,11 +771,11 @@ stats_print_atexit(void)
 				 * and bin locks in the opposite order,
 				 * deadlocks may result.
 				 */
-				malloc_mutex_lock(&arena->lock);
+				malloc_mutex_lock(tsdn, &arena->lock);
 				ql_foreach(tcache, &arena->tcache_ql, link) {
-					tcache_stats_merge(tcache, arena);
+					tcache_stats_merge(tsdn, tcache, arena);
 				}
-				malloc_mutex_unlock(&arena->lock);
+				malloc_mutex_unlock(tsdn, &arena->lock);
 			}
 		}
 	}
@@ -747,6 +812,20 @@ malloc_ncpus(void)
 	SYSTEM_INFO si;
 	GetSystemInfo(&si);
 	result = si.dwNumberOfProcessors;
+#elif defined(JEMALLOC_GLIBC_MALLOC_HOOK) && defined(CPU_COUNT)
+	/*
+	 * glibc >= 2.6 has the CPU_COUNT macro.
+	 *
+	 * glibc's sysconf() uses isspace().  glibc allocates for the first time
+	 * *before* setting up the isspace tables.  Therefore we need a
+	 * different method to get the number of CPUs.
+	 */
+	{
+		cpu_set_t set;
+
+		pthread_getaffinity_np(pthread_self(), sizeof(set), &set);
+		result = CPU_COUNT(&set);
+	}
 #else
 	result = sysconf(_SC_NPROCESSORS_ONLN);
 #endif
@@ -838,6 +917,26 @@ malloc_conf_error(const char *msg, const char *k, size_t klen, const char *v,
 	    (int)vlen, v);
 }
 
+static void
+malloc_slow_flag_init(void)
+{
+	/*
+	 * Combine the runtime options into malloc_slow for fast path.  Called
+	 * after processing all the options.
+	 */
+	malloc_slow_flags |= (opt_junk_alloc ? flag_opt_junk_alloc : 0)
+	    | (opt_junk_free ? flag_opt_junk_free : 0)
+	    | (opt_quarantine ? flag_opt_quarantine : 0)
+	    | (opt_zero ? flag_opt_zero : 0)
+	    | (opt_utrace ? flag_opt_utrace : 0)
+	    | (opt_xmalloc ? flag_opt_xmalloc : 0);
+
+	if (config_valgrind)
+		malloc_slow_flags |= (in_valgrind ? flag_in_valgrind : 0);
+
+	malloc_slow = (malloc_slow_flags != 0);
+}
+
 static void
 malloc_conf_init(void)
 {
@@ -864,10 +963,13 @@ malloc_conf_init(void)
 			opt_tcache = false;
 	}
 
-	for (i = 0; i < 3; i++) {
+	for (i = 0; i < 4; i++) {
 		/* Get runtime configuration. */
 		switch (i) {
 		case 0:
+			opts = config_malloc_conf;
+			break;
+		case 1:
 			if (je_malloc_conf != NULL) {
 				/*
 				 * Use options that were compiled into the
@@ -880,8 +982,8 @@ malloc_conf_init(void)
 				opts = buf;
 			}
 			break;
-		case 1: {
-			int linklen = 0;
+		case 2: {
+			ssize_t linklen = 0;
 #ifndef _WIN32
 			int saved_errno = errno;
 			const char *linkname =
@@ -907,7 +1009,7 @@ malloc_conf_init(void)
 			buf[linklen] = '\0';
 			opts = buf;
 			break;
-		} case 2: {
+		} case 3: {
 			const char *envname =
 #ifdef JEMALLOC_PREFIX
 			    JEMALLOC_CPREFIX"MALLOC_CONF"
@@ -954,7 +1056,11 @@ malloc_conf_init(void)
 				if (cont)				\
 					continue;			\
 			}
-#define	CONF_HANDLE_SIZE_T(o, n, min, max, clip)			\
+#define	CONF_MIN_no(um, min)	false
+#define	CONF_MIN_yes(um, min)	((um) < (min))
+#define	CONF_MAX_no(um, max)	false
+#define	CONF_MAX_yes(um, max)	((um) > (max))
+#define	CONF_HANDLE_T_U(t, o, n, min, max, check_min, check_max, clip)	\
 			if (CONF_MATCH(n)) {				\
 				uintmax_t um;				\
 				char *end;				\
@@ -967,24 +1073,35 @@ malloc_conf_init(void)
 					    "Invalid conf value",	\
 					    k, klen, v, vlen);		\
 				} else if (clip) {			\
-					if ((min) != 0 && um < (min))	\
-						o = (min);		\
-					else if (um > (max))		\
-						o = (max);		\
+					if (CONF_MIN_##check_min(um,	\
+					    (min)))			\
+						o = (t)(min);		\
+					else if (CONF_MAX_##check_max(	\
+					    um, (max)))			\
+						o = (t)(max);		\
 					else				\
-						o = um;			\
+						o = (t)um;		\
 				} else {				\
-					if (((min) != 0 && um < (min))	\
-					    || um > (max)) {		\
+					if (CONF_MIN_##check_min(um,	\
+					    (min)) ||			\
+					    CONF_MAX_##check_max(um,	\
+					    (max))) {			\
 						malloc_conf_error(	\
 						    "Out-of-range "	\
 						    "conf value",	\
 						    k, klen, v, vlen);	\
 					} else				\
-						o = um;			\
+						o = (t)um;		\
 				}					\
 				continue;				\
 			}
+#define	CONF_HANDLE_UNSIGNED(o, n, min, max, check_min, check_max,	\
+    clip)								\
+			CONF_HANDLE_T_U(unsigned, o, n, min, max,	\
+			    check_min, check_max, clip)
+#define	CONF_HANDLE_SIZE_T(o, n, min, max, check_min, check_max, clip)	\
+			CONF_HANDLE_T_U(size_t, o, n, min, max,		\
+			    check_min, check_max, clip)
 #define	CONF_HANDLE_SSIZE_T(o, n, min, max)				\
 			if (CONF_MATCH(n)) {				\
 				long l;					\
@@ -1027,7 +1144,7 @@ malloc_conf_init(void)
 			 */
 			CONF_HANDLE_SIZE_T(opt_lg_chunk, "lg_chunk", LG_PAGE +
 			    LG_SIZE_CLASS_GROUP + (config_fill ? 2 : 1),
-			    (sizeof(size_t) << 3) - 1, true)
+			    (sizeof(size_t) << 3) - 1, yes, yes, true)
 			if (strncmp("dss", k, klen) == 0) {
 				int i;
 				bool match = false;
@@ -1052,17 +1169,47 @@ malloc_conf_init(void)
 				}
 				continue;
 			}
-			CONF_HANDLE_SIZE_T(opt_narenas, "narenas", 1,
-			    SIZE_T_MAX, false)
+			CONF_HANDLE_UNSIGNED(opt_narenas, "narenas", 1,
+			    UINT_MAX, yes, no, false)
+			if (strncmp("purge", k, klen) == 0) {
+				int i;
+				bool match = false;
+				for (i = 0; i < purge_mode_limit; i++) {
+					if (strncmp(purge_mode_names[i], v,
+					    vlen) == 0) {
+						opt_purge = (purge_mode_t)i;
+						match = true;
+						break;
+					}
+				}
+				if (!match) {
+					malloc_conf_error("Invalid conf value",
+					    k, klen, v, vlen);
+				}
+				continue;
+			}
 			CONF_HANDLE_SSIZE_T(opt_lg_dirty_mult, "lg_dirty_mult",
 			    -1, (sizeof(size_t) << 3) - 1)
+			CONF_HANDLE_SSIZE_T(opt_decay_time, "decay_time", -1,
+			    NSTIME_SEC_MAX);
 			CONF_HANDLE_BOOL(opt_stats_print, "stats_print", true)
 			if (config_fill) {
 				if (CONF_MATCH("junk")) {
 					if (CONF_MATCH_VALUE("true")) {
-						opt_junk = "true";
-						opt_junk_alloc = opt_junk_free =
-						    true;
+						if (config_valgrind &&
+						    unlikely(in_valgrind)) {
+							malloc_conf_error(
+							"Deallocation-time "
+							"junk filling cannot "
+							"be enabled while "
+							"running inside "
+							"Valgrind", k, klen, v,
+							vlen);
+						} else {
+							opt_junk = "true";
+							opt_junk_alloc = true;
+							opt_junk_free = true;
+						}
 					} else if (CONF_MATCH_VALUE("false")) {
 						opt_junk = "false";
 						opt_junk_alloc = opt_junk_free =
@@ -1072,9 +1219,20 @@ malloc_conf_init(void)
 						opt_junk_alloc = true;
 						opt_junk_free = false;
 					} else if (CONF_MATCH_VALUE("free")) {
-						opt_junk = "free";
-						opt_junk_alloc = false;
-						opt_junk_free = true;
+						if (config_valgrind &&
+						    unlikely(in_valgrind)) {
+							malloc_conf_error(
+							"Deallocation-time "
+							"junk filling cannot "
+							"be enabled while "
+							"running inside "
+							"Valgrind", k, klen, v,
+							vlen);
+						} else {
+							opt_junk = "free";
+							opt_junk_alloc = false;
+							opt_junk_free = true;
+						}
 					} else {
 						malloc_conf_error(
 						    "Invalid conf value", k,
@@ -1083,7 +1241,7 @@ malloc_conf_init(void)
 					continue;
 				}
 				CONF_HANDLE_SIZE_T(opt_quarantine, "quarantine",
-				    0, SIZE_T_MAX, false)
+				    0, SIZE_T_MAX, no, no, false)
 				CONF_HANDLE_BOOL(opt_redzone, "redzone", true)
 				CONF_HANDLE_BOOL(opt_zero, "zero", true)
 			}
@@ -1120,8 +1278,8 @@ malloc_conf_init(void)
 				CONF_HANDLE_BOOL(opt_prof_thread_active_init,
 				    "prof_thread_active_init", true)
 				CONF_HANDLE_SIZE_T(opt_lg_prof_sample,
-				    "lg_prof_sample", 0,
-				    (sizeof(uint64_t) << 3) - 1, true)
+				    "lg_prof_sample", 0, (sizeof(uint64_t) << 3)
+				    - 1, no, yes, true)
 				CONF_HANDLE_BOOL(opt_prof_accum, "prof_accum",
 				    true)
 				CONF_HANDLE_SSIZE_T(opt_lg_prof_interval,
@@ -1137,7 +1295,14 @@ malloc_conf_init(void)
 			malloc_conf_error("Invalid conf pair", k, klen, v,
 			    vlen);
 #undef CONF_MATCH
+#undef CONF_MATCH_VALUE
 #undef CONF_HANDLE_BOOL
+#undef CONF_MIN_no
+#undef CONF_MIN_yes
+#undef CONF_MAX_no
+#undef CONF_MAX_yes
+#undef CONF_HANDLE_T_U
+#undef CONF_HANDLE_UNSIGNED
 #undef CONF_HANDLE_SIZE_T
 #undef CONF_HANDLE_SSIZE_T
 #undef CONF_HANDLE_CHAR_P
@@ -1145,7 +1310,6 @@ malloc_conf_init(void)
 	}
 }
 
-/* init_lock must be held. */
 static bool
 malloc_init_hard_needed(void)
 {
@@ -1161,11 +1325,14 @@ malloc_init_hard_needed(void)
 	}
 #ifdef JEMALLOC_THREADED_INIT
 	if (malloc_initializer != NO_INITIALIZER && !IS_INITIALIZER) {
+		spin_t spinner;
+
 		/* Busy-wait until the initializing thread completes. */
+		spin_init(&spinner);
 		do {
-			malloc_mutex_unlock(&init_lock);
-			CPU_SPINWAIT;
-			malloc_mutex_lock(&init_lock);
+			malloc_mutex_unlock(TSDN_NULL, &init_lock);
+			spin_adaptive(&spinner);
+			malloc_mutex_lock(TSDN_NULL, &init_lock);
 		} while (!malloc_initialized());
 		return (false);
 	}
@@ -1173,9 +1340,8 @@ malloc_init_hard_needed(void)
 	return (true);
 }
 
-/* init_lock must be held. */
 static bool
-malloc_init_hard_a0_locked(void)
+malloc_init_hard_a0_locked()
 {
 
 	malloc_initializer = INITIALIZER;
@@ -1191,6 +1357,7 @@ malloc_init_hard_a0_locked(void)
 				abort();
 		}
 	}
+	pages_boot();
 	if (base_boot())
 		return (true);
 	if (chunk_boot())
@@ -1199,26 +1366,28 @@ malloc_init_hard_a0_locked(void)
 		return (true);
 	if (config_prof)
 		prof_boot1();
-	if (arena_boot())
-		return (true);
-	if (config_tcache && tcache_boot())
+	arena_boot();
+	if (config_tcache && tcache_boot(TSDN_NULL))
 		return (true);
-	if (malloc_mutex_init(&arenas_lock))
+	if (malloc_mutex_init(&arenas_lock, "arenas", WITNESS_RANK_ARENAS))
 		return (true);
 	/*
 	 * Create enough scaffolding to allow recursive allocation in
 	 * malloc_ncpus().
 	 */
-	narenas_total = narenas_auto = 1;
+	narenas_auto = 1;
+	narenas_total_set(narenas_auto);
 	arenas = &a0;
 	memset(arenas, 0, sizeof(arena_t *) * narenas_auto);
 	/*
 	 * Initialize one arena here.  The rest are lazily created in
 	 * arena_choose_hard().
 	 */
-	if (arena_init(0) == NULL)
+	if (arena_init(TSDN_NULL, 0) == NULL)
 		return (true);
+
 	malloc_init_state = malloc_init_a0_initialized;
+
 	return (false);
 }
 
@@ -1227,45 +1396,42 @@ malloc_init_hard_a0(void)
 {
 	bool ret;
 
-	malloc_mutex_lock(&init_lock);
+	malloc_mutex_lock(TSDN_NULL, &init_lock);
 	ret = malloc_init_hard_a0_locked();
-	malloc_mutex_unlock(&init_lock);
+	malloc_mutex_unlock(TSDN_NULL, &init_lock);
 	return (ret);
 }
 
-/*
- * Initialize data structures which may trigger recursive allocation.
- *
- * init_lock must be held.
- */
-static void
+/* Initialize data structures which may trigger recursive allocation. */
+static bool
 malloc_init_hard_recursible(void)
 {
 
 	malloc_init_state = malloc_init_recursible;
-	malloc_mutex_unlock(&init_lock);
 
 	ncpus = malloc_ncpus();
 
-#if (!defined(JEMALLOC_MUTEX_INIT_CB) && !defined(JEMALLOC_ZONE) \
-    && !defined(_WIN32) && !defined(__native_client__))
-	/* LinuxThreads's pthread_atfork() allocates. */
+#if (defined(JEMALLOC_HAVE_PTHREAD_ATFORK) && !defined(JEMALLOC_MUTEX_INIT_CB) \
+    && !defined(JEMALLOC_ZONE) && !defined(_WIN32) && \
+    !defined(__native_client__))
+	/* LinuxThreads' pthread_atfork() allocates. */
 	if (pthread_atfork(jemalloc_prefork, jemalloc_postfork_parent,
 	    jemalloc_postfork_child) != 0) {
 		malloc_write("<jemalloc>: Error in pthread_atfork()\n");
 		if (opt_abort)
 			abort();
+		return (true);
 	}
 #endif
-	malloc_mutex_lock(&init_lock);
+
+	return (false);
 }
 
-/* init_lock must be held. */
 static bool
-malloc_init_hard_finish(void)
+malloc_init_hard_finish(tsdn_t *tsdn)
 {
 
-	if (mutex_boot())
+	if (malloc_mutex_boot())
 		return (true);
 
 	if (opt_narenas == 0) {
@@ -1280,68 +1446,69 @@ malloc_init_hard_finish(void)
 	}
 	narenas_auto = opt_narenas;
 	/*
-	 * Make sure that the arenas array can be allocated.  In practice, this
-	 * limit is enough to allow the allocator to function, but the ctl
-	 * machinery will fail to allocate memory at far lower limits.
+	 * Limit the number of arenas to the indexing range of MALLOCX_ARENA().
 	 */
-	if (narenas_auto > chunksize / sizeof(arena_t *)) {
-		narenas_auto = chunksize / sizeof(arena_t *);
+	if (narenas_auto > MALLOCX_ARENA_MAX) {
+		narenas_auto = MALLOCX_ARENA_MAX;
 		malloc_printf("<jemalloc>: Reducing narenas to limit (%d)\n",
 		    narenas_auto);
 	}
-	narenas_total = narenas_auto;
+	narenas_total_set(narenas_auto);
 
 	/* Allocate and initialize arenas. */
-	arenas = (arena_t **)base_alloc(sizeof(arena_t *) * narenas_total);
+	arenas = (arena_t **)base_alloc(tsdn, sizeof(arena_t *) *
+	    (MALLOCX_ARENA_MAX+1));
 	if (arenas == NULL)
 		return (true);
-	/*
-	 * Zero the array.  In practice, this should always be pre-zeroed,
-	 * since it was just mmap()ed, but let's be sure.
-	 */
-	memset(arenas, 0, sizeof(arena_t *) * narenas_total);
 	/* Copy the pointer to the one arena that was already initialized. */
-	arenas[0] = a0;
+	arena_set(0, a0);
 
 	malloc_init_state = malloc_init_initialized;
+	malloc_slow_flag_init();
+
 	return (false);
 }
 
 static bool
 malloc_init_hard(void)
 {
+	tsd_t *tsd;
 
 #if defined(_WIN32) && _WIN32_WINNT < 0x0600
 	_init_init_lock();
 #endif
-	malloc_mutex_lock(&init_lock);
+	malloc_mutex_lock(TSDN_NULL, &init_lock);
 	if (!malloc_init_hard_needed()) {
-		malloc_mutex_unlock(&init_lock);
+		malloc_mutex_unlock(TSDN_NULL, &init_lock);
 		return (false);
 	}
 
 	if (malloc_init_state != malloc_init_a0_initialized &&
 	    malloc_init_hard_a0_locked()) {
-		malloc_mutex_unlock(&init_lock);
+		malloc_mutex_unlock(TSDN_NULL, &init_lock);
 		return (true);
 	}
-	if (malloc_tsd_boot0()) {
-		malloc_mutex_unlock(&init_lock);
+
+	malloc_mutex_unlock(TSDN_NULL, &init_lock);
+	/* Recursive allocation relies on functional tsd. */
+	tsd = malloc_tsd_boot0();
+	if (tsd == NULL)
 		return (true);
-	}
-	if (config_prof && prof_boot2()) {
-		malloc_mutex_unlock(&init_lock);
+	if (malloc_init_hard_recursible())
 		return (true);
-	}
+	malloc_mutex_lock(tsd_tsdn(tsd), &init_lock);
 
-	malloc_init_hard_recursible();
+	if (config_prof && prof_boot2(tsd)) {
+		malloc_mutex_unlock(tsd_tsdn(tsd), &init_lock);
+		return (true);
+	}
 
-	if (malloc_init_hard_finish()) {
-		malloc_mutex_unlock(&init_lock);
+	if (malloc_init_hard_finish(tsd_tsdn(tsd))) {
+		malloc_mutex_unlock(tsd_tsdn(tsd), &init_lock);
 		return (true);
 	}
 
-	malloc_mutex_unlock(&init_lock);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &init_lock);
 	malloc_tsd_boot1();
 	return (false);
 }
@@ -1355,61 +1522,104 @@ malloc_init_hard(void)
  */
 
 static void *
-imalloc_prof_sample(tsd_t *tsd, size_t usize, prof_tctx_t *tctx)
+ialloc_prof_sample(tsd_t *tsd, size_t usize, szind_t ind, bool zero,
+    prof_tctx_t *tctx, bool slow_path)
 {
 	void *p;
 
 	if (tctx == NULL)
 		return (NULL);
 	if (usize <= SMALL_MAXCLASS) {
-		p = imalloc(tsd, LARGE_MINCLASS);
+		szind_t ind_large = size2index(LARGE_MINCLASS);
+		p = ialloc(tsd, LARGE_MINCLASS, ind_large, zero, slow_path);
 		if (p == NULL)
 			return (NULL);
-		arena_prof_promoted(p, usize);
+		arena_prof_promoted(tsd_tsdn(tsd), p, usize);
 	} else
-		p = imalloc(tsd, usize);
+		p = ialloc(tsd, usize, ind, zero, slow_path);
 
 	return (p);
 }
 
 JEMALLOC_ALWAYS_INLINE_C void *
-imalloc_prof(tsd_t *tsd, size_t usize)
+ialloc_prof(tsd_t *tsd, size_t usize, szind_t ind, bool zero, bool slow_path)
 {
 	void *p;
 	prof_tctx_t *tctx;
 
 	tctx = prof_alloc_prep(tsd, usize, prof_active_get_unlocked(), true);
 	if (unlikely((uintptr_t)tctx != (uintptr_t)1U))
-		p = imalloc_prof_sample(tsd, usize, tctx);
+		p = ialloc_prof_sample(tsd, usize, ind, zero, tctx, slow_path);
 	else
-		p = imalloc(tsd, usize);
+		p = ialloc(tsd, usize, ind, zero, slow_path);
 	if (unlikely(p == NULL)) {
 		prof_alloc_rollback(tsd, tctx, true);
 		return (NULL);
 	}
-	prof_malloc(p, usize, tctx);
+	prof_malloc(tsd_tsdn(tsd), p, usize, tctx);
 
 	return (p);
 }
 
+/*
+ * ialloc_body() is inlined so that fast and slow paths are generated separately
+ * with statically known slow_path.
+ *
+ * This function guarantees that *tsdn is non-NULL on success.
+ */
 JEMALLOC_ALWAYS_INLINE_C void *
-imalloc_body(size_t size, tsd_t **tsd, size_t *usize)
+ialloc_body(size_t size, bool zero, tsdn_t **tsdn, size_t *usize,
+    bool slow_path)
 {
+	tsd_t *tsd;
+	szind_t ind;
 
-	if (unlikely(malloc_init()))
+	if (slow_path && unlikely(malloc_init())) {
+		*tsdn = NULL;
 		return (NULL);
-	*tsd = tsd_fetch();
+	}
 
-	if (config_prof && opt_prof) {
-		*usize = s2u(size);
-		if (unlikely(*usize == 0))
-			return (NULL);
-		return (imalloc_prof(*tsd, *usize));
+	tsd = tsd_fetch();
+	*tsdn = tsd_tsdn(tsd);
+	witness_assert_lockless(tsd_tsdn(tsd));
+
+	ind = size2index(size);
+	if (unlikely(ind >= NSIZES))
+		return (NULL);
+
+	if (config_stats || (config_prof && opt_prof) || (slow_path &&
+	    config_valgrind && unlikely(in_valgrind))) {
+		*usize = index2size(ind);
+		assert(*usize > 0 && *usize <= HUGE_MAXCLASS);
 	}
 
-	if (config_stats || (config_valgrind && unlikely(in_valgrind)))
-		*usize = s2u(size);
-	return (imalloc(*tsd, size));
+	if (config_prof && opt_prof)
+		return (ialloc_prof(tsd, *usize, ind, zero, slow_path));
+
+	return (ialloc(tsd, size, ind, zero, slow_path));
+}
+
+JEMALLOC_ALWAYS_INLINE_C void
+ialloc_post_check(void *ret, tsdn_t *tsdn, size_t usize, const char *func,
+    bool update_errno, bool slow_path)
+{
+
+	assert(!tsdn_null(tsdn) || ret == NULL);
+
+	if (unlikely(ret == NULL)) {
+		if (slow_path && config_xmalloc && unlikely(opt_xmalloc)) {
+			malloc_printf("<jemalloc>: Error in %s(): out of "
+			    "memory\n", func);
+			abort();
+		}
+		if (update_errno)
+			set_errno(ENOMEM);
+	}
+	if (config_stats && likely(ret != NULL)) {
+		assert(usize == isalloc(tsdn, ret, config_prof));
+		*tsd_thread_allocatedp_get(tsdn_tsd(tsdn)) += usize;
+	}
+	witness_assert_lockless(tsdn);
 }
 
 JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN
@@ -1418,27 +1628,22 @@ JEMALLOC_ATTR(malloc) JEMALLOC_ALLOC_SIZE(1)
 je_malloc(size_t size)
 {
 	void *ret;
-	tsd_t *tsd;
+	tsdn_t *tsdn;
 	size_t usize JEMALLOC_CC_SILENCE_INIT(0);
 
 	if (size == 0)
 		size = 1;
 
-	ret = imalloc_body(size, &tsd, &usize);
-	if (unlikely(ret == NULL)) {
-		if (config_xmalloc && unlikely(opt_xmalloc)) {
-			malloc_write("<jemalloc>: Error in malloc(): "
-			    "out of memory\n");
-			abort();
-		}
-		set_errno(ENOMEM);
-	}
-	if (config_stats && likely(ret != NULL)) {
-		assert(usize == isalloc(ret, config_prof));
-		*tsd_thread_allocatedp_get(tsd) += usize;
+	if (likely(!malloc_slow)) {
+		ret = ialloc_body(size, false, &tsdn, &usize, false);
+		ialloc_post_check(ret, tsdn, usize, "malloc", true, false);
+	} else {
+		ret = ialloc_body(size, false, &tsdn, &usize, true);
+		ialloc_post_check(ret, tsdn, usize, "malloc", true, true);
+		UTRACE(0, size, ret);
+		JEMALLOC_VALGRIND_MALLOC(ret != NULL, tsdn, ret, usize, false);
 	}
-	UTRACE(0, size, ret);
-	JEMALLOC_VALGRIND_MALLOC(ret != NULL, ret, usize, false);
+
 	return (ret);
 }
 
@@ -1455,7 +1660,7 @@ imemalign_prof_sample(tsd_t *tsd, size_t alignment, size_t usize,
 		p = ipalloc(tsd, LARGE_MINCLASS, alignment, false);
 		if (p == NULL)
 			return (NULL);
-		arena_prof_promoted(p, usize);
+		arena_prof_promoted(tsd_tsdn(tsd), p, usize);
 	} else
 		p = ipalloc(tsd, usize, alignment, false);
 
@@ -1477,7 +1682,7 @@ imemalign_prof(tsd_t *tsd, size_t alignment, size_t usize)
 		prof_alloc_rollback(tsd, tctx, true);
 		return (NULL);
 	}
-	prof_malloc(p, usize, tctx);
+	prof_malloc(tsd_tsdn(tsd), p, usize, tctx);
 
 	return (p);
 }
@@ -1494,10 +1699,12 @@ imemalign(void **memptr, size_t alignment, size_t size, size_t min_alignment)
 	assert(min_alignment != 0);
 
 	if (unlikely(malloc_init())) {
+		tsd = NULL;
 		result = NULL;
 		goto label_oom;
 	}
 	tsd = tsd_fetch();
+	witness_assert_lockless(tsd_tsdn(tsd));
 	if (size == 0)
 		size = 1;
 
@@ -1515,7 +1722,7 @@ imemalign(void **memptr, size_t alignment, size_t size, size_t min_alignment)
 	}
 
 	usize = sa2u(size, alignment);
-	if (unlikely(usize == 0)) {
+	if (unlikely(usize == 0 || usize > HUGE_MAXCLASS)) {
 		result = NULL;
 		goto label_oom;
 	}
@@ -1532,10 +1739,13 @@ imemalign(void **memptr, size_t alignment, size_t size, size_t min_alignment)
 	ret = 0;
 label_return:
 	if (config_stats && likely(result != NULL)) {
-		assert(usize == isalloc(result, config_prof));
+		assert(usize == isalloc(tsd_tsdn(tsd), result, config_prof));
 		*tsd_thread_allocatedp_get(tsd) += usize;
 	}
 	UTRACE(0, size, result);
+	JEMALLOC_VALGRIND_MALLOC(result != NULL, tsd_tsdn(tsd), result, usize,
+	    false);
+	witness_assert_lockless(tsd_tsdn(tsd));
 	return (ret);
 label_oom:
 	assert(result == NULL);
@@ -1545,6 +1755,7 @@ label_oom:
 		abort();
 	}
 	ret = ENOMEM;
+	witness_assert_lockless(tsd_tsdn(tsd));
 	goto label_return;
 }
 
@@ -1552,9 +1763,10 @@ JEMALLOC_EXPORT int JEMALLOC_NOTHROW
 JEMALLOC_ATTR(nonnull(1))
 je_posix_memalign(void **memptr, size_t alignment, size_t size)
 {
-	int ret = imemalign(memptr, alignment, size, sizeof(void *));
-	JEMALLOC_VALGRIND_MALLOC(ret == 0, *memptr, isalloc(*memptr,
-	    config_prof), false);
+	int ret;
+
+	ret = imemalign(memptr, alignment, size, sizeof(void *));
+
 	return (ret);
 }
 
@@ -1570,47 +1782,8 @@ je_aligned_alloc(size_t alignment, size_t size)
 		ret = NULL;
 		set_errno(err);
 	}
-	JEMALLOC_VALGRIND_MALLOC(err == 0, ret, isalloc(ret, config_prof),
-	    false);
-	return (ret);
-}
-
-static void *
-icalloc_prof_sample(tsd_t *tsd, size_t usize, prof_tctx_t *tctx)
-{
-	void *p;
-
-	if (tctx == NULL)
-		return (NULL);
-	if (usize <= SMALL_MAXCLASS) {
-		p = icalloc(tsd, LARGE_MINCLASS);
-		if (p == NULL)
-			return (NULL);
-		arena_prof_promoted(p, usize);
-	} else
-		p = icalloc(tsd, usize);
-
-	return (p);
-}
-
-JEMALLOC_ALWAYS_INLINE_C void *
-icalloc_prof(tsd_t *tsd, size_t usize)
-{
-	void *p;
-	prof_tctx_t *tctx;
-
-	tctx = prof_alloc_prep(tsd, usize, prof_active_get_unlocked(), true);
-	if (unlikely((uintptr_t)tctx != (uintptr_t)1U))
-		p = icalloc_prof_sample(tsd, usize, tctx);
-	else
-		p = icalloc(tsd, usize);
-	if (unlikely(p == NULL)) {
-		prof_alloc_rollback(tsd, tctx, true);
-		return (NULL);
-	}
-	prof_malloc(p, usize, tctx);
 
-	return (p);
+	return (ret);
 }
 
 JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN
@@ -1619,65 +1792,35 @@ JEMALLOC_ATTR(malloc) JEMALLOC_ALLOC_SIZE2(1, 2)
 je_calloc(size_t num, size_t size)
 {
 	void *ret;
-	tsd_t *tsd;
+	tsdn_t *tsdn;
 	size_t num_size;
 	size_t usize JEMALLOC_CC_SILENCE_INIT(0);
 
-	if (unlikely(malloc_init())) {
-		num_size = 0;
-		ret = NULL;
-		goto label_return;
-	}
-	tsd = tsd_fetch();
-
 	num_size = num * size;
 	if (unlikely(num_size == 0)) {
 		if (num == 0 || size == 0)
 			num_size = 1;
-		else {
-			ret = NULL;
-			goto label_return;
-		}
+		else
+			num_size = HUGE_MAXCLASS + 1; /* Trigger OOM. */
 	/*
 	 * Try to avoid division here.  We know that it isn't possible to
 	 * overflow during multiplication if neither operand uses any of the
 	 * most significant half of the bits in a size_t.
 	 */
 	} else if (unlikely(((num | size) & (SIZE_T_MAX << (sizeof(size_t) <<
-	    2))) && (num_size / size != num))) {
-		/* size_t overflow. */
-		ret = NULL;
-		goto label_return;
-	}
+	    2))) && (num_size / size != num)))
+		num_size = HUGE_MAXCLASS + 1; /* size_t overflow. */
 
-	if (config_prof && opt_prof) {
-		usize = s2u(num_size);
-		if (unlikely(usize == 0)) {
-			ret = NULL;
-			goto label_return;
-		}
-		ret = icalloc_prof(tsd, usize);
+	if (likely(!malloc_slow)) {
+		ret = ialloc_body(num_size, true, &tsdn, &usize, false);
+		ialloc_post_check(ret, tsdn, usize, "calloc", true, false);
 	} else {
-		if (config_stats || (config_valgrind && unlikely(in_valgrind)))
-			usize = s2u(num_size);
-		ret = icalloc(tsd, num_size);
+		ret = ialloc_body(num_size, true, &tsdn, &usize, true);
+		ialloc_post_check(ret, tsdn, usize, "calloc", true, true);
+		UTRACE(0, num_size, ret);
+		JEMALLOC_VALGRIND_MALLOC(ret != NULL, tsdn, ret, usize, true);
 	}
 
-label_return:
-	if (unlikely(ret == NULL)) {
-		if (config_xmalloc && unlikely(opt_xmalloc)) {
-			malloc_write("<jemalloc>: Error in calloc(): out of "
-			    "memory\n");
-			abort();
-		}
-		set_errno(ENOMEM);
-	}
-	if (config_stats && likely(ret != NULL)) {
-		assert(usize == isalloc(ret, config_prof));
-		*tsd_thread_allocatedp_get(tsd) += usize;
-	}
-	UTRACE(0, num_size, ret);
-	JEMALLOC_VALGRIND_MALLOC(ret != NULL, ret, usize, true);
 	return (ret);
 }
 
@@ -1693,7 +1836,7 @@ irealloc_prof_sample(tsd_t *tsd, void *old_ptr, size_t old_usize, size_t usize,
 		p = iralloc(tsd, old_ptr, old_usize, LARGE_MINCLASS, 0, false);
 		if (p == NULL)
 			return (NULL);
-		arena_prof_promoted(p, usize);
+		arena_prof_promoted(tsd_tsdn(tsd), p, usize);
 	} else
 		p = iralloc(tsd, old_ptr, old_usize, usize, 0, false);
 
@@ -1708,7 +1851,7 @@ irealloc_prof(tsd_t *tsd, void *old_ptr, size_t old_usize, size_t usize)
 	prof_tctx_t *old_tctx, *tctx;
 
 	prof_active = prof_active_get_unlocked();
-	old_tctx = prof_tctx_get(old_ptr);
+	old_tctx = prof_tctx_get(tsd_tsdn(tsd), old_ptr);
 	tctx = prof_alloc_prep(tsd, usize, prof_active, true);
 	if (unlikely((uintptr_t)tctx != (uintptr_t)1U))
 		p = irealloc_prof_sample(tsd, old_ptr, old_usize, usize, tctx);
@@ -1725,32 +1868,41 @@ irealloc_prof(tsd_t *tsd, void *old_ptr, size_t old_usize, size_t usize)
 }
 
 JEMALLOC_INLINE_C void
-ifree(tsd_t *tsd, void *ptr, tcache_t *tcache)
+ifree(tsd_t *tsd, void *ptr, tcache_t *tcache, bool slow_path)
 {
 	size_t usize;
 	UNUSED size_t rzsize JEMALLOC_CC_SILENCE_INIT(0);
 
+	witness_assert_lockless(tsd_tsdn(tsd));
+
 	assert(ptr != NULL);
 	assert(malloc_initialized() || IS_INITIALIZER);
 
 	if (config_prof && opt_prof) {
-		usize = isalloc(ptr, config_prof);
+		usize = isalloc(tsd_tsdn(tsd), ptr, config_prof);
 		prof_free(tsd, ptr, usize);
 	} else if (config_stats || config_valgrind)
-		usize = isalloc(ptr, config_prof);
+		usize = isalloc(tsd_tsdn(tsd), ptr, config_prof);
 	if (config_stats)
 		*tsd_thread_deallocatedp_get(tsd) += usize;
-	if (config_valgrind && unlikely(in_valgrind))
-		rzsize = p2rz(ptr);
-	iqalloc(tsd, ptr, tcache);
-	JEMALLOC_VALGRIND_FREE(ptr, rzsize);
+
+	if (likely(!slow_path))
+		iqalloc(tsd, ptr, tcache, false);
+	else {
+		if (config_valgrind && unlikely(in_valgrind))
+			rzsize = p2rz(tsd_tsdn(tsd), ptr);
+		iqalloc(tsd, ptr, tcache, true);
+		JEMALLOC_VALGRIND_FREE(ptr, rzsize);
+	}
 }
 
 JEMALLOC_INLINE_C void
-isfree(tsd_t *tsd, void *ptr, size_t usize, tcache_t *tcache)
+isfree(tsd_t *tsd, void *ptr, size_t usize, tcache_t *tcache, bool slow_path)
 {
 	UNUSED size_t rzsize JEMALLOC_CC_SILENCE_INIT(0);
 
+	witness_assert_lockless(tsd_tsdn(tsd));
+
 	assert(ptr != NULL);
 	assert(malloc_initialized() || IS_INITIALIZER);
 
@@ -1759,8 +1911,8 @@ isfree(tsd_t *tsd, void *ptr, size_t usize, tcache_t *tcache)
 	if (config_stats)
 		*tsd_thread_deallocatedp_get(tsd) += usize;
 	if (config_valgrind && unlikely(in_valgrind))
-		rzsize = p2rz(ptr);
-	isqalloc(tsd, ptr, usize, tcache);
+		rzsize = p2rz(tsd_tsdn(tsd), ptr);
+	isqalloc(tsd, ptr, usize, tcache, slow_path);
 	JEMALLOC_VALGRIND_FREE(ptr, rzsize);
 }
 
@@ -1770,44 +1922,57 @@ JEMALLOC_ALLOC_SIZE(2)
 je_realloc(void *ptr, size_t size)
 {
 	void *ret;
-	tsd_t *tsd JEMALLOC_CC_SILENCE_INIT(NULL);
+	tsdn_t *tsdn JEMALLOC_CC_SILENCE_INIT(NULL);
 	size_t usize JEMALLOC_CC_SILENCE_INIT(0);
 	size_t old_usize = 0;
 	UNUSED size_t old_rzsize JEMALLOC_CC_SILENCE_INIT(0);
 
 	if (unlikely(size == 0)) {
 		if (ptr != NULL) {
+			tsd_t *tsd;
+
 			/* realloc(ptr, 0) is equivalent to free(ptr). */
 			UTRACE(ptr, 0, 0);
 			tsd = tsd_fetch();
-			ifree(tsd, ptr, tcache_get(tsd, false));
+			ifree(tsd, ptr, tcache_get(tsd, false), true);
 			return (NULL);
 		}
 		size = 1;
 	}
 
 	if (likely(ptr != NULL)) {
+		tsd_t *tsd;
+
 		assert(malloc_initialized() || IS_INITIALIZER);
 		malloc_thread_init();
 		tsd = tsd_fetch();
 
-		old_usize = isalloc(ptr, config_prof);
-		if (config_valgrind && unlikely(in_valgrind))
-			old_rzsize = config_prof ? p2rz(ptr) : u2rz(old_usize);
+		witness_assert_lockless(tsd_tsdn(tsd));
+
+		old_usize = isalloc(tsd_tsdn(tsd), ptr, config_prof);
+		if (config_valgrind && unlikely(in_valgrind)) {
+			old_rzsize = config_prof ? p2rz(tsd_tsdn(tsd), ptr) :
+			    u2rz(old_usize);
+		}
 
 		if (config_prof && opt_prof) {
 			usize = s2u(size);
-			ret = unlikely(usize == 0) ? NULL : irealloc_prof(tsd,
-			    ptr, old_usize, usize);
+			ret = unlikely(usize == 0 || usize > HUGE_MAXCLASS) ?
+			    NULL : irealloc_prof(tsd, ptr, old_usize, usize);
 		} else {
 			if (config_stats || (config_valgrind &&
 			    unlikely(in_valgrind)))
 				usize = s2u(size);
 			ret = iralloc(tsd, ptr, old_usize, size, 0, false);
 		}
+		tsdn = tsd_tsdn(tsd);
 	} else {
 		/* realloc(NULL, size) is equivalent to malloc(size). */
-		ret = imalloc_body(size, &tsd, &usize);
+		if (likely(!malloc_slow))
+			ret = ialloc_body(size, false, &tsdn, &usize, false);
+		else
+			ret = ialloc_body(size, false, &tsdn, &usize, true);
+		assert(!tsdn_null(tsdn) || ret == NULL);
 	}
 
 	if (unlikely(ret == NULL)) {
@@ -1819,13 +1984,17 @@ je_realloc(void *ptr, size_t size)
 		set_errno(ENOMEM);
 	}
 	if (config_stats && likely(ret != NULL)) {
-		assert(usize == isalloc(ret, config_prof));
+		tsd_t *tsd;
+
+		assert(usize == isalloc(tsdn, ret, config_prof));
+		tsd = tsdn_tsd(tsdn);
 		*tsd_thread_allocatedp_get(tsd) += usize;
 		*tsd_thread_deallocatedp_get(tsd) += old_usize;
 	}
 	UTRACE(ptr, size, ret);
-	JEMALLOC_VALGRIND_REALLOC(true, ret, usize, true, ptr, old_usize,
-	    old_rzsize, true, false);
+	JEMALLOC_VALGRIND_REALLOC(maybe, tsdn, ret, usize, maybe, ptr,
+	    old_usize, old_rzsize, maybe, false);
+	witness_assert_lockless(tsdn);
 	return (ret);
 }
 
@@ -1836,7 +2005,12 @@ je_free(void *ptr)
 	UTRACE(ptr, 0, 0);
 	if (likely(ptr != NULL)) {
 		tsd_t *tsd = tsd_fetch();
-		ifree(tsd, ptr, tcache_get(tsd, false));
+		witness_assert_lockless(tsd_tsdn(tsd));
+		if (likely(!malloc_slow))
+			ifree(tsd, ptr, tcache_get(tsd, false), false);
+		else
+			ifree(tsd, ptr, tcache_get(tsd, false), true);
+		witness_assert_lockless(tsd_tsdn(tsd));
 	}
 }
 
@@ -1857,7 +2031,6 @@ je_memalign(size_t alignment, size_t size)
 	void *ret JEMALLOC_CC_SILENCE_INIT(NULL);
 	if (unlikely(imemalign(&ret, alignment, size, 1) != 0))
 		ret = NULL;
-	JEMALLOC_VALGRIND_MALLOC(ret != NULL, ret, size, false);
 	return (ret);
 }
 #endif
@@ -1871,7 +2044,6 @@ je_valloc(size_t size)
 	void *ret JEMALLOC_CC_SILENCE_INIT(NULL);
 	if (unlikely(imemalign(&ret, PAGE, size, 1) != 0))
 		ret = NULL;
-	JEMALLOC_VALGRIND_MALLOC(ret != NULL, ret, size, false);
 	return (ret);
 }
 #endif
@@ -1901,6 +2073,29 @@ JEMALLOC_EXPORT void *(*__realloc_hook)(void *ptr, size_t size) = je_realloc;
 JEMALLOC_EXPORT void *(*__memalign_hook)(size_t alignment, size_t size) =
     je_memalign;
 # endif
+
+#ifdef CPU_COUNT
+/*
+ * To enable static linking with glibc, the libc specific malloc interface must
+ * be implemented also, so none of glibc's malloc.o functions are added to the
+ * link.
+ */
+#define	ALIAS(je_fn)	__attribute__((alias (#je_fn), used))
+/* To force macro expansion of je_ prefix before stringification. */
+#define	PREALIAS(je_fn)  ALIAS(je_fn)
+void	*__libc_malloc(size_t size) PREALIAS(je_malloc);
+void	__libc_free(void* ptr) PREALIAS(je_free);
+void	*__libc_realloc(void* ptr, size_t size) PREALIAS(je_realloc);
+void	*__libc_calloc(size_t n, size_t size) PREALIAS(je_calloc);
+void	*__libc_memalign(size_t align, size_t s) PREALIAS(je_memalign);
+void	*__libc_valloc(size_t size) PREALIAS(je_valloc);
+int	__posix_memalign(void** r, size_t a, size_t s)
+    PREALIAS(je_posix_memalign);
+#undef PREALIAS
+#undef ALIAS
+
+#endif
+
 #endif
 
 /*
@@ -1912,7 +2107,7 @@ JEMALLOC_EXPORT void *(*__memalign_hook)(size_t alignment, size_t size) =
  */
 
 JEMALLOC_ALWAYS_INLINE_C bool
-imallocx_flags_decode_hard(tsd_t *tsd, size_t size, int flags, size_t *usize,
+imallocx_flags_decode(tsd_t *tsd, size_t size, int flags, size_t *usize,
     size_t *alignment, bool *zero, tcache_t **tcache, arena_t **arena)
 {
 
@@ -1923,7 +2118,8 @@ imallocx_flags_decode_hard(tsd_t *tsd, size_t size, int flags, size_t *usize,
 		*alignment = MALLOCX_ALIGN_GET_SPECIFIED(flags);
 		*usize = sa2u(size, *alignment);
 	}
-	assert(*usize != 0);
+	if (unlikely(*usize == 0 || *usize > HUGE_MAXCLASS))
+		return (true);
 	*zero = MALLOCX_ZERO_GET(flags);
 	if ((flags & MALLOCX_TCACHE_MASK) != 0) {
 		if ((flags & MALLOCX_TCACHE_MASK) == MALLOCX_TCACHE_NONE)
@@ -1934,7 +2130,7 @@ imallocx_flags_decode_hard(tsd_t *tsd, size_t size, int flags, size_t *usize,
 		*tcache = tcache_get(tsd, true);
 	if ((flags & MALLOCX_ARENA_MASK) != 0) {
 		unsigned arena_ind = MALLOCX_ARENA_GET(flags);
-		*arena = arena_get(tsd, arena_ind, true, true);
+		*arena = arena_get(tsd_tsdn(tsd), arena_ind, true);
 		if (unlikely(*arena == NULL))
 			return (true);
 	} else
@@ -1942,59 +2138,44 @@ imallocx_flags_decode_hard(tsd_t *tsd, size_t size, int flags, size_t *usize,
 	return (false);
 }
 
-JEMALLOC_ALWAYS_INLINE_C bool
-imallocx_flags_decode(tsd_t *tsd, size_t size, int flags, size_t *usize,
-    size_t *alignment, bool *zero, tcache_t **tcache, arena_t **arena)
-{
-
-	if (likely(flags == 0)) {
-		*usize = s2u(size);
-		assert(*usize != 0);
-		*alignment = 0;
-		*zero = false;
-		*tcache = tcache_get(tsd, true);
-		*arena = NULL;
-		return (false);
-	} else {
-		return (imallocx_flags_decode_hard(tsd, size, flags, usize,
-		    alignment, zero, tcache, arena));
-	}
-}
-
 JEMALLOC_ALWAYS_INLINE_C void *
-imallocx_flags(tsd_t *tsd, size_t usize, size_t alignment, bool zero,
-    tcache_t *tcache, arena_t *arena)
+imallocx_flags(tsdn_t *tsdn, size_t usize, size_t alignment, bool zero,
+    tcache_t *tcache, arena_t *arena, bool slow_path)
 {
+	szind_t ind;
 
 	if (unlikely(alignment != 0))
-		return (ipalloct(tsd, usize, alignment, zero, tcache, arena));
-	if (unlikely(zero))
-		return (icalloct(tsd, usize, tcache, arena));
-	return (imalloct(tsd, usize, tcache, arena));
+		return (ipalloct(tsdn, usize, alignment, zero, tcache, arena));
+	ind = size2index(usize);
+	assert(ind < NSIZES);
+	return (iallocztm(tsdn, usize, ind, zero, tcache, false, arena,
+	    slow_path));
 }
 
 static void *
-imallocx_prof_sample(tsd_t *tsd, size_t usize, size_t alignment, bool zero,
-    tcache_t *tcache, arena_t *arena)
+imallocx_prof_sample(tsdn_t *tsdn, size_t usize, size_t alignment, bool zero,
+    tcache_t *tcache, arena_t *arena, bool slow_path)
 {
 	void *p;
 
 	if (usize <= SMALL_MAXCLASS) {
 		assert(((alignment == 0) ? s2u(LARGE_MINCLASS) :
 		    sa2u(LARGE_MINCLASS, alignment)) == LARGE_MINCLASS);
-		p = imallocx_flags(tsd, LARGE_MINCLASS, alignment, zero, tcache,
-		    arena);
+		p = imallocx_flags(tsdn, LARGE_MINCLASS, alignment, zero,
+		    tcache, arena, slow_path);
 		if (p == NULL)
 			return (NULL);
-		arena_prof_promoted(p, usize);
-	} else
-		p = imallocx_flags(tsd, usize, alignment, zero, tcache, arena);
+		arena_prof_promoted(tsdn, p, usize);
+	} else {
+		p = imallocx_flags(tsdn, usize, alignment, zero, tcache, arena,
+		    slow_path);
+	}
 
 	return (p);
 }
 
 JEMALLOC_ALWAYS_INLINE_C void *
-imallocx_prof(tsd_t *tsd, size_t size, int flags, size_t *usize)
+imallocx_prof(tsd_t *tsd, size_t size, int flags, size_t *usize, bool slow_path)
 {
 	void *p;
 	size_t alignment;
@@ -2007,25 +2188,27 @@ imallocx_prof(tsd_t *tsd, size_t size, int flags, size_t *usize)
 	    &zero, &tcache, &arena)))
 		return (NULL);
 	tctx = prof_alloc_prep(tsd, *usize, prof_active_get_unlocked(), true);
-	if (likely((uintptr_t)tctx == (uintptr_t)1U))
-		p = imallocx_flags(tsd, *usize, alignment, zero, tcache, arena);
-	else if ((uintptr_t)tctx > (uintptr_t)1U) {
-		p = imallocx_prof_sample(tsd, *usize, alignment, zero, tcache,
-		    arena);
+	if (likely((uintptr_t)tctx == (uintptr_t)1U)) {
+		p = imallocx_flags(tsd_tsdn(tsd), *usize, alignment, zero,
+		    tcache, arena, slow_path);
+	} else if ((uintptr_t)tctx > (uintptr_t)1U) {
+		p = imallocx_prof_sample(tsd_tsdn(tsd), *usize, alignment, zero,
+		    tcache, arena, slow_path);
 	} else
 		p = NULL;
 	if (unlikely(p == NULL)) {
 		prof_alloc_rollback(tsd, tctx, true);
 		return (NULL);
 	}
-	prof_malloc(p, *usize, tctx);
+	prof_malloc(tsd_tsdn(tsd), p, *usize, tctx);
 
 	assert(alignment == 0 || ((uintptr_t)p & (alignment - 1)) == ZU(0));
 	return (p);
 }
 
 JEMALLOC_ALWAYS_INLINE_C void *
-imallocx_no_prof(tsd_t *tsd, size_t size, int flags, size_t *usize)
+imallocx_no_prof(tsd_t *tsd, size_t size, int flags, size_t *usize,
+    bool slow_path)
 {
 	void *p;
 	size_t alignment;
@@ -2033,56 +2216,78 @@ imallocx_no_prof(tsd_t *tsd, size_t size, int flags, size_t *usize)
 	tcache_t *tcache;
 	arena_t *arena;
 
-	if (likely(flags == 0)) {
-		if (config_stats || (config_valgrind && unlikely(in_valgrind)))
-			*usize = s2u(size);
-		return (imalloc(tsd, size));
-	}
-
-	if (unlikely(imallocx_flags_decode_hard(tsd, size, flags, usize,
-	    &alignment, &zero, &tcache, &arena)))
+	if (unlikely(imallocx_flags_decode(tsd, size, flags, usize, &alignment,
+	    &zero, &tcache, &arena)))
 		return (NULL);
-	p = imallocx_flags(tsd, *usize, alignment, zero, tcache, arena);
+	p = imallocx_flags(tsd_tsdn(tsd), *usize, alignment, zero, tcache,
+	    arena, slow_path);
 	assert(alignment == 0 || ((uintptr_t)p & (alignment - 1)) == ZU(0));
 	return (p);
 }
 
+/* This function guarantees that *tsdn is non-NULL on success. */
+JEMALLOC_ALWAYS_INLINE_C void *
+imallocx_body(size_t size, int flags, tsdn_t **tsdn, size_t *usize,
+    bool slow_path)
+{
+	tsd_t *tsd;
+
+	if (slow_path && unlikely(malloc_init())) {
+		*tsdn = NULL;
+		return (NULL);
+	}
+
+	tsd = tsd_fetch();
+	*tsdn = tsd_tsdn(tsd);
+	witness_assert_lockless(tsd_tsdn(tsd));
+
+	if (likely(flags == 0)) {
+		szind_t ind = size2index(size);
+		if (unlikely(ind >= NSIZES))
+			return (NULL);
+		if (config_stats || (config_prof && opt_prof) || (slow_path &&
+		    config_valgrind && unlikely(in_valgrind))) {
+			*usize = index2size(ind);
+			assert(*usize > 0 && *usize <= HUGE_MAXCLASS);
+		}
+
+		if (config_prof && opt_prof) {
+			return (ialloc_prof(tsd, *usize, ind, false,
+			    slow_path));
+		}
+
+		return (ialloc(tsd, size, ind, false, slow_path));
+	}
+
+	if (config_prof && opt_prof)
+		return (imallocx_prof(tsd, size, flags, usize, slow_path));
+
+	return (imallocx_no_prof(tsd, size, flags, usize, slow_path));
+}
+
 JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN
 void JEMALLOC_NOTHROW *
 JEMALLOC_ATTR(malloc) JEMALLOC_ALLOC_SIZE(1)
 je_mallocx(size_t size, int flags)
 {
-	tsd_t *tsd;
+	tsdn_t *tsdn;
 	void *p;
 	size_t usize;
 
 	assert(size != 0);
 
-	if (unlikely(malloc_init()))
-		goto label_oom;
-	tsd = tsd_fetch();
-
-	if (config_prof && opt_prof)
-		p = imallocx_prof(tsd, size, flags, &usize);
-	else
-		p = imallocx_no_prof(tsd, size, flags, &usize);
-	if (unlikely(p == NULL))
-		goto label_oom;
-
-	if (config_stats) {
-		assert(usize == isalloc(p, config_prof));
-		*tsd_thread_allocatedp_get(tsd) += usize;
+	if (likely(!malloc_slow)) {
+		p = imallocx_body(size, flags, &tsdn, &usize, false);
+		ialloc_post_check(p, tsdn, usize, "mallocx", false, false);
+	} else {
+		p = imallocx_body(size, flags, &tsdn, &usize, true);
+		ialloc_post_check(p, tsdn, usize, "mallocx", false, true);
+		UTRACE(0, size, p);
+		JEMALLOC_VALGRIND_MALLOC(p != NULL, tsdn, p, usize,
+		    MALLOCX_ZERO_GET(flags));
 	}
-	UTRACE(0, size, p);
-	JEMALLOC_VALGRIND_MALLOC(true, p, usize, MALLOCX_ZERO_GET(flags));
+
 	return (p);
-label_oom:
-	if (config_xmalloc && unlikely(opt_xmalloc)) {
-		malloc_write("<jemalloc>: Error in mallocx(): out of memory\n");
-		abort();
-	}
-	UTRACE(0, size, 0);
-	return (NULL);
 }
 
 static void *
@@ -2099,7 +2304,7 @@ irallocx_prof_sample(tsd_t *tsd, void *old_ptr, size_t old_usize,
 		    zero, tcache, arena);
 		if (p == NULL)
 			return (NULL);
-		arena_prof_promoted(p, usize);
+		arena_prof_promoted(tsd_tsdn(tsd), p, usize);
 	} else {
 		p = iralloct(tsd, old_ptr, old_usize, usize, alignment, zero,
 		    tcache, arena);
@@ -2118,8 +2323,8 @@ irallocx_prof(tsd_t *tsd, void *old_ptr, size_t old_usize, size_t size,
 	prof_tctx_t *old_tctx, *tctx;
 
 	prof_active = prof_active_get_unlocked();
-	old_tctx = prof_tctx_get(old_ptr);
-	tctx = prof_alloc_prep(tsd, *usize, prof_active, true);
+	old_tctx = prof_tctx_get(tsd_tsdn(tsd), old_ptr);
+	tctx = prof_alloc_prep(tsd, *usize, prof_active, false);
 	if (unlikely((uintptr_t)tctx != (uintptr_t)1U)) {
 		p = irallocx_prof_sample(tsd, old_ptr, old_usize, *usize,
 		    alignment, zero, tcache, arena, tctx);
@@ -2128,7 +2333,7 @@ irallocx_prof(tsd_t *tsd, void *old_ptr, size_t old_usize, size_t size,
 		    tcache, arena);
 	}
 	if (unlikely(p == NULL)) {
-		prof_alloc_rollback(tsd, tctx, true);
+		prof_alloc_rollback(tsd, tctx, false);
 		return (NULL);
 	}
 
@@ -2141,9 +2346,9 @@ irallocx_prof(tsd_t *tsd, void *old_ptr, size_t old_usize, size_t size,
 		 * be the same as the current usize because of in-place large
 		 * reallocation.  Therefore, query the actual value of usize.
 		 */
-		*usize = isalloc(p, config_prof);
+		*usize = isalloc(tsd_tsdn(tsd), p, config_prof);
 	}
-	prof_realloc(tsd, p, *usize, tctx, prof_active, true, old_ptr,
+	prof_realloc(tsd, p, *usize, tctx, prof_active, false, old_ptr,
 	    old_usize, old_tctx);
 
 	return (p);
@@ -2169,10 +2374,11 @@ je_rallocx(void *ptr, size_t size, int flags)
 	assert(malloc_initialized() || IS_INITIALIZER);
 	malloc_thread_init();
 	tsd = tsd_fetch();
+	witness_assert_lockless(tsd_tsdn(tsd));
 
 	if (unlikely((flags & MALLOCX_ARENA_MASK) != 0)) {
 		unsigned arena_ind = MALLOCX_ARENA_GET(flags);
-		arena = arena_get(tsd, arena_ind, true, true);
+		arena = arena_get(tsd_tsdn(tsd), arena_ind, true);
 		if (unlikely(arena == NULL))
 			goto label_oom;
 	} else
@@ -2186,13 +2392,14 @@ je_rallocx(void *ptr, size_t size, int flags)
 	} else
 		tcache = tcache_get(tsd, true);
 
-	old_usize = isalloc(ptr, config_prof);
+	old_usize = isalloc(tsd_tsdn(tsd), ptr, config_prof);
 	if (config_valgrind && unlikely(in_valgrind))
 		old_rzsize = u2rz(old_usize);
 
 	if (config_prof && opt_prof) {
 		usize = (alignment == 0) ? s2u(size) : sa2u(size, alignment);
-		assert(usize != 0);
+		if (unlikely(usize == 0 || usize > HUGE_MAXCLASS))
+			goto label_oom;
 		p = irallocx_prof(tsd, ptr, old_usize, size, alignment, &usize,
 		    zero, tcache, arena);
 		if (unlikely(p == NULL))
@@ -2203,7 +2410,7 @@ je_rallocx(void *ptr, size_t size, int flags)
 		if (unlikely(p == NULL))
 			goto label_oom;
 		if (config_stats || (config_valgrind && unlikely(in_valgrind)))
-			usize = isalloc(p, config_prof);
+			usize = isalloc(tsd_tsdn(tsd), p, config_prof);
 	}
 	assert(alignment == 0 || ((uintptr_t)p & (alignment - 1)) == ZU(0));
 
@@ -2212,8 +2419,9 @@ je_rallocx(void *ptr, size_t size, int flags)
 		*tsd_thread_deallocatedp_get(tsd) += old_usize;
 	}
 	UTRACE(ptr, size, p);
-	JEMALLOC_VALGRIND_REALLOC(true, p, usize, false, ptr, old_usize,
-	    old_rzsize, false, zero);
+	JEMALLOC_VALGRIND_REALLOC(maybe, tsd_tsdn(tsd), p, usize, no, ptr,
+	    old_usize, old_rzsize, no, zero);
+	witness_assert_lockless(tsd_tsdn(tsd));
 	return (p);
 label_oom:
 	if (config_xmalloc && unlikely(opt_xmalloc)) {
@@ -2221,31 +2429,33 @@ label_oom:
 		abort();
 	}
 	UTRACE(ptr, size, 0);
+	witness_assert_lockless(tsd_tsdn(tsd));
 	return (NULL);
 }
 
 JEMALLOC_ALWAYS_INLINE_C size_t
-ixallocx_helper(void *ptr, size_t old_usize, size_t size, size_t extra,
-    size_t alignment, bool zero)
+ixallocx_helper(tsdn_t *tsdn, void *ptr, size_t old_usize, size_t size,
+    size_t extra, size_t alignment, bool zero)
 {
 	size_t usize;
 
-	if (ixalloc(ptr, old_usize, size, extra, alignment, zero))
+	if (ixalloc(tsdn, ptr, old_usize, size, extra, alignment, zero))
 		return (old_usize);
-	usize = isalloc(ptr, config_prof);
+	usize = isalloc(tsdn, ptr, config_prof);
 
 	return (usize);
 }
 
 static size_t
-ixallocx_prof_sample(void *ptr, size_t old_usize, size_t size, size_t extra,
-    size_t alignment, bool zero, prof_tctx_t *tctx)
+ixallocx_prof_sample(tsdn_t *tsdn, void *ptr, size_t old_usize, size_t size,
+    size_t extra, size_t alignment, bool zero, prof_tctx_t *tctx)
 {
 	size_t usize;
 
 	if (tctx == NULL)
 		return (old_usize);
-	usize = ixallocx_helper(ptr, old_usize, size, extra, alignment, zero);
+	usize = ixallocx_helper(tsdn, ptr, old_usize, size, extra, alignment,
+	    zero);
 
 	return (usize);
 }
@@ -2259,23 +2469,36 @@ ixallocx_prof(tsd_t *tsd, void *ptr, size_t old_usize, size_t size,
 	prof_tctx_t *old_tctx, *tctx;
 
 	prof_active = prof_active_get_unlocked();
-	old_tctx = prof_tctx_get(ptr);
+	old_tctx = prof_tctx_get(tsd_tsdn(tsd), ptr);
 	/*
 	 * usize isn't knowable before ixalloc() returns when extra is non-zero.
 	 * Therefore, compute its maximum possible value and use that in
 	 * prof_alloc_prep() to decide whether to capture a backtrace.
 	 * prof_realloc() will use the actual usize to decide whether to sample.
 	 */
-	usize_max = (alignment == 0) ? s2u(size+extra) : sa2u(size+extra,
-	    alignment);
-	assert(usize_max != 0);
+	if (alignment == 0) {
+		usize_max = s2u(size+extra);
+		assert(usize_max > 0 && usize_max <= HUGE_MAXCLASS);
+	} else {
+		usize_max = sa2u(size+extra, alignment);
+		if (unlikely(usize_max == 0 || usize_max > HUGE_MAXCLASS)) {
+			/*
+			 * usize_max is out of range, and chances are that
+			 * allocation will fail, but use the maximum possible
+			 * value and carry on with prof_alloc_prep(), just in
+			 * case allocation succeeds.
+			 */
+			usize_max = HUGE_MAXCLASS;
+		}
+	}
 	tctx = prof_alloc_prep(tsd, usize_max, prof_active, false);
+
 	if (unlikely((uintptr_t)tctx != (uintptr_t)1U)) {
-		usize = ixallocx_prof_sample(ptr, old_usize, size, extra,
-		    alignment, zero, tctx);
+		usize = ixallocx_prof_sample(tsd_tsdn(tsd), ptr, old_usize,
+		    size, extra, alignment, zero, tctx);
 	} else {
-		usize = ixallocx_helper(ptr, old_usize, size, extra, alignment,
-		    zero);
+		usize = ixallocx_helper(tsd_tsdn(tsd), ptr, old_usize, size,
+		    extra, alignment, zero);
 	}
 	if (usize == old_usize) {
 		prof_alloc_rollback(tsd, tctx, false);
@@ -2302,18 +2525,25 @@ je_xallocx(void *ptr, size_t size, size_t extra, int flags)
 	assert(malloc_initialized() || IS_INITIALIZER);
 	malloc_thread_init();
 	tsd = tsd_fetch();
+	witness_assert_lockless(tsd_tsdn(tsd));
 
-	old_usize = isalloc(ptr, config_prof);
+	old_usize = isalloc(tsd_tsdn(tsd), ptr, config_prof);
 
-	/* Clamp extra if necessary to avoid (size + extra) overflow. */
-	if (unlikely(size + extra > HUGE_MAXCLASS)) {
-		/* Check for size overflow. */
-		if (unlikely(size > HUGE_MAXCLASS)) {
-			usize = old_usize;
-			goto label_not_resized;
-		}
-		extra = HUGE_MAXCLASS - size;
+	/*
+	 * The API explicitly absolves itself of protecting against (size +
+	 * extra) numerical overflow, but we may need to clamp extra to avoid
+	 * exceeding HUGE_MAXCLASS.
+	 *
+	 * Ordinarily, size limit checking is handled deeper down, but here we
+	 * have to check as part of (size + extra) clamping, since we need the
+	 * clamped value in the above helper functions.
+	 */
+	if (unlikely(size > HUGE_MAXCLASS)) {
+		usize = old_usize;
+		goto label_not_resized;
 	}
+	if (unlikely(HUGE_MAXCLASS - size < extra))
+		extra = HUGE_MAXCLASS - size;
 
 	if (config_valgrind && unlikely(in_valgrind))
 		old_rzsize = u2rz(old_usize);
@@ -2322,8 +2552,8 @@ je_xallocx(void *ptr, size_t size, size_t extra, int flags)
 		usize = ixallocx_prof(tsd, ptr, old_usize, size, extra,
 		    alignment, zero);
 	} else {
-		usize = ixallocx_helper(ptr, old_usize, size, extra, alignment,
-		    zero);
+		usize = ixallocx_helper(tsd_tsdn(tsd), ptr, old_usize, size,
+		    extra, alignment, zero);
 	}
 	if (unlikely(usize == old_usize))
 		goto label_not_resized;
@@ -2332,10 +2562,11 @@ je_xallocx(void *ptr, size_t size, size_t extra, int flags)
 		*tsd_thread_allocatedp_get(tsd) += usize;
 		*tsd_thread_deallocatedp_get(tsd) += old_usize;
 	}
-	JEMALLOC_VALGRIND_REALLOC(false, ptr, usize, false, ptr, old_usize,
-	    old_rzsize, false, zero);
+	JEMALLOC_VALGRIND_REALLOC(no, tsd_tsdn(tsd), ptr, usize, no, ptr,
+	    old_usize, old_rzsize, no, zero);
 label_not_resized:
 	UTRACE(ptr, size, ptr);
+	witness_assert_lockless(tsd_tsdn(tsd));
 	return (usize);
 }
 
@@ -2344,15 +2575,20 @@ JEMALLOC_ATTR(pure)
 je_sallocx(const void *ptr, int flags)
 {
 	size_t usize;
+	tsdn_t *tsdn;
 
 	assert(malloc_initialized() || IS_INITIALIZER);
 	malloc_thread_init();
 
+	tsdn = tsdn_fetch();
+	witness_assert_lockless(tsdn);
+
 	if (config_ivsalloc)
-		usize = ivsalloc(ptr, config_prof);
+		usize = ivsalloc(tsdn, ptr, config_prof);
 	else
-		usize = isalloc(ptr, config_prof);
+		usize = isalloc(tsdn, ptr, config_prof);
 
+	witness_assert_lockless(tsdn);
 	return (usize);
 }
 
@@ -2366,6 +2602,7 @@ je_dallocx(void *ptr, int flags)
 	assert(malloc_initialized() || IS_INITIALIZER);
 
 	tsd = tsd_fetch();
+	witness_assert_lockless(tsd_tsdn(tsd));
 	if (unlikely((flags & MALLOCX_TCACHE_MASK) != 0)) {
 		if ((flags & MALLOCX_TCACHE_MASK) == MALLOCX_TCACHE_NONE)
 			tcache = NULL;
@@ -2375,19 +2612,25 @@ je_dallocx(void *ptr, int flags)
 		tcache = tcache_get(tsd, false);
 
 	UTRACE(ptr, 0, 0);
-	ifree(tsd_fetch(), ptr, tcache);
+	if (likely(!malloc_slow))
+		ifree(tsd, ptr, tcache, false);
+	else
+		ifree(tsd, ptr, tcache, true);
+	witness_assert_lockless(tsd_tsdn(tsd));
 }
 
 JEMALLOC_ALWAYS_INLINE_C size_t
-inallocx(size_t size, int flags)
+inallocx(tsdn_t *tsdn, size_t size, int flags)
 {
 	size_t usize;
 
+	witness_assert_lockless(tsdn);
+
 	if (likely((flags & MALLOCX_LG_ALIGN_MASK) == 0))
 		usize = s2u(size);
 	else
 		usize = sa2u(size, MALLOCX_ALIGN_GET_SPECIFIED(flags));
-	assert(usize != 0);
+	witness_assert_lockless(tsdn);
 	return (usize);
 }
 
@@ -2400,10 +2643,11 @@ je_sdallocx(void *ptr, size_t size, int flags)
 
 	assert(ptr != NULL);
 	assert(malloc_initialized() || IS_INITIALIZER);
-	usize = inallocx(size, flags);
-	assert(usize == isalloc(ptr, config_prof));
-
 	tsd = tsd_fetch();
+	usize = inallocx(tsd_tsdn(tsd), size, flags);
+	assert(usize == isalloc(tsd_tsdn(tsd), ptr, config_prof));
+
+	witness_assert_lockless(tsd_tsdn(tsd));
 	if (unlikely((flags & MALLOCX_TCACHE_MASK) != 0)) {
 		if ((flags & MALLOCX_TCACHE_MASK) == MALLOCX_TCACHE_NONE)
 			tcache = NULL;
@@ -2413,75 +2657,116 @@ je_sdallocx(void *ptr, size_t size, int flags)
 		tcache = tcache_get(tsd, false);
 
 	UTRACE(ptr, 0, 0);
-	isfree(tsd, ptr, usize, tcache);
+	if (likely(!malloc_slow))
+		isfree(tsd, ptr, usize, tcache, false);
+	else
+		isfree(tsd, ptr, usize, tcache, true);
+	witness_assert_lockless(tsd_tsdn(tsd));
 }
 
 JEMALLOC_EXPORT size_t JEMALLOC_NOTHROW
 JEMALLOC_ATTR(pure)
 je_nallocx(size_t size, int flags)
 {
+	size_t usize;
+	tsdn_t *tsdn;
 
 	assert(size != 0);
 
 	if (unlikely(malloc_init()))
 		return (0);
 
-	return (inallocx(size, flags));
+	tsdn = tsdn_fetch();
+	witness_assert_lockless(tsdn);
+
+	usize = inallocx(tsdn, size, flags);
+	if (unlikely(usize > HUGE_MAXCLASS))
+		return (0);
+
+	witness_assert_lockless(tsdn);
+	return (usize);
 }
 
 JEMALLOC_EXPORT int JEMALLOC_NOTHROW
 je_mallctl(const char *name, void *oldp, size_t *oldlenp, void *newp,
     size_t newlen)
 {
+	int ret;
+	tsd_t *tsd;
 
 	if (unlikely(malloc_init()))
 		return (EAGAIN);
 
-	return (ctl_byname(name, oldp, oldlenp, newp, newlen));
+	tsd = tsd_fetch();
+	witness_assert_lockless(tsd_tsdn(tsd));
+	ret = ctl_byname(tsd, name, oldp, oldlenp, newp, newlen);
+	witness_assert_lockless(tsd_tsdn(tsd));
+	return (ret);
 }
 
 JEMALLOC_EXPORT int JEMALLOC_NOTHROW
 je_mallctlnametomib(const char *name, size_t *mibp, size_t *miblenp)
 {
+	int ret;
+	tsdn_t *tsdn;
 
 	if (unlikely(malloc_init()))
 		return (EAGAIN);
 
-	return (ctl_nametomib(name, mibp, miblenp));
+	tsdn = tsdn_fetch();
+	witness_assert_lockless(tsdn);
+	ret = ctl_nametomib(tsdn, name, mibp, miblenp);
+	witness_assert_lockless(tsdn);
+	return (ret);
 }
 
 JEMALLOC_EXPORT int JEMALLOC_NOTHROW
 je_mallctlbymib(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
   void *newp, size_t newlen)
 {
+	int ret;
+	tsd_t *tsd;
 
 	if (unlikely(malloc_init()))
 		return (EAGAIN);
 
-	return (ctl_bymib(mib, miblen, oldp, oldlenp, newp, newlen));
+	tsd = tsd_fetch();
+	witness_assert_lockless(tsd_tsdn(tsd));
+	ret = ctl_bymib(tsd, mib, miblen, oldp, oldlenp, newp, newlen);
+	witness_assert_lockless(tsd_tsdn(tsd));
+	return (ret);
 }
 
 JEMALLOC_EXPORT void JEMALLOC_NOTHROW
 je_malloc_stats_print(void (*write_cb)(void *, const char *), void *cbopaque,
     const char *opts)
 {
+	tsdn_t *tsdn;
 
+	tsdn = tsdn_fetch();
+	witness_assert_lockless(tsdn);
 	stats_print(write_cb, cbopaque, opts);
+	witness_assert_lockless(tsdn);
 }
 
 JEMALLOC_EXPORT size_t JEMALLOC_NOTHROW
 je_malloc_usable_size(JEMALLOC_USABLE_SIZE_CONST void *ptr)
 {
 	size_t ret;
+	tsdn_t *tsdn;
 
 	assert(malloc_initialized() || IS_INITIALIZER);
 	malloc_thread_init();
 
+	tsdn = tsdn_fetch();
+	witness_assert_lockless(tsdn);
+
 	if (config_ivsalloc)
-		ret = ivsalloc(ptr, config_prof);
+		ret = ivsalloc(tsdn, ptr, config_prof);
 	else
-		ret = (ptr == NULL) ? 0 : isalloc(ptr, config_prof);
+		ret = (ptr == NULL) ? 0 : isalloc(tsdn, ptr, config_prof);
 
+	witness_assert_lockless(tsdn);
 	return (ret);
 }
 
@@ -2507,6 +2792,7 @@ je_malloc_usable_size(JEMALLOC_USABLE_SIZE_CONST void *ptr)
  * to trigger the deadlock described above, but doing so would involve forking
  * via a library constructor that runs before jemalloc's runs.
  */
+#ifndef JEMALLOC_JET
 JEMALLOC_ATTR(constructor)
 static void
 jemalloc_constructor(void)
@@ -2514,6 +2800,7 @@ jemalloc_constructor(void)
 
 	malloc_init();
 }
+#endif
 
 #ifndef JEMALLOC_MUTEX_INIT_CB
 void
@@ -2523,7 +2810,9 @@ JEMALLOC_EXPORT void
 _malloc_prefork(void)
 #endif
 {
-	unsigned i;
+	tsd_t *tsd;
+	unsigned i, j, narenas;
+	arena_t *arena;
 
 #ifdef JEMALLOC_MUTEX_INIT_CB
 	if (!malloc_initialized())
@@ -2531,16 +2820,40 @@ _malloc_prefork(void)
 #endif
 	assert(malloc_initialized());
 
+	tsd = tsd_fetch();
+
+	narenas = narenas_total_get();
+
+	witness_prefork(tsd);
 	/* Acquire all mutexes in a safe order. */
-	ctl_prefork();
-	prof_prefork();
-	malloc_mutex_prefork(&arenas_lock);
-	for (i = 0; i < narenas_total; i++) {
-		if (arenas[i] != NULL)
-			arena_prefork(arenas[i]);
+	ctl_prefork(tsd_tsdn(tsd));
+	malloc_mutex_prefork(tsd_tsdn(tsd), &arenas_lock);
+	prof_prefork0(tsd_tsdn(tsd));
+	for (i = 0; i < 3; i++) {
+		for (j = 0; j < narenas; j++) {
+			if ((arena = arena_get(tsd_tsdn(tsd), j, false)) !=
+			    NULL) {
+				switch (i) {
+				case 0:
+					arena_prefork0(tsd_tsdn(tsd), arena);
+					break;
+				case 1:
+					arena_prefork1(tsd_tsdn(tsd), arena);
+					break;
+				case 2:
+					arena_prefork2(tsd_tsdn(tsd), arena);
+					break;
+				default: not_reached();
+				}
+			}
+		}
+	}
+	base_prefork(tsd_tsdn(tsd));
+	for (i = 0; i < narenas; i++) {
+		if ((arena = arena_get(tsd_tsdn(tsd), i, false)) != NULL)
+			arena_prefork3(tsd_tsdn(tsd), arena);
 	}
-	chunk_prefork();
-	base_prefork();
+	prof_prefork1(tsd_tsdn(tsd));
 }
 
 #ifndef JEMALLOC_MUTEX_INIT_CB
@@ -2551,7 +2864,8 @@ JEMALLOC_EXPORT void
 _malloc_postfork(void)
 #endif
 {
-	unsigned i;
+	tsd_t *tsd;
+	unsigned i, narenas;
 
 #ifdef JEMALLOC_MUTEX_INIT_CB
 	if (!malloc_initialized())
@@ -2559,35 +2873,44 @@ _malloc_postfork(void)
 #endif
 	assert(malloc_initialized());
 
+	tsd = tsd_fetch();
+
+	witness_postfork_parent(tsd);
 	/* Release all mutexes, now that fork() has completed. */
-	base_postfork_parent();
-	chunk_postfork_parent();
-	for (i = 0; i < narenas_total; i++) {
-		if (arenas[i] != NULL)
-			arena_postfork_parent(arenas[i]);
+	base_postfork_parent(tsd_tsdn(tsd));
+	for (i = 0, narenas = narenas_total_get(); i < narenas; i++) {
+		arena_t *arena;
+
+		if ((arena = arena_get(tsd_tsdn(tsd), i, false)) != NULL)
+			arena_postfork_parent(tsd_tsdn(tsd), arena);
 	}
-	malloc_mutex_postfork_parent(&arenas_lock);
-	prof_postfork_parent();
-	ctl_postfork_parent();
+	prof_postfork_parent(tsd_tsdn(tsd));
+	malloc_mutex_postfork_parent(tsd_tsdn(tsd), &arenas_lock);
+	ctl_postfork_parent(tsd_tsdn(tsd));
 }
 
 void
 jemalloc_postfork_child(void)
 {
-	unsigned i;
+	tsd_t *tsd;
+	unsigned i, narenas;
 
 	assert(malloc_initialized());
 
+	tsd = tsd_fetch();
+
+	witness_postfork_child(tsd);
 	/* Release all mutexes, now that fork() has completed. */
-	base_postfork_child();
-	chunk_postfork_child();
-	for (i = 0; i < narenas_total; i++) {
-		if (arenas[i] != NULL)
-			arena_postfork_child(arenas[i]);
+	base_postfork_child(tsd_tsdn(tsd));
+	for (i = 0, narenas = narenas_total_get(); i < narenas; i++) {
+		arena_t *arena;
+
+		if ((arena = arena_get(tsd_tsdn(tsd), i, false)) != NULL)
+			arena_postfork_child(tsd_tsdn(tsd), arena);
 	}
-	malloc_mutex_postfork_child(&arenas_lock);
-	prof_postfork_child();
-	ctl_postfork_child();
+	prof_postfork_child(tsd_tsdn(tsd));
+	malloc_mutex_postfork_child(tsd_tsdn(tsd), &arenas_lock);
+	ctl_postfork_child(tsd_tsdn(tsd));
 }
 
 /******************************************************************************/
@@ -2607,9 +2930,10 @@ je_get_defrag_hint(void* ptr, int *bin_util, int *run_util) {
         if (likely((mapbits & CHUNK_MAP_LARGE) == 0)) { /* indication that this is not a LARGE alloc */
             arena_t *arena = extent_node_arena_get(&chunk->node);
             size_t rpages_ind = pageind - arena_mapbits_small_runind_get(chunk, pageind);
-            arena_run_t *run = &arena_miscelm_get(chunk, rpages_ind)->run;
+            arena_run_t *run = &arena_miscelm_get_mutable(chunk, rpages_ind)->run;
             arena_bin_t *bin = &arena->bins[run->binind];
-            malloc_mutex_lock(&bin->lock);
+            tsd_t *tsd = tsd_fetch();
+            malloc_mutex_lock(tsd_tsdn(tsd), &bin->lock);
             /* runs that are in the same chunk in as the current chunk, are likely to be the next currun */
             if (chunk != (arena_chunk_t *)CHUNK_ADDR2BASE(bin->runcur)) {
                 arena_bin_info_t *bin_info = &arena_bin_info[run->binind];
@@ -2618,7 +2942,7 @@ je_get_defrag_hint(void* ptr, int *bin_util, int *run_util) {
                 *run_util = ((bin_info->nregs - run->nfree)<<16) / bin_info->nregs;
                 defrag = 1;
             }
-            malloc_mutex_unlock(&bin->lock);
+	    malloc_mutex_unlock(tsd_tsdn(tsd), &bin->lock);
         }
     }
     return defrag;
diff --git a/deps/jemalloc/src/mutex.c b/deps/jemalloc/src/mutex.c
index 2d47af976..6333e73d6 100644
--- a/deps/jemalloc/src/mutex.c
+++ b/deps/jemalloc/src/mutex.c
@@ -69,7 +69,7 @@ JEMALLOC_EXPORT int	_pthread_mutex_init_calloc_cb(pthread_mutex_t *mutex,
 #endif
 
 bool
-malloc_mutex_init(malloc_mutex_t *mutex)
+malloc_mutex_init(malloc_mutex_t *mutex, const char *name, witness_rank_t rank)
 {
 
 #ifdef _WIN32
@@ -80,6 +80,8 @@ malloc_mutex_init(malloc_mutex_t *mutex)
 	    _CRT_SPINCOUNT))
 		return (true);
 #  endif
+#elif (defined(JEMALLOC_OS_UNFAIR_LOCK))
+	mutex->lock = OS_UNFAIR_LOCK_INIT;
 #elif (defined(JEMALLOC_OSSPIN))
 	mutex->lock = 0;
 #elif (defined(JEMALLOC_MUTEX_INIT_CB))
@@ -103,31 +105,34 @@ malloc_mutex_init(malloc_mutex_t *mutex)
 	}
 	pthread_mutexattr_destroy(&attr);
 #endif
+	if (config_debug)
+		witness_init(&mutex->witness, name, rank, NULL);
 	return (false);
 }
 
 void
-malloc_mutex_prefork(malloc_mutex_t *mutex)
+malloc_mutex_prefork(tsdn_t *tsdn, malloc_mutex_t *mutex)
 {
 
-	malloc_mutex_lock(mutex);
+	malloc_mutex_lock(tsdn, mutex);
 }
 
 void
-malloc_mutex_postfork_parent(malloc_mutex_t *mutex)
+malloc_mutex_postfork_parent(tsdn_t *tsdn, malloc_mutex_t *mutex)
 {
 
-	malloc_mutex_unlock(mutex);
+	malloc_mutex_unlock(tsdn, mutex);
 }
 
 void
-malloc_mutex_postfork_child(malloc_mutex_t *mutex)
+malloc_mutex_postfork_child(tsdn_t *tsdn, malloc_mutex_t *mutex)
 {
 
 #ifdef JEMALLOC_MUTEX_INIT_CB
-	malloc_mutex_unlock(mutex);
+	malloc_mutex_unlock(tsdn, mutex);
 #else
-	if (malloc_mutex_init(mutex)) {
+	if (malloc_mutex_init(mutex, mutex->witness.name,
+	    mutex->witness.rank)) {
 		malloc_printf("<jemalloc>: Error re-initializing mutex in "
 		    "child\n");
 		if (opt_abort)
@@ -137,7 +142,7 @@ malloc_mutex_postfork_child(malloc_mutex_t *mutex)
 }
 
 bool
-mutex_boot(void)
+malloc_mutex_boot(void)
 {
 
 #ifdef JEMALLOC_MUTEX_INIT_CB
diff --git a/deps/jemalloc/src/nstime.c b/deps/jemalloc/src/nstime.c
new file mode 100644
index 000000000..0948e29fa
--- /dev/null
+++ b/deps/jemalloc/src/nstime.c
@@ -0,0 +1,194 @@
+#include "jemalloc/internal/jemalloc_internal.h"
+
+#define	BILLION	UINT64_C(1000000000)
+
+void
+nstime_init(nstime_t *time, uint64_t ns)
+{
+
+	time->ns = ns;
+}
+
+void
+nstime_init2(nstime_t *time, uint64_t sec, uint64_t nsec)
+{
+
+	time->ns = sec * BILLION + nsec;
+}
+
+uint64_t
+nstime_ns(const nstime_t *time)
+{
+
+	return (time->ns);
+}
+
+uint64_t
+nstime_sec(const nstime_t *time)
+{
+
+	return (time->ns / BILLION);
+}
+
+uint64_t
+nstime_nsec(const nstime_t *time)
+{
+
+	return (time->ns % BILLION);
+}
+
+void
+nstime_copy(nstime_t *time, const nstime_t *source)
+{
+
+	*time = *source;
+}
+
+int
+nstime_compare(const nstime_t *a, const nstime_t *b)
+{
+
+	return ((a->ns > b->ns) - (a->ns < b->ns));
+}
+
+void
+nstime_add(nstime_t *time, const nstime_t *addend)
+{
+
+	assert(UINT64_MAX - time->ns >= addend->ns);
+
+	time->ns += addend->ns;
+}
+
+void
+nstime_subtract(nstime_t *time, const nstime_t *subtrahend)
+{
+
+	assert(nstime_compare(time, subtrahend) >= 0);
+
+	time->ns -= subtrahend->ns;
+}
+
+void
+nstime_imultiply(nstime_t *time, uint64_t multiplier)
+{
+
+	assert((((time->ns | multiplier) & (UINT64_MAX << (sizeof(uint64_t) <<
+	    2))) == 0) || ((time->ns * multiplier) / multiplier == time->ns));
+
+	time->ns *= multiplier;
+}
+
+void
+nstime_idivide(nstime_t *time, uint64_t divisor)
+{
+
+	assert(divisor != 0);
+
+	time->ns /= divisor;
+}
+
+uint64_t
+nstime_divide(const nstime_t *time, const nstime_t *divisor)
+{
+
+	assert(divisor->ns != 0);
+
+	return (time->ns / divisor->ns);
+}
+
+#ifdef _WIN32
+#  define NSTIME_MONOTONIC true
+static void
+nstime_get(nstime_t *time)
+{
+	FILETIME ft;
+	uint64_t ticks_100ns;
+
+	GetSystemTimeAsFileTime(&ft);
+	ticks_100ns = (((uint64_t)ft.dwHighDateTime) << 32) | ft.dwLowDateTime;
+
+	nstime_init(time, ticks_100ns * 100);
+}
+#elif JEMALLOC_HAVE_CLOCK_MONOTONIC_COARSE
+#  define NSTIME_MONOTONIC true
+static void
+nstime_get(nstime_t *time)
+{
+	struct timespec ts;
+
+	clock_gettime(CLOCK_MONOTONIC_COARSE, &ts);
+	nstime_init2(time, ts.tv_sec, ts.tv_nsec);
+}
+#elif JEMALLOC_HAVE_CLOCK_MONOTONIC
+#  define NSTIME_MONOTONIC true
+static void
+nstime_get(nstime_t *time)
+{
+	struct timespec ts;
+
+	clock_gettime(CLOCK_MONOTONIC, &ts);
+	nstime_init2(time, ts.tv_sec, ts.tv_nsec);
+}
+#elif JEMALLOC_HAVE_MACH_ABSOLUTE_TIME
+#  define NSTIME_MONOTONIC true
+static void
+nstime_get(nstime_t *time)
+{
+
+	nstime_init(time, mach_absolute_time());
+}
+#else
+#  define NSTIME_MONOTONIC false
+static void
+nstime_get(nstime_t *time)
+{
+	struct timeval tv;
+
+	gettimeofday(&tv, NULL);
+	nstime_init2(time, tv.tv_sec, tv.tv_usec * 1000);
+}
+#endif
+
+#ifdef JEMALLOC_JET
+#undef nstime_monotonic
+#define	nstime_monotonic JEMALLOC_N(n_nstime_monotonic)
+#endif
+bool
+nstime_monotonic(void)
+{
+
+	return (NSTIME_MONOTONIC);
+#undef NSTIME_MONOTONIC
+}
+#ifdef JEMALLOC_JET
+#undef nstime_monotonic
+#define	nstime_monotonic JEMALLOC_N(nstime_monotonic)
+nstime_monotonic_t *nstime_monotonic = JEMALLOC_N(n_nstime_monotonic);
+#endif
+
+#ifdef JEMALLOC_JET
+#undef nstime_update
+#define	nstime_update JEMALLOC_N(n_nstime_update)
+#endif
+bool
+nstime_update(nstime_t *time)
+{
+	nstime_t old_time;
+
+	nstime_copy(&old_time, time);
+	nstime_get(time);
+
+	/* Handle non-monotonic clocks. */
+	if (unlikely(nstime_compare(&old_time, time) > 0)) {
+		nstime_copy(time, &old_time);
+		return (true);
+	}
+
+	return (false);
+}
+#ifdef JEMALLOC_JET
+#undef nstime_update
+#define	nstime_update JEMALLOC_N(nstime_update)
+nstime_update_t *nstime_update = JEMALLOC_N(n_nstime_update);
+#endif
diff --git a/deps/jemalloc/src/pages.c b/deps/jemalloc/src/pages.c
index 83a167f67..5f0c9669d 100644
--- a/deps/jemalloc/src/pages.c
+++ b/deps/jemalloc/src/pages.c
@@ -1,29 +1,49 @@
 #define	JEMALLOC_PAGES_C_
 #include "jemalloc/internal/jemalloc_internal.h"
 
+#ifdef JEMALLOC_SYSCTL_VM_OVERCOMMIT
+#include <sys/sysctl.h>
+#endif
+
+/******************************************************************************/
+/* Data. */
+
+#ifndef _WIN32
+#  define PAGES_PROT_COMMIT (PROT_READ | PROT_WRITE)
+#  define PAGES_PROT_DECOMMIT (PROT_NONE)
+static int	mmap_flags;
+#endif
+static bool	os_overcommits;
+
 /******************************************************************************/
 
 void *
-pages_map(void *addr, size_t size)
+pages_map(void *addr, size_t size, bool *commit)
 {
 	void *ret;
 
 	assert(size != 0);
 
+	if (os_overcommits)
+		*commit = true;
+
 #ifdef _WIN32
 	/*
 	 * If VirtualAlloc can't allocate at the given address when one is
 	 * given, it fails and returns NULL.
 	 */
-	ret = VirtualAlloc(addr, size, MEM_COMMIT | MEM_RESERVE,
+	ret = VirtualAlloc(addr, size, MEM_RESERVE | (*commit ? MEM_COMMIT : 0),
 	    PAGE_READWRITE);
 #else
 	/*
 	 * We don't use MAP_FIXED here, because it can cause the *replacement*
 	 * of existing mappings, and we only want to create new mappings.
 	 */
-	ret = mmap(addr, size, PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANON,
-	    -1, 0);
+	{
+		int prot = *commit ? PAGES_PROT_COMMIT : PAGES_PROT_DECOMMIT;
+
+		ret = mmap(addr, size, prot, mmap_flags, -1, 0);
+	}
 	assert(ret != NULL);
 
 	if (ret == MAP_FAILED)
@@ -67,7 +87,8 @@ pages_unmap(void *addr, size_t size)
 }
 
 void *
-pages_trim(void *addr, size_t alloc_size, size_t leadsize, size_t size)
+pages_trim(void *addr, size_t alloc_size, size_t leadsize, size_t size,
+    bool *commit)
 {
 	void *ret = (void *)((uintptr_t)addr + leadsize);
 
@@ -77,7 +98,7 @@ pages_trim(void *addr, size_t alloc_size, size_t leadsize, size_t size)
 		void *new_addr;
 
 		pages_unmap(addr, alloc_size);
-		new_addr = pages_map(ret, size);
+		new_addr = pages_map(ret, size, commit);
 		if (new_addr == ret)
 			return (ret);
 		if (new_addr)
@@ -101,17 +122,17 @@ static bool
 pages_commit_impl(void *addr, size_t size, bool commit)
 {
 
-#ifndef _WIN32
-	/*
-	 * The following decommit/commit implementation is functional, but
-	 * always disabled because it doesn't add value beyong improved
-	 * debugging (at the cost of extra system calls) on systems that
-	 * overcommit.
-	 */
-	if (false) {
-		int prot = commit ? (PROT_READ | PROT_WRITE) : PROT_NONE;
-		void *result = mmap(addr, size, prot, MAP_PRIVATE | MAP_ANON |
-		    MAP_FIXED, -1, 0);
+	if (os_overcommits)
+		return (true);
+
+#ifdef _WIN32
+	return (commit ? (addr != VirtualAlloc(addr, size, MEM_COMMIT,
+	    PAGE_READWRITE)) : (!VirtualFree(addr, size, MEM_DECOMMIT)));
+#else
+	{
+		int prot = commit ? PAGES_PROT_COMMIT : PAGES_PROT_DECOMMIT;
+		void *result = mmap(addr, size, prot, mmap_flags | MAP_FIXED,
+		    -1, 0);
 		if (result == MAP_FAILED)
 			return (true);
 		if (result != addr) {
@@ -125,7 +146,6 @@ pages_commit_impl(void *addr, size_t size, bool commit)
 		return (false);
 	}
 #endif
-	return (true);
 }
 
 bool
@@ -150,15 +170,16 @@ pages_purge(void *addr, size_t size)
 #ifdef _WIN32
 	VirtualAlloc(addr, size, MEM_RESET, PAGE_READWRITE);
 	unzeroed = true;
-#elif defined(JEMALLOC_HAVE_MADVISE)
-#  ifdef JEMALLOC_PURGE_MADVISE_DONTNEED
-#    define JEMALLOC_MADV_PURGE MADV_DONTNEED
-#    define JEMALLOC_MADV_ZEROS true
-#  elif defined(JEMALLOC_PURGE_MADVISE_FREE)
+#elif (defined(JEMALLOC_PURGE_MADVISE_FREE) || \
+    defined(JEMALLOC_PURGE_MADVISE_DONTNEED))
+#  if defined(JEMALLOC_PURGE_MADVISE_FREE)
 #    define JEMALLOC_MADV_PURGE MADV_FREE
 #    define JEMALLOC_MADV_ZEROS false
+#  elif defined(JEMALLOC_PURGE_MADVISE_DONTNEED)
+#    define JEMALLOC_MADV_PURGE MADV_DONTNEED
+#    define JEMALLOC_MADV_ZEROS true
 #  else
-#    error "No madvise(2) flag defined for purging unused dirty pages."
+#    error No madvise(2) flag defined for purging unused dirty pages
 #  endif
 	int err = madvise(addr, size, JEMALLOC_MADV_PURGE);
 	unzeroed = (!JEMALLOC_MADV_ZEROS || err != 0);
@@ -171,3 +192,111 @@ pages_purge(void *addr, size_t size)
 	return (unzeroed);
 }
 
+bool
+pages_huge(void *addr, size_t size)
+{
+
+	assert(PAGE_ADDR2BASE(addr) == addr);
+	assert(PAGE_CEILING(size) == size);
+
+#ifdef JEMALLOC_THP
+	return (madvise(addr, size, MADV_HUGEPAGE) != 0);
+#else
+	return (false);
+#endif
+}
+
+bool
+pages_nohuge(void *addr, size_t size)
+{
+
+	assert(PAGE_ADDR2BASE(addr) == addr);
+	assert(PAGE_CEILING(size) == size);
+
+#ifdef JEMALLOC_THP
+	return (madvise(addr, size, MADV_NOHUGEPAGE) != 0);
+#else
+	return (false);
+#endif
+}
+
+#ifdef JEMALLOC_SYSCTL_VM_OVERCOMMIT
+static bool
+os_overcommits_sysctl(void)
+{
+	int vm_overcommit;
+	size_t sz;
+
+	sz = sizeof(vm_overcommit);
+	if (sysctlbyname("vm.overcommit", &vm_overcommit, &sz, NULL, 0) != 0)
+		return (false); /* Error. */
+
+	return ((vm_overcommit & 0x3) == 0);
+}
+#endif
+
+#ifdef JEMALLOC_PROC_SYS_VM_OVERCOMMIT_MEMORY
+/*
+ * Use syscall(2) rather than {open,read,close}(2) when possible to avoid
+ * reentry during bootstrapping if another library has interposed system call
+ * wrappers.
+ */
+static bool
+os_overcommits_proc(void)
+{
+	int fd;
+	char buf[1];
+	ssize_t nread;
+
+#if defined(JEMALLOC_USE_SYSCALL) && defined(SYS_open)
+	fd = (int)syscall(SYS_open, "/proc/sys/vm/overcommit_memory", O_RDONLY);
+#else
+	fd = open("/proc/sys/vm/overcommit_memory", O_RDONLY);
+#endif
+	if (fd == -1)
+		return (false); /* Error. */
+
+#if defined(JEMALLOC_USE_SYSCALL) && defined(SYS_read)
+	nread = (ssize_t)syscall(SYS_read, fd, &buf, sizeof(buf));
+#else
+	nread = read(fd, &buf, sizeof(buf));
+#endif
+
+#if defined(JEMALLOC_USE_SYSCALL) && defined(SYS_close)
+	syscall(SYS_close, fd);
+#else
+	close(fd);
+#endif
+
+	if (nread < 1)
+		return (false); /* Error. */
+	/*
+	 * /proc/sys/vm/overcommit_memory meanings:
+	 * 0: Heuristic overcommit.
+	 * 1: Always overcommit.
+	 * 2: Never overcommit.
+	 */
+	return (buf[0] == '0' || buf[0] == '1');
+}
+#endif
+
+void
+pages_boot(void)
+{
+
+#ifndef _WIN32
+	mmap_flags = MAP_PRIVATE | MAP_ANON;
+#endif
+
+#ifdef JEMALLOC_SYSCTL_VM_OVERCOMMIT
+	os_overcommits = os_overcommits_sysctl();
+#elif defined(JEMALLOC_PROC_SYS_VM_OVERCOMMIT_MEMORY)
+	os_overcommits = os_overcommits_proc();
+#  ifdef MAP_NORESERVE
+	if (os_overcommits)
+		mmap_flags |= MAP_NORESERVE;
+#  endif
+#else
+	os_overcommits = false;
+#endif
+}
diff --git a/deps/jemalloc/src/prng.c b/deps/jemalloc/src/prng.c
new file mode 100644
index 000000000..76646a2a4
--- /dev/null
+++ b/deps/jemalloc/src/prng.c
@@ -0,0 +1,2 @@
+#define	JEMALLOC_PRNG_C_
+#include "jemalloc/internal/jemalloc_internal.h"
diff --git a/deps/jemalloc/src/prof.c b/deps/jemalloc/src/prof.c
index 5d2b9598f..c89dade1f 100644
--- a/deps/jemalloc/src/prof.c
+++ b/deps/jemalloc/src/prof.c
@@ -109,7 +109,7 @@ static char		prof_dump_buf[
     1
 #endif
 ];
-static unsigned		prof_dump_buf_end;
+static size_t		prof_dump_buf_end;
 static int		prof_dump_fd;
 
 /* Do not dump any profiles until bootstrapping is complete. */
@@ -121,13 +121,13 @@ static bool		prof_booted = false;
  * definition.
  */
 
-static bool	prof_tctx_should_destroy(prof_tctx_t *tctx);
+static bool	prof_tctx_should_destroy(tsdn_t *tsdn, prof_tctx_t *tctx);
 static void	prof_tctx_destroy(tsd_t *tsd, prof_tctx_t *tctx);
-static bool	prof_tdata_should_destroy(prof_tdata_t *tdata,
+static bool	prof_tdata_should_destroy(tsdn_t *tsdn, prof_tdata_t *tdata,
     bool even_if_attached);
 static void	prof_tdata_destroy(tsd_t *tsd, prof_tdata_t *tdata,
     bool even_if_attached);
-static char	*prof_thread_name_alloc(tsd_t *tsd, const char *thread_name);
+static char	*prof_thread_name_alloc(tsdn_t *tsdn, const char *thread_name);
 
 /******************************************************************************/
 /* Red-black trees. */
@@ -213,22 +213,23 @@ prof_alloc_rollback(tsd_t *tsd, prof_tctx_t *tctx, bool updated)
 	}
 
 	if ((uintptr_t)tctx > (uintptr_t)1U) {
-		malloc_mutex_lock(tctx->tdata->lock);
+		malloc_mutex_lock(tsd_tsdn(tsd), tctx->tdata->lock);
 		tctx->prepared = false;
-		if (prof_tctx_should_destroy(tctx))
+		if (prof_tctx_should_destroy(tsd_tsdn(tsd), tctx))
 			prof_tctx_destroy(tsd, tctx);
 		else
-			malloc_mutex_unlock(tctx->tdata->lock);
+			malloc_mutex_unlock(tsd_tsdn(tsd), tctx->tdata->lock);
 	}
 }
 
 void
-prof_malloc_sample_object(const void *ptr, size_t usize, prof_tctx_t *tctx)
+prof_malloc_sample_object(tsdn_t *tsdn, const void *ptr, size_t usize,
+    prof_tctx_t *tctx)
 {
 
-	prof_tctx_set(ptr, usize, tctx);
+	prof_tctx_set(tsdn, ptr, usize, tctx);
 
-	malloc_mutex_lock(tctx->tdata->lock);
+	malloc_mutex_lock(tsdn, tctx->tdata->lock);
 	tctx->cnts.curobjs++;
 	tctx->cnts.curbytes += usize;
 	if (opt_prof_accum) {
@@ -236,23 +237,23 @@ prof_malloc_sample_object(const void *ptr, size_t usize, prof_tctx_t *tctx)
 		tctx->cnts.accumbytes += usize;
 	}
 	tctx->prepared = false;
-	malloc_mutex_unlock(tctx->tdata->lock);
+	malloc_mutex_unlock(tsdn, tctx->tdata->lock);
 }
 
 void
 prof_free_sampled_object(tsd_t *tsd, size_t usize, prof_tctx_t *tctx)
 {
 
-	malloc_mutex_lock(tctx->tdata->lock);
+	malloc_mutex_lock(tsd_tsdn(tsd), tctx->tdata->lock);
 	assert(tctx->cnts.curobjs > 0);
 	assert(tctx->cnts.curbytes >= usize);
 	tctx->cnts.curobjs--;
 	tctx->cnts.curbytes -= usize;
 
-	if (prof_tctx_should_destroy(tctx))
+	if (prof_tctx_should_destroy(tsd_tsdn(tsd), tctx))
 		prof_tctx_destroy(tsd, tctx);
 	else
-		malloc_mutex_unlock(tctx->tdata->lock);
+		malloc_mutex_unlock(tsd_tsdn(tsd), tctx->tdata->lock);
 }
 
 void
@@ -277,7 +278,7 @@ prof_enter(tsd_t *tsd, prof_tdata_t *tdata)
 		tdata->enq = true;
 	}
 
-	malloc_mutex_lock(&bt2gctx_mtx);
+	malloc_mutex_lock(tsd_tsdn(tsd), &bt2gctx_mtx);
 }
 
 JEMALLOC_INLINE_C void
@@ -287,7 +288,7 @@ prof_leave(tsd_t *tsd, prof_tdata_t *tdata)
 	cassert(config_prof);
 	assert(tdata == prof_tdata_get(tsd, false));
 
-	malloc_mutex_unlock(&bt2gctx_mtx);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &bt2gctx_mtx);
 
 	if (tdata != NULL) {
 		bool idump, gdump;
@@ -300,9 +301,9 @@ prof_leave(tsd_t *tsd, prof_tdata_t *tdata)
 		tdata->enq_gdump = false;
 
 		if (idump)
-			prof_idump();
+			prof_idump(tsd_tsdn(tsd));
 		if (gdump)
-			prof_gdump();
+			prof_gdump(tsd_tsdn(tsd));
 	}
 }
 
@@ -546,14 +547,15 @@ prof_tdata_mutex_choose(uint64_t thr_uid)
 }
 
 static prof_gctx_t *
-prof_gctx_create(tsd_t *tsd, prof_bt_t *bt)
+prof_gctx_create(tsdn_t *tsdn, prof_bt_t *bt)
 {
 	/*
 	 * Create a single allocation that has space for vec of length bt->len.
 	 */
-	prof_gctx_t *gctx = (prof_gctx_t *)iallocztm(tsd, offsetof(prof_gctx_t,
-	    vec) + (bt->len * sizeof(void *)), false, tcache_get(tsd, true),
-	    true, NULL);
+	size_t size = offsetof(prof_gctx_t, vec) + (bt->len * sizeof(void *));
+	prof_gctx_t *gctx = (prof_gctx_t *)iallocztm(tsdn, size,
+	    size2index(size), false, NULL, true, arena_get(TSDN_NULL, 0, true),
+	    true);
 	if (gctx == NULL)
 		return (NULL);
 	gctx->lock = prof_gctx_mutex_choose();
@@ -585,7 +587,7 @@ prof_gctx_try_destroy(tsd_t *tsd, prof_tdata_t *tdata_self, prof_gctx_t *gctx,
 	 * into this function.
 	 */
 	prof_enter(tsd, tdata_self);
-	malloc_mutex_lock(gctx->lock);
+	malloc_mutex_lock(tsd_tsdn(tsd), gctx->lock);
 	assert(gctx->nlimbo != 0);
 	if (tctx_tree_empty(&gctx->tctxs) && gctx->nlimbo == 1) {
 		/* Remove gctx from bt2gctx. */
@@ -593,24 +595,25 @@ prof_gctx_try_destroy(tsd_t *tsd, prof_tdata_t *tdata_self, prof_gctx_t *gctx,
 			not_reached();
 		prof_leave(tsd, tdata_self);
 		/* Destroy gctx. */
-		malloc_mutex_unlock(gctx->lock);
-		idalloctm(tsd, gctx, tcache_get(tsd, false), true);
+		malloc_mutex_unlock(tsd_tsdn(tsd), gctx->lock);
+		idalloctm(tsd_tsdn(tsd), gctx, NULL, true, true);
 	} else {
 		/*
 		 * Compensate for increment in prof_tctx_destroy() or
 		 * prof_lookup().
 		 */
 		gctx->nlimbo--;
-		malloc_mutex_unlock(gctx->lock);
+		malloc_mutex_unlock(tsd_tsdn(tsd), gctx->lock);
 		prof_leave(tsd, tdata_self);
 	}
 }
 
-/* tctx->tdata->lock must be held. */
 static bool
-prof_tctx_should_destroy(prof_tctx_t *tctx)
+prof_tctx_should_destroy(tsdn_t *tsdn, prof_tctx_t *tctx)
 {
 
+	malloc_mutex_assert_owner(tsdn, tctx->tdata->lock);
+
 	if (opt_prof_accum)
 		return (false);
 	if (tctx->cnts.curobjs != 0)
@@ -633,7 +636,6 @@ prof_gctx_should_destroy(prof_gctx_t *gctx)
 	return (true);
 }
 
-/* tctx->tdata->lock is held upon entry, and released before return. */
 static void
 prof_tctx_destroy(tsd_t *tsd, prof_tctx_t *tctx)
 {
@@ -641,6 +643,8 @@ prof_tctx_destroy(tsd_t *tsd, prof_tctx_t *tctx)
 	prof_gctx_t *gctx = tctx->gctx;
 	bool destroy_tdata, destroy_tctx, destroy_gctx;
 
+	malloc_mutex_assert_owner(tsd_tsdn(tsd), tctx->tdata->lock);
+
 	assert(tctx->cnts.curobjs == 0);
 	assert(tctx->cnts.curbytes == 0);
 	assert(!opt_prof_accum);
@@ -648,10 +652,10 @@ prof_tctx_destroy(tsd_t *tsd, prof_tctx_t *tctx)
 	assert(tctx->cnts.accumbytes == 0);
 
 	ckh_remove(tsd, &tdata->bt2tctx, &gctx->bt, NULL, NULL);
-	destroy_tdata = prof_tdata_should_destroy(tdata, false);
-	malloc_mutex_unlock(tdata->lock);
+	destroy_tdata = prof_tdata_should_destroy(tsd_tsdn(tsd), tdata, false);
+	malloc_mutex_unlock(tsd_tsdn(tsd), tdata->lock);
 
-	malloc_mutex_lock(gctx->lock);
+	malloc_mutex_lock(tsd_tsdn(tsd), gctx->lock);
 	switch (tctx->state) {
 	case prof_tctx_state_nominal:
 		tctx_tree_remove(&gctx->tctxs, tctx);
@@ -691,17 +695,19 @@ prof_tctx_destroy(tsd_t *tsd, prof_tctx_t *tctx)
 		destroy_tctx = false;
 		destroy_gctx = false;
 	}
-	malloc_mutex_unlock(gctx->lock);
+	malloc_mutex_unlock(tsd_tsdn(tsd), gctx->lock);
 	if (destroy_gctx) {
 		prof_gctx_try_destroy(tsd, prof_tdata_get(tsd, false), gctx,
 		    tdata);
 	}
 
+	malloc_mutex_assert_not_owner(tsd_tsdn(tsd), tctx->tdata->lock);
+
 	if (destroy_tdata)
 		prof_tdata_destroy(tsd, tdata, false);
 
 	if (destroy_tctx)
-		idalloctm(tsd, tctx, tcache_get(tsd, false), true);
+		idalloctm(tsd_tsdn(tsd), tctx, NULL, true, true);
 }
 
 static bool
@@ -721,7 +727,7 @@ prof_lookup_global(tsd_t *tsd, prof_bt_t *bt, prof_tdata_t *tdata,
 	prof_enter(tsd, tdata);
 	if (ckh_search(&bt2gctx, bt, &btkey.v, &gctx.v)) {
 		/* bt has never been seen before.  Insert it. */
-		gctx.p = prof_gctx_create(tsd, bt);
+		gctx.p = prof_gctx_create(tsd_tsdn(tsd), bt);
 		if (gctx.v == NULL) {
 			prof_leave(tsd, tdata);
 			return (true);
@@ -730,7 +736,7 @@ prof_lookup_global(tsd_t *tsd, prof_bt_t *bt, prof_tdata_t *tdata,
 		if (ckh_insert(tsd, &bt2gctx, btkey.v, gctx.v)) {
 			/* OOM. */
 			prof_leave(tsd, tdata);
-			idalloctm(tsd, gctx.v, tcache_get(tsd, false), true);
+			idalloctm(tsd_tsdn(tsd), gctx.v, NULL, true, true);
 			return (true);
 		}
 		new_gctx = true;
@@ -739,9 +745,9 @@ prof_lookup_global(tsd_t *tsd, prof_bt_t *bt, prof_tdata_t *tdata,
 		 * Increment nlimbo, in order to avoid a race condition with
 		 * prof_tctx_destroy()/prof_gctx_try_destroy().
 		 */
-		malloc_mutex_lock(gctx.p->lock);
+		malloc_mutex_lock(tsd_tsdn(tsd), gctx.p->lock);
 		gctx.p->nlimbo++;
-		malloc_mutex_unlock(gctx.p->lock);
+		malloc_mutex_unlock(tsd_tsdn(tsd), gctx.p->lock);
 		new_gctx = false;
 	}
 	prof_leave(tsd, tdata);
@@ -768,13 +774,12 @@ prof_lookup(tsd_t *tsd, prof_bt_t *bt)
 	if (tdata == NULL)
 		return (NULL);
 
-	malloc_mutex_lock(tdata->lock);
+	malloc_mutex_lock(tsd_tsdn(tsd), tdata->lock);
 	not_found = ckh_search(&tdata->bt2tctx, bt, NULL, &ret.v);
 	if (!not_found) /* Note double negative! */
 		ret.p->prepared = true;
-	malloc_mutex_unlock(tdata->lock);
+	malloc_mutex_unlock(tsd_tsdn(tsd), tdata->lock);
 	if (not_found) {
-		tcache_t *tcache;
 		void *btkey;
 		prof_gctx_t *gctx;
 		bool new_gctx, error;
@@ -788,9 +793,9 @@ prof_lookup(tsd_t *tsd, prof_bt_t *bt)
 			return (NULL);
 
 		/* Link a prof_tctx_t into gctx for this thread. */
-		tcache = tcache_get(tsd, true);
-		ret.v = iallocztm(tsd, sizeof(prof_tctx_t), false, tcache, true,
-		    NULL);
+		ret.v = iallocztm(tsd_tsdn(tsd), sizeof(prof_tctx_t),
+		    size2index(sizeof(prof_tctx_t)), false, NULL, true,
+		    arena_ichoose(tsd, NULL), true);
 		if (ret.p == NULL) {
 			if (new_gctx)
 				prof_gctx_try_destroy(tsd, tdata, gctx, tdata);
@@ -804,41 +809,41 @@ prof_lookup(tsd_t *tsd, prof_bt_t *bt)
 		ret.p->tctx_uid = tdata->tctx_uid_next++;
 		ret.p->prepared = true;
 		ret.p->state = prof_tctx_state_initializing;
-		malloc_mutex_lock(tdata->lock);
+		malloc_mutex_lock(tsd_tsdn(tsd), tdata->lock);
 		error = ckh_insert(tsd, &tdata->bt2tctx, btkey, ret.v);
-		malloc_mutex_unlock(tdata->lock);
+		malloc_mutex_unlock(tsd_tsdn(tsd), tdata->lock);
 		if (error) {
 			if (new_gctx)
 				prof_gctx_try_destroy(tsd, tdata, gctx, tdata);
-			idalloctm(tsd, ret.v, tcache, true);
+			idalloctm(tsd_tsdn(tsd), ret.v, NULL, true, true);
 			return (NULL);
 		}
-		malloc_mutex_lock(gctx->lock);
+		malloc_mutex_lock(tsd_tsdn(tsd), gctx->lock);
 		ret.p->state = prof_tctx_state_nominal;
 		tctx_tree_insert(&gctx->tctxs, ret.p);
 		gctx->nlimbo--;
-		malloc_mutex_unlock(gctx->lock);
+		malloc_mutex_unlock(tsd_tsdn(tsd), gctx->lock);
 	}
 
 	return (ret.p);
 }
 
+/*
+ * The bodies of this function and prof_leakcheck() are compiled out unless heap
+ * profiling is enabled, so that it is possible to compile jemalloc with
+ * floating point support completely disabled.  Avoiding floating point code is
+ * important on memory-constrained systems, but it also enables a workaround for
+ * versions of glibc that don't properly save/restore floating point registers
+ * during dynamic lazy symbol loading (which internally calls into whatever
+ * malloc implementation happens to be integrated into the application).  Note
+ * that some compilers (e.g.  gcc 4.8) may use floating point registers for fast
+ * memory moves, so jemalloc must be compiled with such optimizations disabled
+ * (e.g.
+ * -mno-sse) in order for the workaround to be complete.
+ */
 void
 prof_sample_threshold_update(prof_tdata_t *tdata)
 {
-	/*
-	 * The body of this function is compiled out unless heap profiling is
-	 * enabled, so that it is possible to compile jemalloc with floating
-	 * point support completely disabled.  Avoiding floating point code is
-	 * important on memory-constrained systems, but it also enables a
-	 * workaround for versions of glibc that don't properly save/restore
-	 * floating point registers during dynamic lazy symbol loading (which
-	 * internally calls into whatever malloc implementation happens to be
-	 * integrated into the application).  Note that some compilers (e.g.
-	 * gcc 4.8) may use floating point registers for fast memory moves, so
-	 * jemalloc must be compiled with such optimizations disabled (e.g.
-	 * -mno-sse) in order for the workaround to be complete.
-	 */
 #ifdef JEMALLOC_PROF
 	uint64_t r;
 	double u;
@@ -869,8 +874,7 @@ prof_sample_threshold_update(prof_tdata_t *tdata)
 	 *   pp 500
 	 *   (http://luc.devroye.org/rnbookindex.html)
 	 */
-	prng64(r, 53, tdata->prng_state, UINT64_C(6364136223846793005),
-	    UINT64_C(1442695040888963407));
+	r = prng_lg_range_u64(&tdata->prng_state, 53);
 	u = (double)r * (1.0/9007199254740992.0L);
 	tdata->bytes_until_sample = (uint64_t)(log(u) /
 	    log(1.0 - (1.0 / (double)((uint64_t)1U << lg_prof_sample))))
@@ -893,11 +897,13 @@ size_t
 prof_tdata_count(void)
 {
 	size_t tdata_count = 0;
+	tsdn_t *tsdn;
 
-	malloc_mutex_lock(&tdatas_mtx);
+	tsdn = tsdn_fetch();
+	malloc_mutex_lock(tsdn, &tdatas_mtx);
 	tdata_tree_iter(&tdatas, NULL, prof_tdata_count_iter,
 	    (void *)&tdata_count);
-	malloc_mutex_unlock(&tdatas_mtx);
+	malloc_mutex_unlock(tsdn, &tdatas_mtx);
 
 	return (tdata_count);
 }
@@ -916,9 +922,9 @@ prof_bt_count(void)
 	if (tdata == NULL)
 		return (0);
 
-	malloc_mutex_lock(&bt2gctx_mtx);
+	malloc_mutex_lock(tsd_tsdn(tsd), &bt2gctx_mtx);
 	bt_count = ckh_count(&bt2gctx);
-	malloc_mutex_unlock(&bt2gctx_mtx);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &bt2gctx_mtx);
 
 	return (bt_count);
 }
@@ -988,7 +994,7 @@ prof_dump_close(bool propagate_err)
 static bool
 prof_dump_write(bool propagate_err, const char *s)
 {
-	unsigned i, slen, n;
+	size_t i, slen, n;
 
 	cassert(config_prof);
 
@@ -1031,20 +1037,21 @@ prof_dump_printf(bool propagate_err, const char *format, ...)
 	return (ret);
 }
 
-/* tctx->tdata->lock is held. */
 static void
-prof_tctx_merge_tdata(prof_tctx_t *tctx, prof_tdata_t *tdata)
+prof_tctx_merge_tdata(tsdn_t *tsdn, prof_tctx_t *tctx, prof_tdata_t *tdata)
 {
 
-	malloc_mutex_lock(tctx->gctx->lock);
+	malloc_mutex_assert_owner(tsdn, tctx->tdata->lock);
+
+	malloc_mutex_lock(tsdn, tctx->gctx->lock);
 
 	switch (tctx->state) {
 	case prof_tctx_state_initializing:
-		malloc_mutex_unlock(tctx->gctx->lock);
+		malloc_mutex_unlock(tsdn, tctx->gctx->lock);
 		return;
 	case prof_tctx_state_nominal:
 		tctx->state = prof_tctx_state_dumping;
-		malloc_mutex_unlock(tctx->gctx->lock);
+		malloc_mutex_unlock(tsdn, tctx->gctx->lock);
 
 		memcpy(&tctx->dump_cnts, &tctx->cnts, sizeof(prof_cnt_t));
 
@@ -1063,11 +1070,12 @@ prof_tctx_merge_tdata(prof_tctx_t *tctx, prof_tdata_t *tdata)
 	}
 }
 
-/* gctx->lock is held. */
 static void
-prof_tctx_merge_gctx(prof_tctx_t *tctx, prof_gctx_t *gctx)
+prof_tctx_merge_gctx(tsdn_t *tsdn, prof_tctx_t *tctx, prof_gctx_t *gctx)
 {
 
+	malloc_mutex_assert_owner(tsdn, gctx->lock);
+
 	gctx->cnt_summed.curobjs += tctx->dump_cnts.curobjs;
 	gctx->cnt_summed.curbytes += tctx->dump_cnts.curbytes;
 	if (opt_prof_accum) {
@@ -1076,10 +1084,12 @@ prof_tctx_merge_gctx(prof_tctx_t *tctx, prof_gctx_t *gctx)
 	}
 }
 
-/* tctx->gctx is held. */
 static prof_tctx_t *
 prof_tctx_merge_iter(prof_tctx_tree_t *tctxs, prof_tctx_t *tctx, void *arg)
 {
+	tsdn_t *tsdn = (tsdn_t *)arg;
+
+	malloc_mutex_assert_owner(tsdn, tctx->gctx->lock);
 
 	switch (tctx->state) {
 	case prof_tctx_state_nominal:
@@ -1087,7 +1097,7 @@ prof_tctx_merge_iter(prof_tctx_tree_t *tctxs, prof_tctx_t *tctx, void *arg)
 		break;
 	case prof_tctx_state_dumping:
 	case prof_tctx_state_purgatory:
-		prof_tctx_merge_gctx(tctx, tctx->gctx);
+		prof_tctx_merge_gctx(tsdn, tctx, tctx->gctx);
 		break;
 	default:
 		not_reached();
@@ -1096,11 +1106,18 @@ prof_tctx_merge_iter(prof_tctx_tree_t *tctxs, prof_tctx_t *tctx, void *arg)
 	return (NULL);
 }
 
-/* gctx->lock is held. */
+struct prof_tctx_dump_iter_arg_s {
+	tsdn_t	*tsdn;
+	bool	propagate_err;
+};
+
 static prof_tctx_t *
-prof_tctx_dump_iter(prof_tctx_tree_t *tctxs, prof_tctx_t *tctx, void *arg)
+prof_tctx_dump_iter(prof_tctx_tree_t *tctxs, prof_tctx_t *tctx, void *opaque)
 {
-	bool propagate_err = *(bool *)arg;
+	struct prof_tctx_dump_iter_arg_s *arg =
+	    (struct prof_tctx_dump_iter_arg_s *)opaque;
+
+	malloc_mutex_assert_owner(arg->tsdn, tctx->gctx->lock);
 
 	switch (tctx->state) {
 	case prof_tctx_state_initializing:
@@ -1109,7 +1126,7 @@ prof_tctx_dump_iter(prof_tctx_tree_t *tctxs, prof_tctx_t *tctx, void *arg)
 		break;
 	case prof_tctx_state_dumping:
 	case prof_tctx_state_purgatory:
-		if (prof_dump_printf(propagate_err,
+		if (prof_dump_printf(arg->propagate_err,
 		    "  t%"FMTu64": %"FMTu64": %"FMTu64" [%"FMTu64": "
 		    "%"FMTu64"]\n", tctx->thr_uid, tctx->dump_cnts.curobjs,
 		    tctx->dump_cnts.curbytes, tctx->dump_cnts.accumobjs,
@@ -1122,12 +1139,14 @@ prof_tctx_dump_iter(prof_tctx_tree_t *tctxs, prof_tctx_t *tctx, void *arg)
 	return (NULL);
 }
 
-/* tctx->gctx is held. */
 static prof_tctx_t *
 prof_tctx_finish_iter(prof_tctx_tree_t *tctxs, prof_tctx_t *tctx, void *arg)
 {
+	tsdn_t *tsdn = (tsdn_t *)arg;
 	prof_tctx_t *ret;
 
+	malloc_mutex_assert_owner(tsdn, tctx->gctx->lock);
+
 	switch (tctx->state) {
 	case prof_tctx_state_nominal:
 		/* New since dumping started; ignore. */
@@ -1148,12 +1167,12 @@ label_return:
 }
 
 static void
-prof_dump_gctx_prep(prof_gctx_t *gctx, prof_gctx_tree_t *gctxs)
+prof_dump_gctx_prep(tsdn_t *tsdn, prof_gctx_t *gctx, prof_gctx_tree_t *gctxs)
 {
 
 	cassert(config_prof);
 
-	malloc_mutex_lock(gctx->lock);
+	malloc_mutex_lock(tsdn, gctx->lock);
 
 	/*
 	 * Increment nlimbo so that gctx won't go away before dump.
@@ -1165,19 +1184,26 @@ prof_dump_gctx_prep(prof_gctx_t *gctx, prof_gctx_tree_t *gctxs)
 
 	memset(&gctx->cnt_summed, 0, sizeof(prof_cnt_t));
 
-	malloc_mutex_unlock(gctx->lock);
+	malloc_mutex_unlock(tsdn, gctx->lock);
 }
 
+struct prof_gctx_merge_iter_arg_s {
+	tsdn_t	*tsdn;
+	size_t	leak_ngctx;
+};
+
 static prof_gctx_t *
-prof_gctx_merge_iter(prof_gctx_tree_t *gctxs, prof_gctx_t *gctx, void *arg)
+prof_gctx_merge_iter(prof_gctx_tree_t *gctxs, prof_gctx_t *gctx, void *opaque)
 {
-	size_t *leak_ngctx = (size_t *)arg;
+	struct prof_gctx_merge_iter_arg_s *arg =
+	    (struct prof_gctx_merge_iter_arg_s *)opaque;
 
-	malloc_mutex_lock(gctx->lock);
-	tctx_tree_iter(&gctx->tctxs, NULL, prof_tctx_merge_iter, NULL);
+	malloc_mutex_lock(arg->tsdn, gctx->lock);
+	tctx_tree_iter(&gctx->tctxs, NULL, prof_tctx_merge_iter,
+	    (void *)arg->tsdn);
 	if (gctx->cnt_summed.curobjs != 0)
-		(*leak_ngctx)++;
-	malloc_mutex_unlock(gctx->lock);
+		arg->leak_ngctx++;
+	malloc_mutex_unlock(arg->tsdn, gctx->lock);
 
 	return (NULL);
 }
@@ -1196,7 +1222,7 @@ prof_gctx_finish(tsd_t *tsd, prof_gctx_tree_t *gctxs)
 	 */
 	while ((gctx = gctx_tree_first(gctxs)) != NULL) {
 		gctx_tree_remove(gctxs, gctx);
-		malloc_mutex_lock(gctx->lock);
+		malloc_mutex_lock(tsd_tsdn(tsd), gctx->lock);
 		{
 			prof_tctx_t *next;
 
@@ -1204,14 +1230,15 @@ prof_gctx_finish(tsd_t *tsd, prof_gctx_tree_t *gctxs)
 			do {
 				prof_tctx_t *to_destroy =
 				    tctx_tree_iter(&gctx->tctxs, next,
-				    prof_tctx_finish_iter, NULL);
+				    prof_tctx_finish_iter,
+				    (void *)tsd_tsdn(tsd));
 				if (to_destroy != NULL) {
 					next = tctx_tree_next(&gctx->tctxs,
 					    to_destroy);
 					tctx_tree_remove(&gctx->tctxs,
 					    to_destroy);
-					idalloctm(tsd, to_destroy,
-					    tcache_get(tsd, false), true);
+					idalloctm(tsd_tsdn(tsd), to_destroy,
+					    NULL, true, true);
 				} else
 					next = NULL;
 			} while (next != NULL);
@@ -1219,19 +1246,26 @@ prof_gctx_finish(tsd_t *tsd, prof_gctx_tree_t *gctxs)
 		gctx->nlimbo--;
 		if (prof_gctx_should_destroy(gctx)) {
 			gctx->nlimbo++;
-			malloc_mutex_unlock(gctx->lock);
+			malloc_mutex_unlock(tsd_tsdn(tsd), gctx->lock);
 			prof_gctx_try_destroy(tsd, tdata, gctx, tdata);
 		} else
-			malloc_mutex_unlock(gctx->lock);
+			malloc_mutex_unlock(tsd_tsdn(tsd), gctx->lock);
 	}
 }
 
+struct prof_tdata_merge_iter_arg_s {
+	tsdn_t		*tsdn;
+	prof_cnt_t	cnt_all;
+};
+
 static prof_tdata_t *
-prof_tdata_merge_iter(prof_tdata_tree_t *tdatas, prof_tdata_t *tdata, void *arg)
+prof_tdata_merge_iter(prof_tdata_tree_t *tdatas, prof_tdata_t *tdata,
+    void *opaque)
 {
-	prof_cnt_t *cnt_all = (prof_cnt_t *)arg;
+	struct prof_tdata_merge_iter_arg_s *arg =
+	    (struct prof_tdata_merge_iter_arg_s *)opaque;
 
-	malloc_mutex_lock(tdata->lock);
+	malloc_mutex_lock(arg->tsdn, tdata->lock);
 	if (!tdata->expired) {
 		size_t tabind;
 		union {
@@ -1243,17 +1277,17 @@ prof_tdata_merge_iter(prof_tdata_tree_t *tdatas, prof_tdata_t *tdata, void *arg)
 		memset(&tdata->cnt_summed, 0, sizeof(prof_cnt_t));
 		for (tabind = 0; !ckh_iter(&tdata->bt2tctx, &tabind, NULL,
 		    &tctx.v);)
-			prof_tctx_merge_tdata(tctx.p, tdata);
+			prof_tctx_merge_tdata(arg->tsdn, tctx.p, tdata);
 
-		cnt_all->curobjs += tdata->cnt_summed.curobjs;
-		cnt_all->curbytes += tdata->cnt_summed.curbytes;
+		arg->cnt_all.curobjs += tdata->cnt_summed.curobjs;
+		arg->cnt_all.curbytes += tdata->cnt_summed.curbytes;
 		if (opt_prof_accum) {
-			cnt_all->accumobjs += tdata->cnt_summed.accumobjs;
-			cnt_all->accumbytes += tdata->cnt_summed.accumbytes;
+			arg->cnt_all.accumobjs += tdata->cnt_summed.accumobjs;
+			arg->cnt_all.accumbytes += tdata->cnt_summed.accumbytes;
 		}
 	} else
 		tdata->dumping = false;
-	malloc_mutex_unlock(tdata->lock);
+	malloc_mutex_unlock(arg->tsdn, tdata->lock);
 
 	return (NULL);
 }
@@ -1282,7 +1316,7 @@ prof_tdata_dump_iter(prof_tdata_tree_t *tdatas, prof_tdata_t *tdata, void *arg)
 #define	prof_dump_header JEMALLOC_N(prof_dump_header_impl)
 #endif
 static bool
-prof_dump_header(bool propagate_err, const prof_cnt_t *cnt_all)
+prof_dump_header(tsdn_t *tsdn, bool propagate_err, const prof_cnt_t *cnt_all)
 {
 	bool ret;
 
@@ -1293,10 +1327,10 @@ prof_dump_header(bool propagate_err, const prof_cnt_t *cnt_all)
 	    cnt_all->curbytes, cnt_all->accumobjs, cnt_all->accumbytes))
 		return (true);
 
-	malloc_mutex_lock(&tdatas_mtx);
+	malloc_mutex_lock(tsdn, &tdatas_mtx);
 	ret = (tdata_tree_iter(&tdatas, NULL, prof_tdata_dump_iter,
 	    (void *)&propagate_err) != NULL);
-	malloc_mutex_unlock(&tdatas_mtx);
+	malloc_mutex_unlock(tsdn, &tdatas_mtx);
 	return (ret);
 }
 #ifdef JEMALLOC_JET
@@ -1305,15 +1339,16 @@ prof_dump_header(bool propagate_err, const prof_cnt_t *cnt_all)
 prof_dump_header_t *prof_dump_header = JEMALLOC_N(prof_dump_header_impl);
 #endif
 
-/* gctx->lock is held. */
 static bool
-prof_dump_gctx(bool propagate_err, prof_gctx_t *gctx, const prof_bt_t *bt,
-    prof_gctx_tree_t *gctxs)
+prof_dump_gctx(tsdn_t *tsdn, bool propagate_err, prof_gctx_t *gctx,
+    const prof_bt_t *bt, prof_gctx_tree_t *gctxs)
 {
 	bool ret;
 	unsigned i;
+	struct prof_tctx_dump_iter_arg_s prof_tctx_dump_iter_arg;
 
 	cassert(config_prof);
+	malloc_mutex_assert_owner(tsdn, gctx->lock);
 
 	/* Avoid dumping such gctx's that have no useful data. */
 	if ((!opt_prof_accum && gctx->cnt_summed.curobjs == 0) ||
@@ -1347,8 +1382,10 @@ prof_dump_gctx(bool propagate_err, prof_gctx_t *gctx, const prof_bt_t *bt,
 		goto label_return;
 	}
 
+	prof_tctx_dump_iter_arg.tsdn = tsdn;
+	prof_tctx_dump_iter_arg.propagate_err = propagate_err;
 	if (tctx_tree_iter(&gctx->tctxs, NULL, prof_tctx_dump_iter,
-	    (void *)&propagate_err) != NULL) {
+	    (void *)&prof_tctx_dump_iter_arg) != NULL) {
 		ret = true;
 		goto label_return;
 	}
@@ -1358,6 +1395,7 @@ label_return:
 	return (ret);
 }
 
+#ifndef _WIN32
 JEMALLOC_FORMAT_PRINTF(1, 2)
 static int
 prof_open_maps(const char *format, ...)
@@ -1373,6 +1411,18 @@ prof_open_maps(const char *format, ...)
 
 	return (mfd);
 }
+#endif
+
+static int
+prof_getpid(void)
+{
+
+#ifdef _WIN32
+	return (GetCurrentProcessId());
+#else
+	return (getpid());
+#endif
+}
 
 static bool
 prof_dump_maps(bool propagate_err)
@@ -1383,9 +1433,11 @@ prof_dump_maps(bool propagate_err)
 	cassert(config_prof);
 #ifdef __FreeBSD__
 	mfd = prof_open_maps("/proc/curproc/map");
+#elif defined(_WIN32)
+	mfd = -1; // Not implemented
 #else
 	{
-		int pid = getpid();
+		int pid = prof_getpid();
 
 		mfd = prof_open_maps("/proc/%d/task/%d/maps", pid, pid);
 		if (mfd == -1)
@@ -1426,39 +1478,66 @@ label_return:
 	return (ret);
 }
 
+/*
+ * See prof_sample_threshold_update() comment for why the body of this function
+ * is conditionally compiled.
+ */
 static void
 prof_leakcheck(const prof_cnt_t *cnt_all, size_t leak_ngctx,
     const char *filename)
 {
 
+#ifdef JEMALLOC_PROF
+	/*
+	 * Scaling is equivalent AdjustSamples() in jeprof, but the result may
+	 * differ slightly from what jeprof reports, because here we scale the
+	 * summary values, whereas jeprof scales each context individually and
+	 * reports the sums of the scaled values.
+	 */
 	if (cnt_all->curbytes != 0) {
-		malloc_printf("<jemalloc>: Leak summary: %"FMTu64" byte%s, %"
-		    FMTu64" object%s, %zu context%s\n",
-		    cnt_all->curbytes, (cnt_all->curbytes != 1) ? "s" : "",
-		    cnt_all->curobjs, (cnt_all->curobjs != 1) ? "s" : "",
-		    leak_ngctx, (leak_ngctx != 1) ? "s" : "");
+		double sample_period = (double)((uint64_t)1 << lg_prof_sample);
+		double ratio = (((double)cnt_all->curbytes) /
+		    (double)cnt_all->curobjs) / sample_period;
+		double scale_factor = 1.0 / (1.0 - exp(-ratio));
+		uint64_t curbytes = (uint64_t)round(((double)cnt_all->curbytes)
+		    * scale_factor);
+		uint64_t curobjs = (uint64_t)round(((double)cnt_all->curobjs) *
+		    scale_factor);
+
+		malloc_printf("<jemalloc>: Leak approximation summary: ~%"FMTu64
+		    " byte%s, ~%"FMTu64" object%s, >= %zu context%s\n",
+		    curbytes, (curbytes != 1) ? "s" : "", curobjs, (curobjs !=
+		    1) ? "s" : "", leak_ngctx, (leak_ngctx != 1) ? "s" : "");
 		malloc_printf(
 		    "<jemalloc>: Run jeprof on \"%s\" for leak detail\n",
 		    filename);
 	}
+#endif
 }
 
+struct prof_gctx_dump_iter_arg_s {
+	tsdn_t	*tsdn;
+	bool	propagate_err;
+};
+
 static prof_gctx_t *
-prof_gctx_dump_iter(prof_gctx_tree_t *gctxs, prof_gctx_t *gctx, void *arg)
+prof_gctx_dump_iter(prof_gctx_tree_t *gctxs, prof_gctx_t *gctx, void *opaque)
 {
 	prof_gctx_t *ret;
-	bool propagate_err = *(bool *)arg;
+	struct prof_gctx_dump_iter_arg_s *arg =
+	    (struct prof_gctx_dump_iter_arg_s *)opaque;
 
-	malloc_mutex_lock(gctx->lock);
+	malloc_mutex_lock(arg->tsdn, gctx->lock);
 
-	if (prof_dump_gctx(propagate_err, gctx, &gctx->bt, gctxs)) {
+	if (prof_dump_gctx(arg->tsdn, arg->propagate_err, gctx, &gctx->bt,
+	    gctxs)) {
 		ret = gctx;
 		goto label_return;
 	}
 
 	ret = NULL;
 label_return:
-	malloc_mutex_unlock(gctx->lock);
+	malloc_mutex_unlock(arg->tsdn, gctx->lock);
 	return (ret);
 }
 
@@ -1466,13 +1545,14 @@ static bool
 prof_dump(tsd_t *tsd, bool propagate_err, const char *filename, bool leakcheck)
 {
 	prof_tdata_t *tdata;
-	prof_cnt_t cnt_all;
+	struct prof_tdata_merge_iter_arg_s prof_tdata_merge_iter_arg;
 	size_t tabind;
 	union {
 		prof_gctx_t	*p;
 		void		*v;
 	} gctx;
-	size_t leak_ngctx;
+	struct prof_gctx_merge_iter_arg_s prof_gctx_merge_iter_arg;
+	struct prof_gctx_dump_iter_arg_s prof_gctx_dump_iter_arg;
 	prof_gctx_tree_t gctxs;
 
 	cassert(config_prof);
@@ -1481,7 +1561,7 @@ prof_dump(tsd_t *tsd, bool propagate_err, const char *filename, bool leakcheck)
 	if (tdata == NULL)
 		return (true);
 
-	malloc_mutex_lock(&prof_dump_mtx);
+	malloc_mutex_lock(tsd_tsdn(tsd), &prof_dump_mtx);
 	prof_enter(tsd, tdata);
 
 	/*
@@ -1490,20 +1570,24 @@ prof_dump(tsd_t *tsd, bool propagate_err, const char *filename, bool leakcheck)
 	 */
 	gctx_tree_new(&gctxs);
 	for (tabind = 0; !ckh_iter(&bt2gctx, &tabind, NULL, &gctx.v);)
-		prof_dump_gctx_prep(gctx.p, &gctxs);
+		prof_dump_gctx_prep(tsd_tsdn(tsd), gctx.p, &gctxs);
 
 	/*
 	 * Iterate over tdatas, and for the non-expired ones snapshot their tctx
 	 * stats and merge them into the associated gctx's.
 	 */
-	memset(&cnt_all, 0, sizeof(prof_cnt_t));
-	malloc_mutex_lock(&tdatas_mtx);
-	tdata_tree_iter(&tdatas, NULL, prof_tdata_merge_iter, (void *)&cnt_all);
-	malloc_mutex_unlock(&tdatas_mtx);
+	prof_tdata_merge_iter_arg.tsdn = tsd_tsdn(tsd);
+	memset(&prof_tdata_merge_iter_arg.cnt_all, 0, sizeof(prof_cnt_t));
+	malloc_mutex_lock(tsd_tsdn(tsd), &tdatas_mtx);
+	tdata_tree_iter(&tdatas, NULL, prof_tdata_merge_iter,
+	    (void *)&prof_tdata_merge_iter_arg);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &tdatas_mtx);
 
 	/* Merge tctx stats into gctx's. */
-	leak_ngctx = 0;
-	gctx_tree_iter(&gctxs, NULL, prof_gctx_merge_iter, (void *)&leak_ngctx);
+	prof_gctx_merge_iter_arg.tsdn = tsd_tsdn(tsd);
+	prof_gctx_merge_iter_arg.leak_ngctx = 0;
+	gctx_tree_iter(&gctxs, NULL, prof_gctx_merge_iter,
+	    (void *)&prof_gctx_merge_iter_arg);
 
 	prof_leave(tsd, tdata);
 
@@ -1512,12 +1596,15 @@ prof_dump(tsd_t *tsd, bool propagate_err, const char *filename, bool leakcheck)
 		goto label_open_close_error;
 
 	/* Dump profile header. */
-	if (prof_dump_header(propagate_err, &cnt_all))
+	if (prof_dump_header(tsd_tsdn(tsd), propagate_err,
+	    &prof_tdata_merge_iter_arg.cnt_all))
 		goto label_write_error;
 
 	/* Dump per gctx profile stats. */
+	prof_gctx_dump_iter_arg.tsdn = tsd_tsdn(tsd);
+	prof_gctx_dump_iter_arg.propagate_err = propagate_err;
 	if (gctx_tree_iter(&gctxs, NULL, prof_gctx_dump_iter,
-	    (void *)&propagate_err) != NULL)
+	    (void *)&prof_gctx_dump_iter_arg) != NULL)
 		goto label_write_error;
 
 	/* Dump /proc/<pid>/maps if possible. */
@@ -1528,17 +1615,18 @@ prof_dump(tsd_t *tsd, bool propagate_err, const char *filename, bool leakcheck)
 		goto label_open_close_error;
 
 	prof_gctx_finish(tsd, &gctxs);
-	malloc_mutex_unlock(&prof_dump_mtx);
-
-	if (leakcheck)
-		prof_leakcheck(&cnt_all, leak_ngctx, filename);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &prof_dump_mtx);
 
+	if (leakcheck) {
+		prof_leakcheck(&prof_tdata_merge_iter_arg.cnt_all,
+		    prof_gctx_merge_iter_arg.leak_ngctx, filename);
+	}
 	return (false);
 label_write_error:
 	prof_dump_close(propagate_err);
 label_open_close_error:
 	prof_gctx_finish(tsd, &gctxs);
-	malloc_mutex_unlock(&prof_dump_mtx);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &prof_dump_mtx);
 	return (true);
 }
 
@@ -1554,12 +1642,12 @@ prof_dump_filename(char *filename, char v, uint64_t vseq)
 	        /* "<prefix>.<pid>.<seq>.v<vseq>.heap" */
 		malloc_snprintf(filename, DUMP_FILENAME_BUFSIZE,
 		    "%s.%d.%"FMTu64".%c%"FMTu64".heap",
-		    opt_prof_prefix, (int)getpid(), prof_dump_seq, v, vseq);
+		    opt_prof_prefix, prof_getpid(), prof_dump_seq, v, vseq);
 	} else {
 	        /* "<prefix>.<pid>.<seq>.<v>.heap" */
 		malloc_snprintf(filename, DUMP_FILENAME_BUFSIZE,
 		    "%s.%d.%"FMTu64".%c.heap",
-		    opt_prof_prefix, (int)getpid(), prof_dump_seq, v);
+		    opt_prof_prefix, prof_getpid(), prof_dump_seq, v);
 	}
 	prof_dump_seq++;
 }
@@ -1578,23 +1666,23 @@ prof_fdump(void)
 		return;
 	tsd = tsd_fetch();
 
-	malloc_mutex_lock(&prof_dump_seq_mtx);
+	malloc_mutex_lock(tsd_tsdn(tsd), &prof_dump_seq_mtx);
 	prof_dump_filename(filename, 'f', VSEQ_INVALID);
-	malloc_mutex_unlock(&prof_dump_seq_mtx);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &prof_dump_seq_mtx);
 	prof_dump(tsd, false, filename, opt_prof_leak);
 }
 
 void
-prof_idump(void)
+prof_idump(tsdn_t *tsdn)
 {
 	tsd_t *tsd;
 	prof_tdata_t *tdata;
 
 	cassert(config_prof);
 
-	if (!prof_booted)
+	if (!prof_booted || tsdn_null(tsdn))
 		return;
-	tsd = tsd_fetch();
+	tsd = tsdn_tsd(tsdn);
 	tdata = prof_tdata_get(tsd, false);
 	if (tdata == NULL)
 		return;
@@ -1605,50 +1693,48 @@ prof_idump(void)
 
 	if (opt_prof_prefix[0] != '\0') {
 		char filename[PATH_MAX + 1];
-		malloc_mutex_lock(&prof_dump_seq_mtx);
+		malloc_mutex_lock(tsd_tsdn(tsd), &prof_dump_seq_mtx);
 		prof_dump_filename(filename, 'i', prof_dump_iseq);
 		prof_dump_iseq++;
-		malloc_mutex_unlock(&prof_dump_seq_mtx);
+		malloc_mutex_unlock(tsd_tsdn(tsd), &prof_dump_seq_mtx);
 		prof_dump(tsd, false, filename, false);
 	}
 }
 
 bool
-prof_mdump(const char *filename)
+prof_mdump(tsd_t *tsd, const char *filename)
 {
-	tsd_t *tsd;
 	char filename_buf[DUMP_FILENAME_BUFSIZE];
 
 	cassert(config_prof);
 
 	if (!opt_prof || !prof_booted)
 		return (true);
-	tsd = tsd_fetch();
 
 	if (filename == NULL) {
 		/* No filename specified, so automatically generate one. */
 		if (opt_prof_prefix[0] == '\0')
 			return (true);
-		malloc_mutex_lock(&prof_dump_seq_mtx);
+		malloc_mutex_lock(tsd_tsdn(tsd), &prof_dump_seq_mtx);
 		prof_dump_filename(filename_buf, 'm', prof_dump_mseq);
 		prof_dump_mseq++;
-		malloc_mutex_unlock(&prof_dump_seq_mtx);
+		malloc_mutex_unlock(tsd_tsdn(tsd), &prof_dump_seq_mtx);
 		filename = filename_buf;
 	}
 	return (prof_dump(tsd, true, filename, false));
 }
 
 void
-prof_gdump(void)
+prof_gdump(tsdn_t *tsdn)
 {
 	tsd_t *tsd;
 	prof_tdata_t *tdata;
 
 	cassert(config_prof);
 
-	if (!prof_booted)
+	if (!prof_booted || tsdn_null(tsdn))
 		return;
-	tsd = tsd_fetch();
+	tsd = tsdn_tsd(tsdn);
 	tdata = prof_tdata_get(tsd, false);
 	if (tdata == NULL)
 		return;
@@ -1659,10 +1745,10 @@ prof_gdump(void)
 
 	if (opt_prof_prefix[0] != '\0') {
 		char filename[DUMP_FILENAME_BUFSIZE];
-		malloc_mutex_lock(&prof_dump_seq_mtx);
+		malloc_mutex_lock(tsdn, &prof_dump_seq_mtx);
 		prof_dump_filename(filename, 'u', prof_dump_useq);
 		prof_dump_useq++;
-		malloc_mutex_unlock(&prof_dump_seq_mtx);
+		malloc_mutex_unlock(tsdn, &prof_dump_seq_mtx);
 		prof_dump(tsd, false, filename, false);
 	}
 }
@@ -1691,14 +1777,14 @@ prof_bt_keycomp(const void *k1, const void *k2)
 }
 
 JEMALLOC_INLINE_C uint64_t
-prof_thr_uid_alloc(void)
+prof_thr_uid_alloc(tsdn_t *tsdn)
 {
 	uint64_t thr_uid;
 
-	malloc_mutex_lock(&next_thr_uid_mtx);
+	malloc_mutex_lock(tsdn, &next_thr_uid_mtx);
 	thr_uid = next_thr_uid;
 	next_thr_uid++;
-	malloc_mutex_unlock(&next_thr_uid_mtx);
+	malloc_mutex_unlock(tsdn, &next_thr_uid_mtx);
 
 	return (thr_uid);
 }
@@ -1708,14 +1794,13 @@ prof_tdata_init_impl(tsd_t *tsd, uint64_t thr_uid, uint64_t thr_discrim,
     char *thread_name, bool active)
 {
 	prof_tdata_t *tdata;
-	tcache_t *tcache;
 
 	cassert(config_prof);
 
 	/* Initialize an empty cache for this thread. */
-	tcache = tcache_get(tsd, true);
-	tdata = (prof_tdata_t *)iallocztm(tsd, sizeof(prof_tdata_t), false,
-	    tcache, true, NULL);
+	tdata = (prof_tdata_t *)iallocztm(tsd_tsdn(tsd), sizeof(prof_tdata_t),
+	    size2index(sizeof(prof_tdata_t)), false, NULL, true,
+	    arena_get(TSDN_NULL, 0, true), true);
 	if (tdata == NULL)
 		return (NULL);
 
@@ -1727,9 +1812,9 @@ prof_tdata_init_impl(tsd_t *tsd, uint64_t thr_uid, uint64_t thr_discrim,
 	tdata->expired = false;
 	tdata->tctx_uid_next = 0;
 
-	if (ckh_new(tsd, &tdata->bt2tctx, PROF_CKH_MINITEMS,
-	    prof_bt_hash, prof_bt_keycomp)) {
-		idalloctm(tsd, tdata, tcache, true);
+	if (ckh_new(tsd, &tdata->bt2tctx, PROF_CKH_MINITEMS, prof_bt_hash,
+	    prof_bt_keycomp)) {
+		idalloctm(tsd_tsdn(tsd), tdata, NULL, true, true);
 		return (NULL);
 	}
 
@@ -1743,9 +1828,9 @@ prof_tdata_init_impl(tsd_t *tsd, uint64_t thr_uid, uint64_t thr_discrim,
 	tdata->dumping = false;
 	tdata->active = active;
 
-	malloc_mutex_lock(&tdatas_mtx);
+	malloc_mutex_lock(tsd_tsdn(tsd), &tdatas_mtx);
 	tdata_tree_insert(&tdatas, tdata);
-	malloc_mutex_unlock(&tdatas_mtx);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &tdatas_mtx);
 
 	return (tdata);
 }
@@ -1754,13 +1839,12 @@ prof_tdata_t *
 prof_tdata_init(tsd_t *tsd)
 {
 
-	return (prof_tdata_init_impl(tsd, prof_thr_uid_alloc(), 0, NULL,
-	    prof_thread_active_init_get()));
+	return (prof_tdata_init_impl(tsd, prof_thr_uid_alloc(tsd_tsdn(tsd)), 0,
+	    NULL, prof_thread_active_init_get(tsd_tsdn(tsd))));
 }
 
-/* tdata->lock must be held. */
 static bool
-prof_tdata_should_destroy(prof_tdata_t *tdata, bool even_if_attached)
+prof_tdata_should_destroy_unlocked(prof_tdata_t *tdata, bool even_if_attached)
 {
 
 	if (tdata->attached && !even_if_attached)
@@ -1770,32 +1854,40 @@ prof_tdata_should_destroy(prof_tdata_t *tdata, bool even_if_attached)
 	return (true);
 }
 
-/* tdatas_mtx must be held. */
+static bool
+prof_tdata_should_destroy(tsdn_t *tsdn, prof_tdata_t *tdata,
+    bool even_if_attached)
+{
+
+	malloc_mutex_assert_owner(tsdn, tdata->lock);
+
+	return (prof_tdata_should_destroy_unlocked(tdata, even_if_attached));
+}
+
 static void
 prof_tdata_destroy_locked(tsd_t *tsd, prof_tdata_t *tdata,
     bool even_if_attached)
 {
-	tcache_t *tcache;
 
-	assert(prof_tdata_should_destroy(tdata, even_if_attached));
-	assert(tsd_prof_tdata_get(tsd) != tdata);
+	malloc_mutex_assert_owner(tsd_tsdn(tsd), &tdatas_mtx);
 
 	tdata_tree_remove(&tdatas, tdata);
 
-	tcache = tcache_get(tsd, false);
+	assert(prof_tdata_should_destroy_unlocked(tdata, even_if_attached));
+
 	if (tdata->thread_name != NULL)
-		idalloctm(tsd, tdata->thread_name, tcache, true);
+		idalloctm(tsd_tsdn(tsd), tdata->thread_name, NULL, true, true);
 	ckh_delete(tsd, &tdata->bt2tctx);
-	idalloctm(tsd, tdata, tcache, true);
+	idalloctm(tsd_tsdn(tsd), tdata, NULL, true, true);
 }
 
 static void
 prof_tdata_destroy(tsd_t *tsd, prof_tdata_t *tdata, bool even_if_attached)
 {
 
-	malloc_mutex_lock(&tdatas_mtx);
+	malloc_mutex_lock(tsd_tsdn(tsd), &tdatas_mtx);
 	prof_tdata_destroy_locked(tsd, tdata, even_if_attached);
-	malloc_mutex_unlock(&tdatas_mtx);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &tdatas_mtx);
 }
 
 static void
@@ -1803,9 +1895,10 @@ prof_tdata_detach(tsd_t *tsd, prof_tdata_t *tdata)
 {
 	bool destroy_tdata;
 
-	malloc_mutex_lock(tdata->lock);
+	malloc_mutex_lock(tsd_tsdn(tsd), tdata->lock);
 	if (tdata->attached) {
-		destroy_tdata = prof_tdata_should_destroy(tdata, true);
+		destroy_tdata = prof_tdata_should_destroy(tsd_tsdn(tsd), tdata,
+		    true);
 		/*
 		 * Only detach if !destroy_tdata, because detaching would allow
 		 * another thread to win the race to destroy tdata.
@@ -1815,7 +1908,7 @@ prof_tdata_detach(tsd_t *tsd, prof_tdata_t *tdata)
 		tsd_prof_tdata_set(tsd, NULL);
 	} else
 		destroy_tdata = false;
-	malloc_mutex_unlock(tdata->lock);
+	malloc_mutex_unlock(tsd_tsdn(tsd), tdata->lock);
 	if (destroy_tdata)
 		prof_tdata_destroy(tsd, tdata, true);
 }
@@ -1826,7 +1919,7 @@ prof_tdata_reinit(tsd_t *tsd, prof_tdata_t *tdata)
 	uint64_t thr_uid = tdata->thr_uid;
 	uint64_t thr_discrim = tdata->thr_discrim + 1;
 	char *thread_name = (tdata->thread_name != NULL) ?
-	    prof_thread_name_alloc(tsd, tdata->thread_name) : NULL;
+	    prof_thread_name_alloc(tsd_tsdn(tsd), tdata->thread_name) : NULL;
 	bool active = tdata->active;
 
 	prof_tdata_detach(tsd, tdata);
@@ -1835,18 +1928,18 @@ prof_tdata_reinit(tsd_t *tsd, prof_tdata_t *tdata)
 }
 
 static bool
-prof_tdata_expire(prof_tdata_t *tdata)
+prof_tdata_expire(tsdn_t *tsdn, prof_tdata_t *tdata)
 {
 	bool destroy_tdata;
 
-	malloc_mutex_lock(tdata->lock);
+	malloc_mutex_lock(tsdn, tdata->lock);
 	if (!tdata->expired) {
 		tdata->expired = true;
 		destroy_tdata = tdata->attached ? false :
-		    prof_tdata_should_destroy(tdata, false);
+		    prof_tdata_should_destroy(tsdn, tdata, false);
 	} else
 		destroy_tdata = false;
-	malloc_mutex_unlock(tdata->lock);
+	malloc_mutex_unlock(tsdn, tdata->lock);
 
 	return (destroy_tdata);
 }
@@ -1854,8 +1947,9 @@ prof_tdata_expire(prof_tdata_t *tdata)
 static prof_tdata_t *
 prof_tdata_reset_iter(prof_tdata_tree_t *tdatas, prof_tdata_t *tdata, void *arg)
 {
+	tsdn_t *tsdn = (tsdn_t *)arg;
 
-	return (prof_tdata_expire(tdata) ? tdata : NULL);
+	return (prof_tdata_expire(tsdn, tdata) ? tdata : NULL);
 }
 
 void
@@ -1865,15 +1959,15 @@ prof_reset(tsd_t *tsd, size_t lg_sample)
 
 	assert(lg_sample < (sizeof(uint64_t) << 3));
 
-	malloc_mutex_lock(&prof_dump_mtx);
-	malloc_mutex_lock(&tdatas_mtx);
+	malloc_mutex_lock(tsd_tsdn(tsd), &prof_dump_mtx);
+	malloc_mutex_lock(tsd_tsdn(tsd), &tdatas_mtx);
 
 	lg_prof_sample = lg_sample;
 
 	next = NULL;
 	do {
 		prof_tdata_t *to_destroy = tdata_tree_iter(&tdatas, next,
-		    prof_tdata_reset_iter, NULL);
+		    prof_tdata_reset_iter, (void *)tsd);
 		if (to_destroy != NULL) {
 			next = tdata_tree_next(&tdatas, to_destroy);
 			prof_tdata_destroy_locked(tsd, to_destroy, false);
@@ -1881,8 +1975,8 @@ prof_reset(tsd_t *tsd, size_t lg_sample)
 			next = NULL;
 	} while (next != NULL);
 
-	malloc_mutex_unlock(&tdatas_mtx);
-	malloc_mutex_unlock(&prof_dump_mtx);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &tdatas_mtx);
+	malloc_mutex_unlock(tsd_tsdn(tsd), &prof_dump_mtx);
 }
 
 void
@@ -1899,35 +1993,33 @@ prof_tdata_cleanup(tsd_t *tsd)
 }
 
 bool
-prof_active_get(void)
+prof_active_get(tsdn_t *tsdn)
 {
 	bool prof_active_current;
 
-	malloc_mutex_lock(&prof_active_mtx);
+	malloc_mutex_lock(tsdn, &prof_active_mtx);
 	prof_active_current = prof_active;
-	malloc_mutex_unlock(&prof_active_mtx);
+	malloc_mutex_unlock(tsdn, &prof_active_mtx);
 	return (prof_active_current);
 }
 
 bool
-prof_active_set(bool active)
+prof_active_set(tsdn_t *tsdn, bool active)
 {
 	bool prof_active_old;
 
-	malloc_mutex_lock(&prof_active_mtx);
+	malloc_mutex_lock(tsdn, &prof_active_mtx);
 	prof_active_old = prof_active;
 	prof_active = active;
-	malloc_mutex_unlock(&prof_active_mtx);
+	malloc_mutex_unlock(tsdn, &prof_active_mtx);
 	return (prof_active_old);
 }
 
 const char *
-prof_thread_name_get(void)
+prof_thread_name_get(tsd_t *tsd)
 {
-	tsd_t *tsd;
 	prof_tdata_t *tdata;
 
-	tsd = tsd_fetch();
 	tdata = prof_tdata_get(tsd, true);
 	if (tdata == NULL)
 		return ("");
@@ -1935,7 +2027,7 @@ prof_thread_name_get(void)
 }
 
 static char *
-prof_thread_name_alloc(tsd_t *tsd, const char *thread_name)
+prof_thread_name_alloc(tsdn_t *tsdn, const char *thread_name)
 {
 	char *ret;
 	size_t size;
@@ -1947,7 +2039,8 @@ prof_thread_name_alloc(tsd_t *tsd, const char *thread_name)
 	if (size == 1)
 		return ("");
 
-	ret = iallocztm(tsd, size, false, tcache_get(tsd, true), true, NULL);
+	ret = iallocztm(tsdn, size, size2index(size), false, NULL, true,
+	    arena_get(TSDN_NULL, 0, true), true);
 	if (ret == NULL)
 		return (NULL);
 	memcpy(ret, thread_name, size);
@@ -1974,13 +2067,12 @@ prof_thread_name_set(tsd_t *tsd, const char *thread_name)
 			return (EFAULT);
 	}
 
-	s = prof_thread_name_alloc(tsd, thread_name);
+	s = prof_thread_name_alloc(tsd_tsdn(tsd), thread_name);
 	if (s == NULL)
 		return (EAGAIN);
 
 	if (tdata->thread_name != NULL) {
-		idalloctm(tsd, tdata->thread_name, tcache_get(tsd, false),
-		    true);
+		idalloctm(tsd_tsdn(tsd), tdata->thread_name, NULL, true, true);
 		tdata->thread_name = NULL;
 	}
 	if (strlen(s) > 0)
@@ -1989,12 +2081,10 @@ prof_thread_name_set(tsd_t *tsd, const char *thread_name)
 }
 
 bool
-prof_thread_active_get(void)
+prof_thread_active_get(tsd_t *tsd)
 {
-	tsd_t *tsd;
 	prof_tdata_t *tdata;
 
-	tsd = tsd_fetch();
 	tdata = prof_tdata_get(tsd, true);
 	if (tdata == NULL)
 		return (false);
@@ -2002,12 +2092,10 @@ prof_thread_active_get(void)
 }
 
 bool
-prof_thread_active_set(bool active)
+prof_thread_active_set(tsd_t *tsd, bool active)
 {
-	tsd_t *tsd;
 	prof_tdata_t *tdata;
 
-	tsd = tsd_fetch();
 	tdata = prof_tdata_get(tsd, true);
 	if (tdata == NULL)
 		return (true);
@@ -2016,48 +2104,48 @@ prof_thread_active_set(bool active)
 }
 
 bool
-prof_thread_active_init_get(void)
+prof_thread_active_init_get(tsdn_t *tsdn)
 {
 	bool active_init;
 
-	malloc_mutex_lock(&prof_thread_active_init_mtx);
+	malloc_mutex_lock(tsdn, &prof_thread_active_init_mtx);
 	active_init = prof_thread_active_init;
-	malloc_mutex_unlock(&prof_thread_active_init_mtx);
+	malloc_mutex_unlock(tsdn, &prof_thread_active_init_mtx);
 	return (active_init);
 }
 
 bool
-prof_thread_active_init_set(bool active_init)
+prof_thread_active_init_set(tsdn_t *tsdn, bool active_init)
 {
 	bool active_init_old;
 
-	malloc_mutex_lock(&prof_thread_active_init_mtx);
+	malloc_mutex_lock(tsdn, &prof_thread_active_init_mtx);
 	active_init_old = prof_thread_active_init;
 	prof_thread_active_init = active_init;
-	malloc_mutex_unlock(&prof_thread_active_init_mtx);
+	malloc_mutex_unlock(tsdn, &prof_thread_active_init_mtx);
 	return (active_init_old);
 }
 
 bool
-prof_gdump_get(void)
+prof_gdump_get(tsdn_t *tsdn)
 {
 	bool prof_gdump_current;
 
-	malloc_mutex_lock(&prof_gdump_mtx);
+	malloc_mutex_lock(tsdn, &prof_gdump_mtx);
 	prof_gdump_current = prof_gdump_val;
-	malloc_mutex_unlock(&prof_gdump_mtx);
+	malloc_mutex_unlock(tsdn, &prof_gdump_mtx);
 	return (prof_gdump_current);
 }
 
 bool
-prof_gdump_set(bool gdump)
+prof_gdump_set(tsdn_t *tsdn, bool gdump)
 {
 	bool prof_gdump_old;
 
-	malloc_mutex_lock(&prof_gdump_mtx);
+	malloc_mutex_lock(tsdn, &prof_gdump_mtx);
 	prof_gdump_old = prof_gdump_val;
 	prof_gdump_val = gdump;
-	malloc_mutex_unlock(&prof_gdump_mtx);
+	malloc_mutex_unlock(tsdn, &prof_gdump_mtx);
 	return (prof_gdump_old);
 }
 
@@ -2098,47 +2186,54 @@ prof_boot1(void)
 }
 
 bool
-prof_boot2(void)
+prof_boot2(tsd_t *tsd)
 {
 
 	cassert(config_prof);
 
 	if (opt_prof) {
-		tsd_t *tsd;
 		unsigned i;
 
 		lg_prof_sample = opt_lg_prof_sample;
 
 		prof_active = opt_prof_active;
-		if (malloc_mutex_init(&prof_active_mtx))
+		if (malloc_mutex_init(&prof_active_mtx, "prof_active",
+		    WITNESS_RANK_PROF_ACTIVE))
 			return (true);
 
 		prof_gdump_val = opt_prof_gdump;
-		if (malloc_mutex_init(&prof_gdump_mtx))
+		if (malloc_mutex_init(&prof_gdump_mtx, "prof_gdump",
+		    WITNESS_RANK_PROF_GDUMP))
 			return (true);
 
 		prof_thread_active_init = opt_prof_thread_active_init;
-		if (malloc_mutex_init(&prof_thread_active_init_mtx))
+		if (malloc_mutex_init(&prof_thread_active_init_mtx,
+		    "prof_thread_active_init",
+		    WITNESS_RANK_PROF_THREAD_ACTIVE_INIT))
 			return (true);
 
-		tsd = tsd_fetch();
 		if (ckh_new(tsd, &bt2gctx, PROF_CKH_MINITEMS, prof_bt_hash,
 		    prof_bt_keycomp))
 			return (true);
-		if (malloc_mutex_init(&bt2gctx_mtx))
+		if (malloc_mutex_init(&bt2gctx_mtx, "prof_bt2gctx",
+		    WITNESS_RANK_PROF_BT2GCTX))
 			return (true);
 
 		tdata_tree_new(&tdatas);
-		if (malloc_mutex_init(&tdatas_mtx))
+		if (malloc_mutex_init(&tdatas_mtx, "prof_tdatas",
+		    WITNESS_RANK_PROF_TDATAS))
 			return (true);
 
 		next_thr_uid = 0;
-		if (malloc_mutex_init(&next_thr_uid_mtx))
+		if (malloc_mutex_init(&next_thr_uid_mtx, "prof_next_thr_uid",
+		    WITNESS_RANK_PROF_NEXT_THR_UID))
 			return (true);
 
-		if (malloc_mutex_init(&prof_dump_seq_mtx))
+		if (malloc_mutex_init(&prof_dump_seq_mtx, "prof_dump_seq",
+		    WITNESS_RANK_PROF_DUMP_SEQ))
 			return (true);
-		if (malloc_mutex_init(&prof_dump_mtx))
+		if (malloc_mutex_init(&prof_dump_mtx, "prof_dump",
+		    WITNESS_RANK_PROF_DUMP))
 			return (true);
 
 		if (opt_prof_final && opt_prof_prefix[0] != '\0' &&
@@ -2148,21 +2243,23 @@ prof_boot2(void)
 				abort();
 		}
 
-		gctx_locks = (malloc_mutex_t *)base_alloc(PROF_NCTX_LOCKS *
-		    sizeof(malloc_mutex_t));
+		gctx_locks = (malloc_mutex_t *)base_alloc(tsd_tsdn(tsd),
+		    PROF_NCTX_LOCKS * sizeof(malloc_mutex_t));
 		if (gctx_locks == NULL)
 			return (true);
 		for (i = 0; i < PROF_NCTX_LOCKS; i++) {
-			if (malloc_mutex_init(&gctx_locks[i]))
+			if (malloc_mutex_init(&gctx_locks[i], "prof_gctx",
+			    WITNESS_RANK_PROF_GCTX))
 				return (true);
 		}
 
-		tdata_locks = (malloc_mutex_t *)base_alloc(PROF_NTDATA_LOCKS *
-		    sizeof(malloc_mutex_t));
+		tdata_locks = (malloc_mutex_t *)base_alloc(tsd_tsdn(tsd),
+		    PROF_NTDATA_LOCKS * sizeof(malloc_mutex_t));
 		if (tdata_locks == NULL)
 			return (true);
 		for (i = 0; i < PROF_NTDATA_LOCKS; i++) {
-			if (malloc_mutex_init(&tdata_locks[i]))
+			if (malloc_mutex_init(&tdata_locks[i], "prof_tdata",
+			    WITNESS_RANK_PROF_TDATA))
 				return (true);
 		}
 	}
@@ -2181,56 +2278,77 @@ prof_boot2(void)
 }
 
 void
-prof_prefork(void)
+prof_prefork0(tsdn_t *tsdn)
 {
 
 	if (opt_prof) {
 		unsigned i;
 
-		malloc_mutex_prefork(&tdatas_mtx);
-		malloc_mutex_prefork(&bt2gctx_mtx);
-		malloc_mutex_prefork(&next_thr_uid_mtx);
-		malloc_mutex_prefork(&prof_dump_seq_mtx);
-		for (i = 0; i < PROF_NCTX_LOCKS; i++)
-			malloc_mutex_prefork(&gctx_locks[i]);
+		malloc_mutex_prefork(tsdn, &prof_dump_mtx);
+		malloc_mutex_prefork(tsdn, &bt2gctx_mtx);
+		malloc_mutex_prefork(tsdn, &tdatas_mtx);
 		for (i = 0; i < PROF_NTDATA_LOCKS; i++)
-			malloc_mutex_prefork(&tdata_locks[i]);
+			malloc_mutex_prefork(tsdn, &tdata_locks[i]);
+		for (i = 0; i < PROF_NCTX_LOCKS; i++)
+			malloc_mutex_prefork(tsdn, &gctx_locks[i]);
 	}
 }
 
 void
-prof_postfork_parent(void)
+prof_prefork1(tsdn_t *tsdn)
+{
+
+	if (opt_prof) {
+		malloc_mutex_prefork(tsdn, &prof_active_mtx);
+		malloc_mutex_prefork(tsdn, &prof_dump_seq_mtx);
+		malloc_mutex_prefork(tsdn, &prof_gdump_mtx);
+		malloc_mutex_prefork(tsdn, &next_thr_uid_mtx);
+		malloc_mutex_prefork(tsdn, &prof_thread_active_init_mtx);
+	}
+}
+
+void
+prof_postfork_parent(tsdn_t *tsdn)
 {
 
 	if (opt_prof) {
 		unsigned i;
 
-		for (i = 0; i < PROF_NTDATA_LOCKS; i++)
-			malloc_mutex_postfork_parent(&tdata_locks[i]);
+		malloc_mutex_postfork_parent(tsdn,
+		    &prof_thread_active_init_mtx);
+		malloc_mutex_postfork_parent(tsdn, &next_thr_uid_mtx);
+		malloc_mutex_postfork_parent(tsdn, &prof_gdump_mtx);
+		malloc_mutex_postfork_parent(tsdn, &prof_dump_seq_mtx);
+		malloc_mutex_postfork_parent(tsdn, &prof_active_mtx);
 		for (i = 0; i < PROF_NCTX_LOCKS; i++)
-			malloc_mutex_postfork_parent(&gctx_locks[i]);
-		malloc_mutex_postfork_parent(&prof_dump_seq_mtx);
-		malloc_mutex_postfork_parent(&next_thr_uid_mtx);
-		malloc_mutex_postfork_parent(&bt2gctx_mtx);
-		malloc_mutex_postfork_parent(&tdatas_mtx);
+			malloc_mutex_postfork_parent(tsdn, &gctx_locks[i]);
+		for (i = 0; i < PROF_NTDATA_LOCKS; i++)
+			malloc_mutex_postfork_parent(tsdn, &tdata_locks[i]);
+		malloc_mutex_postfork_parent(tsdn, &tdatas_mtx);
+		malloc_mutex_postfork_parent(tsdn, &bt2gctx_mtx);
+		malloc_mutex_postfork_parent(tsdn, &prof_dump_mtx);
 	}
 }
 
 void
-prof_postfork_child(void)
+prof_postfork_child(tsdn_t *tsdn)
 {
 
 	if (opt_prof) {
 		unsigned i;
 
-		for (i = 0; i < PROF_NTDATA_LOCKS; i++)
-			malloc_mutex_postfork_child(&tdata_locks[i]);
+		malloc_mutex_postfork_child(tsdn, &prof_thread_active_init_mtx);
+		malloc_mutex_postfork_child(tsdn, &next_thr_uid_mtx);
+		malloc_mutex_postfork_child(tsdn, &prof_gdump_mtx);
+		malloc_mutex_postfork_child(tsdn, &prof_dump_seq_mtx);
+		malloc_mutex_postfork_child(tsdn, &prof_active_mtx);
 		for (i = 0; i < PROF_NCTX_LOCKS; i++)
-			malloc_mutex_postfork_child(&gctx_locks[i]);
-		malloc_mutex_postfork_child(&prof_dump_seq_mtx);
-		malloc_mutex_postfork_child(&next_thr_uid_mtx);
-		malloc_mutex_postfork_child(&bt2gctx_mtx);
-		malloc_mutex_postfork_child(&tdatas_mtx);
+			malloc_mutex_postfork_child(tsdn, &gctx_locks[i]);
+		for (i = 0; i < PROF_NTDATA_LOCKS; i++)
+			malloc_mutex_postfork_child(tsdn, &tdata_locks[i]);
+		malloc_mutex_postfork_child(tsdn, &tdatas_mtx);
+		malloc_mutex_postfork_child(tsdn, &bt2gctx_mtx);
+		malloc_mutex_postfork_child(tsdn, &prof_dump_mtx);
 	}
 }
 
diff --git a/deps/jemalloc/src/quarantine.c b/deps/jemalloc/src/quarantine.c
index 6c43dfcaa..18903fb5c 100644
--- a/deps/jemalloc/src/quarantine.c
+++ b/deps/jemalloc/src/quarantine.c
@@ -13,22 +13,22 @@
 /* Function prototypes for non-inline static functions. */
 
 static quarantine_t	*quarantine_grow(tsd_t *tsd, quarantine_t *quarantine);
-static void	quarantine_drain_one(tsd_t *tsd, quarantine_t *quarantine);
-static void	quarantine_drain(tsd_t *tsd, quarantine_t *quarantine,
+static void	quarantine_drain_one(tsdn_t *tsdn, quarantine_t *quarantine);
+static void	quarantine_drain(tsdn_t *tsdn, quarantine_t *quarantine,
     size_t upper_bound);
 
 /******************************************************************************/
 
 static quarantine_t *
-quarantine_init(tsd_t *tsd, size_t lg_maxobjs)
+quarantine_init(tsdn_t *tsdn, size_t lg_maxobjs)
 {
 	quarantine_t *quarantine;
+	size_t size;
 
-	assert(tsd_nominal(tsd));
-
-	quarantine = (quarantine_t *)iallocztm(tsd, offsetof(quarantine_t, objs)
-	    + ((ZU(1) << lg_maxobjs) * sizeof(quarantine_obj_t)), false,
-	    tcache_get(tsd, true), true, NULL);
+	size = offsetof(quarantine_t, objs) + ((ZU(1) << lg_maxobjs) *
+	    sizeof(quarantine_obj_t));
+	quarantine = (quarantine_t *)iallocztm(tsdn, size, size2index(size),
+	    false, NULL, true, arena_get(TSDN_NULL, 0, true), true);
 	if (quarantine == NULL)
 		return (NULL);
 	quarantine->curbytes = 0;
@@ -47,7 +47,7 @@ quarantine_alloc_hook_work(tsd_t *tsd)
 	if (!tsd_nominal(tsd))
 		return;
 
-	quarantine = quarantine_init(tsd, LG_MAXOBJS_INIT);
+	quarantine = quarantine_init(tsd_tsdn(tsd), LG_MAXOBJS_INIT);
 	/*
 	 * Check again whether quarantine has been initialized, because
 	 * quarantine_init() may have triggered recursive initialization.
@@ -55,7 +55,7 @@ quarantine_alloc_hook_work(tsd_t *tsd)
 	if (tsd_quarantine_get(tsd) == NULL)
 		tsd_quarantine_set(tsd, quarantine);
 	else
-		idalloctm(tsd, quarantine, tcache_get(tsd, false), true);
+		idalloctm(tsd_tsdn(tsd), quarantine, NULL, true, true);
 }
 
 static quarantine_t *
@@ -63,9 +63,9 @@ quarantine_grow(tsd_t *tsd, quarantine_t *quarantine)
 {
 	quarantine_t *ret;
 
-	ret = quarantine_init(tsd, quarantine->lg_maxobjs + 1);
+	ret = quarantine_init(tsd_tsdn(tsd), quarantine->lg_maxobjs + 1);
 	if (ret == NULL) {
-		quarantine_drain_one(tsd, quarantine);
+		quarantine_drain_one(tsd_tsdn(tsd), quarantine);
 		return (quarantine);
 	}
 
@@ -87,18 +87,18 @@ quarantine_grow(tsd_t *tsd, quarantine_t *quarantine)
 		memcpy(&ret->objs[ncopy_a], quarantine->objs, ncopy_b *
 		    sizeof(quarantine_obj_t));
 	}
-	idalloctm(tsd, quarantine, tcache_get(tsd, false), true);
+	idalloctm(tsd_tsdn(tsd), quarantine, NULL, true, true);
 
 	tsd_quarantine_set(tsd, ret);
 	return (ret);
 }
 
 static void
-quarantine_drain_one(tsd_t *tsd, quarantine_t *quarantine)
+quarantine_drain_one(tsdn_t *tsdn, quarantine_t *quarantine)
 {
 	quarantine_obj_t *obj = &quarantine->objs[quarantine->first];
-	assert(obj->usize == isalloc(obj->ptr, config_prof));
-	idalloctm(tsd, obj->ptr, NULL, false);
+	assert(obj->usize == isalloc(tsdn, obj->ptr, config_prof));
+	idalloctm(tsdn, obj->ptr, NULL, false, true);
 	quarantine->curbytes -= obj->usize;
 	quarantine->curobjs--;
 	quarantine->first = (quarantine->first + 1) & ((ZU(1) <<
@@ -106,24 +106,24 @@ quarantine_drain_one(tsd_t *tsd, quarantine_t *quarantine)
 }
 
 static void
-quarantine_drain(tsd_t *tsd, quarantine_t *quarantine, size_t upper_bound)
+quarantine_drain(tsdn_t *tsdn, quarantine_t *quarantine, size_t upper_bound)
 {
 
 	while (quarantine->curbytes > upper_bound && quarantine->curobjs > 0)
-		quarantine_drain_one(tsd, quarantine);
+		quarantine_drain_one(tsdn, quarantine);
 }
 
 void
 quarantine(tsd_t *tsd, void *ptr)
 {
 	quarantine_t *quarantine;
-	size_t usize = isalloc(ptr, config_prof);
+	size_t usize = isalloc(tsd_tsdn(tsd), ptr, config_prof);
 
 	cassert(config_fill);
 	assert(opt_quarantine);
 
 	if ((quarantine = tsd_quarantine_get(tsd)) == NULL) {
-		idalloctm(tsd, ptr, NULL, false);
+		idalloctm(tsd_tsdn(tsd), ptr, NULL, false, true);
 		return;
 	}
 	/*
@@ -133,7 +133,7 @@ quarantine(tsd_t *tsd, void *ptr)
 	if (quarantine->curbytes + usize > opt_quarantine) {
 		size_t upper_bound = (opt_quarantine >= usize) ? opt_quarantine
 		    - usize : 0;
-		quarantine_drain(tsd, quarantine, upper_bound);
+		quarantine_drain(tsd_tsdn(tsd), quarantine, upper_bound);
 	}
 	/* Grow the quarantine ring buffer if it's full. */
 	if (quarantine->curobjs == (ZU(1) << quarantine->lg_maxobjs))
@@ -158,11 +158,11 @@ quarantine(tsd_t *tsd, void *ptr)
 			    && usize <= SMALL_MAXCLASS)
 				arena_quarantine_junk_small(ptr, usize);
 			else
-				memset(ptr, 0x5a, usize);
+				memset(ptr, JEMALLOC_FREE_JUNK, usize);
 		}
 	} else {
 		assert(quarantine->curbytes == 0);
-		idalloctm(tsd, ptr, NULL, false);
+		idalloctm(tsd_tsdn(tsd), ptr, NULL, false, true);
 	}
 }
 
@@ -176,8 +176,8 @@ quarantine_cleanup(tsd_t *tsd)
 
 	quarantine = tsd_quarantine_get(tsd);
 	if (quarantine != NULL) {
-		quarantine_drain(tsd, quarantine, 0);
-		idalloctm(tsd, quarantine, tcache_get(tsd, false), true);
+		quarantine_drain(tsd_tsdn(tsd), quarantine, 0);
+		idalloctm(tsd_tsdn(tsd), quarantine, NULL, true, true);
 		tsd_quarantine_set(tsd, NULL);
 	}
 }
diff --git a/deps/jemalloc/src/rtree.c b/deps/jemalloc/src/rtree.c
index af0d97e75..f2e2997d5 100644
--- a/deps/jemalloc/src/rtree.c
+++ b/deps/jemalloc/src/rtree.c
@@ -15,6 +15,8 @@ rtree_new(rtree_t *rtree, unsigned bits, rtree_node_alloc_t *alloc,
 {
 	unsigned bits_in_leaf, height, i;
 
+	assert(RTREE_HEIGHT_MAX == ((ZU(1) << (LG_SIZEOF_PTR+3)) /
+	    RTREE_BITS_PER_LEVEL));
 	assert(bits > 0 && bits <= (sizeof(uintptr_t) << 3));
 
 	bits_in_leaf = (bits % RTREE_BITS_PER_LEVEL) == 0 ? RTREE_BITS_PER_LEVEL
@@ -94,12 +96,15 @@ rtree_node_init(rtree_t *rtree, unsigned level, rtree_node_elm_t **elmp)
 	rtree_node_elm_t *node;
 
 	if (atomic_cas_p((void **)elmp, NULL, RTREE_NODE_INITIALIZING)) {
+		spin_t spinner;
+
 		/*
 		 * Another thread is already in the process of initializing.
 		 * Spin-wait until initialization is complete.
 		 */
+		spin_init(&spinner);
 		do {
-			CPU_SPINWAIT;
+			spin_adaptive(&spinner);
 			node = atomic_read_p((void **)elmp);
 		} while (node == RTREE_NODE_INITIALIZING);
 	} else {
@@ -123,5 +128,5 @@ rtree_node_elm_t *
 rtree_child_read_hard(rtree_t *rtree, rtree_node_elm_t *elm, unsigned level)
 {
 
-	return (rtree_node_init(rtree, level, &elm->child));
+	return (rtree_node_init(rtree, level+1, &elm->child));
 }
diff --git a/deps/jemalloc/src/spin.c b/deps/jemalloc/src/spin.c
new file mode 100644
index 000000000..5242d95aa
--- /dev/null
+++ b/deps/jemalloc/src/spin.c
@@ -0,0 +1,2 @@
+#define	JEMALLOC_SPIN_C_
+#include "jemalloc/internal/jemalloc_internal.h"
diff --git a/deps/jemalloc/src/stats.c b/deps/jemalloc/src/stats.c
old mode 100644
new mode 100755
index 154c3e74c..1360f3bd0
--- a/deps/jemalloc/src/stats.c
+++ b/deps/jemalloc/src/stats.c
@@ -3,7 +3,7 @@
 
 #define	CTL_GET(n, v, t) do {						\
 	size_t sz = sizeof(t);						\
-	xmallctl(n, v, &sz, NULL, 0);					\
+	xmallctl(n, (void *)v, &sz, NULL, 0);				\
 } while (0)
 
 #define	CTL_M2_GET(n, i, v, t) do {					\
@@ -12,7 +12,7 @@
 	size_t sz = sizeof(t);						\
 	xmallctlnametomib(n, mib, &miblen);				\
 	mib[2] = (i);							\
-	xmallctlbymib(mib, miblen, v, &sz, NULL, 0);			\
+	xmallctlbymib(mib, miblen, (void *)v, &sz, NULL, 0);		\
 } while (0)
 
 #define	CTL_M2_M4_GET(n, i, j, v, t) do {				\
@@ -22,7 +22,7 @@
 	xmallctlnametomib(n, mib, &miblen);				\
 	mib[2] = (i);							\
 	mib[4] = (j);							\
-	xmallctlbymib(mib, miblen, v, &sz, NULL, 0);			\
+	xmallctlbymib(mib, miblen, (void *)v, &sz, NULL, 0);		\
 } while (0)
 
 /******************************************************************************/
@@ -32,86 +32,107 @@ bool	opt_stats_print = false;
 
 size_t	stats_cactive = 0;
 
-/******************************************************************************/
-/* Function prototypes for non-inline static functions. */
-
-static void	stats_arena_bins_print(void (*write_cb)(void *, const char *),
-    void *cbopaque, unsigned i);
-static void	stats_arena_lruns_print(void (*write_cb)(void *, const char *),
-    void *cbopaque, unsigned i);
-static void	stats_arena_hchunks_print(
-    void (*write_cb)(void *, const char *), void *cbopaque, unsigned i);
-static void	stats_arena_print(void (*write_cb)(void *, const char *),
-    void *cbopaque, unsigned i, bool bins, bool large, bool huge);
-
 /******************************************************************************/
 
 static void
 stats_arena_bins_print(void (*write_cb)(void *, const char *), void *cbopaque,
-    unsigned i)
+    bool json, bool large, bool huge, unsigned i)
 {
 	size_t page;
-	bool config_tcache, in_gap;
+	bool config_tcache, in_gap, in_gap_prev;
 	unsigned nbins, j;
 
 	CTL_GET("arenas.page", &page, size_t);
 
-	CTL_GET("config.tcache", &config_tcache, bool);
-	if (config_tcache) {
+	CTL_GET("arenas.nbins", &nbins, unsigned);
+	if (json) {
 		malloc_cprintf(write_cb, cbopaque,
-		    "bins:           size ind    allocated      nmalloc"
-		    "      ndalloc    nrequests      curregs      curruns regs"
-		    " pgs  util       nfills     nflushes      newruns"
-		    "       reruns\n");
+		    "\t\t\t\t\"bins\": [\n");
 	} else {
-		malloc_cprintf(write_cb, cbopaque,
-		    "bins:           size ind    allocated      nmalloc"
-		    "      ndalloc    nrequests      curregs      curruns regs"
-		    " pgs  util      newruns       reruns\n");
+		CTL_GET("config.tcache", &config_tcache, bool);
+		if (config_tcache) {
+			malloc_cprintf(write_cb, cbopaque,
+			    "bins:           size ind    allocated      nmalloc"
+			    "      ndalloc    nrequests      curregs"
+			    "      curruns regs pgs  util       nfills"
+			    "     nflushes      newruns       reruns\n");
+		} else {
+			malloc_cprintf(write_cb, cbopaque,
+			    "bins:           size ind    allocated      nmalloc"
+			    "      ndalloc    nrequests      curregs"
+			    "      curruns regs pgs  util      newruns"
+			    "       reruns\n");
+		}
 	}
-	CTL_GET("arenas.nbins", &nbins, unsigned);
 	for (j = 0, in_gap = false; j < nbins; j++) {
 		uint64_t nruns;
+		size_t reg_size, run_size, curregs;
+		size_t curruns;
+		uint32_t nregs;
+		uint64_t nmalloc, ndalloc, nrequests, nfills, nflushes;
+		uint64_t nreruns;
 
 		CTL_M2_M4_GET("stats.arenas.0.bins.0.nruns", i, j, &nruns,
 		    uint64_t);
-		if (nruns == 0)
-			in_gap = true;
-		else {
-			size_t reg_size, run_size, curregs, availregs, milli;
-			size_t curruns;
-			uint32_t nregs;
-			uint64_t nmalloc, ndalloc, nrequests, nfills, nflushes;
-			uint64_t reruns;
-			char util[6]; /* "x.yyy". */
+		in_gap_prev = in_gap;
+		in_gap = (nruns == 0);
 
-			if (in_gap) {
-				malloc_cprintf(write_cb, cbopaque,
-				    "                     ---\n");
-				in_gap = false;
-			}
-			CTL_M2_GET("arenas.bin.0.size", j, &reg_size, size_t);
-			CTL_M2_GET("arenas.bin.0.nregs", j, &nregs, uint32_t);
-			CTL_M2_GET("arenas.bin.0.run_size", j, &run_size,
-			    size_t);
-			CTL_M2_M4_GET("stats.arenas.0.bins.0.nmalloc", i, j,
-			    &nmalloc, uint64_t);
-			CTL_M2_M4_GET("stats.arenas.0.bins.0.ndalloc", i, j,
-			    &ndalloc, uint64_t);
-			CTL_M2_M4_GET("stats.arenas.0.bins.0.curregs", i, j,
-			    &curregs, size_t);
-			CTL_M2_M4_GET("stats.arenas.0.bins.0.nrequests", i, j,
-			    &nrequests, uint64_t);
+		if (!json && in_gap_prev && !in_gap) {
+			malloc_cprintf(write_cb, cbopaque,
+			    "                     ---\n");
+		}
+
+		CTL_M2_GET("arenas.bin.0.size", j, &reg_size, size_t);
+		CTL_M2_GET("arenas.bin.0.nregs", j, &nregs, uint32_t);
+		CTL_M2_GET("arenas.bin.0.run_size", j, &run_size, size_t);
+
+		CTL_M2_M4_GET("stats.arenas.0.bins.0.nmalloc", i, j, &nmalloc,
+		    uint64_t);
+		CTL_M2_M4_GET("stats.arenas.0.bins.0.ndalloc", i, j, &ndalloc,
+		    uint64_t);
+		CTL_M2_M4_GET("stats.arenas.0.bins.0.curregs", i, j, &curregs,
+		    size_t);
+		CTL_M2_M4_GET("stats.arenas.0.bins.0.nrequests", i, j,
+		    &nrequests, uint64_t);
+		if (config_tcache) {
+			CTL_M2_M4_GET("stats.arenas.0.bins.0.nfills", i, j,
+			    &nfills, uint64_t);
+			CTL_M2_M4_GET("stats.arenas.0.bins.0.nflushes", i, j,
+			    &nflushes, uint64_t);
+		}
+		CTL_M2_M4_GET("stats.arenas.0.bins.0.nreruns", i, j, &nreruns,
+		    uint64_t);
+		CTL_M2_M4_GET("stats.arenas.0.bins.0.curruns", i, j, &curruns,
+		    size_t);
+
+		if (json) {
+			malloc_cprintf(write_cb, cbopaque,
+			    "\t\t\t\t\t{\n"
+			    "\t\t\t\t\t\t\"nmalloc\": %"FMTu64",\n"
+			    "\t\t\t\t\t\t\"ndalloc\": %"FMTu64",\n"
+			    "\t\t\t\t\t\t\"curregs\": %zu,\n"
+			    "\t\t\t\t\t\t\"nrequests\": %"FMTu64",\n",
+			    nmalloc,
+			    ndalloc,
+			    curregs,
+			    nrequests);
 			if (config_tcache) {
-				CTL_M2_M4_GET("stats.arenas.0.bins.0.nfills", i,
-				    j, &nfills, uint64_t);
-				CTL_M2_M4_GET("stats.arenas.0.bins.0.nflushes",
-				    i, j, &nflushes, uint64_t);
+				malloc_cprintf(write_cb, cbopaque,
+				    "\t\t\t\t\t\t\"nfills\": %"FMTu64",\n"
+				    "\t\t\t\t\t\t\"nflushes\": %"FMTu64",\n",
+				    nfills,
+				    nflushes);
 			}
-			CTL_M2_M4_GET("stats.arenas.0.bins.0.nreruns", i, j,
-			    &reruns, uint64_t);
-			CTL_M2_M4_GET("stats.arenas.0.bins.0.curruns", i, j,
-			    &curruns, size_t);
+			malloc_cprintf(write_cb, cbopaque,
+			    "\t\t\t\t\t\t\"nreruns\": %"FMTu64",\n"
+			    "\t\t\t\t\t\t\"curruns\": %zu\n"
+			    "\t\t\t\t\t}%s\n",
+			    nreruns,
+			    curruns,
+			    (j + 1 < nbins) ? "," : "");
+		} else if (!in_gap) {
+			size_t availregs, milli;
+			char util[6]; /* "x.yyy". */
 
 			availregs = nregs * curruns;
 			milli = (availregs != 0) ? (1000 * curregs) / availregs
@@ -138,7 +159,7 @@ stats_arena_bins_print(void (*write_cb)(void *, const char *), void *cbopaque,
 				    reg_size, j, curregs * reg_size, nmalloc,
 				    ndalloc, nrequests, curregs, curruns, nregs,
 				    run_size / page, util, nfills, nflushes,
-				    nruns, reruns);
+				    nruns, nreruns);
 			} else {
 				malloc_cprintf(write_cb, cbopaque,
 				    "%20zu %3u %12zu %12"FMTu64
@@ -147,28 +168,38 @@ stats_arena_bins_print(void (*write_cb)(void *, const char *), void *cbopaque,
 				    " %12"FMTu64"\n",
 				    reg_size, j, curregs * reg_size, nmalloc,
 				    ndalloc, nrequests, curregs, curruns, nregs,
-				    run_size / page, util, nruns, reruns);
+				    run_size / page, util, nruns, nreruns);
 			}
 		}
 	}
-	if (in_gap) {
+	if (json) {
 		malloc_cprintf(write_cb, cbopaque,
-		    "                     ---\n");
+		    "\t\t\t\t]%s\n", (large || huge) ? "," : "");
+	} else {
+		if (in_gap) {
+			malloc_cprintf(write_cb, cbopaque,
+			    "                     ---\n");
+		}
 	}
 }
 
 static void
 stats_arena_lruns_print(void (*write_cb)(void *, const char *), void *cbopaque,
-    unsigned i)
+    bool json, bool huge, unsigned i)
 {
 	unsigned nbins, nlruns, j;
-	bool in_gap;
+	bool in_gap, in_gap_prev;
 
-	malloc_cprintf(write_cb, cbopaque,
-	    "large:          size ind    allocated      nmalloc      ndalloc"
-	    "    nrequests      curruns\n");
 	CTL_GET("arenas.nbins", &nbins, unsigned);
 	CTL_GET("arenas.nlruns", &nlruns, unsigned);
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\"lruns\": [\n");
+	} else {
+		malloc_cprintf(write_cb, cbopaque,
+		    "large:          size ind    allocated      nmalloc"
+		    "      ndalloc    nrequests      curruns\n");
+	}
 	for (j = 0, in_gap = false; j < nlruns; j++) {
 		uint64_t nmalloc, ndalloc, nrequests;
 		size_t run_size, curruns;
@@ -179,17 +210,25 @@ stats_arena_lruns_print(void (*write_cb)(void *, const char *), void *cbopaque,
 		    uint64_t);
 		CTL_M2_M4_GET("stats.arenas.0.lruns.0.nrequests", i, j,
 		    &nrequests, uint64_t);
-		if (nrequests == 0)
-			in_gap = true;
-		else {
-			CTL_M2_GET("arenas.lrun.0.size", j, &run_size, size_t);
-			CTL_M2_M4_GET("stats.arenas.0.lruns.0.curruns", i, j,
-			    &curruns, size_t);
-			if (in_gap) {
-				malloc_cprintf(write_cb, cbopaque,
-				    "                     ---\n");
-				in_gap = false;
-			}
+		in_gap_prev = in_gap;
+		in_gap = (nrequests == 0);
+
+		if (!json && in_gap_prev && !in_gap) {
+			malloc_cprintf(write_cb, cbopaque,
+			    "                     ---\n");
+		}
+
+		CTL_M2_GET("arenas.lrun.0.size", j, &run_size, size_t);
+		CTL_M2_M4_GET("stats.arenas.0.lruns.0.curruns", i, j, &curruns,
+		    size_t);
+		if (json) {
+			malloc_cprintf(write_cb, cbopaque,
+			    "\t\t\t\t\t{\n"
+			    "\t\t\t\t\t\t\"curruns\": %zu\n"
+			    "\t\t\t\t\t}%s\n",
+			    curruns,
+			    (j + 1 < nlruns) ? "," : "");
+		} else if (!in_gap) {
 			malloc_cprintf(write_cb, cbopaque,
 			    "%20zu %3u %12zu %12"FMTu64" %12"FMTu64
 			    " %12"FMTu64" %12zu\n",
@@ -197,25 +236,35 @@ stats_arena_lruns_print(void (*write_cb)(void *, const char *), void *cbopaque,
 			    ndalloc, nrequests, curruns);
 		}
 	}
-	if (in_gap) {
+	if (json) {
 		malloc_cprintf(write_cb, cbopaque,
-		    "                     ---\n");
+		    "\t\t\t\t]%s\n", huge ? "," : "");
+	} else {
+		if (in_gap) {
+			malloc_cprintf(write_cb, cbopaque,
+			    "                     ---\n");
+		}
 	}
 }
 
 static void
 stats_arena_hchunks_print(void (*write_cb)(void *, const char *),
-    void *cbopaque, unsigned i)
+    void *cbopaque, bool json, unsigned i)
 {
 	unsigned nbins, nlruns, nhchunks, j;
-	bool in_gap;
+	bool in_gap, in_gap_prev;
 
-	malloc_cprintf(write_cb, cbopaque,
-	    "huge:           size ind    allocated      nmalloc      ndalloc"
-	    "    nrequests   curhchunks\n");
 	CTL_GET("arenas.nbins", &nbins, unsigned);
 	CTL_GET("arenas.nlruns", &nlruns, unsigned);
 	CTL_GET("arenas.nhchunks", &nhchunks, unsigned);
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\"hchunks\": [\n");
+	} else {
+		malloc_cprintf(write_cb, cbopaque,
+		    "huge:           size ind    allocated      nmalloc"
+		    "      ndalloc    nrequests   curhchunks\n");
+	}
 	for (j = 0, in_gap = false; j < nhchunks; j++) {
 		uint64_t nmalloc, ndalloc, nrequests;
 		size_t hchunk_size, curhchunks;
@@ -226,18 +275,25 @@ stats_arena_hchunks_print(void (*write_cb)(void *, const char *),
 		    &ndalloc, uint64_t);
 		CTL_M2_M4_GET("stats.arenas.0.hchunks.0.nrequests", i, j,
 		    &nrequests, uint64_t);
-		if (nrequests == 0)
-			in_gap = true;
-		else {
-			CTL_M2_GET("arenas.hchunk.0.size", j, &hchunk_size,
-			    size_t);
-			CTL_M2_M4_GET("stats.arenas.0.hchunks.0.curhchunks", i,
-			    j, &curhchunks, size_t);
-			if (in_gap) {
-				malloc_cprintf(write_cb, cbopaque,
-				    "                     ---\n");
-				in_gap = false;
-			}
+		in_gap_prev = in_gap;
+		in_gap = (nrequests == 0);
+
+		if (!json && in_gap_prev && !in_gap) {
+			malloc_cprintf(write_cb, cbopaque,
+			    "                     ---\n");
+		}
+
+		CTL_M2_GET("arenas.hchunk.0.size", j, &hchunk_size, size_t);
+		CTL_M2_M4_GET("stats.arenas.0.hchunks.0.curhchunks", i, j,
+		    &curhchunks, size_t);
+		if (json) {
+			malloc_cprintf(write_cb, cbopaque,
+			    "\t\t\t\t\t{\n"
+			    "\t\t\t\t\t\t\"curhchunks\": %zu\n"
+			    "\t\t\t\t\t}%s\n",
+			    curhchunks,
+			    (j + 1 < nhchunks) ? "," : "");
+		} else if (!in_gap) {
 			malloc_cprintf(write_cb, cbopaque,
 			    "%20zu %3u %12zu %12"FMTu64" %12"FMTu64
 			    " %12"FMTu64" %12zu\n",
@@ -246,20 +302,25 @@ stats_arena_hchunks_print(void (*write_cb)(void *, const char *),
 			    nrequests, curhchunks);
 		}
 	}
-	if (in_gap) {
+	if (json) {
 		malloc_cprintf(write_cb, cbopaque,
-		    "                     ---\n");
+		    "\t\t\t\t]\n");
+	} else {
+		if (in_gap) {
+			malloc_cprintf(write_cb, cbopaque,
+			    "                     ---\n");
+		}
 	}
 }
 
 static void
 stats_arena_print(void (*write_cb)(void *, const char *), void *cbopaque,
-    unsigned i, bool bins, bool large, bool huge)
+    bool json, unsigned i, bool bins, bool large, bool huge)
 {
 	unsigned nthreads;
 	const char *dss;
-	ssize_t lg_dirty_mult;
-	size_t page, pactive, pdirty, mapped;
+	ssize_t lg_dirty_mult, decay_time;
+	size_t page, pactive, pdirty, mapped, retained;
 	size_t metadata_mapped, metadata_allocated;
 	uint64_t npurge, nmadvise, purged;
 	size_t small_allocated;
@@ -272,240 +333,435 @@ stats_arena_print(void (*write_cb)(void *, const char *), void *cbopaque,
 	CTL_GET("arenas.page", &page, size_t);
 
 	CTL_M2_GET("stats.arenas.0.nthreads", i, &nthreads, unsigned);
-	malloc_cprintf(write_cb, cbopaque,
-	    "assigned threads: %u\n", nthreads);
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\"nthreads\": %u,\n", nthreads);
+	} else {
+		malloc_cprintf(write_cb, cbopaque,
+		    "assigned threads: %u\n", nthreads);
+	}
+
 	CTL_M2_GET("stats.arenas.0.dss", i, &dss, const char *);
-	malloc_cprintf(write_cb, cbopaque, "dss allocation precedence: %s\n",
-	    dss);
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\"dss\": \"%s\",\n", dss);
+	} else {
+		malloc_cprintf(write_cb, cbopaque,
+		    "dss allocation precedence: %s\n", dss);
+	}
+
 	CTL_M2_GET("stats.arenas.0.lg_dirty_mult", i, &lg_dirty_mult, ssize_t);
-	if (lg_dirty_mult >= 0) {
+	if (json) {
 		malloc_cprintf(write_cb, cbopaque,
-		    "min active:dirty page ratio: %u:1\n",
-		    (1U << lg_dirty_mult));
+		    "\t\t\t\t\"lg_dirty_mult\": %zd,\n", lg_dirty_mult);
 	} else {
+		if (opt_purge == purge_mode_ratio) {
+			if (lg_dirty_mult >= 0) {
+				malloc_cprintf(write_cb, cbopaque,
+				    "min active:dirty page ratio: %u:1\n",
+				    (1U << lg_dirty_mult));
+			} else {
+				malloc_cprintf(write_cb, cbopaque,
+				    "min active:dirty page ratio: N/A\n");
+			}
+		}
+	}
+
+	CTL_M2_GET("stats.arenas.0.decay_time", i, &decay_time, ssize_t);
+	if (json) {
 		malloc_cprintf(write_cb, cbopaque,
-		    "min active:dirty page ratio: N/A\n");
+		    "\t\t\t\t\"decay_time\": %zd,\n", decay_time);
+	} else {
+		if (opt_purge == purge_mode_decay) {
+			if (decay_time >= 0) {
+				malloc_cprintf(write_cb, cbopaque,
+				    "decay time: %zd\n", decay_time);
+			} else {
+				malloc_cprintf(write_cb, cbopaque,
+				    "decay time: N/A\n");
+			}
+		}
 	}
+
 	CTL_M2_GET("stats.arenas.0.pactive", i, &pactive, size_t);
 	CTL_M2_GET("stats.arenas.0.pdirty", i, &pdirty, size_t);
 	CTL_M2_GET("stats.arenas.0.npurge", i, &npurge, uint64_t);
 	CTL_M2_GET("stats.arenas.0.nmadvise", i, &nmadvise, uint64_t);
 	CTL_M2_GET("stats.arenas.0.purged", i, &purged, uint64_t);
-	malloc_cprintf(write_cb, cbopaque,
-	    "dirty pages: %zu:%zu active:dirty, %"FMTu64" sweep%s, %"FMTu64
-	    " madvise%s, %"FMTu64" purged\n", pactive, pdirty, npurge, npurge ==
-	    1 ? "" : "s", nmadvise, nmadvise == 1 ? "" : "s", purged);
-
-	malloc_cprintf(write_cb, cbopaque,
-	    "                            allocated      nmalloc      ndalloc"
-	    "    nrequests\n");
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\"pactive\": %zu,\n", pactive);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\"pdirty\": %zu,\n", pdirty);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\"npurge\": %"FMTu64",\n", npurge);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\"nmadvise\": %"FMTu64",\n", nmadvise);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\"purged\": %"FMTu64",\n", purged);
+	} else {
+		malloc_cprintf(write_cb, cbopaque,
+		    "purging: dirty: %zu, sweeps: %"FMTu64", madvises: %"FMTu64
+		    ", purged: %"FMTu64"\n", pdirty, npurge, nmadvise, purged);
+	}
+
 	CTL_M2_GET("stats.arenas.0.small.allocated", i, &small_allocated,
 	    size_t);
 	CTL_M2_GET("stats.arenas.0.small.nmalloc", i, &small_nmalloc, uint64_t);
 	CTL_M2_GET("stats.arenas.0.small.ndalloc", i, &small_ndalloc, uint64_t);
 	CTL_M2_GET("stats.arenas.0.small.nrequests", i, &small_nrequests,
 	    uint64_t);
-	malloc_cprintf(write_cb, cbopaque,
-	    "small:                   %12zu %12"FMTu64" %12"FMTu64
-	    " %12"FMTu64"\n",
-	    small_allocated, small_nmalloc, small_ndalloc, small_nrequests);
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\"small\": {\n");
+
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\t\"allocated\": %zu,\n", small_allocated);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\t\"nmalloc\": %"FMTu64",\n", small_nmalloc);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\t\"ndalloc\": %"FMTu64",\n", small_ndalloc);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\t\"nrequests\": %"FMTu64"\n", small_nrequests);
+
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t},\n");
+	} else {
+		malloc_cprintf(write_cb, cbopaque,
+		    "                            allocated      nmalloc"
+		    "      ndalloc    nrequests\n");
+		malloc_cprintf(write_cb, cbopaque,
+		    "small:                   %12zu %12"FMTu64" %12"FMTu64
+		    " %12"FMTu64"\n",
+		    small_allocated, small_nmalloc, small_ndalloc,
+		    small_nrequests);
+	}
+
 	CTL_M2_GET("stats.arenas.0.large.allocated", i, &large_allocated,
 	    size_t);
 	CTL_M2_GET("stats.arenas.0.large.nmalloc", i, &large_nmalloc, uint64_t);
 	CTL_M2_GET("stats.arenas.0.large.ndalloc", i, &large_ndalloc, uint64_t);
 	CTL_M2_GET("stats.arenas.0.large.nrequests", i, &large_nrequests,
 	    uint64_t);
-	malloc_cprintf(write_cb, cbopaque,
-	    "large:                   %12zu %12"FMTu64" %12"FMTu64
-	    " %12"FMTu64"\n",
-	    large_allocated, large_nmalloc, large_ndalloc, large_nrequests);
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\"large\": {\n");
+
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\t\"allocated\": %zu,\n", large_allocated);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\t\"nmalloc\": %"FMTu64",\n", large_nmalloc);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\t\"ndalloc\": %"FMTu64",\n", large_ndalloc);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\t\"nrequests\": %"FMTu64"\n", large_nrequests);
+
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t},\n");
+	} else {
+		malloc_cprintf(write_cb, cbopaque,
+		    "large:                   %12zu %12"FMTu64" %12"FMTu64
+		    " %12"FMTu64"\n",
+		    large_allocated, large_nmalloc, large_ndalloc,
+		    large_nrequests);
+	}
+
 	CTL_M2_GET("stats.arenas.0.huge.allocated", i, &huge_allocated, size_t);
 	CTL_M2_GET("stats.arenas.0.huge.nmalloc", i, &huge_nmalloc, uint64_t);
 	CTL_M2_GET("stats.arenas.0.huge.ndalloc", i, &huge_ndalloc, uint64_t);
 	CTL_M2_GET("stats.arenas.0.huge.nrequests", i, &huge_nrequests,
 	    uint64_t);
-	malloc_cprintf(write_cb, cbopaque,
-	    "huge:                    %12zu %12"FMTu64" %12"FMTu64
-	    " %12"FMTu64"\n",
-	    huge_allocated, huge_nmalloc, huge_ndalloc, huge_nrequests);
-	malloc_cprintf(write_cb, cbopaque,
-	    "total:                   %12zu %12"FMTu64" %12"FMTu64
-	    " %12"FMTu64"\n",
-	    small_allocated + large_allocated + huge_allocated,
-	    small_nmalloc + large_nmalloc + huge_nmalloc,
-	    small_ndalloc + large_ndalloc + huge_ndalloc,
-	    small_nrequests + large_nrequests + huge_nrequests);
-	malloc_cprintf(write_cb, cbopaque,
-	    "active:                  %12zu\n", pactive * page);
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\"huge\": {\n");
+
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\t\"allocated\": %zu,\n", huge_allocated);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\t\"nmalloc\": %"FMTu64",\n", huge_nmalloc);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\t\"ndalloc\": %"FMTu64",\n", huge_ndalloc);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\t\"nrequests\": %"FMTu64"\n", huge_nrequests);
+
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t},\n");
+	} else {
+		malloc_cprintf(write_cb, cbopaque,
+		    "huge:                    %12zu %12"FMTu64" %12"FMTu64
+		    " %12"FMTu64"\n",
+		    huge_allocated, huge_nmalloc, huge_ndalloc, huge_nrequests);
+		malloc_cprintf(write_cb, cbopaque,
+		    "total:                   %12zu %12"FMTu64" %12"FMTu64
+		    " %12"FMTu64"\n",
+		    small_allocated + large_allocated + huge_allocated,
+		    small_nmalloc + large_nmalloc + huge_nmalloc,
+		    small_ndalloc + large_ndalloc + huge_ndalloc,
+		    small_nrequests + large_nrequests + huge_nrequests);
+	}
+	if (!json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "active:                  %12zu\n", pactive * page);
+	}
+
 	CTL_M2_GET("stats.arenas.0.mapped", i, &mapped, size_t);
-	malloc_cprintf(write_cb, cbopaque,
-	    "mapped:                  %12zu\n", mapped);
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\"mapped\": %zu,\n", mapped);
+	} else {
+		malloc_cprintf(write_cb, cbopaque,
+		    "mapped:                  %12zu\n", mapped);
+	}
+
+	CTL_M2_GET("stats.arenas.0.retained", i, &retained, size_t);
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\"retained\": %zu,\n", retained);
+	} else {
+		malloc_cprintf(write_cb, cbopaque,
+		    "retained:                %12zu\n", retained);
+	}
+
 	CTL_M2_GET("stats.arenas.0.metadata.mapped", i, &metadata_mapped,
 	    size_t);
 	CTL_M2_GET("stats.arenas.0.metadata.allocated", i, &metadata_allocated,
 	    size_t);
-	malloc_cprintf(write_cb, cbopaque,
-	    "metadata: mapped: %zu, allocated: %zu\n",
-	    metadata_mapped, metadata_allocated);
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\"metadata\": {\n");
+
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\t\"mapped\": %zu,\n", metadata_mapped);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t\t\"allocated\": %zu\n", metadata_allocated);
+
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\t},\n");
+	} else {
+		malloc_cprintf(write_cb, cbopaque,
+		    "metadata: mapped: %zu, allocated: %zu\n",
+		    metadata_mapped, metadata_allocated);
+	}
 
-	if (bins)
-		stats_arena_bins_print(write_cb, cbopaque, i);
+	if (bins) {
+		stats_arena_bins_print(write_cb, cbopaque, json, large, huge,
+		    i);
+	}
 	if (large)
-		stats_arena_lruns_print(write_cb, cbopaque, i);
+		stats_arena_lruns_print(write_cb, cbopaque, json, huge, i);
 	if (huge)
-		stats_arena_hchunks_print(write_cb, cbopaque, i);
+		stats_arena_hchunks_print(write_cb, cbopaque, json, i);
 }
 
-void
-stats_print(void (*write_cb)(void *, const char *), void *cbopaque,
-    const char *opts)
+static void
+stats_general_print(void (*write_cb)(void *, const char *), void *cbopaque,
+    bool json, bool merged, bool unmerged)
 {
-	int err;
-	uint64_t epoch;
-	size_t u64sz;
-	bool general = true;
-	bool merged = true;
-	bool unmerged = true;
-	bool bins = true;
-	bool large = true;
-	bool huge = true;
+	const char *cpv;
+	bool bv;
+	unsigned uv;
+	uint32_t u32v;
+	uint64_t u64v;
+	ssize_t ssv;
+	size_t sv, bsz, usz, ssz, sssz, cpsz;
 
-	/*
-	 * Refresh stats, in case mallctl() was called by the application.
-	 *
-	 * Check for OOM here, since refreshing the ctl cache can trigger
-	 * allocation.  In practice, none of the subsequent mallctl()-related
-	 * calls in this function will cause OOM if this one succeeds.
-	 * */
-	epoch = 1;
-	u64sz = sizeof(uint64_t);
-	err = je_mallctl("epoch", &epoch, &u64sz, &epoch, sizeof(uint64_t));
-	if (err != 0) {
-		if (err == EAGAIN) {
-			malloc_write("<jemalloc>: Memory allocation failure in "
-			    "mallctl(\"epoch\", ...)\n");
-			return;
-		}
-		malloc_write("<jemalloc>: Failure in mallctl(\"epoch\", "
-		    "...)\n");
-		abort();
-	}
+	bsz = sizeof(bool);
+	usz = sizeof(unsigned);
+	ssz = sizeof(size_t);
+	sssz = sizeof(ssize_t);
+	cpsz = sizeof(const char *);
 
-	if (opts != NULL) {
-		unsigned i;
+	CTL_GET("version", &cpv, const char *);
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		"\t\t\"version\": \"%s\",\n", cpv);
+	} else
+		malloc_cprintf(write_cb, cbopaque, "Version: %s\n", cpv);
 
-		for (i = 0; opts[i] != '\0'; i++) {
-			switch (opts[i]) {
-			case 'g':
-				general = false;
-				break;
-			case 'm':
-				merged = false;
-				break;
-			case 'a':
-				unmerged = false;
-				break;
-			case 'b':
-				bins = false;
-				break;
-			case 'l':
-				large = false;
-				break;
-			case 'h':
-				huge = false;
-				break;
-			default:;
-			}
-		}
+	/* config. */
+#define	CONFIG_WRITE_BOOL_JSON(n, c)					\
+	if (json) {							\
+		CTL_GET("config."#n, &bv, bool);			\
+		malloc_cprintf(write_cb, cbopaque,			\
+		    "\t\t\t\""#n"\": %s%s\n", bv ? "true" : "false",	\
+		    (c));						\
 	}
 
-	malloc_cprintf(write_cb, cbopaque,
-	    "___ Begin jemalloc statistics ___\n");
-	if (general) {
-		const char *cpv;
-		bool bv;
-		unsigned uv;
-		ssize_t ssv;
-		size_t sv, bsz, ssz, sssz, cpsz;
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\"config\": {\n");
+	}
 
-		bsz = sizeof(bool);
-		ssz = sizeof(size_t);
-		sssz = sizeof(ssize_t);
-		cpsz = sizeof(const char *);
+	CONFIG_WRITE_BOOL_JSON(cache_oblivious, ",")
 
-		CTL_GET("version", &cpv, const char *);
-		malloc_cprintf(write_cb, cbopaque, "Version: %s\n", cpv);
-		CTL_GET("config.debug", &bv, bool);
+	CTL_GET("config.debug", &bv, bool);
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"debug\": %s,\n", bv ? "true" : "false");
+	} else {
 		malloc_cprintf(write_cb, cbopaque, "Assertions %s\n",
 		    bv ? "enabled" : "disabled");
+	}
+
+	CONFIG_WRITE_BOOL_JSON(fill, ",")
+	CONFIG_WRITE_BOOL_JSON(lazy_lock, ",")
+
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"malloc_conf\": \"%s\",\n",
+		    config_malloc_conf);
+	} else {
+		malloc_cprintf(write_cb, cbopaque,
+		    "config.malloc_conf: \"%s\"\n", config_malloc_conf);
+	}
 
-#define	OPT_WRITE_BOOL(n)						\
-		if (je_mallctl("opt."#n, &bv, &bsz, NULL, 0) == 0) {	\
+	CONFIG_WRITE_BOOL_JSON(munmap, ",")
+	CONFIG_WRITE_BOOL_JSON(prof, ",")
+	CONFIG_WRITE_BOOL_JSON(prof_libgcc, ",")
+	CONFIG_WRITE_BOOL_JSON(prof_libunwind, ",")
+	CONFIG_WRITE_BOOL_JSON(stats, ",")
+	CONFIG_WRITE_BOOL_JSON(tcache, ",")
+	CONFIG_WRITE_BOOL_JSON(tls, ",")
+	CONFIG_WRITE_BOOL_JSON(utrace, ",")
+	CONFIG_WRITE_BOOL_JSON(valgrind, ",")
+	CONFIG_WRITE_BOOL_JSON(xmalloc, "")
+
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t},\n");
+	}
+#undef CONFIG_WRITE_BOOL_JSON
+
+	/* opt. */
+#define	OPT_WRITE_BOOL(n, c)						\
+	if (je_mallctl("opt."#n, (void *)&bv, &bsz, NULL, 0) == 0) {	\
+		if (json) {						\
+			malloc_cprintf(write_cb, cbopaque,		\
+			    "\t\t\t\""#n"\": %s%s\n", bv ? "true" :	\
+			    "false", (c));				\
+		} else {						\
 			malloc_cprintf(write_cb, cbopaque,		\
 			    "  opt."#n": %s\n", bv ? "true" : "false");	\
-		}
-#define	OPT_WRITE_BOOL_MUTABLE(n, m) {					\
-		bool bv2;						\
-		if (je_mallctl("opt."#n, &bv, &bsz, NULL, 0) == 0 &&	\
-		    je_mallctl(#m, &bv2, &bsz, NULL, 0) == 0) {		\
+		}							\
+	}
+#define	OPT_WRITE_BOOL_MUTABLE(n, m, c) {				\
+	bool bv2;							\
+	if (je_mallctl("opt."#n, (void *)&bv, &bsz, NULL, 0) == 0 &&	\
+	    je_mallctl(#m, &bv2, (void *)&bsz, NULL, 0) == 0) {		\
+		if (json) {						\
+			malloc_cprintf(write_cb, cbopaque,		\
+			    "\t\t\t\""#n"\": %s%s\n", bv ? "true" :	\
+			    "false", (c));				\
+		} else {						\
 			malloc_cprintf(write_cb, cbopaque,		\
 			    "  opt."#n": %s ("#m": %s)\n", bv ? "true"	\
 			    : "false", bv2 ? "true" : "false");		\
 		}							\
+	}								\
 }
-#define	OPT_WRITE_SIZE_T(n)						\
-		if (je_mallctl("opt."#n, &sv, &ssz, NULL, 0) == 0) {	\
+#define	OPT_WRITE_UNSIGNED(n, c)					\
+	if (je_mallctl("opt."#n, (void *)&uv, &usz, NULL, 0) == 0) {	\
+		if (json) {						\
+			malloc_cprintf(write_cb, cbopaque,		\
+			    "\t\t\t\""#n"\": %u%s\n", uv, (c));		\
+		} else {						\
+			malloc_cprintf(write_cb, cbopaque,		\
+			"  opt."#n": %u\n", uv);			\
+		}							\
+	}
+#define	OPT_WRITE_SIZE_T(n, c)						\
+	if (je_mallctl("opt."#n, (void *)&sv, &ssz, NULL, 0) == 0) {	\
+		if (json) {						\
+			malloc_cprintf(write_cb, cbopaque,		\
+			    "\t\t\t\""#n"\": %zu%s\n", sv, (c));	\
+		} else {						\
 			malloc_cprintf(write_cb, cbopaque,		\
 			"  opt."#n": %zu\n", sv);			\
-		}
-#define	OPT_WRITE_SSIZE_T(n)						\
-		if (je_mallctl("opt."#n, &ssv, &sssz, NULL, 0) == 0) {	\
+		}							\
+	}
+#define	OPT_WRITE_SSIZE_T(n, c)						\
+	if (je_mallctl("opt."#n, (void *)&ssv, &sssz, NULL, 0) == 0) {	\
+		if (json) {						\
+			malloc_cprintf(write_cb, cbopaque,		\
+			    "\t\t\t\""#n"\": %zd%s\n", ssv, (c));	\
+		} else {						\
 			malloc_cprintf(write_cb, cbopaque,		\
 			    "  opt."#n": %zd\n", ssv);			\
-		}
-#define	OPT_WRITE_SSIZE_T_MUTABLE(n, m) {				\
-		ssize_t ssv2;						\
-		if (je_mallctl("opt."#n, &ssv, &sssz, NULL, 0) == 0 &&	\
-		    je_mallctl(#m, &ssv2, &sssz, NULL, 0) == 0) {	\
+		}							\
+	}
+#define	OPT_WRITE_SSIZE_T_MUTABLE(n, m, c) {				\
+	ssize_t ssv2;							\
+	if (je_mallctl("opt."#n, (void *)&ssv, &sssz, NULL, 0) == 0 &&	\
+	    je_mallctl(#m, (void *)&ssv2, &sssz, NULL, 0) == 0) {	\
+		if (json) {						\
+			malloc_cprintf(write_cb, cbopaque,		\
+			    "\t\t\t\""#n"\": %zd%s\n", ssv, (c));	\
+		} else {						\
 			malloc_cprintf(write_cb, cbopaque,		\
 			    "  opt."#n": %zd ("#m": %zd)\n",		\
 			    ssv, ssv2);					\
 		}							\
+	}								\
 }
-#define	OPT_WRITE_CHAR_P(n)						\
-		if (je_mallctl("opt."#n, &cpv, &cpsz, NULL, 0) == 0) {	\
+#define	OPT_WRITE_CHAR_P(n, c)						\
+	if (je_mallctl("opt."#n, (void *)&cpv, &cpsz, NULL, 0) == 0) {	\
+		if (json) {						\
+			malloc_cprintf(write_cb, cbopaque,		\
+			    "\t\t\t\""#n"\": \"%s\"%s\n", cpv, (c));	\
+		} else {						\
 			malloc_cprintf(write_cb, cbopaque,		\
 			    "  opt."#n": \"%s\"\n", cpv);		\
-		}
+		}							\
+	}
 
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\"opt\": {\n");
+	} else {
 		malloc_cprintf(write_cb, cbopaque,
 		    "Run-time option settings:\n");
-		OPT_WRITE_BOOL(abort)
-		OPT_WRITE_SIZE_T(lg_chunk)
-		OPT_WRITE_CHAR_P(dss)
-		OPT_WRITE_SIZE_T(narenas)
-		OPT_WRITE_SSIZE_T_MUTABLE(lg_dirty_mult, arenas.lg_dirty_mult)
-		OPT_WRITE_BOOL(stats_print)
-		OPT_WRITE_CHAR_P(junk)
-		OPT_WRITE_SIZE_T(quarantine)
-		OPT_WRITE_BOOL(redzone)
-		OPT_WRITE_BOOL(zero)
-		OPT_WRITE_BOOL(utrace)
-		OPT_WRITE_BOOL(valgrind)
-		OPT_WRITE_BOOL(xmalloc)
-		OPT_WRITE_BOOL(tcache)
-		OPT_WRITE_SSIZE_T(lg_tcache_max)
-		OPT_WRITE_BOOL(prof)
-		OPT_WRITE_CHAR_P(prof_prefix)
-		OPT_WRITE_BOOL_MUTABLE(prof_active, prof.active)
-		OPT_WRITE_BOOL_MUTABLE(prof_thread_active_init,
-		    prof.thread_active_init)
-		OPT_WRITE_SSIZE_T(lg_prof_sample)
-		OPT_WRITE_BOOL(prof_accum)
-		OPT_WRITE_SSIZE_T(lg_prof_interval)
-		OPT_WRITE_BOOL(prof_gdump)
-		OPT_WRITE_BOOL(prof_final)
-		OPT_WRITE_BOOL(prof_leak)
+	}
+	OPT_WRITE_BOOL(abort, ",")
+	OPT_WRITE_SIZE_T(lg_chunk, ",")
+	OPT_WRITE_CHAR_P(dss, ",")
+	OPT_WRITE_UNSIGNED(narenas, ",")
+	OPT_WRITE_CHAR_P(purge, ",")
+	if (json || opt_purge == purge_mode_ratio) {
+		OPT_WRITE_SSIZE_T_MUTABLE(lg_dirty_mult,
+		    arenas.lg_dirty_mult, ",")
+	}
+	if (json || opt_purge == purge_mode_decay) {
+		OPT_WRITE_SSIZE_T_MUTABLE(decay_time, arenas.decay_time, ",")
+	}
+	OPT_WRITE_CHAR_P(junk, ",")
+	OPT_WRITE_SIZE_T(quarantine, ",")
+	OPT_WRITE_BOOL(redzone, ",")
+	OPT_WRITE_BOOL(zero, ",")
+	OPT_WRITE_BOOL(utrace, ",")
+	OPT_WRITE_BOOL(xmalloc, ",")
+	OPT_WRITE_BOOL(tcache, ",")
+	OPT_WRITE_SSIZE_T(lg_tcache_max, ",")
+	OPT_WRITE_BOOL(prof, ",")
+	OPT_WRITE_CHAR_P(prof_prefix, ",")
+	OPT_WRITE_BOOL_MUTABLE(prof_active, prof.active, ",")
+	OPT_WRITE_BOOL_MUTABLE(prof_thread_active_init, prof.thread_active_init,
+	    ",")
+	OPT_WRITE_SSIZE_T_MUTABLE(lg_prof_sample, prof.lg_sample, ",")
+	OPT_WRITE_BOOL(prof_accum, ",")
+	OPT_WRITE_SSIZE_T(lg_prof_interval, ",")
+	OPT_WRITE_BOOL(prof_gdump, ",")
+	OPT_WRITE_BOOL(prof_final, ",")
+	OPT_WRITE_BOOL(prof_leak, ",")
+	/*
+	 * stats_print is always emitted, so as long as stats_print comes last
+	 * it's safe to unconditionally omit the comma here (rather than having
+	 * to conditionally omit it elsewhere depending on configuration).
+	 */
+	OPT_WRITE_BOOL(stats_print, "")
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t},\n");
+	}
 
 #undef OPT_WRITE_BOOL
 #undef OPT_WRITE_BOOL_MUTABLE
@@ -513,128 +769,386 @@ stats_print(void (*write_cb)(void *, const char *), void *cbopaque,
 #undef OPT_WRITE_SSIZE_T
 #undef OPT_WRITE_CHAR_P
 
-		malloc_cprintf(write_cb, cbopaque, "CPUs: %u\n", ncpus);
+	/* arenas. */
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\"arenas\": {\n");
+	}
 
-		CTL_GET("arenas.narenas", &uv, unsigned);
+	CTL_GET("arenas.narenas", &uv, unsigned);
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"narenas\": %u,\n", uv);
+	} else
 		malloc_cprintf(write_cb, cbopaque, "Arenas: %u\n", uv);
 
-		malloc_cprintf(write_cb, cbopaque, "Pointer size: %zu\n",
-		    sizeof(void *));
+	CTL_GET("arenas.lg_dirty_mult", &ssv, ssize_t);
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"lg_dirty_mult\": %zd,\n", ssv);
+	} else if (opt_purge == purge_mode_ratio) {
+		if (ssv >= 0) {
+			malloc_cprintf(write_cb, cbopaque,
+			    "Min active:dirty page ratio per arena: "
+			    "%u:1\n", (1U << ssv));
+		} else {
+			malloc_cprintf(write_cb, cbopaque,
+			    "Min active:dirty page ratio per arena: "
+			    "N/A\n");
+		}
+	}
+	CTL_GET("arenas.decay_time", &ssv, ssize_t);
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"decay_time\": %zd,\n", ssv);
+	} else if (opt_purge == purge_mode_decay) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "Unused dirty page decay time: %zd%s\n",
+		    ssv, (ssv < 0) ? " (no decay)" : "");
+	}
 
-		CTL_GET("arenas.quantum", &sv, size_t);
-		malloc_cprintf(write_cb, cbopaque, "Quantum size: %zu\n",
-		    sv);
+	CTL_GET("arenas.quantum", &sv, size_t);
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"quantum\": %zu,\n", sv);
+	} else
+		malloc_cprintf(write_cb, cbopaque, "Quantum size: %zu\n", sv);
 
-		CTL_GET("arenas.page", &sv, size_t);
+	CTL_GET("arenas.page", &sv, size_t);
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"page\": %zu,\n", sv);
+	} else
 		malloc_cprintf(write_cb, cbopaque, "Page size: %zu\n", sv);
 
-		CTL_GET("arenas.lg_dirty_mult", &ssv, ssize_t);
-		if (ssv >= 0) {
+	if (je_mallctl("arenas.tcache_max", (void *)&sv, &ssz, NULL, 0) == 0) {
+		if (json) {
 			malloc_cprintf(write_cb, cbopaque,
-			    "Min active:dirty page ratio per arena: %u:1\n",
-			    (1U << ssv));
+			    "\t\t\t\"tcache_max\": %zu,\n", sv);
 		} else {
 			malloc_cprintf(write_cb, cbopaque,
-			    "Min active:dirty page ratio per arena: N/A\n");
+			    "Maximum thread-cached size class: %zu\n", sv);
 		}
-		if (je_mallctl("arenas.tcache_max", &sv, &ssz, NULL, 0) == 0) {
+	}
+
+	if (json) {
+		unsigned nbins, nlruns, nhchunks, i;
+
+		CTL_GET("arenas.nbins", &nbins, unsigned);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"nbins\": %u,\n", nbins);
+
+		CTL_GET("arenas.nhbins", &uv, unsigned);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"nhbins\": %u,\n", uv);
+
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"bin\": [\n");
+		for (i = 0; i < nbins; i++) {
 			malloc_cprintf(write_cb, cbopaque,
-			    "Maximum thread-cached size class: %zu\n", sv);
+			    "\t\t\t\t{\n");
+
+			CTL_M2_GET("arenas.bin.0.size", i, &sv, size_t);
+			malloc_cprintf(write_cb, cbopaque,
+			    "\t\t\t\t\t\"size\": %zu,\n", sv);
+
+			CTL_M2_GET("arenas.bin.0.nregs", i, &u32v, uint32_t);
+			malloc_cprintf(write_cb, cbopaque,
+			    "\t\t\t\t\t\"nregs\": %"FMTu32",\n", u32v);
+
+			CTL_M2_GET("arenas.bin.0.run_size", i, &sv, size_t);
+			malloc_cprintf(write_cb, cbopaque,
+			    "\t\t\t\t\t\"run_size\": %zu\n", sv);
+
+			malloc_cprintf(write_cb, cbopaque,
+			    "\t\t\t\t}%s\n", (i + 1 < nbins) ? "," : "");
 		}
-		if (je_mallctl("opt.prof", &bv, &bsz, NULL, 0) == 0 && bv) {
-			CTL_GET("prof.lg_sample", &sv, size_t);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t],\n");
+
+		CTL_GET("arenas.nlruns", &nlruns, unsigned);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"nlruns\": %u,\n", nlruns);
+
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"lrun\": [\n");
+		for (i = 0; i < nlruns; i++) {
 			malloc_cprintf(write_cb, cbopaque,
-			    "Average profile sample interval: %"FMTu64
-			    " (2^%zu)\n", (((uint64_t)1U) << sv), sv);
+			    "\t\t\t\t{\n");
 
-			CTL_GET("opt.lg_prof_interval", &ssv, ssize_t);
-			if (ssv >= 0) {
-				malloc_cprintf(write_cb, cbopaque,
-				    "Average profile dump interval: %"FMTu64
-				    " (2^%zd)\n",
-				    (((uint64_t)1U) << ssv), ssv);
-			} else {
-				malloc_cprintf(write_cb, cbopaque,
-				    "Average profile dump interval: N/A\n");
-			}
+			CTL_M2_GET("arenas.lrun.0.size", i, &sv, size_t);
+			malloc_cprintf(write_cb, cbopaque,
+			    "\t\t\t\t\t\"size\": %zu\n", sv);
+
+			malloc_cprintf(write_cb, cbopaque,
+			    "\t\t\t\t}%s\n", (i + 1 < nlruns) ? "," : "");
 		}
-		CTL_GET("opt.lg_chunk", &sv, size_t);
 		malloc_cprintf(write_cb, cbopaque,
-		    "Chunk size: %zu (2^%zu)\n", (ZU(1) << sv), sv);
+		    "\t\t\t],\n");
+
+		CTL_GET("arenas.nhchunks", &nhchunks, unsigned);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"nhchunks\": %u,\n", nhchunks);
+
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"hchunk\": [\n");
+		for (i = 0; i < nhchunks; i++) {
+			malloc_cprintf(write_cb, cbopaque,
+			    "\t\t\t\t{\n");
+
+			CTL_M2_GET("arenas.hchunk.0.size", i, &sv, size_t);
+			malloc_cprintf(write_cb, cbopaque,
+			    "\t\t\t\t\t\"size\": %zu\n", sv);
+
+			malloc_cprintf(write_cb, cbopaque,
+			    "\t\t\t\t}%s\n", (i + 1 < nhchunks) ? "," : "");
+		}
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t]\n");
+
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t},\n");
 	}
 
-	if (config_stats) {
-		size_t *cactive;
-		size_t allocated, active, metadata, resident, mapped;
+	/* prof. */
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\"prof\": {\n");
 
-		CTL_GET("stats.cactive", &cactive, size_t *);
-		CTL_GET("stats.allocated", &allocated, size_t);
-		CTL_GET("stats.active", &active, size_t);
-		CTL_GET("stats.metadata", &metadata, size_t);
-		CTL_GET("stats.resident", &resident, size_t);
-		CTL_GET("stats.mapped", &mapped, size_t);
+		CTL_GET("prof.thread_active_init", &bv, bool);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"thread_active_init\": %s,\n", bv ? "true" :
+		    "false");
+
+		CTL_GET("prof.active", &bv, bool);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"active\": %s,\n", bv ? "true" : "false");
+
+		CTL_GET("prof.gdump", &bv, bool);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"gdump\": %s,\n", bv ? "true" : "false");
+
+		CTL_GET("prof.interval", &u64v, uint64_t);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"interval\": %"FMTu64",\n", u64v);
+
+		CTL_GET("prof.lg_sample", &ssv, ssize_t);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"lg_sample\": %zd\n", ssv);
+
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t}%s\n", (config_stats || merged || unmerged) ? "," :
+		    "");
+	}
+}
+
+static void
+stats_print_helper(void (*write_cb)(void *, const char *), void *cbopaque,
+    bool json, bool merged, bool unmerged, bool bins, bool large, bool huge)
+{
+	size_t *cactive;
+	size_t allocated, active, metadata, resident, mapped, retained;
+
+	CTL_GET("stats.cactive", &cactive, size_t *);
+	CTL_GET("stats.allocated", &allocated, size_t);
+	CTL_GET("stats.active", &active, size_t);
+	CTL_GET("stats.metadata", &metadata, size_t);
+	CTL_GET("stats.resident", &resident, size_t);
+	CTL_GET("stats.mapped", &mapped, size_t);
+	CTL_GET("stats.retained", &retained, size_t);
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\"stats\": {\n");
+
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"cactive\": %zu,\n", atomic_read_z(cactive));
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"allocated\": %zu,\n", allocated);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"active\": %zu,\n", active);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"metadata\": %zu,\n", metadata);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"resident\": %zu,\n", resident);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"mapped\": %zu,\n", mapped);
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t\t\"retained\": %zu\n", retained);
+
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t\t}%s\n", (merged || unmerged) ? "," : "");
+	} else {
 		malloc_cprintf(write_cb, cbopaque,
 		    "Allocated: %zu, active: %zu, metadata: %zu,"
-		    " resident: %zu, mapped: %zu\n",
-		    allocated, active, metadata, resident, mapped);
+		    " resident: %zu, mapped: %zu, retained: %zu\n",
+		    allocated, active, metadata, resident, mapped, retained);
 		malloc_cprintf(write_cb, cbopaque,
 		    "Current active ceiling: %zu\n",
 		    atomic_read_z(cactive));
+	}
 
-		if (merged) {
-			unsigned narenas;
-
-			CTL_GET("arenas.narenas", &narenas, unsigned);
-			{
-				VARIABLE_ARRAY(bool, initialized, narenas);
-				size_t isz;
-				unsigned i, ninitialized;
-
-				isz = sizeof(bool) * narenas;
-				xmallctl("arenas.initialized", initialized,
-				    &isz, NULL, 0);
-				for (i = ninitialized = 0; i < narenas; i++) {
-					if (initialized[i])
-						ninitialized++;
-				}
+	if (merged || unmerged) {
+		unsigned narenas;
+
+		if (json) {
+			malloc_cprintf(write_cb, cbopaque,
+			    "\t\t\"stats.arenas\": {\n");
+		}
+
+		CTL_GET("arenas.narenas", &narenas, unsigned);
+		{
+			VARIABLE_ARRAY(bool, initialized, narenas);
+			size_t isz;
+			unsigned i, j, ninitialized;
+
+			isz = sizeof(bool) * narenas;
+			xmallctl("arenas.initialized", (void *)initialized,
+			    &isz, NULL, 0);
+			for (i = ninitialized = 0; i < narenas; i++) {
+				if (initialized[i])
+					ninitialized++;
+			}
 
-				if (ninitialized > 1 || !unmerged) {
-					/* Print merged arena stats. */
+			/* Merged stats. */
+			if (merged && (ninitialized > 1 || !unmerged)) {
+				/* Print merged arena stats. */
+				if (json) {
+					malloc_cprintf(write_cb, cbopaque,
+					    "\t\t\t\"merged\": {\n");
+				} else {
 					malloc_cprintf(write_cb, cbopaque,
 					    "\nMerged arenas stats:\n");
+				}
+				stats_arena_print(write_cb, cbopaque, json,
+				    narenas, bins, large, huge);
+				if (json) {
+					malloc_cprintf(write_cb, cbopaque,
+					    "\t\t\t}%s\n", (ninitialized > 1) ?
+					    "," : "");
+				}
+			}
+
+			/* Unmerged stats. */
+			for (i = j = 0; i < narenas; i++) {
+				if (initialized[i]) {
+					if (json) {
+						j++;
+						malloc_cprintf(write_cb,
+						    cbopaque,
+						    "\t\t\t\"%u\": {\n", i);
+					} else {
+						malloc_cprintf(write_cb,
+						    cbopaque, "\narenas[%u]:\n",
+						    i);
+					}
 					stats_arena_print(write_cb, cbopaque,
-					    narenas, bins, large, huge);
+					    json, i, bins, large, huge);
+					if (json) {
+						malloc_cprintf(write_cb,
+						    cbopaque,
+						    "\t\t\t}%s\n", (j <
+						    ninitialized) ? "," : "");
+					}
 				}
 			}
 		}
 
-		if (unmerged) {
-			unsigned narenas;
+		if (json) {
+			malloc_cprintf(write_cb, cbopaque,
+			    "\t\t}\n");
+		}
+	}
+}
 
-			/* Print stats for each arena. */
+void
+stats_print(void (*write_cb)(void *, const char *), void *cbopaque,
+    const char *opts)
+{
+	int err;
+	uint64_t epoch;
+	size_t u64sz;
+	bool json = false;
+	bool general = true;
+	bool merged = true;
+	bool unmerged = true;
+	bool bins = true;
+	bool large = true;
+	bool huge = true;
 
-			CTL_GET("arenas.narenas", &narenas, unsigned);
-			{
-				VARIABLE_ARRAY(bool, initialized, narenas);
-				size_t isz;
-				unsigned i;
+	/*
+	 * Refresh stats, in case mallctl() was called by the application.
+	 *
+	 * Check for OOM here, since refreshing the ctl cache can trigger
+	 * allocation.  In practice, none of the subsequent mallctl()-related
+	 * calls in this function will cause OOM if this one succeeds.
+	 * */
+	epoch = 1;
+	u64sz = sizeof(uint64_t);
+	err = je_mallctl("epoch", (void *)&epoch, &u64sz, (void *)&epoch,
+	    sizeof(uint64_t));
+	if (err != 0) {
+		if (err == EAGAIN) {
+			malloc_write("<jemalloc>: Memory allocation failure in "
+			    "mallctl(\"epoch\", ...)\n");
+			return;
+		}
+		malloc_write("<jemalloc>: Failure in mallctl(\"epoch\", "
+		    "...)\n");
+		abort();
+	}
 
-				isz = sizeof(bool) * narenas;
-				xmallctl("arenas.initialized", initialized,
-				    &isz, NULL, 0);
+	if (opts != NULL) {
+		unsigned i;
 
-				for (i = 0; i < narenas; i++) {
-					if (initialized[i]) {
-						malloc_cprintf(write_cb,
-						    cbopaque,
-						    "\narenas[%u]:\n", i);
-						stats_arena_print(write_cb,
-						    cbopaque, i, bins, large,
-						    huge);
-					}
-				}
+		for (i = 0; opts[i] != '\0'; i++) {
+			switch (opts[i]) {
+			case 'J':
+				json = true;
+				break;
+			case 'g':
+				general = false;
+				break;
+			case 'm':
+				merged = false;
+				break;
+			case 'a':
+				unmerged = false;
+				break;
+			case 'b':
+				bins = false;
+				break;
+			case 'l':
+				large = false;
+				break;
+			case 'h':
+				huge = false;
+				break;
+			default:;
 			}
 		}
 	}
-	malloc_cprintf(write_cb, cbopaque, "--- End jemalloc statistics ---\n");
+
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "{\n"
+		    "\t\"jemalloc\": {\n");
+	} else {
+		malloc_cprintf(write_cb, cbopaque,
+		    "___ Begin jemalloc statistics ___\n");
+	}
+
+	if (general)
+		stats_general_print(write_cb, cbopaque, json, merged, unmerged);
+	if (config_stats) {
+		stats_print_helper(write_cb, cbopaque, json, merged, unmerged,
+		    bins, large, huge);
+	}
+	if (json) {
+		malloc_cprintf(write_cb, cbopaque,
+		    "\t}\n"
+		    "}\n");
+	} else {
+		malloc_cprintf(write_cb, cbopaque,
+		    "--- End jemalloc statistics ---\n");
+	}
 }
diff --git a/deps/jemalloc/src/tcache.c b/deps/jemalloc/src/tcache.c
old mode 100644
new mode 100755
index fdafd0c62..21540ff46
--- a/deps/jemalloc/src/tcache.c
+++ b/deps/jemalloc/src/tcache.c
@@ -10,7 +10,7 @@ ssize_t	opt_lg_tcache_max = LG_TCACHE_MAXCLASS_DEFAULT;
 tcache_bin_info_t	*tcache_bin_info;
 static unsigned		stack_nelms; /* Total stack elms per tcache. */
 
-size_t			nhbins;
+unsigned		nhbins;
 size_t			tcache_maxclass;
 
 tcaches_t		*tcaches;
@@ -23,10 +23,11 @@ static tcaches_t	*tcaches_avail;
 
 /******************************************************************************/
 
-size_t	tcache_salloc(const void *ptr)
+size_t
+tcache_salloc(tsdn_t *tsdn, const void *ptr)
 {
 
-	return (arena_salloc(ptr, false));
+	return (arena_salloc(tsdn, ptr, false));
 }
 
 void
@@ -67,20 +68,19 @@ tcache_event_hard(tsd_t *tsd, tcache_t *tcache)
 	tcache->next_gc_bin++;
 	if (tcache->next_gc_bin == nhbins)
 		tcache->next_gc_bin = 0;
-	tcache->ev_cnt = 0;
 }
 
 void *
-tcache_alloc_small_hard(tsd_t *tsd, arena_t *arena, tcache_t *tcache,
-    tcache_bin_t *tbin, szind_t binind)
+tcache_alloc_small_hard(tsdn_t *tsdn, arena_t *arena, tcache_t *tcache,
+    tcache_bin_t *tbin, szind_t binind, bool *tcache_success)
 {
 	void *ret;
 
-	arena_tcache_fill_small(arena, tbin, binind, config_prof ?
+	arena_tcache_fill_small(tsdn, arena, tbin, binind, config_prof ?
 	    tcache->prof_accumbytes : 0);
 	if (config_prof)
 		tcache->prof_accumbytes = 0;
-	ret = tcache_alloc_easy(tbin);
+	ret = tcache_alloc_easy(tbin, tcache_success);
 
 	return (ret);
 }
@@ -102,17 +102,18 @@ tcache_bin_flush_small(tsd_t *tsd, tcache_t *tcache, tcache_bin_t *tbin,
 	for (nflush = tbin->ncached - rem; nflush > 0; nflush = ndeferred) {
 		/* Lock the arena bin associated with the first object. */
 		arena_chunk_t *chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(
-		    tbin->avail[0]);
+		    *(tbin->avail - 1));
 		arena_t *bin_arena = extent_node_arena_get(&chunk->node);
 		arena_bin_t *bin = &bin_arena->bins[binind];
 
 		if (config_prof && bin_arena == arena) {
-			if (arena_prof_accum(arena, tcache->prof_accumbytes))
-				prof_idump();
+			if (arena_prof_accum(tsd_tsdn(tsd), arena,
+			    tcache->prof_accumbytes))
+				prof_idump(tsd_tsdn(tsd));
 			tcache->prof_accumbytes = 0;
 		}
 
-		malloc_mutex_lock(&bin->lock);
+		malloc_mutex_lock(tsd_tsdn(tsd), &bin->lock);
 		if (config_stats && bin_arena == arena) {
 			assert(!merged_stats);
 			merged_stats = true;
@@ -122,16 +123,16 @@ tcache_bin_flush_small(tsd_t *tsd, tcache_t *tcache, tcache_bin_t *tbin,
 		}
 		ndeferred = 0;
 		for (i = 0; i < nflush; i++) {
-			ptr = tbin->avail[i];
+			ptr = *(tbin->avail - 1 - i);
 			assert(ptr != NULL);
 			chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(ptr);
 			if (extent_node_arena_get(&chunk->node) == bin_arena) {
 				size_t pageind = ((uintptr_t)ptr -
 				    (uintptr_t)chunk) >> LG_PAGE;
 				arena_chunk_map_bits_t *bitselm =
-				    arena_bitselm_get(chunk, pageind);
-				arena_dalloc_bin_junked_locked(bin_arena, chunk,
-				    ptr, bitselm);
+				    arena_bitselm_get_mutable(chunk, pageind);
+				arena_dalloc_bin_junked_locked(tsd_tsdn(tsd),
+				    bin_arena, chunk, ptr, bitselm);
 			} else {
 				/*
 				 * This object was allocated via a different
@@ -139,11 +140,12 @@ tcache_bin_flush_small(tsd_t *tsd, tcache_t *tcache, tcache_bin_t *tbin,
 				 * locked.  Stash the object, so that it can be
 				 * handled in a future pass.
 				 */
-				tbin->avail[ndeferred] = ptr;
+				*(tbin->avail - 1 - ndeferred) = ptr;
 				ndeferred++;
 			}
 		}
-		malloc_mutex_unlock(&bin->lock);
+		malloc_mutex_unlock(tsd_tsdn(tsd), &bin->lock);
+		arena_decay_ticks(tsd_tsdn(tsd), bin_arena, nflush - ndeferred);
 	}
 	if (config_stats && !merged_stats) {
 		/*
@@ -151,15 +153,15 @@ tcache_bin_flush_small(tsd_t *tsd, tcache_t *tcache, tcache_bin_t *tbin,
 		 * arena, so the stats didn't get merged.  Manually do so now.
 		 */
 		arena_bin_t *bin = &arena->bins[binind];
-		malloc_mutex_lock(&bin->lock);
+		malloc_mutex_lock(tsd_tsdn(tsd), &bin->lock);
 		bin->stats.nflushes++;
 		bin->stats.nrequests += tbin->tstats.nrequests;
 		tbin->tstats.nrequests = 0;
-		malloc_mutex_unlock(&bin->lock);
+		malloc_mutex_unlock(tsd_tsdn(tsd), &bin->lock);
 	}
 
-	memmove(tbin->avail, &tbin->avail[tbin->ncached - rem],
-	    rem * sizeof(void *));
+	memmove(tbin->avail - rem, tbin->avail - tbin->ncached, rem *
+	    sizeof(void *));
 	tbin->ncached = rem;
 	if ((int)tbin->ncached < tbin->low_water)
 		tbin->low_water = tbin->ncached;
@@ -182,13 +184,13 @@ tcache_bin_flush_large(tsd_t *tsd, tcache_bin_t *tbin, szind_t binind,
 	for (nflush = tbin->ncached - rem; nflush > 0; nflush = ndeferred) {
 		/* Lock the arena associated with the first object. */
 		arena_chunk_t *chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(
-		    tbin->avail[0]);
+		    *(tbin->avail - 1));
 		arena_t *locked_arena = extent_node_arena_get(&chunk->node);
 		UNUSED bool idump;
 
 		if (config_prof)
 			idump = false;
-		malloc_mutex_lock(&locked_arena->lock);
+		malloc_mutex_lock(tsd_tsdn(tsd), &locked_arena->lock);
 		if ((config_prof || config_stats) && locked_arena == arena) {
 			if (config_prof) {
 				idump = arena_prof_accum_locked(arena,
@@ -206,13 +208,13 @@ tcache_bin_flush_large(tsd_t *tsd, tcache_bin_t *tbin, szind_t binind,
 		}
 		ndeferred = 0;
 		for (i = 0; i < nflush; i++) {
-			ptr = tbin->avail[i];
+			ptr = *(tbin->avail - 1 - i);
 			assert(ptr != NULL);
 			chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(ptr);
 			if (extent_node_arena_get(&chunk->node) ==
 			    locked_arena) {
-				arena_dalloc_large_junked_locked(locked_arena,
-				    chunk, ptr);
+				arena_dalloc_large_junked_locked(tsd_tsdn(tsd),
+				    locked_arena, chunk, ptr);
 			} else {
 				/*
 				 * This object was allocated via a different
@@ -220,62 +222,56 @@ tcache_bin_flush_large(tsd_t *tsd, tcache_bin_t *tbin, szind_t binind,
 				 * Stash the object, so that it can be handled
 				 * in a future pass.
 				 */
-				tbin->avail[ndeferred] = ptr;
+				*(tbin->avail - 1 - ndeferred) = ptr;
 				ndeferred++;
 			}
 		}
-		malloc_mutex_unlock(&locked_arena->lock);
+		malloc_mutex_unlock(tsd_tsdn(tsd), &locked_arena->lock);
 		if (config_prof && idump)
-			prof_idump();
+			prof_idump(tsd_tsdn(tsd));
+		arena_decay_ticks(tsd_tsdn(tsd), locked_arena, nflush -
+		    ndeferred);
 	}
 	if (config_stats && !merged_stats) {
 		/*
 		 * The flush loop didn't happen to flush to this thread's
 		 * arena, so the stats didn't get merged.  Manually do so now.
 		 */
-		malloc_mutex_lock(&arena->lock);
+		malloc_mutex_lock(tsd_tsdn(tsd), &arena->lock);
 		arena->stats.nrequests_large += tbin->tstats.nrequests;
 		arena->stats.lstats[binind - NBINS].nrequests +=
 		    tbin->tstats.nrequests;
 		tbin->tstats.nrequests = 0;
-		malloc_mutex_unlock(&arena->lock);
+		malloc_mutex_unlock(tsd_tsdn(tsd), &arena->lock);
 	}
 
-	memmove(tbin->avail, &tbin->avail[tbin->ncached - rem],
-	    rem * sizeof(void *));
+	memmove(tbin->avail - rem, tbin->avail - tbin->ncached, rem *
+	    sizeof(void *));
 	tbin->ncached = rem;
 	if ((int)tbin->ncached < tbin->low_water)
 		tbin->low_water = tbin->ncached;
 }
 
-void
-tcache_arena_associate(tcache_t *tcache, arena_t *arena)
+static void
+tcache_arena_associate(tsdn_t *tsdn, tcache_t *tcache, arena_t *arena)
 {
 
 	if (config_stats) {
 		/* Link into list of extant tcaches. */
-		malloc_mutex_lock(&arena->lock);
+		malloc_mutex_lock(tsdn, &arena->lock);
 		ql_elm_new(tcache, link);
 		ql_tail_insert(&arena->tcache_ql, tcache, link);
-		malloc_mutex_unlock(&arena->lock);
+		malloc_mutex_unlock(tsdn, &arena->lock);
 	}
 }
 
-void
-tcache_arena_reassociate(tcache_t *tcache, arena_t *oldarena, arena_t *newarena)
-{
-
-	tcache_arena_dissociate(tcache, oldarena);
-	tcache_arena_associate(tcache, newarena);
-}
-
-void
-tcache_arena_dissociate(tcache_t *tcache, arena_t *arena)
+static void
+tcache_arena_dissociate(tsdn_t *tsdn, tcache_t *tcache, arena_t *arena)
 {
 
 	if (config_stats) {
 		/* Unlink from list of extant tcaches. */
-		malloc_mutex_lock(&arena->lock);
+		malloc_mutex_lock(tsdn, &arena->lock);
 		if (config_debug) {
 			bool in_ql = false;
 			tcache_t *iter;
@@ -288,11 +284,20 @@ tcache_arena_dissociate(tcache_t *tcache, arena_t *arena)
 			assert(in_ql);
 		}
 		ql_remove(&arena->tcache_ql, tcache, link);
-		tcache_stats_merge(tcache, arena);
-		malloc_mutex_unlock(&arena->lock);
+		tcache_stats_merge(tsdn, tcache, arena);
+		malloc_mutex_unlock(tsdn, &arena->lock);
 	}
 }
 
+void
+tcache_arena_reassociate(tsdn_t *tsdn, tcache_t *tcache, arena_t *oldarena,
+    arena_t *newarena)
+{
+
+	tcache_arena_dissociate(tsdn, tcache, oldarena);
+	tcache_arena_associate(tsdn, tcache, newarena);
+}
+
 tcache_t *
 tcache_get_hard(tsd_t *tsd)
 {
@@ -306,11 +311,11 @@ tcache_get_hard(tsd_t *tsd)
 	arena = arena_choose(tsd, NULL);
 	if (unlikely(arena == NULL))
 		return (NULL);
-	return (tcache_create(tsd, arena));
+	return (tcache_create(tsd_tsdn(tsd), arena));
 }
 
 tcache_t *
-tcache_create(tsd_t *tsd, arena_t *arena)
+tcache_create(tsdn_t *tsdn, arena_t *arena)
 {
 	tcache_t *tcache;
 	size_t size, stack_offset;
@@ -324,18 +329,26 @@ tcache_create(tsd_t *tsd, arena_t *arena)
 	/* Avoid false cacheline sharing. */
 	size = sa2u(size, CACHELINE);
 
-	tcache = ipallocztm(tsd, size, CACHELINE, true, false, true, a0get());
+	tcache = ipallocztm(tsdn, size, CACHELINE, true, NULL, true,
+	    arena_get(TSDN_NULL, 0, true));
 	if (tcache == NULL)
 		return (NULL);
 
-	tcache_arena_associate(tcache, arena);
+	tcache_arena_associate(tsdn, tcache, arena);
+
+	ticker_init(&tcache->gc_ticker, TCACHE_GC_INCR);
 
 	assert((TCACHE_NSLOTS_SMALL_MAX & 1U) == 0);
 	for (i = 0; i < nhbins; i++) {
 		tcache->tbins[i].lg_fill_div = 1;
+		stack_offset += tcache_bin_info[i].ncached_max * sizeof(void *);
+		/*
+		 * avail points past the available space.  Allocations will
+		 * access the slots toward higher addresses (for the benefit of
+		 * prefetch).
+		 */
 		tcache->tbins[i].avail = (void **)((uintptr_t)tcache +
 		    (uintptr_t)stack_offset);
-		stack_offset += tcache_bin_info[i].ncached_max * sizeof(void *);
 	}
 
 	return (tcache);
@@ -348,7 +361,7 @@ tcache_destroy(tsd_t *tsd, tcache_t *tcache)
 	unsigned i;
 
 	arena = arena_choose(tsd, NULL);
-	tcache_arena_dissociate(tcache, arena);
+	tcache_arena_dissociate(tsd_tsdn(tsd), tcache, arena);
 
 	for (i = 0; i < NBINS; i++) {
 		tcache_bin_t *tbin = &tcache->tbins[i];
@@ -356,9 +369,9 @@ tcache_destroy(tsd_t *tsd, tcache_t *tcache)
 
 		if (config_stats && tbin->tstats.nrequests != 0) {
 			arena_bin_t *bin = &arena->bins[i];
-			malloc_mutex_lock(&bin->lock);
+			malloc_mutex_lock(tsd_tsdn(tsd), &bin->lock);
 			bin->stats.nrequests += tbin->tstats.nrequests;
-			malloc_mutex_unlock(&bin->lock);
+			malloc_mutex_unlock(tsd_tsdn(tsd), &bin->lock);
 		}
 	}
 
@@ -367,19 +380,19 @@ tcache_destroy(tsd_t *tsd, tcache_t *tcache)
 		tcache_bin_flush_large(tsd, tbin, i, 0, tcache);
 
 		if (config_stats && tbin->tstats.nrequests != 0) {
-			malloc_mutex_lock(&arena->lock);
+			malloc_mutex_lock(tsd_tsdn(tsd), &arena->lock);
 			arena->stats.nrequests_large += tbin->tstats.nrequests;
 			arena->stats.lstats[i - NBINS].nrequests +=
 			    tbin->tstats.nrequests;
-			malloc_mutex_unlock(&arena->lock);
+			malloc_mutex_unlock(tsd_tsdn(tsd), &arena->lock);
 		}
 	}
 
 	if (config_prof && tcache->prof_accumbytes > 0 &&
-	    arena_prof_accum(arena, tcache->prof_accumbytes))
-		prof_idump();
+	    arena_prof_accum(tsd_tsdn(tsd), arena, tcache->prof_accumbytes))
+		prof_idump(tsd_tsdn(tsd));
 
-	idalloctm(tsd, tcache, false, true);
+	idalloctm(tsd_tsdn(tsd), tcache, NULL, true, true);
 }
 
 void
@@ -403,21 +416,22 @@ tcache_enabled_cleanup(tsd_t *tsd)
 	/* Do nothing. */
 }
 
-/* Caller must own arena->lock. */
 void
-tcache_stats_merge(tcache_t *tcache, arena_t *arena)
+tcache_stats_merge(tsdn_t *tsdn, tcache_t *tcache, arena_t *arena)
 {
 	unsigned i;
 
 	cassert(config_stats);
 
+	malloc_mutex_assert_owner(tsdn, &arena->lock);
+
 	/* Merge and reset tcache stats. */
 	for (i = 0; i < NBINS; i++) {
 		arena_bin_t *bin = &arena->bins[i];
 		tcache_bin_t *tbin = &tcache->tbins[i];
-		malloc_mutex_lock(&bin->lock);
+		malloc_mutex_lock(tsdn, &bin->lock);
 		bin->stats.nrequests += tbin->tstats.nrequests;
-		malloc_mutex_unlock(&bin->lock);
+		malloc_mutex_unlock(tsdn, &bin->lock);
 		tbin->tstats.nrequests = 0;
 	}
 
@@ -433,11 +447,12 @@ tcache_stats_merge(tcache_t *tcache, arena_t *arena)
 bool
 tcaches_create(tsd_t *tsd, unsigned *r_ind)
 {
+	arena_t *arena;
 	tcache_t *tcache;
 	tcaches_t *elm;
 
 	if (tcaches == NULL) {
-		tcaches = base_alloc(sizeof(tcache_t *) *
+		tcaches = base_alloc(tsd_tsdn(tsd), sizeof(tcache_t *) *
 		    (MALLOCX_TCACHE_MAX+1));
 		if (tcaches == NULL)
 			return (true);
@@ -445,7 +460,10 @@ tcaches_create(tsd_t *tsd, unsigned *r_ind)
 
 	if (tcaches_avail == NULL && tcaches_past > MALLOCX_TCACHE_MAX)
 		return (true);
-	tcache = tcache_create(tsd, a0get());
+	arena = arena_ichoose(tsd, NULL);
+	if (unlikely(arena == NULL))
+		return (true);
+	tcache = tcache_create(tsd_tsdn(tsd), arena);
 	if (tcache == NULL)
 		return (true);
 
@@ -453,7 +471,7 @@ tcaches_create(tsd_t *tsd, unsigned *r_ind)
 		elm = tcaches_avail;
 		tcaches_avail = tcaches_avail->next;
 		elm->tcache = tcache;
-		*r_ind = elm - tcaches;
+		*r_ind = (unsigned)(elm - tcaches);
 	} else {
 		elm = &tcaches[tcaches_past];
 		elm->tcache = tcache;
@@ -491,7 +509,7 @@ tcaches_destroy(tsd_t *tsd, unsigned ind)
 }
 
 bool
-tcache_boot(void)
+tcache_boot(tsdn_t *tsdn)
 {
 	unsigned i;
 
@@ -499,17 +517,17 @@ tcache_boot(void)
 	 * If necessary, clamp opt_lg_tcache_max, now that large_maxclass is
 	 * known.
 	 */
-	if (opt_lg_tcache_max < 0 || (1U << opt_lg_tcache_max) < SMALL_MAXCLASS)
+	if (opt_lg_tcache_max < 0 || (ZU(1) << opt_lg_tcache_max) < SMALL_MAXCLASS)
 		tcache_maxclass = SMALL_MAXCLASS;
-	else if ((1U << opt_lg_tcache_max) > large_maxclass)
+	else if ((ZU(1) << opt_lg_tcache_max) > large_maxclass)
 		tcache_maxclass = large_maxclass;
 	else
-		tcache_maxclass = (1U << opt_lg_tcache_max);
+		tcache_maxclass = (ZU(1) << opt_lg_tcache_max);
 
 	nhbins = size2index(tcache_maxclass) + 1;
 
 	/* Initialize tcache_bin_info. */
-	tcache_bin_info = (tcache_bin_info_t *)base_alloc(nhbins *
+	tcache_bin_info = (tcache_bin_info_t *)base_alloc(tsdn, nhbins *
 	    sizeof(tcache_bin_info_t));
 	if (tcache_bin_info == NULL)
 		return (true);
diff --git a/deps/jemalloc/src/ticker.c b/deps/jemalloc/src/ticker.c
new file mode 100644
index 000000000..db0902404
--- /dev/null
+++ b/deps/jemalloc/src/ticker.c
@@ -0,0 +1,2 @@
+#define	JEMALLOC_TICKER_C_
+#include "jemalloc/internal/jemalloc_internal.h"
diff --git a/deps/jemalloc/src/tsd.c b/deps/jemalloc/src/tsd.c
index 9ffe9afef..ec69a51c3 100644
--- a/deps/jemalloc/src/tsd.c
+++ b/deps/jemalloc/src/tsd.c
@@ -77,7 +77,7 @@ tsd_cleanup(void *arg)
 		/* Do nothing. */
 		break;
 	case tsd_state_nominal:
-#define O(n, t)								\
+#define	O(n, t)								\
 		n##_cleanup(tsd);
 MALLOC_TSD
 #undef O
@@ -106,15 +106,17 @@ MALLOC_TSD
 	}
 }
 
-bool
+tsd_t *
 malloc_tsd_boot0(void)
 {
+	tsd_t *tsd;
 
 	ncleanups = 0;
 	if (tsd_boot0())
-		return (true);
-	*tsd_arenas_cache_bypassp_get(tsd_fetch()) = true;
-	return (false);
+		return (NULL);
+	tsd = tsd_fetch();
+	*tsd_arenas_tdata_bypassp_get(tsd) = true;
+	return (tsd);
 }
 
 void
@@ -122,7 +124,7 @@ malloc_tsd_boot1(void)
 {
 
 	tsd_boot1();
-	*tsd_arenas_cache_bypassp_get(tsd_fetch()) = false;
+	*tsd_arenas_tdata_bypassp_get(tsd_fetch()) = false;
 }
 
 #ifdef _WIN32
@@ -148,13 +150,15 @@ _tls_callback(HINSTANCE hinstDLL, DWORD fdwReason, LPVOID lpvReserved)
 #ifdef _MSC_VER
 #  ifdef _M_IX86
 #    pragma comment(linker, "/INCLUDE:__tls_used")
+#    pragma comment(linker, "/INCLUDE:_tls_callback")
 #  else
 #    pragma comment(linker, "/INCLUDE:_tls_used")
+#    pragma comment(linker, "/INCLUDE:tls_callback")
 #  endif
 #  pragma section(".CRT$XLY",long,read)
 #endif
 JEMALLOC_SECTION(".CRT$XLY") JEMALLOC_ATTR(used)
-static BOOL	(WINAPI *const tls_callback)(HINSTANCE hinstDLL,
+BOOL	(WINAPI *const tls_callback)(HINSTANCE hinstDLL,
     DWORD fdwReason, LPVOID lpvReserved) = _tls_callback;
 #endif
 
@@ -167,10 +171,10 @@ tsd_init_check_recursion(tsd_init_head_t *head, tsd_init_block_t *block)
 	tsd_init_block_t *iter;
 
 	/* Check whether this thread has already inserted into the list. */
-	malloc_mutex_lock(&head->lock);
+	malloc_mutex_lock(TSDN_NULL, &head->lock);
 	ql_foreach(iter, &head->blocks, link) {
 		if (iter->thread == self) {
-			malloc_mutex_unlock(&head->lock);
+			malloc_mutex_unlock(TSDN_NULL, &head->lock);
 			return (iter->data);
 		}
 	}
@@ -178,7 +182,7 @@ tsd_init_check_recursion(tsd_init_head_t *head, tsd_init_block_t *block)
 	ql_elm_new(block, link);
 	block->thread = self;
 	ql_tail_insert(&head->blocks, block, link);
-	malloc_mutex_unlock(&head->lock);
+	malloc_mutex_unlock(TSDN_NULL, &head->lock);
 	return (NULL);
 }
 
@@ -186,8 +190,8 @@ void
 tsd_init_finish(tsd_init_head_t *head, tsd_init_block_t *block)
 {
 
-	malloc_mutex_lock(&head->lock);
+	malloc_mutex_lock(TSDN_NULL, &head->lock);
 	ql_remove(&head->blocks, block, link);
-	malloc_mutex_unlock(&head->lock);
+	malloc_mutex_unlock(TSDN_NULL, &head->lock);
 }
 #endif
diff --git a/deps/jemalloc/src/util.c b/deps/jemalloc/src/util.c
old mode 100644
new mode 100755
index 4cb0d6c1e..dd8c23630
--- a/deps/jemalloc/src/util.c
+++ b/deps/jemalloc/src/util.c
@@ -1,3 +1,7 @@
+/*
+ * Define simple versions of assertion macros that won't recurse in case
+ * of assertion failures in malloc_*printf().
+ */
 #define	assert(e) do {							\
 	if (config_debug && !(e)) {					\
 		malloc_write("<jemalloc>: Failed assertion\n");		\
@@ -10,6 +14,7 @@
 		malloc_write("<jemalloc>: Unreachable code reached\n");	\
 		abort();						\
 	}								\
+	unreachable();							\
 } while (0)
 
 #define	not_implemented() do {						\
@@ -44,15 +49,19 @@ static void
 wrtmessage(void *cbopaque, const char *s)
 {
 
-#ifdef SYS_write
+#if defined(JEMALLOC_USE_SYSCALL) && defined(SYS_write)
 	/*
 	 * Use syscall(2) rather than write(2) when possible in order to avoid
 	 * the possibility of memory allocation within libc.  This is necessary
 	 * on FreeBSD; most operating systems do not have this problem though.
+	 *
+	 * syscall() returns long or int, depending on platform, so capture the
+	 * unused result in the widest plausible type to avoid compiler
+	 * warnings.
 	 */
-	UNUSED int result = syscall(SYS_write, STDERR_FILENO, s, strlen(s));
+	UNUSED long result = syscall(SYS_write, STDERR_FILENO, s, strlen(s));
 #else
-	UNUSED int result = write(STDERR_FILENO, s, strlen(s));
+	UNUSED ssize_t result = write(STDERR_FILENO, s, strlen(s));
 #endif
 }
 
@@ -82,7 +91,7 @@ buferror(int err, char *buf, size_t buflen)
 
 #ifdef _WIN32
 	FormatMessageA(FORMAT_MESSAGE_FROM_SYSTEM, NULL, err, 0,
-	    (LPSTR)buf, buflen, NULL);
+	    (LPSTR)buf, (DWORD)buflen, NULL);
 	return (0);
 #elif defined(__GLIBC__) && defined(_GNU_SOURCE)
 	char *b = strerror_r(err, buf, buflen);
@@ -191,7 +200,7 @@ malloc_strtoumax(const char *restrict nptr, char **restrict endptr, int base)
 		p++;
 	}
 	if (neg)
-		ret = -ret;
+		ret = (uintmax_t)(-((intmax_t)ret));
 
 	if (p == ns) {
 		/* No conversion performed. */
@@ -306,10 +315,9 @@ x2s(uintmax_t x, bool alt_form, bool uppercase, char *s, size_t *slen_p)
 	return (s);
 }
 
-int
+size_t
 malloc_vsnprintf(char *str, size_t size, const char *format, va_list ap)
 {
-	int ret;
 	size_t i;
 	const char *f;
 
@@ -400,6 +408,8 @@ malloc_vsnprintf(char *str, size_t size, const char *format, va_list ap)
 			int prec = -1;
 			int width = -1;
 			unsigned char len = '?';
+			char *s;
+			size_t slen;
 
 			f++;
 			/* Flags. */
@@ -490,8 +500,6 @@ malloc_vsnprintf(char *str, size_t size, const char *format, va_list ap)
 			}
 			/* Conversion specifier. */
 			switch (*f) {
-				char *s;
-				size_t slen;
 			case '%':
 				/* %% */
 				APPEND_C(*f);
@@ -577,20 +585,19 @@ malloc_vsnprintf(char *str, size_t size, const char *format, va_list ap)
 		str[i] = '\0';
 	else
 		str[size - 1] = '\0';
-	ret = i;
 
 #undef APPEND_C
 #undef APPEND_S
 #undef APPEND_PADDED_S
 #undef GET_ARG_NUMERIC
-	return (ret);
+	return (i);
 }
 
 JEMALLOC_FORMAT_PRINTF(3, 4)
-int
+size_t
 malloc_snprintf(char *str, size_t size, const char *format, ...)
 {
-	int ret;
+	size_t ret;
 	va_list ap;
 
 	va_start(ap, format);
@@ -648,3 +655,12 @@ malloc_printf(const char *format, ...)
 	malloc_vcprintf(NULL, NULL, format, ap);
 	va_end(ap);
 }
+
+/*
+ * Restore normal assertion macros, in order to make it possible to compile all
+ * C files as a single concatenation.
+ */
+#undef assert
+#undef not_reached
+#undef not_implemented
+#include "jemalloc/internal/assert.h"
diff --git a/deps/jemalloc/src/witness.c b/deps/jemalloc/src/witness.c
new file mode 100644
index 000000000..23753f246
--- /dev/null
+++ b/deps/jemalloc/src/witness.c
@@ -0,0 +1,136 @@
+#define	JEMALLOC_WITNESS_C_
+#include "jemalloc/internal/jemalloc_internal.h"
+
+void
+witness_init(witness_t *witness, const char *name, witness_rank_t rank,
+    witness_comp_t *comp)
+{
+
+	witness->name = name;
+	witness->rank = rank;
+	witness->comp = comp;
+}
+
+#ifdef JEMALLOC_JET
+#undef witness_lock_error
+#define	witness_lock_error JEMALLOC_N(n_witness_lock_error)
+#endif
+void
+witness_lock_error(const witness_list_t *witnesses, const witness_t *witness)
+{
+	witness_t *w;
+
+	malloc_printf("<jemalloc>: Lock rank order reversal:");
+	ql_foreach(w, witnesses, link) {
+		malloc_printf(" %s(%u)", w->name, w->rank);
+	}
+	malloc_printf(" %s(%u)\n", witness->name, witness->rank);
+	abort();
+}
+#ifdef JEMALLOC_JET
+#undef witness_lock_error
+#define	witness_lock_error JEMALLOC_N(witness_lock_error)
+witness_lock_error_t *witness_lock_error = JEMALLOC_N(n_witness_lock_error);
+#endif
+
+#ifdef JEMALLOC_JET
+#undef witness_owner_error
+#define	witness_owner_error JEMALLOC_N(n_witness_owner_error)
+#endif
+void
+witness_owner_error(const witness_t *witness)
+{
+
+	malloc_printf("<jemalloc>: Should own %s(%u)\n", witness->name,
+	    witness->rank);
+	abort();
+}
+#ifdef JEMALLOC_JET
+#undef witness_owner_error
+#define	witness_owner_error JEMALLOC_N(witness_owner_error)
+witness_owner_error_t *witness_owner_error = JEMALLOC_N(n_witness_owner_error);
+#endif
+
+#ifdef JEMALLOC_JET
+#undef witness_not_owner_error
+#define	witness_not_owner_error JEMALLOC_N(n_witness_not_owner_error)
+#endif
+void
+witness_not_owner_error(const witness_t *witness)
+{
+
+	malloc_printf("<jemalloc>: Should not own %s(%u)\n", witness->name,
+	    witness->rank);
+	abort();
+}
+#ifdef JEMALLOC_JET
+#undef witness_not_owner_error
+#define	witness_not_owner_error JEMALLOC_N(witness_not_owner_error)
+witness_not_owner_error_t *witness_not_owner_error =
+    JEMALLOC_N(n_witness_not_owner_error);
+#endif
+
+#ifdef JEMALLOC_JET
+#undef witness_lockless_error
+#define	witness_lockless_error JEMALLOC_N(n_witness_lockless_error)
+#endif
+void
+witness_lockless_error(const witness_list_t *witnesses)
+{
+	witness_t *w;
+
+	malloc_printf("<jemalloc>: Should not own any locks:");
+	ql_foreach(w, witnesses, link) {
+		malloc_printf(" %s(%u)", w->name, w->rank);
+	}
+	malloc_printf("\n");
+	abort();
+}
+#ifdef JEMALLOC_JET
+#undef witness_lockless_error
+#define	witness_lockless_error JEMALLOC_N(witness_lockless_error)
+witness_lockless_error_t *witness_lockless_error =
+    JEMALLOC_N(n_witness_lockless_error);
+#endif
+
+void
+witnesses_cleanup(tsd_t *tsd)
+{
+
+	witness_assert_lockless(tsd_tsdn(tsd));
+
+	/* Do nothing. */
+}
+
+void
+witness_fork_cleanup(tsd_t *tsd)
+{
+
+	/* Do nothing. */
+}
+
+void
+witness_prefork(tsd_t *tsd)
+{
+
+	tsd_witness_fork_set(tsd, true);
+}
+
+void
+witness_postfork_parent(tsd_t *tsd)
+{
+
+	tsd_witness_fork_set(tsd, false);
+}
+
+void
+witness_postfork_child(tsd_t *tsd)
+{
+#ifndef JEMALLOC_MUTEX_INIT_CB
+	witness_list_t *witnesses;
+
+	witnesses = tsd_witnessesp_get(tsd);
+	ql_new(witnesses);
+#endif
+	tsd_witness_fork_set(tsd, false);
+}
diff --git a/deps/jemalloc/src/zone.c b/deps/jemalloc/src/zone.c
index 12e1734a9..0571920e4 100644
--- a/deps/jemalloc/src/zone.c
+++ b/deps/jemalloc/src/zone.c
@@ -4,7 +4,7 @@
 #endif
 
 /*
- * The malloc_default_purgeable_zone function is only available on >= 10.6.
+ * The malloc_default_purgeable_zone() function is only available on >= 10.6.
  * We need to check whether it is present at runtime, thus the weak_import.
  */
 extern malloc_zone_t *malloc_default_purgeable_zone(void)
@@ -13,8 +13,9 @@ JEMALLOC_ATTR(weak_import);
 /******************************************************************************/
 /* Data. */
 
-static malloc_zone_t zone;
-static struct malloc_introspection_t zone_introspect;
+static malloc_zone_t *default_zone, *purgeable_zone;
+static malloc_zone_t jemalloc_zone;
+static struct malloc_introspection_t jemalloc_zone_introspect;
 
 /******************************************************************************/
 /* Function prototypes for non-inline static functions. */
@@ -56,7 +57,7 @@ zone_size(malloc_zone_t *zone, void *ptr)
 	 * not work in practice, we must check all pointers to assure that they
 	 * reside within a mapped chunk before determining size.
 	 */
-	return (ivsalloc(ptr, config_prof));
+	return (ivsalloc(tsdn_fetch(), ptr, config_prof));
 }
 
 static void *
@@ -87,7 +88,7 @@ static void
 zone_free(malloc_zone_t *zone, void *ptr)
 {
 
-	if (ivsalloc(ptr, config_prof) != 0) {
+	if (ivsalloc(tsdn_fetch(), ptr, config_prof) != 0) {
 		je_free(ptr);
 		return;
 	}
@@ -99,7 +100,7 @@ static void *
 zone_realloc(malloc_zone_t *zone, void *ptr, size_t size)
 {
 
-	if (ivsalloc(ptr, config_prof) != 0)
+	if (ivsalloc(tsdn_fetch(), ptr, config_prof) != 0)
 		return (je_realloc(ptr, size));
 
 	return (realloc(ptr, size));
@@ -121,9 +122,11 @@ zone_memalign(malloc_zone_t *zone, size_t alignment, size_t size)
 static void
 zone_free_definite_size(malloc_zone_t *zone, void *ptr, size_t size)
 {
+	size_t alloc_size;
 
-	if (ivsalloc(ptr, config_prof) != 0) {
-		assert(ivsalloc(ptr, config_prof) == size);
+	alloc_size = ivsalloc(tsdn_fetch(), ptr, config_prof);
+	if (alloc_size != 0) {
+		assert(alloc_size == size);
 		je_free(ptr);
 		return;
 	}
@@ -162,89 +165,103 @@ static void
 zone_force_unlock(malloc_zone_t *zone)
 {
 
+	/*
+	 * Call jemalloc_postfork_child() rather than
+	 * jemalloc_postfork_parent(), because this function is executed by both
+	 * parent and child.  The parent can tolerate having state
+	 * reinitialized, but the child cannot unlock mutexes that were locked
+	 * by the parent.
+	 */
 	if (isthreaded)
-		jemalloc_postfork_parent();
+		jemalloc_postfork_child();
 }
 
-JEMALLOC_ATTR(constructor)
-void
-register_zone(void)
+static void
+zone_init(void)
 {
 
-	/*
-	 * If something else replaced the system default zone allocator, don't
-	 * register jemalloc's.
-	 */
-	malloc_zone_t *default_zone = malloc_default_zone();
-	malloc_zone_t *purgeable_zone = NULL;
-	if (!default_zone->zone_name ||
-	    strcmp(default_zone->zone_name, "DefaultMallocZone") != 0) {
-		return;
-	}
-
-	zone.size = (void *)zone_size;
-	zone.malloc = (void *)zone_malloc;
-	zone.calloc = (void *)zone_calloc;
-	zone.valloc = (void *)zone_valloc;
-	zone.free = (void *)zone_free;
-	zone.realloc = (void *)zone_realloc;
-	zone.destroy = (void *)zone_destroy;
-	zone.zone_name = "jemalloc_zone";
-	zone.batch_malloc = NULL;
-	zone.batch_free = NULL;
-	zone.introspect = &zone_introspect;
-	zone.version = JEMALLOC_ZONE_VERSION;
+	jemalloc_zone.size = (void *)zone_size;
+	jemalloc_zone.malloc = (void *)zone_malloc;
+	jemalloc_zone.calloc = (void *)zone_calloc;
+	jemalloc_zone.valloc = (void *)zone_valloc;
+	jemalloc_zone.free = (void *)zone_free;
+	jemalloc_zone.realloc = (void *)zone_realloc;
+	jemalloc_zone.destroy = (void *)zone_destroy;
+	jemalloc_zone.zone_name = "jemalloc_zone";
+	jemalloc_zone.batch_malloc = NULL;
+	jemalloc_zone.batch_free = NULL;
+	jemalloc_zone.introspect = &jemalloc_zone_introspect;
+	jemalloc_zone.version = JEMALLOC_ZONE_VERSION;
 #if (JEMALLOC_ZONE_VERSION >= 5)
-	zone.memalign = zone_memalign;
+	jemalloc_zone.memalign = zone_memalign;
 #endif
 #if (JEMALLOC_ZONE_VERSION >= 6)
-	zone.free_definite_size = zone_free_definite_size;
+	jemalloc_zone.free_definite_size = zone_free_definite_size;
 #endif
 #if (JEMALLOC_ZONE_VERSION >= 8)
-	zone.pressure_relief = NULL;
+	jemalloc_zone.pressure_relief = NULL;
 #endif
 
-	zone_introspect.enumerator = NULL;
-	zone_introspect.good_size = (void *)zone_good_size;
-	zone_introspect.check = NULL;
-	zone_introspect.print = NULL;
-	zone_introspect.log = NULL;
-	zone_introspect.force_lock = (void *)zone_force_lock;
-	zone_introspect.force_unlock = (void *)zone_force_unlock;
-	zone_introspect.statistics = NULL;
+	jemalloc_zone_introspect.enumerator = NULL;
+	jemalloc_zone_introspect.good_size = (void *)zone_good_size;
+	jemalloc_zone_introspect.check = NULL;
+	jemalloc_zone_introspect.print = NULL;
+	jemalloc_zone_introspect.log = NULL;
+	jemalloc_zone_introspect.force_lock = (void *)zone_force_lock;
+	jemalloc_zone_introspect.force_unlock = (void *)zone_force_unlock;
+	jemalloc_zone_introspect.statistics = NULL;
 #if (JEMALLOC_ZONE_VERSION >= 6)
-	zone_introspect.zone_locked = NULL;
+	jemalloc_zone_introspect.zone_locked = NULL;
 #endif
 #if (JEMALLOC_ZONE_VERSION >= 7)
-	zone_introspect.enable_discharge_checking = NULL;
-	zone_introspect.disable_discharge_checking = NULL;
-	zone_introspect.discharge = NULL;
-#ifdef __BLOCKS__
-	zone_introspect.enumerate_discharged_pointers = NULL;
-#else
-	zone_introspect.enumerate_unavailable_without_blocks = NULL;
-#endif
+	jemalloc_zone_introspect.enable_discharge_checking = NULL;
+	jemalloc_zone_introspect.disable_discharge_checking = NULL;
+	jemalloc_zone_introspect.discharge = NULL;
+#  ifdef __BLOCKS__
+	jemalloc_zone_introspect.enumerate_discharged_pointers = NULL;
+#  else
+	jemalloc_zone_introspect.enumerate_unavailable_without_blocks = NULL;
+#  endif
 #endif
+}
+
+static malloc_zone_t *
+zone_default_get(void)
+{
+	malloc_zone_t **zones = NULL;
+	unsigned int num_zones = 0;
 
 	/*
-	 * The default purgeable zone is created lazily by OSX's libc.  It uses
-	 * the default zone when it is created for "small" allocations
-	 * (< 15 KiB), but assumes the default zone is a scalable_zone.  This
-	 * obviously fails when the default zone is the jemalloc zone, so
-	 * malloc_default_purgeable_zone is called beforehand so that the
-	 * default purgeable zone is created when the default zone is still
-	 * a scalable_zone.  As purgeable zones only exist on >= 10.6, we need
-	 * to check for the existence of malloc_default_purgeable_zone() at
-	 * run time.
+	 * On OSX 10.12, malloc_default_zone returns a special zone that is not
+	 * present in the list of registered zones. That zone uses a "lite zone"
+	 * if one is present (apparently enabled when malloc stack logging is
+	 * enabled), or the first registered zone otherwise. In practice this
+	 * means unless malloc stack logging is enabled, the first registered
+	 * zone is the default.  So get the list of zones to get the first one,
+	 * instead of relying on malloc_default_zone.
 	 */
-	if (malloc_default_purgeable_zone != NULL)
-		purgeable_zone = malloc_default_purgeable_zone();
+	if (KERN_SUCCESS != malloc_get_all_zones(0, NULL,
+	    (vm_address_t**)&zones, &num_zones)) {
+		/*
+		 * Reset the value in case the failure happened after it was
+		 * set.
+		 */
+		num_zones = 0;
+	}
 
-	/* Register the custom zone.  At this point it won't be the default. */
-	malloc_zone_register(&zone);
+	if (num_zones)
+		return (zones[0]);
+
+	return (malloc_default_zone());
+}
+
+/* As written, this function can only promote jemalloc_zone. */
+static void
+zone_promote(void)
+{
+	malloc_zone_t *zone;
 
 	do {
-		default_zone = malloc_default_zone();
 		/*
 		 * Unregister and reregister the default zone.  On OSX >= 10.6,
 		 * unregistering takes the last registered zone and places it
@@ -255,6 +272,7 @@ register_zone(void)
 		 */
 		malloc_zone_unregister(default_zone);
 		malloc_zone_register(default_zone);
+
 		/*
 		 * On OSX 10.6, having the default purgeable zone appear before
 		 * the default zone makes some things crash because it thinks it
@@ -266,9 +284,47 @@ register_zone(void)
 		 * above, i.e. the default zone.  Registering it again then puts
 		 * it at the end, obviously after the default zone.
 		 */
-		if (purgeable_zone) {
+		if (purgeable_zone != NULL) {
 			malloc_zone_unregister(purgeable_zone);
 			malloc_zone_register(purgeable_zone);
 		}
-	} while (malloc_default_zone() != &zone);
+
+		zone = zone_default_get();
+	} while (zone != &jemalloc_zone);
+}
+
+JEMALLOC_ATTR(constructor)
+void
+zone_register(void)
+{
+
+	/*
+	 * If something else replaced the system default zone allocator, don't
+	 * register jemalloc's.
+	 */
+	default_zone = zone_default_get();
+	if (!default_zone->zone_name || strcmp(default_zone->zone_name,
+	    "DefaultMallocZone") != 0)
+		return;
+
+	/*
+	 * The default purgeable zone is created lazily by OSX's libc.  It uses
+	 * the default zone when it is created for "small" allocations
+	 * (< 15 KiB), but assumes the default zone is a scalable_zone.  This
+	 * obviously fails when the default zone is the jemalloc zone, so
+	 * malloc_default_purgeable_zone() is called beforehand so that the
+	 * default purgeable zone is created when the default zone is still
+	 * a scalable_zone.  As purgeable zones only exist on >= 10.6, we need
+	 * to check for the existence of malloc_default_purgeable_zone() at
+	 * run time.
+	 */
+	purgeable_zone = (malloc_default_purgeable_zone == NULL) ? NULL :
+	    malloc_default_purgeable_zone();
+
+	/* Register the custom zone.  At this point it won't be the default. */
+	zone_init();
+	malloc_zone_register(&jemalloc_zone);
+
+	/* Promote the custom zone to be default. */
+	zone_promote();
 }
diff --git a/deps/jemalloc/test/include/test/jemalloc_test.h.in b/deps/jemalloc/test/include/test/jemalloc_test.h.in
index 455569da4..1f36e4695 100644
--- a/deps/jemalloc/test/include/test/jemalloc_test.h.in
+++ b/deps/jemalloc/test/include/test/jemalloc_test.h.in
@@ -11,7 +11,6 @@
 #ifdef _WIN32
 #  include "msvc_compat/strings.h"
 #endif
-#include <sys/time.h>
 
 #ifdef _WIN32
 #  include <windows.h>
@@ -20,39 +19,6 @@
 #  include <pthread.h>
 #endif
 
-/******************************************************************************/
-/*
- * Define always-enabled assertion macros, so that test assertions execute even
- * if assertions are disabled in the library code.  These definitions must
- * exist prior to including "jemalloc/internal/util.h".
- */
-#define	assert(e) do {							\
-	if (!(e)) {							\
-		malloc_printf(						\
-		    "<jemalloc>: %s:%d: Failed assertion: \"%s\"\n",	\
-		    __FILE__, __LINE__, #e);				\
-		abort();						\
-	}								\
-} while (0)
-
-#define	not_reached() do {						\
-	malloc_printf(							\
-	    "<jemalloc>: %s:%d: Unreachable code reached\n",		\
-	    __FILE__, __LINE__);					\
-	abort();							\
-} while (0)
-
-#define	not_implemented() do {						\
-	malloc_printf("<jemalloc>: %s:%d: Not implemented\n",		\
-	    __FILE__, __LINE__);					\
-	abort();							\
-} while (0)
-
-#define	assert_not_implemented(e) do {					\
-	if (!(e))							\
-		not_implemented();					\
-} while (0)
-
 #include "test/jemalloc_test_defs.h"
 
 #ifdef JEMALLOC_OSSPIN
@@ -87,6 +53,14 @@
 #  include "jemalloc/internal/jemalloc_internal_defs.h"
 #  include "jemalloc/internal/jemalloc_internal_macros.h"
 
+static const bool config_debug =
+#ifdef JEMALLOC_DEBUG
+    true
+#else
+    false
+#endif
+    ;
+
 #  define JEMALLOC_N(n) @private_namespace@##n
 #  include "jemalloc/internal/private_namespace.h"
 
@@ -94,6 +68,7 @@
 #  define JEMALLOC_H_STRUCTS
 #  define JEMALLOC_H_EXTERNS
 #  define JEMALLOC_H_INLINES
+#  include "jemalloc/internal/nstime.h"
 #  include "jemalloc/internal/util.h"
 #  include "jemalloc/internal/qr.h"
 #  include "jemalloc/internal/ql.h"
@@ -149,3 +124,40 @@
 #include "test/thd.h"
 #define	MEXP 19937
 #include "test/SFMT.h"
+
+/******************************************************************************/
+/*
+ * Define always-enabled assertion macros, so that test assertions execute even
+ * if assertions are disabled in the library code.
+ */
+#undef assert
+#undef not_reached
+#undef not_implemented
+#undef assert_not_implemented
+
+#define	assert(e) do {							\
+	if (!(e)) {							\
+		malloc_printf(						\
+		    "<jemalloc>: %s:%d: Failed assertion: \"%s\"\n",	\
+		    __FILE__, __LINE__, #e);				\
+		abort();						\
+	}								\
+} while (0)
+
+#define	not_reached() do {						\
+	malloc_printf(							\
+	    "<jemalloc>: %s:%d: Unreachable code reached\n",		\
+	    __FILE__, __LINE__);					\
+	abort();							\
+} while (0)
+
+#define	not_implemented() do {						\
+	malloc_printf("<jemalloc>: %s:%d: Not implemented\n",		\
+	    __FILE__, __LINE__);					\
+	abort();							\
+} while (0)
+
+#define	assert_not_implemented(e) do {					\
+	if (!(e))							\
+		not_implemented();					\
+} while (0)
diff --git a/deps/jemalloc/test/include/test/mtx.h b/deps/jemalloc/test/include/test/mtx.h
index bbe822f54..58afbc3d1 100644
--- a/deps/jemalloc/test/include/test/mtx.h
+++ b/deps/jemalloc/test/include/test/mtx.h
@@ -8,6 +8,8 @@
 typedef struct {
 #ifdef _WIN32
 	CRITICAL_SECTION	lock;
+#elif (defined(JEMALLOC_OS_UNFAIR_LOCK))
+	os_unfair_lock		lock;
 #elif (defined(JEMALLOC_OSSPIN))
 	OSSpinLock		lock;
 #else
diff --git a/deps/jemalloc/test/include/test/test.h b/deps/jemalloc/test/include/test/test.h
index 3cf901fc4..c8112eb8b 100644
--- a/deps/jemalloc/test/include/test/test.h
+++ b/deps/jemalloc/test/include/test/test.h
@@ -311,6 +311,9 @@ label_test_end:								\
 #define	test(...)							\
 	p_test(__VA_ARGS__, NULL)
 
+#define	test_no_malloc_init(...)					\
+	p_test_no_malloc_init(__VA_ARGS__, NULL)
+
 #define	test_skip_if(e) do {						\
 	if (e) {							\
 		test_skip("%s:%s:%d: Test skipped: (%s)",		\
@@ -324,6 +327,7 @@ void	test_fail(const char *format, ...) JEMALLOC_FORMAT_PRINTF(1, 2);
 
 /* For private use by macros. */
 test_status_t	p_test(test_t *t, ...);
+test_status_t	p_test_no_malloc_init(test_t *t, ...);
 void	p_test_init(const char *name);
 void	p_test_fini(void);
 void	p_test_fail(const char *prefix, const char *message);
diff --git a/deps/jemalloc/test/include/test/timer.h b/deps/jemalloc/test/include/test/timer.h
index a7fefdfd1..ace6191b8 100644
--- a/deps/jemalloc/test/include/test/timer.h
+++ b/deps/jemalloc/test/include/test/timer.h
@@ -1,23 +1,8 @@
 /* Simple timer, for use in benchmark reporting. */
 
-#include <unistd.h>
-#include <sys/time.h>
-
-#define JEMALLOC_CLOCK_GETTIME defined(_POSIX_MONOTONIC_CLOCK) \
-    && _POSIX_MONOTONIC_CLOCK >= 0
-
 typedef struct {
-#ifdef _WIN32
-	FILETIME ft0;
-	FILETIME ft1;
-#elif JEMALLOC_CLOCK_GETTIME
-	struct timespec ts0;
-	struct timespec ts1;
-	int clock_id;
-#else
-	struct timeval tv0;
-	struct timeval tv1;
-#endif
+	nstime_t t0;
+	nstime_t t1;
 } timedelta_t;
 
 void	timer_start(timedelta_t *timer);
diff --git a/deps/jemalloc/test/integration/MALLOCX_ARENA.c b/deps/jemalloc/test/integration/MALLOCX_ARENA.c
old mode 100644
new mode 100755
index 30c203ae6..910a096fd
--- a/deps/jemalloc/test/integration/MALLOCX_ARENA.c
+++ b/deps/jemalloc/test/integration/MALLOCX_ARENA.c
@@ -19,8 +19,8 @@ thd_start(void *arg)
 	size_t sz;
 
 	sz = sizeof(arena_ind);
-	assert_d_eq(mallctl("arenas.extend", &arena_ind, &sz, NULL, 0), 0,
-	    "Error in arenas.extend");
+	assert_d_eq(mallctl("arenas.extend", (void *)&arena_ind, &sz, NULL, 0),
+	    0, "Error in arenas.extend");
 
 	if (thread_ind % 4 != 3) {
 		size_t mib[3];
diff --git a/deps/jemalloc/test/integration/aligned_alloc.c b/deps/jemalloc/test/integration/aligned_alloc.c
index 609001487..58438421d 100644
--- a/deps/jemalloc/test/integration/aligned_alloc.c
+++ b/deps/jemalloc/test/integration/aligned_alloc.c
@@ -1,9 +1,20 @@
 #include "test/jemalloc_test.h"
 
 #define	CHUNK 0x400000
-/* #define MAXALIGN ((size_t)UINT64_C(0x80000000000)) */
-#define	MAXALIGN ((size_t)0x2000000LU)
-#define	NITER 4
+#define	MAXALIGN (((size_t)1) << 23)
+
+/*
+ * On systems which can't merge extents, tests that call this function generate
+ * a lot of dirty memory very quickly.  Purging between cycles mitigates
+ * potential OOM on e.g. 32-bit Windows.
+ */
+static void
+purge(void)
+{
+
+	assert_d_eq(mallctl("arena.0.purge", NULL, NULL, NULL, 0), 0,
+	    "Unexpected mallctl error");
+}
 
 TEST_BEGIN(test_alignment_errors)
 {
@@ -74,6 +85,7 @@ TEST_END
 
 TEST_BEGIN(test_alignment_and_size)
 {
+#define	NITER 4
 	size_t alignment, size, total;
 	unsigned i;
 	void *ps[NITER];
@@ -110,7 +122,9 @@ TEST_BEGIN(test_alignment_and_size)
 				}
 			}
 		}
+		purge();
 	}
+#undef NITER
 }
 TEST_END
 
diff --git a/deps/jemalloc/test/integration/allocated.c b/deps/jemalloc/test/integration/allocated.c
old mode 100644
new mode 100755
index 3630e80ce..6ce145b3e
--- a/deps/jemalloc/test/integration/allocated.c
+++ b/deps/jemalloc/test/integration/allocated.c
@@ -18,14 +18,14 @@ thd_start(void *arg)
 	size_t sz, usize;
 
 	sz = sizeof(a0);
-	if ((err = mallctl("thread.allocated", &a0, &sz, NULL, 0))) {
+	if ((err = mallctl("thread.allocated", (void *)&a0, &sz, NULL, 0))) {
 		if (err == ENOENT)
 			goto label_ENOENT;
 		test_fail("%s(): Error in mallctl(): %s", __func__,
 		    strerror(err));
 	}
 	sz = sizeof(ap0);
-	if ((err = mallctl("thread.allocatedp", &ap0, &sz, NULL, 0))) {
+	if ((err = mallctl("thread.allocatedp", (void *)&ap0, &sz, NULL, 0))) {
 		if (err == ENOENT)
 			goto label_ENOENT;
 		test_fail("%s(): Error in mallctl(): %s", __func__,
@@ -36,14 +36,15 @@ thd_start(void *arg)
 	    "storage");
 
 	sz = sizeof(d0);
-	if ((err = mallctl("thread.deallocated", &d0, &sz, NULL, 0))) {
+	if ((err = mallctl("thread.deallocated", (void *)&d0, &sz, NULL, 0))) {
 		if (err == ENOENT)
 			goto label_ENOENT;
 		test_fail("%s(): Error in mallctl(): %s", __func__,
 		    strerror(err));
 	}
 	sz = sizeof(dp0);
-	if ((err = mallctl("thread.deallocatedp", &dp0, &sz, NULL, 0))) {
+	if ((err = mallctl("thread.deallocatedp", (void *)&dp0, &sz, NULL,
+	    0))) {
 		if (err == ENOENT)
 			goto label_ENOENT;
 		test_fail("%s(): Error in mallctl(): %s", __func__,
@@ -57,9 +58,9 @@ thd_start(void *arg)
 	assert_ptr_not_null(p, "Unexpected malloc() error");
 
 	sz = sizeof(a1);
-	mallctl("thread.allocated", &a1, &sz, NULL, 0);
+	mallctl("thread.allocated", (void *)&a1, &sz, NULL, 0);
 	sz = sizeof(ap1);
-	mallctl("thread.allocatedp", &ap1, &sz, NULL, 0);
+	mallctl("thread.allocatedp", (void *)&ap1, &sz, NULL, 0);
 	assert_u64_eq(*ap1, a1,
 	    "Dereferenced \"thread.allocatedp\" value should equal "
 	    "\"thread.allocated\" value");
@@ -74,9 +75,9 @@ thd_start(void *arg)
 	free(p);
 
 	sz = sizeof(d1);
-	mallctl("thread.deallocated", &d1, &sz, NULL, 0);
+	mallctl("thread.deallocated", (void *)&d1, &sz, NULL, 0);
 	sz = sizeof(dp1);
-	mallctl("thread.deallocatedp", &dp1, &sz, NULL, 0);
+	mallctl("thread.deallocatedp", (void *)&dp1, &sz, NULL, 0);
 	assert_u64_eq(*dp1, d1,
 	    "Dereferenced \"thread.deallocatedp\" value should equal "
 	    "\"thread.deallocated\" value");
diff --git a/deps/jemalloc/test/integration/chunk.c b/deps/jemalloc/test/integration/chunk.c
index af1c9a53e..94cf0025a 100644
--- a/deps/jemalloc/test/integration/chunk.c
+++ b/deps/jemalloc/test/integration/chunk.c
@@ -121,6 +121,10 @@ TEST_BEGIN(test_chunk)
 {
 	void *p;
 	size_t old_size, new_size, large0, large1, huge0, huge1, huge2, sz;
+	unsigned arena_ind;
+	int flags;
+	size_t hooks_mib[3], purge_mib[3];
+	size_t hooks_miblen, purge_miblen;
 	chunk_hooks_t new_hooks = {
 		chunk_alloc,
 		chunk_dalloc,
@@ -132,11 +136,21 @@ TEST_BEGIN(test_chunk)
 	};
 	bool xallocx_success_a, xallocx_success_b, xallocx_success_c;
 
+	sz = sizeof(unsigned);
+	assert_d_eq(mallctl("arenas.extend", (void *)&arena_ind, &sz, NULL, 0),
+	    0, "Unexpected mallctl() failure");
+	flags = MALLOCX_ARENA(arena_ind) | MALLOCX_TCACHE_NONE;
+
 	/* Install custom chunk hooks. */
+	hooks_miblen = sizeof(hooks_mib)/sizeof(size_t);
+	assert_d_eq(mallctlnametomib("arena.0.chunk_hooks", hooks_mib,
+	    &hooks_miblen), 0, "Unexpected mallctlnametomib() failure");
+	hooks_mib[1] = (size_t)arena_ind;
 	old_size = sizeof(chunk_hooks_t);
 	new_size = sizeof(chunk_hooks_t);
-	assert_d_eq(mallctl("arena.0.chunk_hooks", &old_hooks, &old_size,
-	    &new_hooks, new_size), 0, "Unexpected chunk_hooks error");
+	assert_d_eq(mallctlbymib(hooks_mib, hooks_miblen, (void *)&old_hooks,
+	    &old_size, (void *)&new_hooks, new_size), 0,
+	    "Unexpected chunk_hooks error");
 	orig_hooks = old_hooks;
 	assert_ptr_ne(old_hooks.alloc, chunk_alloc, "Unexpected alloc error");
 	assert_ptr_ne(old_hooks.dalloc, chunk_dalloc,
@@ -151,59 +165,63 @@ TEST_BEGIN(test_chunk)
 
 	/* Get large size classes. */
 	sz = sizeof(size_t);
-	assert_d_eq(mallctl("arenas.lrun.0.size", &large0, &sz, NULL, 0), 0,
-	    "Unexpected arenas.lrun.0.size failure");
-	assert_d_eq(mallctl("arenas.lrun.1.size", &large1, &sz, NULL, 0), 0,
-	    "Unexpected arenas.lrun.1.size failure");
+	assert_d_eq(mallctl("arenas.lrun.0.size", (void *)&large0, &sz, NULL,
+	    0), 0, "Unexpected arenas.lrun.0.size failure");
+	assert_d_eq(mallctl("arenas.lrun.1.size", (void *)&large1, &sz, NULL,
+	    0), 0, "Unexpected arenas.lrun.1.size failure");
 
 	/* Get huge size classes. */
-	assert_d_eq(mallctl("arenas.hchunk.0.size", &huge0, &sz, NULL, 0), 0,
-	    "Unexpected arenas.hchunk.0.size failure");
-	assert_d_eq(mallctl("arenas.hchunk.1.size", &huge1, &sz, NULL, 0), 0,
-	    "Unexpected arenas.hchunk.1.size failure");
-	assert_d_eq(mallctl("arenas.hchunk.2.size", &huge2, &sz, NULL, 0), 0,
-	    "Unexpected arenas.hchunk.2.size failure");
+	assert_d_eq(mallctl("arenas.hchunk.0.size", (void *)&huge0, &sz, NULL,
+	    0), 0, "Unexpected arenas.hchunk.0.size failure");
+	assert_d_eq(mallctl("arenas.hchunk.1.size", (void *)&huge1, &sz, NULL,
+	    0), 0, "Unexpected arenas.hchunk.1.size failure");
+	assert_d_eq(mallctl("arenas.hchunk.2.size", (void *)&huge2, &sz, NULL,
+	    0), 0, "Unexpected arenas.hchunk.2.size failure");
 
 	/* Test dalloc/decommit/purge cascade. */
+	purge_miblen = sizeof(purge_mib)/sizeof(size_t);
+	assert_d_eq(mallctlnametomib("arena.0.purge", purge_mib, &purge_miblen),
+	    0, "Unexpected mallctlnametomib() failure");
+	purge_mib[1] = (size_t)arena_ind;
 	do_dalloc = false;
 	do_decommit = false;
-	p = mallocx(huge0 * 2, 0);
+	p = mallocx(huge0 * 2, flags);
 	assert_ptr_not_null(p, "Unexpected mallocx() error");
 	did_dalloc = false;
 	did_decommit = false;
 	did_purge = false;
 	did_split = false;
-	xallocx_success_a = (xallocx(p, huge0, 0, 0) == huge0);
-	assert_d_eq(mallctl("arena.0.purge", NULL, NULL, NULL, 0), 0,
-	    "Unexpected arena.0.purge error");
+	xallocx_success_a = (xallocx(p, huge0, 0, flags) == huge0);
+	assert_d_eq(mallctlbymib(purge_mib, purge_miblen, NULL, NULL, NULL, 0),
+	    0, "Unexpected arena.%u.purge error", arena_ind);
 	if (xallocx_success_a) {
 		assert_true(did_dalloc, "Expected dalloc");
 		assert_false(did_decommit, "Unexpected decommit");
 		assert_true(did_purge, "Expected purge");
 	}
 	assert_true(did_split, "Expected split");
-	dallocx(p, 0);
+	dallocx(p, flags);
 	do_dalloc = true;
 
 	/* Test decommit/commit and observe split/merge. */
 	do_dalloc = false;
 	do_decommit = true;
-	p = mallocx(huge0 * 2, 0);
+	p = mallocx(huge0 * 2, flags);
 	assert_ptr_not_null(p, "Unexpected mallocx() error");
 	did_decommit = false;
 	did_commit = false;
 	did_split = false;
 	did_merge = false;
-	xallocx_success_b = (xallocx(p, huge0, 0, 0) == huge0);
-	assert_d_eq(mallctl("arena.0.purge", NULL, NULL, NULL, 0), 0,
-	    "Unexpected arena.0.purge error");
+	xallocx_success_b = (xallocx(p, huge0, 0, flags) == huge0);
+	assert_d_eq(mallctlbymib(purge_mib, purge_miblen, NULL, NULL, NULL, 0),
+	    0, "Unexpected arena.%u.purge error", arena_ind);
 	if (xallocx_success_b)
 		assert_true(did_split, "Expected split");
-	xallocx_success_c = (xallocx(p, huge0 * 2, 0, 0) == huge0 * 2);
+	xallocx_success_c = (xallocx(p, huge0 * 2, 0, flags) == huge0 * 2);
 	assert_b_eq(did_decommit, did_commit, "Expected decommit/commit match");
 	if (xallocx_success_b && xallocx_success_c)
 		assert_true(did_merge, "Expected merge");
-	dallocx(p, 0);
+	dallocx(p, flags);
 	do_dalloc = true;
 	do_decommit = false;
 
@@ -214,43 +232,43 @@ TEST_BEGIN(test_chunk)
 		 * successful xallocx() from size=huge2 to size=huge1 is
 		 * guaranteed to leave trailing purgeable memory.
 		 */
-		p = mallocx(huge2, 0);
+		p = mallocx(huge2, flags);
 		assert_ptr_not_null(p, "Unexpected mallocx() error");
 		did_purge = false;
-		assert_zu_eq(xallocx(p, huge1, 0, 0), huge1,
+		assert_zu_eq(xallocx(p, huge1, 0, flags), huge1,
 		    "Unexpected xallocx() failure");
 		assert_true(did_purge, "Expected purge");
-		dallocx(p, 0);
+		dallocx(p, flags);
 	}
 
 	/* Test decommit for large allocations. */
 	do_decommit = true;
-	p = mallocx(large1, 0);
+	p = mallocx(large1, flags);
 	assert_ptr_not_null(p, "Unexpected mallocx() error");
-	assert_d_eq(mallctl("arena.0.purge", NULL, NULL, NULL, 0), 0,
-	    "Unexpected arena.0.purge error");
+	assert_d_eq(mallctlbymib(purge_mib, purge_miblen, NULL, NULL, NULL, 0),
+	    0, "Unexpected arena.%u.purge error", arena_ind);
 	did_decommit = false;
-	assert_zu_eq(xallocx(p, large0, 0, 0), large0,
+	assert_zu_eq(xallocx(p, large0, 0, flags), large0,
 	    "Unexpected xallocx() failure");
-	assert_d_eq(mallctl("arena.0.purge", NULL, NULL, NULL, 0), 0,
-	    "Unexpected arena.0.purge error");
+	assert_d_eq(mallctlbymib(purge_mib, purge_miblen, NULL, NULL, NULL, 0),
+	    0, "Unexpected arena.%u.purge error", arena_ind);
 	did_commit = false;
-	assert_zu_eq(xallocx(p, large1, 0, 0), large1,
+	assert_zu_eq(xallocx(p, large1, 0, flags), large1,
 	    "Unexpected xallocx() failure");
 	assert_b_eq(did_decommit, did_commit, "Expected decommit/commit match");
-	dallocx(p, 0);
+	dallocx(p, flags);
 	do_decommit = false;
 
 	/* Make sure non-huge allocation succeeds. */
-	p = mallocx(42, 0);
+	p = mallocx(42, flags);
 	assert_ptr_not_null(p, "Unexpected mallocx() error");
-	dallocx(p, 0);
+	dallocx(p, flags);
 
 	/* Restore chunk hooks. */
-	assert_d_eq(mallctl("arena.0.chunk_hooks", NULL, NULL, &old_hooks,
-	    new_size), 0, "Unexpected chunk_hooks error");
-	assert_d_eq(mallctl("arena.0.chunk_hooks", &old_hooks, &old_size,
-	    NULL, 0), 0, "Unexpected chunk_hooks error");
+	assert_d_eq(mallctlbymib(hooks_mib, hooks_miblen, NULL, NULL,
+	    (void *)&old_hooks, new_size), 0, "Unexpected chunk_hooks error");
+	assert_d_eq(mallctlbymib(hooks_mib, hooks_miblen, (void *)&old_hooks,
+	    &old_size, NULL, 0), 0, "Unexpected chunk_hooks error");
 	assert_ptr_eq(old_hooks.alloc, orig_hooks.alloc,
 	    "Unexpected alloc error");
 	assert_ptr_eq(old_hooks.dalloc, orig_hooks.dalloc,
diff --git a/deps/jemalloc/test/integration/mallocx.c b/deps/jemalloc/test/integration/mallocx.c
old mode 100644
new mode 100755
index 6253175d6..d709eb301
--- a/deps/jemalloc/test/integration/mallocx.c
+++ b/deps/jemalloc/test/integration/mallocx.c
@@ -1,5 +1,9 @@
 #include "test/jemalloc_test.h"
 
+#ifdef JEMALLOC_FILL
+const char *malloc_conf = "junk:false";
+#endif
+
 static unsigned
 get_nsizes_impl(const char *cmd)
 {
@@ -7,7 +11,7 @@ get_nsizes_impl(const char *cmd)
 	size_t z;
 
 	z = sizeof(unsigned);
-	assert_d_eq(mallctl(cmd, &ret, &z, NULL, 0), 0,
+	assert_d_eq(mallctl(cmd, (void *)&ret, &z, NULL, 0), 0,
 	    "Unexpected mallctl(\"%s\", ...) failure", cmd);
 
 	return (ret);
@@ -33,7 +37,7 @@ get_size_impl(const char *cmd, size_t ind)
 	    0, "Unexpected mallctlnametomib(\"%s\", ...) failure", cmd);
 	mib[2] = ind;
 	z = sizeof(size_t);
-	assert_d_eq(mallctlbymib(mib, miblen, &ret, &z, NULL, 0),
+	assert_d_eq(mallctlbymib(mib, miblen, (void *)&ret, &z, NULL, 0),
 	    0, "Unexpected mallctlbymib([\"%s\", %zu], ...) failure", cmd, ind);
 
 	return (ret);
@@ -46,43 +50,84 @@ get_huge_size(size_t ind)
 	return (get_size_impl("arenas.hchunk.0.size", ind));
 }
 
-TEST_BEGIN(test_oom)
+/*
+ * On systems which can't merge extents, tests that call this function generate
+ * a lot of dirty memory very quickly.  Purging between cycles mitigates
+ * potential OOM on e.g. 32-bit Windows.
+ */
+static void
+purge(void)
 {
-	size_t hugemax, size, alignment;
+
+	assert_d_eq(mallctl("arena.0.purge", NULL, NULL, NULL, 0), 0,
+	    "Unexpected mallctl error");
+}
+
+TEST_BEGIN(test_overflow)
+{
+	size_t hugemax;
 
 	hugemax = get_huge_size(get_nhuge()-1);
 
+	assert_ptr_null(mallocx(hugemax+1, 0),
+	    "Expected OOM for mallocx(size=%#zx, 0)", hugemax+1);
+
+	assert_ptr_null(mallocx(ZU(PTRDIFF_MAX)+1, 0),
+	    "Expected OOM for mallocx(size=%#zx, 0)", ZU(PTRDIFF_MAX)+1);
+
+	assert_ptr_null(mallocx(SIZE_T_MAX, 0),
+	    "Expected OOM for mallocx(size=%#zx, 0)", SIZE_T_MAX);
+
+	assert_ptr_null(mallocx(1, MALLOCX_ALIGN(ZU(PTRDIFF_MAX)+1)),
+	    "Expected OOM for mallocx(size=1, MALLOCX_ALIGN(%#zx))",
+	    ZU(PTRDIFF_MAX)+1);
+}
+TEST_END
+
+TEST_BEGIN(test_oom)
+{
+	size_t hugemax;
+	bool oom;
+	void *ptrs[3];
+	unsigned i;
+
 	/*
-	 * It should be impossible to allocate two objects that each consume
-	 * more than half the virtual address space.
+	 * It should be impossible to allocate three objects that each consume
+	 * nearly half the virtual address space.
 	 */
-	{
-		void *p;
-
-		p = mallocx(hugemax, 0);
-		if (p != NULL) {
-			assert_ptr_null(mallocx(hugemax, 0),
-			    "Expected OOM for mallocx(size=%#zx, 0)", hugemax);
-			dallocx(p, 0);
-		}
+	hugemax = get_huge_size(get_nhuge()-1);
+	oom = false;
+	for (i = 0; i < sizeof(ptrs) / sizeof(void *); i++) {
+		ptrs[i] = mallocx(hugemax, 0);
+		if (ptrs[i] == NULL)
+			oom = true;
+	}
+	assert_true(oom,
+	    "Expected OOM during series of calls to mallocx(size=%zu, 0)",
+	    hugemax);
+	for (i = 0; i < sizeof(ptrs) / sizeof(void *); i++) {
+		if (ptrs[i] != NULL)
+			dallocx(ptrs[i], 0);
 	}
+	purge();
 
 #if LG_SIZEOF_PTR == 3
-	size      = ZU(0x8000000000000000);
-	alignment = ZU(0x8000000000000000);
+	assert_ptr_null(mallocx(0x8000000000000000ULL,
+	    MALLOCX_ALIGN(0x8000000000000000ULL)),
+	    "Expected OOM for mallocx()");
+	assert_ptr_null(mallocx(0x8000000000000000ULL,
+	    MALLOCX_ALIGN(0x80000000)),
+	    "Expected OOM for mallocx()");
 #else
-	size      = ZU(0x80000000);
-	alignment = ZU(0x80000000);
+	assert_ptr_null(mallocx(0x80000000UL, MALLOCX_ALIGN(0x80000000UL)),
+	    "Expected OOM for mallocx()");
 #endif
-	assert_ptr_null(mallocx(size, MALLOCX_ALIGN(alignment)),
-	    "Expected OOM for mallocx(size=%#zx, MALLOCX_ALIGN(%#zx)", size,
-	    alignment);
 }
 TEST_END
 
 TEST_BEGIN(test_basic)
 {
-#define	MAXSZ (((size_t)1) << 26)
+#define	MAXSZ (((size_t)1) << 23)
 	size_t sz;
 
 	for (sz = 1; sz < MAXSZ; sz = nallocx(sz, 0) + 1) {
@@ -91,23 +136,28 @@ TEST_BEGIN(test_basic)
 		nsz = nallocx(sz, 0);
 		assert_zu_ne(nsz, 0, "Unexpected nallocx() error");
 		p = mallocx(sz, 0);
-		assert_ptr_not_null(p, "Unexpected mallocx() error");
+		assert_ptr_not_null(p,
+		    "Unexpected mallocx(size=%zx, flags=0) error", sz);
 		rsz = sallocx(p, 0);
 		assert_zu_ge(rsz, sz, "Real size smaller than expected");
 		assert_zu_eq(nsz, rsz, "nallocx()/sallocx() size mismatch");
 		dallocx(p, 0);
 
 		p = mallocx(sz, 0);
-		assert_ptr_not_null(p, "Unexpected mallocx() error");
+		assert_ptr_not_null(p,
+		    "Unexpected mallocx(size=%zx, flags=0) error", sz);
 		dallocx(p, 0);
 
 		nsz = nallocx(sz, MALLOCX_ZERO);
 		assert_zu_ne(nsz, 0, "Unexpected nallocx() error");
 		p = mallocx(sz, MALLOCX_ZERO);
-		assert_ptr_not_null(p, "Unexpected mallocx() error");
+		assert_ptr_not_null(p,
+		    "Unexpected mallocx(size=%zx, flags=MALLOCX_ZERO) error",
+		    nsz);
 		rsz = sallocx(p, 0);
 		assert_zu_eq(nsz, rsz, "nallocx()/sallocx() rsize mismatch");
 		dallocx(p, 0);
+		purge();
 	}
 #undef MAXSZ
 }
@@ -115,7 +165,7 @@ TEST_END
 
 TEST_BEGIN(test_alignment_and_size)
 {
-#define	MAXALIGN (((size_t)1) << 25)
+#define	MAXALIGN (((size_t)1) << 23)
 #define	NITER 4
 	size_t nsz, rsz, sz, alignment, total;
 	unsigned i;
@@ -165,6 +215,7 @@ TEST_BEGIN(test_alignment_and_size)
 				}
 			}
 		}
+		purge();
 	}
 #undef MAXALIGN
 #undef NITER
@@ -176,6 +227,7 @@ main(void)
 {
 
 	return (test(
+	    test_overflow,
 	    test_oom,
 	    test_basic,
 	    test_alignment_and_size));
diff --git a/deps/jemalloc/test/integration/overflow.c b/deps/jemalloc/test/integration/overflow.c
old mode 100644
new mode 100755
index 303d9b2d3..84a35652c
--- a/deps/jemalloc/test/integration/overflow.c
+++ b/deps/jemalloc/test/integration/overflow.c
@@ -8,8 +8,8 @@ TEST_BEGIN(test_overflow)
 	void *p;
 
 	sz = sizeof(unsigned);
-	assert_d_eq(mallctl("arenas.nhchunks", &nhchunks, &sz, NULL, 0), 0,
-	    "Unexpected mallctl() error");
+	assert_d_eq(mallctl("arenas.nhchunks", (void *)&nhchunks, &sz, NULL, 0),
+	    0, "Unexpected mallctl() error");
 
 	miblen = sizeof(mib) / sizeof(size_t);
 	assert_d_eq(mallctlnametomib("arenas.hchunk.0.size", mib, &miblen), 0,
@@ -17,8 +17,8 @@ TEST_BEGIN(test_overflow)
 	mib[2] = nhchunks - 1;
 
 	sz = sizeof(size_t);
-	assert_d_eq(mallctlbymib(mib, miblen, &max_size_class, &sz, NULL, 0), 0,
-	    "Unexpected mallctlbymib() error");
+	assert_d_eq(mallctlbymib(mib, miblen, (void *)&max_size_class, &sz,
+	    NULL, 0), 0, "Unexpected mallctlbymib() error");
 
 	assert_ptr_null(malloc(max_size_class + 1),
 	    "Expected OOM due to over-sized allocation request");
diff --git a/deps/jemalloc/test/integration/posix_memalign.c b/deps/jemalloc/test/integration/posix_memalign.c
index 19741c6cb..e22e10200 100644
--- a/deps/jemalloc/test/integration/posix_memalign.c
+++ b/deps/jemalloc/test/integration/posix_memalign.c
@@ -1,9 +1,20 @@
 #include "test/jemalloc_test.h"
 
 #define	CHUNK 0x400000
-/* #define MAXALIGN ((size_t)UINT64_C(0x80000000000)) */
-#define	MAXALIGN ((size_t)0x2000000LU)
-#define	NITER 4
+#define	MAXALIGN (((size_t)1) << 23)
+
+/*
+ * On systems which can't merge extents, tests that call this function generate
+ * a lot of dirty memory very quickly.  Purging between cycles mitigates
+ * potential OOM on e.g. 32-bit Windows.
+ */
+static void
+purge(void)
+{
+
+	assert_d_eq(mallctl("arena.0.purge", NULL, NULL, NULL, 0), 0,
+	    "Unexpected mallctl error");
+}
 
 TEST_BEGIN(test_alignment_errors)
 {
@@ -66,6 +77,7 @@ TEST_END
 
 TEST_BEGIN(test_alignment_and_size)
 {
+#define	NITER 4
 	size_t alignment, size, total;
 	unsigned i;
 	int err;
@@ -104,7 +116,9 @@ TEST_BEGIN(test_alignment_and_size)
 				}
 			}
 		}
+		purge();
 	}
+#undef NITER
 }
 TEST_END
 
diff --git a/deps/jemalloc/test/integration/rallocx.c b/deps/jemalloc/test/integration/rallocx.c
old mode 100644
new mode 100755
index be1b27b73..506bf1c90
--- a/deps/jemalloc/test/integration/rallocx.c
+++ b/deps/jemalloc/test/integration/rallocx.c
@@ -1,5 +1,51 @@
 #include "test/jemalloc_test.h"
 
+static unsigned
+get_nsizes_impl(const char *cmd)
+{
+	unsigned ret;
+	size_t z;
+
+	z = sizeof(unsigned);
+	assert_d_eq(mallctl(cmd, (void *)&ret, &z, NULL, 0), 0,
+	    "Unexpected mallctl(\"%s\", ...) failure", cmd);
+
+	return (ret);
+}
+
+static unsigned
+get_nhuge(void)
+{
+
+	return (get_nsizes_impl("arenas.nhchunks"));
+}
+
+static size_t
+get_size_impl(const char *cmd, size_t ind)
+{
+	size_t ret;
+	size_t z;
+	size_t mib[4];
+	size_t miblen = 4;
+
+	z = sizeof(size_t);
+	assert_d_eq(mallctlnametomib(cmd, mib, &miblen),
+	    0, "Unexpected mallctlnametomib(\"%s\", ...) failure", cmd);
+	mib[2] = ind;
+	z = sizeof(size_t);
+	assert_d_eq(mallctlbymib(mib, miblen, (void *)&ret, &z, NULL, 0),
+	    0, "Unexpected mallctlbymib([\"%s\", %zu], ...) failure", cmd, ind);
+
+	return (ret);
+}
+
+static size_t
+get_huge_size(size_t ind)
+{
+
+	return (get_size_impl("arenas.hchunk.0.size", ind));
+}
+
 TEST_BEGIN(test_grow_and_shrink)
 {
 	void *p, *q;
@@ -138,22 +184,22 @@ TEST_END
 TEST_BEGIN(test_lg_align_and_zero)
 {
 	void *p, *q;
-	size_t lg_align, sz;
+	unsigned lg_align;
+	size_t sz;
 #define	MAX_LG_ALIGN 25
 #define	MAX_VALIDATE (ZU(1) << 22)
 
-	lg_align = ZU(0);
+	lg_align = 0;
 	p = mallocx(1, MALLOCX_LG_ALIGN(lg_align)|MALLOCX_ZERO);
 	assert_ptr_not_null(p, "Unexpected mallocx() error");
 
 	for (lg_align++; lg_align <= MAX_LG_ALIGN; lg_align++) {
 		q = rallocx(p, 1, MALLOCX_LG_ALIGN(lg_align)|MALLOCX_ZERO);
 		assert_ptr_not_null(q,
-		    "Unexpected rallocx() error for lg_align=%zu", lg_align);
+		    "Unexpected rallocx() error for lg_align=%u", lg_align);
 		assert_ptr_null(
 		    (void *)((uintptr_t)q & ((ZU(1) << lg_align)-1)),
-		    "%p inadequately aligned for lg_align=%zu",
-		    q, lg_align);
+		    "%p inadequately aligned for lg_align=%u", q, lg_align);
 		sz = sallocx(q, 0);
 		if ((sz << 1) <= MAX_VALIDATE) {
 			assert_false(validate_fill(q, 0, 0, sz),
@@ -173,6 +219,33 @@ TEST_BEGIN(test_lg_align_and_zero)
 }
 TEST_END
 
+TEST_BEGIN(test_overflow)
+{
+	size_t hugemax;
+	void *p;
+
+	hugemax = get_huge_size(get_nhuge()-1);
+
+	p = mallocx(1, 0);
+	assert_ptr_not_null(p, "Unexpected mallocx() failure");
+
+	assert_ptr_null(rallocx(p, hugemax+1, 0),
+	    "Expected OOM for rallocx(p, size=%#zx, 0)", hugemax+1);
+
+	assert_ptr_null(rallocx(p, ZU(PTRDIFF_MAX)+1, 0),
+	    "Expected OOM for rallocx(p, size=%#zx, 0)", ZU(PTRDIFF_MAX)+1);
+
+	assert_ptr_null(rallocx(p, SIZE_T_MAX, 0),
+	    "Expected OOM for rallocx(p, size=%#zx, 0)", SIZE_T_MAX);
+
+	assert_ptr_null(rallocx(p, 1, MALLOCX_ALIGN(ZU(PTRDIFF_MAX)+1)),
+	    "Expected OOM for rallocx(p, size=1, MALLOCX_ALIGN(%#zx))",
+	    ZU(PTRDIFF_MAX)+1);
+
+	dallocx(p, 0);
+}
+TEST_END
+
 int
 main(void)
 {
@@ -181,5 +254,6 @@ main(void)
 	    test_grow_and_shrink,
 	    test_zero,
 	    test_align,
-	    test_lg_align_and_zero));
+	    test_lg_align_and_zero,
+	    test_overflow));
 }
diff --git a/deps/jemalloc/test/integration/sdallocx.c b/deps/jemalloc/test/integration/sdallocx.c
index b84817d76..f92e0589c 100644
--- a/deps/jemalloc/test/integration/sdallocx.c
+++ b/deps/jemalloc/test/integration/sdallocx.c
@@ -1,7 +1,7 @@
 #include "test/jemalloc_test.h"
 
-#define	MAXALIGN (((size_t)1) << 25)
-#define	NITER 4
+#define	MAXALIGN (((size_t)1) << 22)
+#define	NITER 3
 
 TEST_BEGIN(test_basic)
 {
diff --git a/deps/jemalloc/test/integration/thread_arena.c b/deps/jemalloc/test/integration/thread_arena.c
old mode 100644
new mode 100755
index 67be53513..7a35a6351
--- a/deps/jemalloc/test/integration/thread_arena.c
+++ b/deps/jemalloc/test/integration/thread_arena.c
@@ -16,8 +16,8 @@ thd_start(void *arg)
 	free(p);
 
 	size = sizeof(arena_ind);
-	if ((err = mallctl("thread.arena", &arena_ind, &size, &main_arena_ind,
-	    sizeof(main_arena_ind)))) {
+	if ((err = mallctl("thread.arena", (void *)&arena_ind, &size,
+	    (void *)&main_arena_ind, sizeof(main_arena_ind)))) {
 		char buf[BUFERROR_BUF];
 
 		buferror(err, buf, sizeof(buf));
@@ -25,7 +25,8 @@ thd_start(void *arg)
 	}
 
 	size = sizeof(arena_ind);
-	if ((err = mallctl("thread.arena", &arena_ind, &size, NULL, 0))) {
+	if ((err = mallctl("thread.arena", (void *)&arena_ind, &size, NULL,
+	    0))) {
 		char buf[BUFERROR_BUF];
 
 		buferror(err, buf, sizeof(buf));
@@ -50,7 +51,8 @@ TEST_BEGIN(test_thread_arena)
 	assert_ptr_not_null(p, "Error in malloc()");
 
 	size = sizeof(arena_ind);
-	if ((err = mallctl("thread.arena", &arena_ind, &size, NULL, 0))) {
+	if ((err = mallctl("thread.arena", (void *)&arena_ind, &size, NULL,
+	    0))) {
 		char buf[BUFERROR_BUF];
 
 		buferror(err, buf, sizeof(buf));
diff --git a/deps/jemalloc/test/integration/thread_tcache_enabled.c b/deps/jemalloc/test/integration/thread_tcache_enabled.c
old mode 100644
new mode 100755
index f4e89c682..2c2825e19
--- a/deps/jemalloc/test/integration/thread_tcache_enabled.c
+++ b/deps/jemalloc/test/integration/thread_tcache_enabled.c
@@ -16,7 +16,8 @@ thd_start(void *arg)
 	bool e0, e1;
 
 	sz = sizeof(bool);
-	if ((err = mallctl("thread.tcache.enabled", &e0, &sz, NULL, 0))) {
+	if ((err = mallctl("thread.tcache.enabled", (void *)&e0, &sz, NULL,
+	    0))) {
 		if (err == ENOENT) {
 			assert_false(config_tcache,
 			    "ENOENT should only be returned if tcache is "
@@ -27,53 +28,53 @@ thd_start(void *arg)
 
 	if (e0) {
 		e1 = false;
-		assert_d_eq(mallctl("thread.tcache.enabled", &e0, &sz, &e1, sz),
-		    0, "Unexpected mallctl() error");
+		assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz,
+		    (void *)&e1, sz), 0, "Unexpected mallctl() error");
 		assert_true(e0, "tcache should be enabled");
 	}
 
 	e1 = true;
-	assert_d_eq(mallctl("thread.tcache.enabled", &e0, &sz, &e1, sz), 0,
-	    "Unexpected mallctl() error");
+	assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz,
+	    (void *)&e1, sz), 0, "Unexpected mallctl() error");
 	assert_false(e0, "tcache should be disabled");
 
 	e1 = true;
-	assert_d_eq(mallctl("thread.tcache.enabled", &e0, &sz, &e1, sz), 0,
-	    "Unexpected mallctl() error");
+	assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz,
+	    (void *)&e1, sz), 0, "Unexpected mallctl() error");
 	assert_true(e0, "tcache should be enabled");
 
 	e1 = false;
-	assert_d_eq(mallctl("thread.tcache.enabled", &e0, &sz, &e1, sz), 0,
-	    "Unexpected mallctl() error");
+	assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz,
+	    (void *)&e1, sz), 0, "Unexpected mallctl() error");
 	assert_true(e0, "tcache should be enabled");
 
 	e1 = false;
-	assert_d_eq(mallctl("thread.tcache.enabled", &e0, &sz, &e1, sz), 0,
-	    "Unexpected mallctl() error");
+	assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz,
+	    (void *)&e1, sz), 0, "Unexpected mallctl() error");
 	assert_false(e0, "tcache should be disabled");
 
 	free(malloc(1));
 	e1 = true;
-	assert_d_eq(mallctl("thread.tcache.enabled", &e0, &sz, &e1, sz), 0,
-	    "Unexpected mallctl() error");
+	assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz,
+	    (void *)&e1, sz), 0, "Unexpected mallctl() error");
 	assert_false(e0, "tcache should be disabled");
 
 	free(malloc(1));
 	e1 = true;
-	assert_d_eq(mallctl("thread.tcache.enabled", &e0, &sz, &e1, sz), 0,
-	    "Unexpected mallctl() error");
+	assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz,
+	    (void *)&e1, sz), 0, "Unexpected mallctl() error");
 	assert_true(e0, "tcache should be enabled");
 
 	free(malloc(1));
 	e1 = false;
-	assert_d_eq(mallctl("thread.tcache.enabled", &e0, &sz, &e1, sz), 0,
-	    "Unexpected mallctl() error");
+	assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz,
+	    (void *)&e1, sz), 0, "Unexpected mallctl() error");
 	assert_true(e0, "tcache should be enabled");
 
 	free(malloc(1));
 	e1 = false;
-	assert_d_eq(mallctl("thread.tcache.enabled", &e0, &sz, &e1, sz), 0,
-	    "Unexpected mallctl() error");
+	assert_d_eq(mallctl("thread.tcache.enabled", (void *)&e0, &sz,
+	    (void *)&e1, sz), 0, "Unexpected mallctl() error");
 	assert_false(e0, "tcache should be disabled");
 
 	free(malloc(1));
diff --git a/deps/jemalloc/test/integration/xallocx.c b/deps/jemalloc/test/integration/xallocx.c
old mode 100644
new mode 100755
index 373625219..67e0a0e71
--- a/deps/jemalloc/test/integration/xallocx.c
+++ b/deps/jemalloc/test/integration/xallocx.c
@@ -1,5 +1,28 @@
 #include "test/jemalloc_test.h"
 
+#ifdef JEMALLOC_FILL
+const char *malloc_conf = "junk:false";
+#endif
+
+/*
+ * Use a separate arena for xallocx() extension/contraction tests so that
+ * internal allocation e.g. by heap profiling can't interpose allocations where
+ * xallocx() would ordinarily be able to extend.
+ */
+static unsigned
+arena_ind(void)
+{
+	static unsigned ind = 0;
+
+	if (ind == 0) {
+		size_t sz = sizeof(ind);
+		assert_d_eq(mallctl("arenas.extend", (void *)&ind, &sz, NULL,
+		    0), 0, "Unexpected mallctl failure creating arena");
+	}
+
+	return (ind);
+}
+
 TEST_BEGIN(test_same_size)
 {
 	void *p;
@@ -55,7 +78,7 @@ get_nsizes_impl(const char *cmd)
 	size_t z;
 
 	z = sizeof(unsigned);
-	assert_d_eq(mallctl(cmd, &ret, &z, NULL, 0), 0,
+	assert_d_eq(mallctl(cmd, (void *)&ret, &z, NULL, 0), 0,
 	    "Unexpected mallctl(\"%s\", ...) failure", cmd);
 
 	return (ret);
@@ -95,7 +118,7 @@ get_size_impl(const char *cmd, size_t ind)
 	    0, "Unexpected mallctlnametomib(\"%s\", ...) failure", cmd);
 	mib[2] = ind;
 	z = sizeof(size_t);
-	assert_d_eq(mallctlbymib(mib, miblen, &ret, &z, NULL, 0),
+	assert_d_eq(mallctlbymib(mib, miblen, (void *)&ret, &z, NULL, 0),
 	    0, "Unexpected mallctlbymib([\"%s\", %zu], ...) failure", cmd, ind);
 
 	return (ret);
@@ -218,6 +241,7 @@ TEST_END
 
 TEST_BEGIN(test_extra_large)
 {
+	int flags = MALLOCX_ARENA(arena_ind());
 	size_t smallmax, large0, large1, large2, huge0, hugemax;
 	void *p;
 
@@ -229,121 +253,122 @@ TEST_BEGIN(test_extra_large)
 	huge0 = get_huge_size(0);
 	hugemax = get_huge_size(get_nhuge()-1);
 
-	p = mallocx(large2, 0);
+	p = mallocx(large2, flags);
 	assert_ptr_not_null(p, "Unexpected mallocx() error");
 
-	assert_zu_eq(xallocx(p, large2, 0, 0), large2,
+	assert_zu_eq(xallocx(p, large2, 0, flags), large2,
 	    "Unexpected xallocx() behavior");
 	/* Test size decrease with zero extra. */
-	assert_zu_eq(xallocx(p, large0, 0, 0), large0,
+	assert_zu_eq(xallocx(p, large0, 0, flags), large0,
 	    "Unexpected xallocx() behavior");
-	assert_zu_eq(xallocx(p, smallmax, 0, 0), large0,
+	assert_zu_eq(xallocx(p, smallmax, 0, flags), large0,
 	    "Unexpected xallocx() behavior");
 
-	assert_zu_eq(xallocx(p, large2, 0, 0), large2,
+	assert_zu_eq(xallocx(p, large2, 0, flags), large2,
 	    "Unexpected xallocx() behavior");
 	/* Test size decrease with non-zero extra. */
-	assert_zu_eq(xallocx(p, large0, large2 - large0, 0), large2,
+	assert_zu_eq(xallocx(p, large0, large2 - large0, flags), large2,
 	    "Unexpected xallocx() behavior");
-	assert_zu_eq(xallocx(p, large1, large2 - large1, 0), large2,
+	assert_zu_eq(xallocx(p, large1, large2 - large1, flags), large2,
 	    "Unexpected xallocx() behavior");
-	assert_zu_eq(xallocx(p, large0, large1 - large0, 0), large1,
+	assert_zu_eq(xallocx(p, large0, large1 - large0, flags), large1,
 	    "Unexpected xallocx() behavior");
-	assert_zu_eq(xallocx(p, smallmax, large0 - smallmax, 0), large0,
+	assert_zu_eq(xallocx(p, smallmax, large0 - smallmax, flags), large0,
 	    "Unexpected xallocx() behavior");
 
-	assert_zu_eq(xallocx(p, large0, 0, 0), large0,
+	assert_zu_eq(xallocx(p, large0, 0, flags), large0,
 	    "Unexpected xallocx() behavior");
 	/* Test size increase with zero extra. */
-	assert_zu_eq(xallocx(p, large2, 0, 0), large2,
+	assert_zu_eq(xallocx(p, large2, 0, flags), large2,
 	    "Unexpected xallocx() behavior");
-	assert_zu_eq(xallocx(p, huge0, 0, 0), large2,
+	assert_zu_eq(xallocx(p, huge0, 0, flags), large2,
 	    "Unexpected xallocx() behavior");
 
-	assert_zu_eq(xallocx(p, large0, 0, 0), large0,
+	assert_zu_eq(xallocx(p, large0, 0, flags), large0,
 	    "Unexpected xallocx() behavior");
 	/* Test size increase with non-zero extra. */
-	assert_zu_lt(xallocx(p, large0, huge0 - large0, 0), huge0,
+	assert_zu_lt(xallocx(p, large0, huge0 - large0, flags), huge0,
 	    "Unexpected xallocx() behavior");
 
-	assert_zu_eq(xallocx(p, large0, 0, 0), large0,
+	assert_zu_eq(xallocx(p, large0, 0, flags), large0,
 	    "Unexpected xallocx() behavior");
 	/* Test size increase with non-zero extra. */
-	assert_zu_eq(xallocx(p, large0, large2 - large0, 0), large2,
+	assert_zu_eq(xallocx(p, large0, large2 - large0, flags), large2,
 	    "Unexpected xallocx() behavior");
 
-	assert_zu_eq(xallocx(p, large2, 0, 0), large2,
+	assert_zu_eq(xallocx(p, large2, 0, flags), large2,
 	    "Unexpected xallocx() behavior");
 	/* Test size+extra overflow. */
-	assert_zu_lt(xallocx(p, large2, hugemax - large2 + 1, 0), huge0,
+	assert_zu_lt(xallocx(p, large2, hugemax - large2 + 1, flags), huge0,
 	    "Unexpected xallocx() behavior");
 
-	dallocx(p, 0);
+	dallocx(p, flags);
 }
 TEST_END
 
 TEST_BEGIN(test_extra_huge)
 {
-	size_t largemax, huge0, huge1, huge2, hugemax;
+	int flags = MALLOCX_ARENA(arena_ind());
+	size_t largemax, huge1, huge2, huge3, hugemax;
 	void *p;
 
 	/* Get size classes. */
 	largemax = get_large_size(get_nlarge()-1);
-	huge0 = get_huge_size(0);
 	huge1 = get_huge_size(1);
 	huge2 = get_huge_size(2);
+	huge3 = get_huge_size(3);
 	hugemax = get_huge_size(get_nhuge()-1);
 
-	p = mallocx(huge2, 0);
+	p = mallocx(huge3, flags);
 	assert_ptr_not_null(p, "Unexpected mallocx() error");
 
-	assert_zu_eq(xallocx(p, huge2, 0, 0), huge2,
+	assert_zu_eq(xallocx(p, huge3, 0, flags), huge3,
 	    "Unexpected xallocx() behavior");
 	/* Test size decrease with zero extra. */
-	assert_zu_ge(xallocx(p, huge0, 0, 0), huge0,
+	assert_zu_ge(xallocx(p, huge1, 0, flags), huge1,
 	    "Unexpected xallocx() behavior");
-	assert_zu_ge(xallocx(p, largemax, 0, 0), huge0,
+	assert_zu_ge(xallocx(p, largemax, 0, flags), huge1,
 	    "Unexpected xallocx() behavior");
 
-	assert_zu_eq(xallocx(p, huge2, 0, 0), huge2,
+	assert_zu_eq(xallocx(p, huge3, 0, flags), huge3,
 	    "Unexpected xallocx() behavior");
 	/* Test size decrease with non-zero extra. */
-	assert_zu_eq(xallocx(p, huge0, huge2 - huge0, 0), huge2,
+	assert_zu_eq(xallocx(p, huge1, huge3 - huge1, flags), huge3,
 	    "Unexpected xallocx() behavior");
-	assert_zu_eq(xallocx(p, huge1, huge2 - huge1, 0), huge2,
+	assert_zu_eq(xallocx(p, huge2, huge3 - huge2, flags), huge3,
 	    "Unexpected xallocx() behavior");
-	assert_zu_eq(xallocx(p, huge0, huge1 - huge0, 0), huge1,
+	assert_zu_eq(xallocx(p, huge1, huge2 - huge1, flags), huge2,
 	    "Unexpected xallocx() behavior");
-	assert_zu_ge(xallocx(p, largemax, huge0 - largemax, 0), huge0,
+	assert_zu_ge(xallocx(p, largemax, huge1 - largemax, flags), huge1,
 	    "Unexpected xallocx() behavior");
 
-	assert_zu_ge(xallocx(p, huge0, 0, 0), huge0,
+	assert_zu_ge(xallocx(p, huge1, 0, flags), huge1,
 	    "Unexpected xallocx() behavior");
 	/* Test size increase with zero extra. */
-	assert_zu_le(xallocx(p, huge2, 0, 0), huge2,
+	assert_zu_le(xallocx(p, huge3, 0, flags), huge3,
 	    "Unexpected xallocx() behavior");
-	assert_zu_le(xallocx(p, hugemax+1, 0, 0), huge2,
+	assert_zu_le(xallocx(p, hugemax+1, 0, flags), huge3,
 	    "Unexpected xallocx() behavior");
 
-	assert_zu_ge(xallocx(p, huge0, 0, 0), huge0,
+	assert_zu_ge(xallocx(p, huge1, 0, flags), huge1,
 	    "Unexpected xallocx() behavior");
 	/* Test size increase with non-zero extra. */
-	assert_zu_le(xallocx(p, huge0, SIZE_T_MAX - huge0, 0), hugemax,
+	assert_zu_le(xallocx(p, huge1, SIZE_T_MAX - huge1, flags), hugemax,
 	    "Unexpected xallocx() behavior");
 
-	assert_zu_ge(xallocx(p, huge0, 0, 0), huge0,
+	assert_zu_ge(xallocx(p, huge1, 0, flags), huge1,
 	    "Unexpected xallocx() behavior");
 	/* Test size increase with non-zero extra. */
-	assert_zu_le(xallocx(p, huge0, huge2 - huge0, 0), huge2,
+	assert_zu_le(xallocx(p, huge1, huge3 - huge1, flags), huge3,
 	    "Unexpected xallocx() behavior");
 
-	assert_zu_eq(xallocx(p, huge2, 0, 0), huge2,
+	assert_zu_eq(xallocx(p, huge3, 0, flags), huge3,
 	    "Unexpected xallocx() behavior");
 	/* Test size+extra overflow. */
-	assert_zu_le(xallocx(p, huge2, hugemax - huge2 + 1, 0), hugemax,
+	assert_zu_le(xallocx(p, huge3, hugemax - huge3 + 1, flags), hugemax,
 	    "Unexpected xallocx() behavior");
 
-	dallocx(p, 0);
+	dallocx(p, flags);
 }
 TEST_END
 
@@ -388,12 +413,13 @@ validate_fill(const void *p, uint8_t c, size_t offset, size_t len)
 static void
 test_zero(size_t szmin, size_t szmax)
 {
+	int flags = MALLOCX_ARENA(arena_ind()) | MALLOCX_ZERO;
 	size_t sz, nsz;
 	void *p;
 #define	FILL_BYTE 0x7aU
 
 	sz = szmax;
-	p = mallocx(sz, MALLOCX_ZERO);
+	p = mallocx(sz, flags);
 	assert_ptr_not_null(p, "Unexpected mallocx() error");
 	assert_false(validate_fill(p, 0x00, 0, sz), "Memory not filled: sz=%zu",
 	    sz);
@@ -408,14 +434,14 @@ test_zero(size_t szmin, size_t szmax)
 
 	/* Shrink in place so that we can expect growing in place to succeed. */
 	sz = szmin;
-	assert_zu_eq(xallocx(p, sz, 0, MALLOCX_ZERO), sz,
+	assert_zu_eq(xallocx(p, sz, 0, flags), sz,
 	    "Unexpected xallocx() error");
 	assert_false(validate_fill(p, FILL_BYTE, 0, sz),
 	    "Memory not filled: sz=%zu", sz);
 
 	for (sz = szmin; sz < szmax; sz = nsz) {
-		nsz = nallocx(sz+1, MALLOCX_ZERO);
-		assert_zu_eq(xallocx(p, sz+1, 0, MALLOCX_ZERO), nsz,
+		nsz = nallocx(sz+1, flags);
+		assert_zu_eq(xallocx(p, sz+1, 0, flags), nsz,
 		    "Unexpected xallocx() failure");
 		assert_false(validate_fill(p, FILL_BYTE, 0, sz),
 		    "Memory not filled: sz=%zu", sz);
@@ -426,7 +452,7 @@ test_zero(size_t szmin, size_t szmax)
 		    "Memory not filled: nsz=%zu", nsz);
 	}
 
-	dallocx(p, 0);
+	dallocx(p, flags);
 }
 
 TEST_BEGIN(test_zero_large)
diff --git a/deps/jemalloc/test/src/mtx.c b/deps/jemalloc/test/src/mtx.c
index 73bd02f6d..8a5dfdd99 100644
--- a/deps/jemalloc/test/src/mtx.c
+++ b/deps/jemalloc/test/src/mtx.c
@@ -11,6 +11,8 @@ mtx_init(mtx_t *mtx)
 #ifdef _WIN32
 	if (!InitializeCriticalSectionAndSpinCount(&mtx->lock, _CRT_SPINCOUNT))
 		return (true);
+#elif (defined(JEMALLOC_OS_UNFAIR_LOCK))
+	mtx->lock = OS_UNFAIR_LOCK_INIT;
 #elif (defined(JEMALLOC_OSSPIN))
 	mtx->lock = 0;
 #else
@@ -33,6 +35,7 @@ mtx_fini(mtx_t *mtx)
 {
 
 #ifdef _WIN32
+#elif (defined(JEMALLOC_OS_UNFAIR_LOCK))
 #elif (defined(JEMALLOC_OSSPIN))
 #else
 	pthread_mutex_destroy(&mtx->lock);
@@ -45,6 +48,8 @@ mtx_lock(mtx_t *mtx)
 
 #ifdef _WIN32
 	EnterCriticalSection(&mtx->lock);
+#elif (defined(JEMALLOC_OS_UNFAIR_LOCK))
+	os_unfair_lock_lock(&mtx->lock);
 #elif (defined(JEMALLOC_OSSPIN))
 	OSSpinLockLock(&mtx->lock);
 #else
@@ -58,6 +63,8 @@ mtx_unlock(mtx_t *mtx)
 
 #ifdef _WIN32
 	LeaveCriticalSection(&mtx->lock);
+#elif (defined(JEMALLOC_OS_UNFAIR_LOCK))
+	os_unfair_lock_unlock(&mtx->lock);
 #elif (defined(JEMALLOC_OSSPIN))
 	OSSpinLockUnlock(&mtx->lock);
 #else
diff --git a/deps/jemalloc/test/src/test.c b/deps/jemalloc/test/src/test.c
index 8173614cf..d70cc7501 100644
--- a/deps/jemalloc/test/src/test.c
+++ b/deps/jemalloc/test/src/test.c
@@ -60,32 +60,30 @@ p_test_fini(void)
 	malloc_printf("%s: %s\n", test_name, test_status_string(test_status));
 }
 
-test_status_t
-p_test(test_t *t, ...)
+static test_status_t
+p_test_impl(bool do_malloc_init, test_t *t, va_list ap)
 {
 	test_status_t ret;
-	va_list ap;
 
-	/*
-	 * Make sure initialization occurs prior to running tests.  Tests are
-	 * special because they may use internal facilities prior to triggering
-	 * initialization as a side effect of calling into the public API.  This
-	 * is a final safety that works even if jemalloc_constructor() doesn't
-	 * run, as for MSVC builds.
-	 */
-	if (nallocx(1, 0) == 0) {
-		malloc_printf("Initialization error");
-		return (test_status_fail);
+	if (do_malloc_init) {
+		/*
+		 * Make sure initialization occurs prior to running tests.
+		 * Tests are special because they may use internal facilities
+		 * prior to triggering initialization as a side effect of
+		 * calling into the public API.
+		 */
+		if (nallocx(1, 0) == 0) {
+			malloc_printf("Initialization error");
+			return (test_status_fail);
+		}
 	}
 
 	ret = test_status_pass;
-	va_start(ap, t);
 	for (; t != NULL; t = va_arg(ap, test_t *)) {
 		t();
 		if (test_status > ret)
 			ret = test_status;
 	}
-	va_end(ap);
 
 	malloc_printf("--- %s: %u/%u, %s: %u/%u, %s: %u/%u ---\n",
 	    test_status_string(test_status_pass),
@@ -98,6 +96,34 @@ p_test(test_t *t, ...)
 	return (ret);
 }
 
+test_status_t
+p_test(test_t *t, ...)
+{
+	test_status_t ret;
+	va_list ap;
+
+	ret = test_status_pass;
+	va_start(ap, t);
+	ret = p_test_impl(true, t, ap);
+	va_end(ap);
+
+	return (ret);
+}
+
+test_status_t
+p_test_no_malloc_init(test_t *t, ...)
+{
+	test_status_t ret;
+	va_list ap;
+
+	ret = test_status_pass;
+	va_start(ap, t);
+	ret = p_test_impl(false, t, ap);
+	va_end(ap);
+
+	return (ret);
+}
+
 void
 p_test_fail(const char *prefix, const char *message)
 {
diff --git a/deps/jemalloc/test/src/timer.c b/deps/jemalloc/test/src/timer.c
index 0c93abaf9..3c7e63a26 100644
--- a/deps/jemalloc/test/src/timer.c
+++ b/deps/jemalloc/test/src/timer.c
@@ -4,50 +4,26 @@ void
 timer_start(timedelta_t *timer)
 {
 
-#ifdef _WIN32
-	GetSystemTimeAsFileTime(&timer->ft0);
-#elif JEMALLOC_CLOCK_GETTIME
-	if (sysconf(_SC_MONOTONIC_CLOCK) <= 0)
-		timer->clock_id = CLOCK_REALTIME;
-	else
-		timer->clock_id = CLOCK_MONOTONIC;
-	clock_gettime(timer->clock_id, &timer->ts0);
-#else
-	gettimeofday(&timer->tv0, NULL);
-#endif
+	nstime_init(&timer->t0, 0);
+	nstime_update(&timer->t0);
 }
 
 void
 timer_stop(timedelta_t *timer)
 {
 
-#ifdef _WIN32
-	GetSystemTimeAsFileTime(&timer->ft0);
-#elif JEMALLOC_CLOCK_GETTIME
-	clock_gettime(timer->clock_id, &timer->ts1);
-#else
-	gettimeofday(&timer->tv1, NULL);
-#endif
+	nstime_copy(&timer->t1, &timer->t0);
+	nstime_update(&timer->t1);
 }
 
 uint64_t
 timer_usec(const timedelta_t *timer)
 {
+	nstime_t delta;
 
-#ifdef _WIN32
-	uint64_t t0, t1;
-	t0 = (((uint64_t)timer->ft0.dwHighDateTime) << 32) |
-	    timer->ft0.dwLowDateTime;
-	t1 = (((uint64_t)timer->ft1.dwHighDateTime) << 32) |
-	    timer->ft1.dwLowDateTime;
-	return ((t1 - t0) / 10);
-#elif JEMALLOC_CLOCK_GETTIME
-	return (((timer->ts1.tv_sec - timer->ts0.tv_sec) * 1000000) +
-	    (timer->ts1.tv_nsec - timer->ts0.tv_nsec) / 1000);
-#else
-	return (((timer->tv1.tv_sec - timer->tv0.tv_sec) * 1000000) +
-	    timer->tv1.tv_usec - timer->tv0.tv_usec);
-#endif
+	nstime_copy(&delta, &timer->t1);
+	nstime_subtract(&delta, &timer->t0);
+	return (nstime_ns(&delta) / 1000);
 }
 
 void
@@ -56,9 +32,8 @@ timer_ratio(timedelta_t *a, timedelta_t *b, char *buf, size_t buflen)
 	uint64_t t0 = timer_usec(a);
 	uint64_t t1 = timer_usec(b);
 	uint64_t mult;
-	unsigned i = 0;
-	unsigned j;
-	int n;
+	size_t i = 0;
+	size_t j, n;
 
 	/* Whole. */
 	n = malloc_snprintf(&buf[i], buflen-i, "%"FMTu64, t0 / t1);
diff --git a/deps/jemalloc/test/stress/microbench.c b/deps/jemalloc/test/stress/microbench.c
index ee39fea7f..7dc45f89c 100644
--- a/deps/jemalloc/test/stress/microbench.c
+++ b/deps/jemalloc/test/stress/microbench.c
@@ -1,7 +1,8 @@
 #include "test/jemalloc_test.h"
 
 JEMALLOC_INLINE_C void
-time_func(timedelta_t *timer, uint64_t nwarmup, uint64_t niter, void (*func)(void))
+time_func(timedelta_t *timer, uint64_t nwarmup, uint64_t niter,
+    void (*func)(void))
 {
 	uint64_t i;
 
diff --git a/deps/jemalloc/test/unit/a0.c b/deps/jemalloc/test/unit/a0.c
new file mode 100644
index 000000000..b9ba45a3d
--- /dev/null
+++ b/deps/jemalloc/test/unit/a0.c
@@ -0,0 +1,19 @@
+#include "test/jemalloc_test.h"
+
+TEST_BEGIN(test_a0)
+{
+	void *p;
+
+	p = a0malloc(1);
+	assert_ptr_not_null(p, "Unexpected a0malloc() error");
+	a0dalloc(p);
+}
+TEST_END
+
+int
+main(void)
+{
+
+	return (test_no_malloc_init(
+	    test_a0));
+}
diff --git a/deps/jemalloc/test/unit/arena_reset.c b/deps/jemalloc/test/unit/arena_reset.c
new file mode 100755
index 000000000..adf9baa5d
--- /dev/null
+++ b/deps/jemalloc/test/unit/arena_reset.c
@@ -0,0 +1,159 @@
+#include "test/jemalloc_test.h"
+
+#ifdef JEMALLOC_PROF
+const char *malloc_conf = "prof:true,lg_prof_sample:0";
+#endif
+
+static unsigned
+get_nsizes_impl(const char *cmd)
+{
+	unsigned ret;
+	size_t z;
+
+	z = sizeof(unsigned);
+	assert_d_eq(mallctl(cmd, (void *)&ret, &z, NULL, 0), 0,
+	    "Unexpected mallctl(\"%s\", ...) failure", cmd);
+
+	return (ret);
+}
+
+static unsigned
+get_nsmall(void)
+{
+
+	return (get_nsizes_impl("arenas.nbins"));
+}
+
+static unsigned
+get_nlarge(void)
+{
+
+	return (get_nsizes_impl("arenas.nlruns"));
+}
+
+static unsigned
+get_nhuge(void)
+{
+
+	return (get_nsizes_impl("arenas.nhchunks"));
+}
+
+static size_t
+get_size_impl(const char *cmd, size_t ind)
+{
+	size_t ret;
+	size_t z;
+	size_t mib[4];
+	size_t miblen = 4;
+
+	z = sizeof(size_t);
+	assert_d_eq(mallctlnametomib(cmd, mib, &miblen),
+	    0, "Unexpected mallctlnametomib(\"%s\", ...) failure", cmd);
+	mib[2] = ind;
+	z = sizeof(size_t);
+	assert_d_eq(mallctlbymib(mib, miblen, (void *)&ret, &z, NULL, 0),
+	    0, "Unexpected mallctlbymib([\"%s\", %zu], ...) failure", cmd, ind);
+
+	return (ret);
+}
+
+static size_t
+get_small_size(size_t ind)
+{
+
+	return (get_size_impl("arenas.bin.0.size", ind));
+}
+
+static size_t
+get_large_size(size_t ind)
+{
+
+	return (get_size_impl("arenas.lrun.0.size", ind));
+}
+
+static size_t
+get_huge_size(size_t ind)
+{
+
+	return (get_size_impl("arenas.hchunk.0.size", ind));
+}
+
+TEST_BEGIN(test_arena_reset)
+{
+#define	NHUGE	4
+	unsigned arena_ind, nsmall, nlarge, nhuge, nptrs, i;
+	size_t sz, miblen;
+	void **ptrs;
+	int flags;
+	size_t mib[3];
+	tsdn_t *tsdn;
+
+	test_skip_if((config_valgrind && unlikely(in_valgrind)) || (config_fill
+	    && unlikely(opt_quarantine)));
+
+	sz = sizeof(unsigned);
+	assert_d_eq(mallctl("arenas.extend", (void *)&arena_ind, &sz, NULL, 0),
+	    0, "Unexpected mallctl() failure");
+
+	flags = MALLOCX_ARENA(arena_ind) | MALLOCX_TCACHE_NONE;
+
+	nsmall = get_nsmall();
+	nlarge = get_nlarge();
+	nhuge = get_nhuge() > NHUGE ? NHUGE : get_nhuge();
+	nptrs = nsmall + nlarge + nhuge;
+	ptrs = (void **)malloc(nptrs * sizeof(void *));
+	assert_ptr_not_null(ptrs, "Unexpected malloc() failure");
+
+	/* Allocate objects with a wide range of sizes. */
+	for (i = 0; i < nsmall; i++) {
+		sz = get_small_size(i);
+		ptrs[i] = mallocx(sz, flags);
+		assert_ptr_not_null(ptrs[i],
+		    "Unexpected mallocx(%zu, %#x) failure", sz, flags);
+	}
+	for (i = 0; i < nlarge; i++) {
+		sz = get_large_size(i);
+		ptrs[nsmall + i] = mallocx(sz, flags);
+		assert_ptr_not_null(ptrs[i],
+		    "Unexpected mallocx(%zu, %#x) failure", sz, flags);
+	}
+	for (i = 0; i < nhuge; i++) {
+		sz = get_huge_size(i);
+		ptrs[nsmall + nlarge + i] = mallocx(sz, flags);
+		assert_ptr_not_null(ptrs[i],
+		    "Unexpected mallocx(%zu, %#x) failure", sz, flags);
+	}
+
+	tsdn = tsdn_fetch();
+
+	/* Verify allocations. */
+	for (i = 0; i < nptrs; i++) {
+		assert_zu_gt(ivsalloc(tsdn, ptrs[i], false), 0,
+		    "Allocation should have queryable size");
+	}
+
+	/* Reset. */
+	miblen = sizeof(mib)/sizeof(size_t);
+	assert_d_eq(mallctlnametomib("arena.0.reset", mib, &miblen), 0,
+	    "Unexpected mallctlnametomib() failure");
+	mib[1] = (size_t)arena_ind;
+	assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, NULL, 0), 0,
+	    "Unexpected mallctlbymib() failure");
+
+	/* Verify allocations no longer exist. */
+	for (i = 0; i < nptrs; i++) {
+		assert_zu_eq(ivsalloc(tsdn, ptrs[i], false), 0,
+		    "Allocation should no longer exist");
+	}
+
+	free(ptrs);
+}
+TEST_END
+
+int
+main(void)
+{
+
+	return (test(
+	    test_arena_reset));
+}
diff --git a/deps/jemalloc/test/unit/bitmap.c b/deps/jemalloc/test/unit/bitmap.c
index 7da583d85..a2dd54630 100644
--- a/deps/jemalloc/test/unit/bitmap.c
+++ b/deps/jemalloc/test/unit/bitmap.c
@@ -6,7 +6,11 @@ TEST_BEGIN(test_bitmap_size)
 
 	prev_size = 0;
 	for (i = 1; i <= BITMAP_MAXBITS; i++) {
-		size_t size = bitmap_size(i);
+		bitmap_info_t binfo;
+		size_t size;
+
+		bitmap_info_init(&binfo, i);
+		size = bitmap_size(&binfo);
 		assert_true(size >= prev_size,
 		    "Bitmap size is smaller than expected");
 		prev_size = size;
@@ -23,8 +27,8 @@ TEST_BEGIN(test_bitmap_init)
 		bitmap_info_init(&binfo, i);
 		{
 			size_t j;
-			bitmap_t *bitmap = (bitmap_t *)malloc(sizeof(bitmap_t) *
-				bitmap_info_ngroups(&binfo));
+			bitmap_t *bitmap = (bitmap_t *)malloc(
+			    bitmap_size(&binfo));
 			bitmap_init(bitmap, &binfo);
 
 			for (j = 0; j < i; j++) {
@@ -46,8 +50,8 @@ TEST_BEGIN(test_bitmap_set)
 		bitmap_info_init(&binfo, i);
 		{
 			size_t j;
-			bitmap_t *bitmap = (bitmap_t *)malloc(sizeof(bitmap_t) *
-				bitmap_info_ngroups(&binfo));
+			bitmap_t *bitmap = (bitmap_t *)malloc(
+			    bitmap_size(&binfo));
 			bitmap_init(bitmap, &binfo);
 
 			for (j = 0; j < i; j++)
@@ -69,8 +73,8 @@ TEST_BEGIN(test_bitmap_unset)
 		bitmap_info_init(&binfo, i);
 		{
 			size_t j;
-			bitmap_t *bitmap = (bitmap_t *)malloc(sizeof(bitmap_t) *
-				bitmap_info_ngroups(&binfo));
+			bitmap_t *bitmap = (bitmap_t *)malloc(
+			    bitmap_size(&binfo));
 			bitmap_init(bitmap, &binfo);
 
 			for (j = 0; j < i; j++)
@@ -97,9 +101,9 @@ TEST_BEGIN(test_bitmap_sfu)
 		bitmap_info_t binfo;
 		bitmap_info_init(&binfo, i);
 		{
-			ssize_t j;
-			bitmap_t *bitmap = (bitmap_t *)malloc(sizeof(bitmap_t) *
-				bitmap_info_ngroups(&binfo));
+			size_t j;
+			bitmap_t *bitmap = (bitmap_t *)malloc(
+			    bitmap_size(&binfo));
 			bitmap_init(bitmap, &binfo);
 
 			/* Iteratively set bits starting at the beginning. */
@@ -115,7 +119,7 @@ TEST_BEGIN(test_bitmap_sfu)
 			 * Iteratively unset bits starting at the end, and
 			 * verify that bitmap_sfu() reaches the unset bits.
 			 */
-			for (j = i - 1; j >= 0; j--) {
+			for (j = i - 1; j < i; j--) { /* (i..0] */
 				bitmap_unset(bitmap, &binfo, j);
 				assert_zd_eq(bitmap_sfu(bitmap, &binfo), j,
 				    "First unset bit should the bit previously "
diff --git a/deps/jemalloc/test/unit/ckh.c b/deps/jemalloc/test/unit/ckh.c
index b11759599..2cbc22688 100644
--- a/deps/jemalloc/test/unit/ckh.c
+++ b/deps/jemalloc/test/unit/ckh.c
@@ -7,8 +7,8 @@ TEST_BEGIN(test_new_delete)
 
 	tsd = tsd_fetch();
 
-	assert_false(ckh_new(tsd, &ckh, 2, ckh_string_hash, ckh_string_keycomp),
-	    "Unexpected ckh_new() error");
+	assert_false(ckh_new(tsd, &ckh, 2, ckh_string_hash,
+	    ckh_string_keycomp), "Unexpected ckh_new() error");
 	ckh_delete(tsd, &ckh);
 
 	assert_false(ckh_new(tsd, &ckh, 3, ckh_pointer_hash,
@@ -32,8 +32,8 @@ TEST_BEGIN(test_count_insert_search_remove)
 
 	tsd = tsd_fetch();
 
-	assert_false(ckh_new(tsd, &ckh, 2, ckh_string_hash, ckh_string_keycomp),
-	    "Unexpected ckh_new() error");
+	assert_false(ckh_new(tsd, &ckh, 2, ckh_string_hash,
+	    ckh_string_keycomp), "Unexpected ckh_new() error");
 	assert_zu_eq(ckh_count(&ckh), 0,
 	    "ckh_count() should return %zu, but it returned %zu", ZU(0),
 	    ckh_count(&ckh));
diff --git a/deps/jemalloc/test/unit/decay.c b/deps/jemalloc/test/unit/decay.c
new file mode 100755
index 000000000..5af8f8074
--- /dev/null
+++ b/deps/jemalloc/test/unit/decay.c
@@ -0,0 +1,374 @@
+#include "test/jemalloc_test.h"
+
+const char *malloc_conf = "purge:decay,decay_time:1";
+
+static nstime_monotonic_t *nstime_monotonic_orig;
+static nstime_update_t *nstime_update_orig;
+
+static unsigned nupdates_mock;
+static nstime_t time_mock;
+static bool monotonic_mock;
+
+static bool
+nstime_monotonic_mock(void)
+{
+
+	return (monotonic_mock);
+}
+
+static bool
+nstime_update_mock(nstime_t *time)
+{
+
+	nupdates_mock++;
+	if (monotonic_mock)
+		nstime_copy(time, &time_mock);
+	return (!monotonic_mock);
+}
+
+TEST_BEGIN(test_decay_ticks)
+{
+	ticker_t *decay_ticker;
+	unsigned tick0, tick1;
+	size_t sz, huge0, large0;
+	void *p;
+
+	test_skip_if(opt_purge != purge_mode_decay);
+
+	decay_ticker = decay_ticker_get(tsd_fetch(), 0);
+	assert_ptr_not_null(decay_ticker,
+	    "Unexpected failure getting decay ticker");
+
+	sz = sizeof(size_t);
+	assert_d_eq(mallctl("arenas.hchunk.0.size", (void *)&huge0, &sz, NULL,
+	    0), 0, "Unexpected mallctl failure");
+	assert_d_eq(mallctl("arenas.lrun.0.size", (void *)&large0, &sz, NULL,
+	    0), 0, "Unexpected mallctl failure");
+
+	/*
+	 * Test the standard APIs using a huge size class, since we can't
+	 * control tcache interactions (except by completely disabling tcache
+	 * for the entire test program).
+	 */
+
+	/* malloc(). */
+	tick0 = ticker_read(decay_ticker);
+	p = malloc(huge0);
+	assert_ptr_not_null(p, "Unexpected malloc() failure");
+	tick1 = ticker_read(decay_ticker);
+	assert_u32_ne(tick1, tick0, "Expected ticker to tick during malloc()");
+	/* free(). */
+	tick0 = ticker_read(decay_ticker);
+	free(p);
+	tick1 = ticker_read(decay_ticker);
+	assert_u32_ne(tick1, tick0, "Expected ticker to tick during free()");
+
+	/* calloc(). */
+	tick0 = ticker_read(decay_ticker);
+	p = calloc(1, huge0);
+	assert_ptr_not_null(p, "Unexpected calloc() failure");
+	tick1 = ticker_read(decay_ticker);
+	assert_u32_ne(tick1, tick0, "Expected ticker to tick during calloc()");
+	free(p);
+
+	/* posix_memalign(). */
+	tick0 = ticker_read(decay_ticker);
+	assert_d_eq(posix_memalign(&p, sizeof(size_t), huge0), 0,
+	    "Unexpected posix_memalign() failure");
+	tick1 = ticker_read(decay_ticker);
+	assert_u32_ne(tick1, tick0,
+	    "Expected ticker to tick during posix_memalign()");
+	free(p);
+
+	/* aligned_alloc(). */
+	tick0 = ticker_read(decay_ticker);
+	p = aligned_alloc(sizeof(size_t), huge0);
+	assert_ptr_not_null(p, "Unexpected aligned_alloc() failure");
+	tick1 = ticker_read(decay_ticker);
+	assert_u32_ne(tick1, tick0,
+	    "Expected ticker to tick during aligned_alloc()");
+	free(p);
+
+	/* realloc(). */
+	/* Allocate. */
+	tick0 = ticker_read(decay_ticker);
+	p = realloc(NULL, huge0);
+	assert_ptr_not_null(p, "Unexpected realloc() failure");
+	tick1 = ticker_read(decay_ticker);
+	assert_u32_ne(tick1, tick0, "Expected ticker to tick during realloc()");
+	/* Reallocate. */
+	tick0 = ticker_read(decay_ticker);
+	p = realloc(p, huge0);
+	assert_ptr_not_null(p, "Unexpected realloc() failure");
+	tick1 = ticker_read(decay_ticker);
+	assert_u32_ne(tick1, tick0, "Expected ticker to tick during realloc()");
+	/* Deallocate. */
+	tick0 = ticker_read(decay_ticker);
+	realloc(p, 0);
+	tick1 = ticker_read(decay_ticker);
+	assert_u32_ne(tick1, tick0, "Expected ticker to tick during realloc()");
+
+	/*
+	 * Test the *allocx() APIs using huge, large, and small size classes,
+	 * with tcache explicitly disabled.
+	 */
+	{
+		unsigned i;
+		size_t allocx_sizes[3];
+		allocx_sizes[0] = huge0;
+		allocx_sizes[1] = large0;
+		allocx_sizes[2] = 1;
+
+		for (i = 0; i < sizeof(allocx_sizes) / sizeof(size_t); i++) {
+			sz = allocx_sizes[i];
+
+			/* mallocx(). */
+			tick0 = ticker_read(decay_ticker);
+			p = mallocx(sz, MALLOCX_TCACHE_NONE);
+			assert_ptr_not_null(p, "Unexpected mallocx() failure");
+			tick1 = ticker_read(decay_ticker);
+			assert_u32_ne(tick1, tick0,
+			    "Expected ticker to tick during mallocx() (sz=%zu)",
+			    sz);
+			/* rallocx(). */
+			tick0 = ticker_read(decay_ticker);
+			p = rallocx(p, sz, MALLOCX_TCACHE_NONE);
+			assert_ptr_not_null(p, "Unexpected rallocx() failure");
+			tick1 = ticker_read(decay_ticker);
+			assert_u32_ne(tick1, tick0,
+			    "Expected ticker to tick during rallocx() (sz=%zu)",
+			    sz);
+			/* xallocx(). */
+			tick0 = ticker_read(decay_ticker);
+			xallocx(p, sz, 0, MALLOCX_TCACHE_NONE);
+			tick1 = ticker_read(decay_ticker);
+			assert_u32_ne(tick1, tick0,
+			    "Expected ticker to tick during xallocx() (sz=%zu)",
+			    sz);
+			/* dallocx(). */
+			tick0 = ticker_read(decay_ticker);
+			dallocx(p, MALLOCX_TCACHE_NONE);
+			tick1 = ticker_read(decay_ticker);
+			assert_u32_ne(tick1, tick0,
+			    "Expected ticker to tick during dallocx() (sz=%zu)",
+			    sz);
+			/* sdallocx(). */
+			p = mallocx(sz, MALLOCX_TCACHE_NONE);
+			assert_ptr_not_null(p, "Unexpected mallocx() failure");
+			tick0 = ticker_read(decay_ticker);
+			sdallocx(p, sz, MALLOCX_TCACHE_NONE);
+			tick1 = ticker_read(decay_ticker);
+			assert_u32_ne(tick1, tick0,
+			    "Expected ticker to tick during sdallocx() "
+			    "(sz=%zu)", sz);
+		}
+	}
+
+	/*
+	 * Test tcache fill/flush interactions for large and small size classes,
+	 * using an explicit tcache.
+	 */
+	if (config_tcache) {
+		unsigned tcache_ind, i;
+		size_t tcache_sizes[2];
+		tcache_sizes[0] = large0;
+		tcache_sizes[1] = 1;
+
+		sz = sizeof(unsigned);
+		assert_d_eq(mallctl("tcache.create", (void *)&tcache_ind, &sz,
+		    NULL, 0), 0, "Unexpected mallctl failure");
+
+		for (i = 0; i < sizeof(tcache_sizes) / sizeof(size_t); i++) {
+			sz = tcache_sizes[i];
+
+			/* tcache fill. */
+			tick0 = ticker_read(decay_ticker);
+			p = mallocx(sz, MALLOCX_TCACHE(tcache_ind));
+			assert_ptr_not_null(p, "Unexpected mallocx() failure");
+			tick1 = ticker_read(decay_ticker);
+			assert_u32_ne(tick1, tick0,
+			    "Expected ticker to tick during tcache fill "
+			    "(sz=%zu)", sz);
+			/* tcache flush. */
+			dallocx(p, MALLOCX_TCACHE(tcache_ind));
+			tick0 = ticker_read(decay_ticker);
+			assert_d_eq(mallctl("tcache.flush", NULL, NULL,
+			    (void *)&tcache_ind, sizeof(unsigned)), 0,
+			    "Unexpected mallctl failure");
+			tick1 = ticker_read(decay_ticker);
+			assert_u32_ne(tick1, tick0,
+			    "Expected ticker to tick during tcache flush "
+			    "(sz=%zu)", sz);
+		}
+	}
+}
+TEST_END
+
+TEST_BEGIN(test_decay_ticker)
+{
+#define	NPS 1024
+	int flags = (MALLOCX_ARENA(0) | MALLOCX_TCACHE_NONE);
+	void *ps[NPS];
+	uint64_t epoch;
+	uint64_t npurge0 = 0;
+	uint64_t npurge1 = 0;
+	size_t sz, large;
+	unsigned i, nupdates0;
+	nstime_t time, decay_time, deadline;
+
+	test_skip_if(opt_purge != purge_mode_decay);
+
+	/*
+	 * Allocate a bunch of large objects, pause the clock, deallocate the
+	 * objects, restore the clock, then [md]allocx() in a tight loop to
+	 * verify the ticker triggers purging.
+	 */
+
+	if (config_tcache) {
+		size_t tcache_max;
+
+		sz = sizeof(size_t);
+		assert_d_eq(mallctl("arenas.tcache_max", (void *)&tcache_max,
+		    &sz, NULL, 0), 0, "Unexpected mallctl failure");
+		large = nallocx(tcache_max + 1, flags);
+	}  else {
+		sz = sizeof(size_t);
+		assert_d_eq(mallctl("arenas.lrun.0.size", (void *)&large, &sz,
+		    NULL, 0), 0, "Unexpected mallctl failure");
+	}
+
+	assert_d_eq(mallctl("arena.0.purge", NULL, NULL, NULL, 0), 0,
+	    "Unexpected mallctl failure");
+	assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch,
+	    sizeof(uint64_t)), 0, "Unexpected mallctl failure");
+	sz = sizeof(uint64_t);
+	assert_d_eq(mallctl("stats.arenas.0.npurge", (void *)&npurge0, &sz,
+	    NULL, 0), config_stats ? 0 : ENOENT, "Unexpected mallctl result");
+
+	for (i = 0; i < NPS; i++) {
+		ps[i] = mallocx(large, flags);
+		assert_ptr_not_null(ps[i], "Unexpected mallocx() failure");
+	}
+
+	nupdates_mock = 0;
+	nstime_init(&time_mock, 0);
+	nstime_update(&time_mock);
+	monotonic_mock = true;
+
+	nstime_monotonic_orig = nstime_monotonic;
+	nstime_update_orig = nstime_update;
+	nstime_monotonic = nstime_monotonic_mock;
+	nstime_update = nstime_update_mock;
+
+	for (i = 0; i < NPS; i++) {
+		dallocx(ps[i], flags);
+		nupdates0 = nupdates_mock;
+		assert_d_eq(mallctl("arena.0.decay", NULL, NULL, NULL, 0), 0,
+		    "Unexpected arena.0.decay failure");
+		assert_u_gt(nupdates_mock, nupdates0,
+		    "Expected nstime_update() to be called");
+	}
+
+	nstime_monotonic = nstime_monotonic_orig;
+	nstime_update = nstime_update_orig;
+
+	nstime_init(&time, 0);
+	nstime_update(&time);
+	nstime_init2(&decay_time, opt_decay_time, 0);
+	nstime_copy(&deadline, &time);
+	nstime_add(&deadline, &decay_time);
+	do {
+		for (i = 0; i < DECAY_NTICKS_PER_UPDATE / 2; i++) {
+			void *p = mallocx(1, flags);
+			assert_ptr_not_null(p, "Unexpected mallocx() failure");
+			dallocx(p, flags);
+		}
+		assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch,
+		    sizeof(uint64_t)), 0, "Unexpected mallctl failure");
+		sz = sizeof(uint64_t);
+		assert_d_eq(mallctl("stats.arenas.0.npurge", (void *)&npurge1,
+		    &sz, NULL, 0), config_stats ? 0 : ENOENT,
+		    "Unexpected mallctl result");
+
+		nstime_update(&time);
+	} while (nstime_compare(&time, &deadline) <= 0 && npurge1 == npurge0);
+
+	if (config_stats)
+		assert_u64_gt(npurge1, npurge0, "Expected purging to occur");
+#undef NPS
+}
+TEST_END
+
+TEST_BEGIN(test_decay_nonmonotonic)
+{
+#define	NPS (SMOOTHSTEP_NSTEPS + 1)
+	int flags = (MALLOCX_ARENA(0) | MALLOCX_TCACHE_NONE);
+	void *ps[NPS];
+	uint64_t epoch;
+	uint64_t npurge0 = 0;
+	uint64_t npurge1 = 0;
+	size_t sz, large0;
+	unsigned i, nupdates0;
+
+	test_skip_if(opt_purge != purge_mode_decay);
+
+	sz = sizeof(size_t);
+	assert_d_eq(mallctl("arenas.lrun.0.size", (void *)&large0, &sz, NULL,
+	    0), 0, "Unexpected mallctl failure");
+
+	assert_d_eq(mallctl("arena.0.purge", NULL, NULL, NULL, 0), 0,
+	    "Unexpected mallctl failure");
+	assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch,
+	    sizeof(uint64_t)), 0, "Unexpected mallctl failure");
+	sz = sizeof(uint64_t);
+	assert_d_eq(mallctl("stats.arenas.0.npurge", (void *)&npurge0, &sz,
+	    NULL, 0), config_stats ? 0 : ENOENT, "Unexpected mallctl result");
+
+	nupdates_mock = 0;
+	nstime_init(&time_mock, 0);
+	nstime_update(&time_mock);
+	monotonic_mock = false;
+
+	nstime_monotonic_orig = nstime_monotonic;
+	nstime_update_orig = nstime_update;
+	nstime_monotonic = nstime_monotonic_mock;
+	nstime_update = nstime_update_mock;
+
+	for (i = 0; i < NPS; i++) {
+		ps[i] = mallocx(large0, flags);
+		assert_ptr_not_null(ps[i], "Unexpected mallocx() failure");
+	}
+
+	for (i = 0; i < NPS; i++) {
+		dallocx(ps[i], flags);
+		nupdates0 = nupdates_mock;
+		assert_d_eq(mallctl("arena.0.decay", NULL, NULL, NULL, 0), 0,
+		    "Unexpected arena.0.decay failure");
+		assert_u_gt(nupdates_mock, nupdates0,
+		    "Expected nstime_update() to be called");
+	}
+
+	assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch,
+	    sizeof(uint64_t)), 0, "Unexpected mallctl failure");
+	sz = sizeof(uint64_t);
+	assert_d_eq(mallctl("stats.arenas.0.npurge", (void *)&npurge1, &sz,
+	    NULL, 0), config_stats ? 0 : ENOENT, "Unexpected mallctl result");
+
+	if (config_stats)
+		assert_u64_eq(npurge0, npurge1, "Unexpected purging occurred");
+
+	nstime_monotonic = nstime_monotonic_orig;
+	nstime_update = nstime_update_orig;
+#undef NPS
+}
+TEST_END
+
+int
+main(void)
+{
+
+	return (test(
+	    test_decay_ticks,
+	    test_decay_ticker,
+	    test_decay_nonmonotonic));
+}
diff --git a/deps/jemalloc/test/unit/fork.c b/deps/jemalloc/test/unit/fork.c
new file mode 100644
index 000000000..c530797c4
--- /dev/null
+++ b/deps/jemalloc/test/unit/fork.c
@@ -0,0 +1,64 @@
+#include "test/jemalloc_test.h"
+
+#ifndef _WIN32
+#include <sys/wait.h>
+#endif
+
+TEST_BEGIN(test_fork)
+{
+#ifndef _WIN32
+	void *p;
+	pid_t pid;
+
+	p = malloc(1);
+	assert_ptr_not_null(p, "Unexpected malloc() failure");
+
+	pid = fork();
+
+	free(p);
+
+	p = malloc(64);
+	assert_ptr_not_null(p, "Unexpected malloc() failure");
+	free(p);
+
+	if (pid == -1) {
+		/* Error. */
+		test_fail("Unexpected fork() failure");
+	} else if (pid == 0) {
+		/* Child. */
+		_exit(0);
+	} else {
+		int status;
+
+		/* Parent. */
+		while (true) {
+			if (waitpid(pid, &status, 0) == -1)
+				test_fail("Unexpected waitpid() failure");
+			if (WIFSIGNALED(status)) {
+				test_fail("Unexpected child termination due to "
+				    "signal %d", WTERMSIG(status));
+				break;
+			}
+			if (WIFEXITED(status)) {
+				if (WEXITSTATUS(status) != 0) {
+					test_fail(
+					    "Unexpected child exit value %d",
+					    WEXITSTATUS(status));
+				}
+				break;
+			}
+		}
+	}
+#else
+	test_skip("fork(2) is irrelevant to Windows");
+#endif
+}
+TEST_END
+
+int
+main(void)
+{
+
+	return (test(
+	    test_fork));
+}
diff --git a/deps/jemalloc/test/unit/hash.c b/deps/jemalloc/test/unit/hash.c
index 77a8cede9..010c9d76f 100644
--- a/deps/jemalloc/test/unit/hash.c
+++ b/deps/jemalloc/test/unit/hash.c
@@ -35,7 +35,7 @@ typedef enum {
 	hash_variant_x64_128
 } hash_variant_t;
 
-static size_t
+static int
 hash_variant_bits(hash_variant_t variant)
 {
 
@@ -59,19 +59,20 @@ hash_variant_string(hash_variant_t variant)
 	}
 }
 
+#define	KEY_SIZE	256
 static void
-hash_variant_verify(hash_variant_t variant)
+hash_variant_verify_key(hash_variant_t variant, uint8_t *key)
 {
-	const size_t hashbytes = hash_variant_bits(variant) / 8;
-	uint8_t key[256];
-	VARIABLE_ARRAY(uint8_t, hashes, hashbytes * 256);
+	const int hashbytes = hash_variant_bits(variant) / 8;
+	const int hashes_size = hashbytes * 256;
+	VARIABLE_ARRAY(uint8_t, hashes, hashes_size);
 	VARIABLE_ARRAY(uint8_t, final, hashbytes);
 	unsigned i;
 	uint32_t computed, expected;
 
-	memset(key, 0, sizeof(key));
-	memset(hashes, 0, sizeof(hashes));
-	memset(final, 0, sizeof(final));
+	memset(key, 0, KEY_SIZE);
+	memset(hashes, 0, hashes_size);
+	memset(final, 0, hashbytes);
 
 	/*
 	 * Hash keys of the form {0}, {0,1}, {0,1,2}, ..., {0,1,...,255} as the
@@ -102,17 +103,17 @@ hash_variant_verify(hash_variant_t variant)
 	/* Hash the result array. */
 	switch (variant) {
 	case hash_variant_x86_32: {
-		uint32_t out = hash_x86_32(hashes, hashbytes*256, 0);
+		uint32_t out = hash_x86_32(hashes, hashes_size, 0);
 		memcpy(final, &out, sizeof(out));
 		break;
 	} case hash_variant_x86_128: {
 		uint64_t out[2];
-		hash_x86_128(hashes, hashbytes*256, 0, out);
+		hash_x86_128(hashes, hashes_size, 0, out);
 		memcpy(final, out, sizeof(out));
 		break;
 	} case hash_variant_x64_128: {
 		uint64_t out[2];
-		hash_x64_128(hashes, hashbytes*256, 0, out);
+		hash_x64_128(hashes, hashes_size, 0, out);
 		memcpy(final, out, sizeof(out));
 		break;
 	} default: not_reached();
@@ -139,6 +140,19 @@ hash_variant_verify(hash_variant_t variant)
 	    hash_variant_string(variant), expected, computed);
 }
 
+static void
+hash_variant_verify(hash_variant_t variant)
+{
+#define	MAX_ALIGN	16
+	uint8_t key[KEY_SIZE + (MAX_ALIGN - 1)];
+	unsigned i;
+
+	for (i = 0; i < MAX_ALIGN; i++)
+		hash_variant_verify_key(variant, &key[i]);
+#undef MAX_ALIGN
+}
+#undef KEY_SIZE
+
 TEST_BEGIN(test_hash_x86_32)
 {
 
diff --git a/deps/jemalloc/test/unit/junk.c b/deps/jemalloc/test/unit/junk.c
index b23dd1e95..460bd524d 100644
--- a/deps/jemalloc/test/unit/junk.c
+++ b/deps/jemalloc/test/unit/junk.c
@@ -29,7 +29,7 @@ arena_dalloc_junk_small_intercept(void *ptr, arena_bin_info_t *bin_info)
 
 	arena_dalloc_junk_small_orig(ptr, bin_info);
 	for (i = 0; i < bin_info->reg_size; i++) {
-		assert_c_eq(((char *)ptr)[i], 0x5a,
+		assert_u_eq(((uint8_t *)ptr)[i], JEMALLOC_FREE_JUNK,
 		    "Missing junk fill for byte %zu/%zu of deallocated region",
 		    i, bin_info->reg_size);
 	}
@@ -44,7 +44,7 @@ arena_dalloc_junk_large_intercept(void *ptr, size_t usize)
 
 	arena_dalloc_junk_large_orig(ptr, usize);
 	for (i = 0; i < usize; i++) {
-		assert_c_eq(((char *)ptr)[i], 0x5a,
+		assert_u_eq(((uint8_t *)ptr)[i], JEMALLOC_FREE_JUNK,
 		    "Missing junk fill for byte %zu/%zu of deallocated region",
 		    i, usize);
 	}
@@ -69,7 +69,7 @@ huge_dalloc_junk_intercept(void *ptr, size_t usize)
 static void
 test_junk(size_t sz_min, size_t sz_max)
 {
-	char *s;
+	uint8_t *s;
 	size_t sz_prev, sz, i;
 
 	if (opt_junk_free) {
@@ -82,23 +82,23 @@ test_junk(size_t sz_min, size_t sz_max)
 	}
 
 	sz_prev = 0;
-	s = (char *)mallocx(sz_min, 0);
+	s = (uint8_t *)mallocx(sz_min, 0);
 	assert_ptr_not_null((void *)s, "Unexpected mallocx() failure");
 
 	for (sz = sallocx(s, 0); sz <= sz_max;
 	    sz_prev = sz, sz = sallocx(s, 0)) {
 		if (sz_prev > 0) {
-			assert_c_eq(s[0], 'a',
+			assert_u_eq(s[0], 'a',
 			    "Previously allocated byte %zu/%zu is corrupted",
 			    ZU(0), sz_prev);
-			assert_c_eq(s[sz_prev-1], 'a',
+			assert_u_eq(s[sz_prev-1], 'a',
 			    "Previously allocated byte %zu/%zu is corrupted",
 			    sz_prev-1, sz_prev);
 		}
 
 		for (i = sz_prev; i < sz; i++) {
 			if (opt_junk_alloc) {
-				assert_c_eq(s[i], 0xa5,
+				assert_u_eq(s[i], JEMALLOC_ALLOC_JUNK,
 				    "Newly allocated byte %zu/%zu isn't "
 				    "junk-filled", i, sz);
 			}
@@ -107,7 +107,7 @@ test_junk(size_t sz_min, size_t sz_max)
 
 		if (xallocx(s, sz+1, 0, 0) == sz) {
 			watch_junking(s);
-			s = (char *)rallocx(s, sz+1, 0);
+			s = (uint8_t *)rallocx(s, sz+1, 0);
 			assert_ptr_not_null((void *)s,
 			    "Unexpected rallocx() failure");
 			assert_true(!opt_junk_free || saw_junking,
@@ -244,7 +244,6 @@ int
 main(void)
 {
 
-	assert(!config_fill || opt_junk_alloc || opt_junk_free);
 	return (test(
 	    test_junk_small,
 	    test_junk_large,
diff --git a/deps/jemalloc/test/unit/junk_alloc.c b/deps/jemalloc/test/unit/junk_alloc.c
index 8db3331d2..a5895b5c0 100644
--- a/deps/jemalloc/test/unit/junk_alloc.c
+++ b/deps/jemalloc/test/unit/junk_alloc.c
@@ -1,3 +1,3 @@
-#define JEMALLOC_TEST_JUNK_OPT "junk:alloc"
+#define	JEMALLOC_TEST_JUNK_OPT "junk:alloc"
 #include "junk.c"
 #undef JEMALLOC_TEST_JUNK_OPT
diff --git a/deps/jemalloc/test/unit/junk_free.c b/deps/jemalloc/test/unit/junk_free.c
index 482a61d07..bb5183c90 100644
--- a/deps/jemalloc/test/unit/junk_free.c
+++ b/deps/jemalloc/test/unit/junk_free.c
@@ -1,3 +1,3 @@
-#define JEMALLOC_TEST_JUNK_OPT "junk:free"
+#define	JEMALLOC_TEST_JUNK_OPT "junk:free"
 #include "junk.c"
 #undef JEMALLOC_TEST_JUNK_OPT
diff --git a/deps/jemalloc/test/unit/mallctl.c b/deps/jemalloc/test/unit/mallctl.c
old mode 100644
new mode 100755
index 31e354ca7..2353c92c1
--- a/deps/jemalloc/test/unit/mallctl.c
+++ b/deps/jemalloc/test/unit/mallctl.c
@@ -12,16 +12,18 @@ TEST_BEGIN(test_mallctl_errors)
 	    EPERM, "mallctl() should return EPERM on attempt to write "
 	    "read-only value");
 
-	assert_d_eq(mallctl("epoch", NULL, NULL, &epoch, sizeof(epoch)-1),
-	    EINVAL, "mallctl() should return EINVAL for input size mismatch");
-	assert_d_eq(mallctl("epoch", NULL, NULL, &epoch, sizeof(epoch)+1),
-	    EINVAL, "mallctl() should return EINVAL for input size mismatch");
+	assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch,
+	    sizeof(epoch)-1), EINVAL,
+	    "mallctl() should return EINVAL for input size mismatch");
+	assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch,
+	    sizeof(epoch)+1), EINVAL,
+	    "mallctl() should return EINVAL for input size mismatch");
 
 	sz = sizeof(epoch)-1;
-	assert_d_eq(mallctl("epoch", &epoch, &sz, NULL, 0), EINVAL,
+	assert_d_eq(mallctl("epoch", (void *)&epoch, &sz, NULL, 0), EINVAL,
 	    "mallctl() should return EINVAL for output size mismatch");
 	sz = sizeof(epoch)+1;
-	assert_d_eq(mallctl("epoch", &epoch, &sz, NULL, 0), EINVAL,
+	assert_d_eq(mallctl("epoch", (void *)&epoch, &sz, NULL, 0), EINVAL,
 	    "mallctl() should return EINVAL for output size mismatch");
 }
 TEST_END
@@ -56,18 +58,20 @@ TEST_BEGIN(test_mallctlbymib_errors)
 	assert_d_eq(mallctlnametomib("epoch", mib, &miblen), 0,
 	    "Unexpected mallctlnametomib() failure");
 
-	assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, &epoch,
+	assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, (void *)&epoch,
 	    sizeof(epoch)-1), EINVAL,
 	    "mallctlbymib() should return EINVAL for input size mismatch");
-	assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, &epoch,
+	assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, (void *)&epoch,
 	    sizeof(epoch)+1), EINVAL,
 	    "mallctlbymib() should return EINVAL for input size mismatch");
 
 	sz = sizeof(epoch)-1;
-	assert_d_eq(mallctlbymib(mib, miblen, &epoch, &sz, NULL, 0), EINVAL,
+	assert_d_eq(mallctlbymib(mib, miblen, (void *)&epoch, &sz, NULL, 0),
+	    EINVAL,
 	    "mallctlbymib() should return EINVAL for output size mismatch");
 	sz = sizeof(epoch)+1;
-	assert_d_eq(mallctlbymib(mib, miblen, &epoch, &sz, NULL, 0), EINVAL,
+	assert_d_eq(mallctlbymib(mib, miblen, (void *)&epoch, &sz, NULL, 0),
+	    EINVAL,
 	    "mallctlbymib() should return EINVAL for output size mismatch");
 }
 TEST_END
@@ -83,18 +87,19 @@ TEST_BEGIN(test_mallctl_read_write)
 	assert_zu_eq(sz, sizeof(old_epoch), "Unexpected output size");
 
 	/* Read. */
-	assert_d_eq(mallctl("epoch", &old_epoch, &sz, NULL, 0), 0,
+	assert_d_eq(mallctl("epoch", (void *)&old_epoch, &sz, NULL, 0), 0,
 	    "Unexpected mallctl() failure");
 	assert_zu_eq(sz, sizeof(old_epoch), "Unexpected output size");
 
 	/* Write. */
-	assert_d_eq(mallctl("epoch", NULL, NULL, &new_epoch, sizeof(new_epoch)),
-	    0, "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&new_epoch,
+	    sizeof(new_epoch)), 0, "Unexpected mallctl() failure");
 	assert_zu_eq(sz, sizeof(old_epoch), "Unexpected output size");
 
 	/* Read+write. */
-	assert_d_eq(mallctl("epoch", &old_epoch, &sz, &new_epoch,
-	    sizeof(new_epoch)), 0, "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("epoch", (void *)&old_epoch, &sz,
+	    (void *)&new_epoch, sizeof(new_epoch)), 0,
+	    "Unexpected mallctl() failure");
 	assert_zu_eq(sz, sizeof(old_epoch), "Unexpected output size");
 }
 TEST_END
@@ -117,29 +122,30 @@ TEST_END
 TEST_BEGIN(test_mallctl_config)
 {
 
-#define	TEST_MALLCTL_CONFIG(config) do {				\
-	bool oldval;							\
+#define	TEST_MALLCTL_CONFIG(config, t) do {				\
+	t oldval;							\
 	size_t sz = sizeof(oldval);					\
-	assert_d_eq(mallctl("config."#config, &oldval, &sz, NULL, 0),	\
-	    0, "Unexpected mallctl() failure");				\
+	assert_d_eq(mallctl("config."#config, (void *)&oldval, &sz,	\
+	    NULL, 0), 0, "Unexpected mallctl() failure");		\
 	assert_b_eq(oldval, config_##config, "Incorrect config value");	\
 	assert_zu_eq(sz, sizeof(oldval), "Unexpected output size");	\
 } while (0)
 
-	TEST_MALLCTL_CONFIG(cache_oblivious);
-	TEST_MALLCTL_CONFIG(debug);
-	TEST_MALLCTL_CONFIG(fill);
-	TEST_MALLCTL_CONFIG(lazy_lock);
-	TEST_MALLCTL_CONFIG(munmap);
-	TEST_MALLCTL_CONFIG(prof);
-	TEST_MALLCTL_CONFIG(prof_libgcc);
-	TEST_MALLCTL_CONFIG(prof_libunwind);
-	TEST_MALLCTL_CONFIG(stats);
-	TEST_MALLCTL_CONFIG(tcache);
-	TEST_MALLCTL_CONFIG(tls);
-	TEST_MALLCTL_CONFIG(utrace);
-	TEST_MALLCTL_CONFIG(valgrind);
-	TEST_MALLCTL_CONFIG(xmalloc);
+	TEST_MALLCTL_CONFIG(cache_oblivious, bool);
+	TEST_MALLCTL_CONFIG(debug, bool);
+	TEST_MALLCTL_CONFIG(fill, bool);
+	TEST_MALLCTL_CONFIG(lazy_lock, bool);
+	TEST_MALLCTL_CONFIG(malloc_conf, const char *);
+	TEST_MALLCTL_CONFIG(munmap, bool);
+	TEST_MALLCTL_CONFIG(prof, bool);
+	TEST_MALLCTL_CONFIG(prof_libgcc, bool);
+	TEST_MALLCTL_CONFIG(prof_libunwind, bool);
+	TEST_MALLCTL_CONFIG(stats, bool);
+	TEST_MALLCTL_CONFIG(tcache, bool);
+	TEST_MALLCTL_CONFIG(tls, bool);
+	TEST_MALLCTL_CONFIG(utrace, bool);
+	TEST_MALLCTL_CONFIG(valgrind, bool);
+	TEST_MALLCTL_CONFIG(xmalloc, bool);
 
 #undef TEST_MALLCTL_CONFIG
 }
@@ -153,7 +159,8 @@ TEST_BEGIN(test_mallctl_opt)
 	t oldval;							\
 	size_t sz = sizeof(oldval);					\
 	int expected = config_##config ? 0 : ENOENT;			\
-	int result = mallctl("opt."#opt, &oldval, &sz, NULL, 0);	\
+	int result = mallctl("opt."#opt, (void *)&oldval, &sz, NULL,	\
+	    0);								\
 	assert_d_eq(result, expected,					\
 	    "Unexpected mallctl() result for opt."#opt);		\
 	assert_zu_eq(sz, sizeof(oldval), "Unexpected output size");	\
@@ -162,8 +169,10 @@ TEST_BEGIN(test_mallctl_opt)
 	TEST_MALLCTL_OPT(bool, abort, always);
 	TEST_MALLCTL_OPT(size_t, lg_chunk, always);
 	TEST_MALLCTL_OPT(const char *, dss, always);
-	TEST_MALLCTL_OPT(size_t, narenas, always);
+	TEST_MALLCTL_OPT(unsigned, narenas, always);
+	TEST_MALLCTL_OPT(const char *, purge, always);
 	TEST_MALLCTL_OPT(ssize_t, lg_dirty_mult, always);
+	TEST_MALLCTL_OPT(ssize_t, decay_time, always);
 	TEST_MALLCTL_OPT(bool, stats_print, always);
 	TEST_MALLCTL_OPT(const char *, junk, fill);
 	TEST_MALLCTL_OPT(size_t, quarantine, fill);
@@ -194,7 +203,7 @@ TEST_BEGIN(test_manpage_example)
 	size_t len, miblen;
 
 	len = sizeof(nbins);
-	assert_d_eq(mallctl("arenas.nbins", &nbins, &len, NULL, 0), 0,
+	assert_d_eq(mallctl("arenas.nbins", (void *)&nbins, &len, NULL, 0), 0,
 	    "Unexpected mallctl() failure");
 
 	miblen = 4;
@@ -205,8 +214,8 @@ TEST_BEGIN(test_manpage_example)
 
 		mib[2] = i;
 		len = sizeof(bin_size);
-		assert_d_eq(mallctlbymib(mib, miblen, &bin_size, &len, NULL, 0),
-		    0, "Unexpected mallctlbymib() failure");
+		assert_d_eq(mallctlbymib(mib, miblen, (void *)&bin_size, &len,
+		    NULL, 0), 0, "Unexpected mallctlbymib() failure");
 		/* Do something with bin_size... */
 	}
 }
@@ -255,25 +264,25 @@ TEST_BEGIN(test_tcache)
 	/* Create tcaches. */
 	for (i = 0; i < NTCACHES; i++) {
 		sz = sizeof(unsigned);
-		assert_d_eq(mallctl("tcache.create", &tis[i], &sz, NULL, 0), 0,
-		    "Unexpected mallctl() failure, i=%u", i);
+		assert_d_eq(mallctl("tcache.create", (void *)&tis[i], &sz, NULL,
+		    0), 0, "Unexpected mallctl() failure, i=%u", i);
 	}
 
 	/* Exercise tcache ID recycling. */
 	for (i = 0; i < NTCACHES; i++) {
-		assert_d_eq(mallctl("tcache.destroy", NULL, NULL, &tis[i],
-		    sizeof(unsigned)), 0, "Unexpected mallctl() failure, i=%u",
-		    i);
+		assert_d_eq(mallctl("tcache.destroy", NULL, NULL,
+		    (void *)&tis[i], sizeof(unsigned)), 0,
+		    "Unexpected mallctl() failure, i=%u", i);
 	}
 	for (i = 0; i < NTCACHES; i++) {
 		sz = sizeof(unsigned);
-		assert_d_eq(mallctl("tcache.create", &tis[i], &sz, NULL, 0), 0,
-		    "Unexpected mallctl() failure, i=%u", i);
+		assert_d_eq(mallctl("tcache.create", (void *)&tis[i], &sz, NULL,
+		    0), 0, "Unexpected mallctl() failure, i=%u", i);
 	}
 
 	/* Flush empty tcaches. */
 	for (i = 0; i < NTCACHES; i++) {
-		assert_d_eq(mallctl("tcache.flush", NULL, NULL, &tis[i],
+		assert_d_eq(mallctl("tcache.flush", NULL, NULL, (void *)&tis[i],
 		    sizeof(unsigned)), 0, "Unexpected mallctl() failure, i=%u",
 		    i);
 	}
@@ -318,16 +327,16 @@ TEST_BEGIN(test_tcache)
 
 	/* Flush some non-empty tcaches. */
 	for (i = 0; i < NTCACHES/2; i++) {
-		assert_d_eq(mallctl("tcache.flush", NULL, NULL, &tis[i],
+		assert_d_eq(mallctl("tcache.flush", NULL, NULL, (void *)&tis[i],
 		    sizeof(unsigned)), 0, "Unexpected mallctl() failure, i=%u",
 		    i);
 	}
 
 	/* Destroy tcaches. */
 	for (i = 0; i < NTCACHES; i++) {
-		assert_d_eq(mallctl("tcache.destroy", NULL, NULL, &tis[i],
-		    sizeof(unsigned)), 0, "Unexpected mallctl() failure, i=%u",
-		    i);
+		assert_d_eq(mallctl("tcache.destroy", NULL, NULL,
+		    (void *)&tis[i], sizeof(unsigned)), 0,
+		    "Unexpected mallctl() failure, i=%u", i);
 	}
 }
 TEST_END
@@ -337,15 +346,17 @@ TEST_BEGIN(test_thread_arena)
 	unsigned arena_old, arena_new, narenas;
 	size_t sz = sizeof(unsigned);
 
-	assert_d_eq(mallctl("arenas.narenas", &narenas, &sz, NULL, 0), 0,
-	    "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("arenas.narenas", (void *)&narenas, &sz, NULL, 0),
+	    0, "Unexpected mallctl() failure");
 	assert_u_eq(narenas, opt_narenas, "Number of arenas incorrect");
 	arena_new = narenas - 1;
-	assert_d_eq(mallctl("thread.arena", &arena_old, &sz, &arena_new,
-	    sizeof(unsigned)), 0, "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("thread.arena", (void *)&arena_old, &sz,
+	    (void *)&arena_new, sizeof(unsigned)), 0,
+	    "Unexpected mallctl() failure");
 	arena_new = 0;
-	assert_d_eq(mallctl("thread.arena", &arena_old, &sz, &arena_new,
-	    sizeof(unsigned)), 0, "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("thread.arena", (void *)&arena_old, &sz,
+	    (void *)&arena_new, sizeof(unsigned)), 0,
+	    "Unexpected mallctl() failure");
 }
 TEST_END
 
@@ -354,17 +365,20 @@ TEST_BEGIN(test_arena_i_lg_dirty_mult)
 	ssize_t lg_dirty_mult, orig_lg_dirty_mult, prev_lg_dirty_mult;
 	size_t sz = sizeof(ssize_t);
 
-	assert_d_eq(mallctl("arena.0.lg_dirty_mult", &orig_lg_dirty_mult, &sz,
-	    NULL, 0), 0, "Unexpected mallctl() failure");
+	test_skip_if(opt_purge != purge_mode_ratio);
+
+	assert_d_eq(mallctl("arena.0.lg_dirty_mult",
+	    (void *)&orig_lg_dirty_mult, &sz, NULL, 0), 0,
+	    "Unexpected mallctl() failure");
 
 	lg_dirty_mult = -2;
 	assert_d_eq(mallctl("arena.0.lg_dirty_mult", NULL, NULL,
-	    &lg_dirty_mult, sizeof(ssize_t)), EFAULT,
+	    (void *)&lg_dirty_mult, sizeof(ssize_t)), EFAULT,
 	    "Unexpected mallctl() success");
 
 	lg_dirty_mult = (sizeof(size_t) << 3);
 	assert_d_eq(mallctl("arena.0.lg_dirty_mult", NULL, NULL,
-	    &lg_dirty_mult, sizeof(ssize_t)), EFAULT,
+	    (void *)&lg_dirty_mult, sizeof(ssize_t)), EFAULT,
 	    "Unexpected mallctl() success");
 
 	for (prev_lg_dirty_mult = orig_lg_dirty_mult, lg_dirty_mult = -1;
@@ -372,15 +386,48 @@ TEST_BEGIN(test_arena_i_lg_dirty_mult)
 	    = lg_dirty_mult, lg_dirty_mult++) {
 		ssize_t old_lg_dirty_mult;
 
-		assert_d_eq(mallctl("arena.0.lg_dirty_mult", &old_lg_dirty_mult,
-		    &sz, &lg_dirty_mult, sizeof(ssize_t)), 0,
-		    "Unexpected mallctl() failure");
+		assert_d_eq(mallctl("arena.0.lg_dirty_mult",
+		    (void *)&old_lg_dirty_mult, &sz, (void *)&lg_dirty_mult,
+		    sizeof(ssize_t)), 0, "Unexpected mallctl() failure");
 		assert_zd_eq(old_lg_dirty_mult, prev_lg_dirty_mult,
 		    "Unexpected old arena.0.lg_dirty_mult");
 	}
 }
 TEST_END
 
+TEST_BEGIN(test_arena_i_decay_time)
+{
+	ssize_t decay_time, orig_decay_time, prev_decay_time;
+	size_t sz = sizeof(ssize_t);
+
+	test_skip_if(opt_purge != purge_mode_decay);
+
+	assert_d_eq(mallctl("arena.0.decay_time", (void *)&orig_decay_time, &sz,
+	    NULL, 0), 0, "Unexpected mallctl() failure");
+
+	decay_time = -2;
+	assert_d_eq(mallctl("arena.0.decay_time", NULL, NULL,
+	    (void *)&decay_time, sizeof(ssize_t)), EFAULT,
+	    "Unexpected mallctl() success");
+
+	decay_time = 0x7fffffff;
+	assert_d_eq(mallctl("arena.0.decay_time", NULL, NULL,
+	    (void *)&decay_time, sizeof(ssize_t)), 0,
+	    "Unexpected mallctl() failure");
+
+	for (prev_decay_time = decay_time, decay_time = -1;
+	    decay_time < 20; prev_decay_time = decay_time, decay_time++) {
+		ssize_t old_decay_time;
+
+		assert_d_eq(mallctl("arena.0.decay_time", (void *)&old_decay_time,
+		    &sz, (void *)&decay_time, sizeof(ssize_t)), 0,
+		    "Unexpected mallctl() failure");
+		assert_zd_eq(old_decay_time, prev_decay_time,
+		    "Unexpected old arena.0.decay_time");
+	}
+}
+TEST_END
+
 TEST_BEGIN(test_arena_i_purge)
 {
 	unsigned narenas;
@@ -391,8 +438,8 @@ TEST_BEGIN(test_arena_i_purge)
 	assert_d_eq(mallctl("arena.0.purge", NULL, NULL, NULL, 0), 0,
 	    "Unexpected mallctl() failure");
 
-	assert_d_eq(mallctl("arenas.narenas", &narenas, &sz, NULL, 0), 0,
-	    "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("arenas.narenas", (void *)&narenas, &sz, NULL, 0),
+	    0, "Unexpected mallctl() failure");
 	assert_d_eq(mallctlnametomib("arena.0.purge", mib, &miblen), 0,
 	    "Unexpected mallctlnametomib() failure");
 	mib[1] = narenas;
@@ -401,6 +448,26 @@ TEST_BEGIN(test_arena_i_purge)
 }
 TEST_END
 
+TEST_BEGIN(test_arena_i_decay)
+{
+	unsigned narenas;
+	size_t sz = sizeof(unsigned);
+	size_t mib[3];
+	size_t miblen = 3;
+
+	assert_d_eq(mallctl("arena.0.decay", NULL, NULL, NULL, 0), 0,
+	    "Unexpected mallctl() failure");
+
+	assert_d_eq(mallctl("arenas.narenas", (void *)&narenas, &sz, NULL, 0),
+	    0, "Unexpected mallctl() failure");
+	assert_d_eq(mallctlnametomib("arena.0.decay", mib, &miblen), 0,
+	    "Unexpected mallctlnametomib() failure");
+	mib[1] = narenas;
+	assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, NULL, 0), 0,
+	    "Unexpected mallctlbymib() failure");
+}
+TEST_END
+
 TEST_BEGIN(test_arena_i_dss)
 {
 	const char *dss_prec_old, *dss_prec_new;
@@ -413,31 +480,35 @@ TEST_BEGIN(test_arena_i_dss)
 	    "Unexpected mallctlnametomib() error");
 
 	dss_prec_new = "disabled";
-	assert_d_eq(mallctlbymib(mib, miblen, &dss_prec_old, &sz, &dss_prec_new,
-	    sizeof(dss_prec_new)), 0, "Unexpected mallctl() failure");
+	assert_d_eq(mallctlbymib(mib, miblen, (void *)&dss_prec_old, &sz,
+	    (void *)&dss_prec_new, sizeof(dss_prec_new)), 0,
+	    "Unexpected mallctl() failure");
 	assert_str_ne(dss_prec_old, "primary",
 	    "Unexpected default for dss precedence");
 
-	assert_d_eq(mallctlbymib(mib, miblen, &dss_prec_new, &sz, &dss_prec_old,
-	    sizeof(dss_prec_old)), 0, "Unexpected mallctl() failure");
-
-	assert_d_eq(mallctlbymib(mib, miblen, &dss_prec_old, &sz, NULL, 0), 0,
+	assert_d_eq(mallctlbymib(mib, miblen, (void *)&dss_prec_new, &sz,
+	    (void *)&dss_prec_old, sizeof(dss_prec_old)), 0,
 	    "Unexpected mallctl() failure");
+
+	assert_d_eq(mallctlbymib(mib, miblen, (void *)&dss_prec_old, &sz, NULL,
+	    0), 0, "Unexpected mallctl() failure");
 	assert_str_ne(dss_prec_old, "primary",
 	    "Unexpected value for dss precedence");
 
 	mib[1] = narenas_total_get();
 	dss_prec_new = "disabled";
-	assert_d_eq(mallctlbymib(mib, miblen, &dss_prec_old, &sz, &dss_prec_new,
-	    sizeof(dss_prec_new)), 0, "Unexpected mallctl() failure");
+	assert_d_eq(mallctlbymib(mib, miblen, (void *)&dss_prec_old, &sz,
+	    (void *)&dss_prec_new, sizeof(dss_prec_new)), 0,
+	    "Unexpected mallctl() failure");
 	assert_str_ne(dss_prec_old, "primary",
 	    "Unexpected default for dss precedence");
 
-	assert_d_eq(mallctlbymib(mib, miblen, &dss_prec_new, &sz, &dss_prec_old,
-	    sizeof(dss_prec_new)), 0, "Unexpected mallctl() failure");
-
-	assert_d_eq(mallctlbymib(mib, miblen, &dss_prec_old, &sz, NULL, 0), 0,
+	assert_d_eq(mallctlbymib(mib, miblen, (void *)&dss_prec_new, &sz,
+	    (void *)&dss_prec_old, sizeof(dss_prec_new)), 0,
 	    "Unexpected mallctl() failure");
+
+	assert_d_eq(mallctlbymib(mib, miblen, (void *)&dss_prec_old, &sz, NULL,
+	    0), 0, "Unexpected mallctl() failure");
 	assert_str_ne(dss_prec_old, "primary",
 	    "Unexpected value for dss precedence");
 }
@@ -448,14 +519,14 @@ TEST_BEGIN(test_arenas_initialized)
 	unsigned narenas;
 	size_t sz = sizeof(narenas);
 
-	assert_d_eq(mallctl("arenas.narenas", &narenas, &sz, NULL, 0), 0,
-	    "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("arenas.narenas", (void *)&narenas, &sz, NULL, 0),
+	    0, "Unexpected mallctl() failure");
 	{
 		VARIABLE_ARRAY(bool, initialized, narenas);
 
 		sz = narenas * sizeof(bool);
-		assert_d_eq(mallctl("arenas.initialized", initialized, &sz,
-		    NULL, 0), 0, "Unexpected mallctl() failure");
+		assert_d_eq(mallctl("arenas.initialized", (void *)initialized,
+		    &sz, NULL, 0), 0, "Unexpected mallctl() failure");
 	}
 }
 TEST_END
@@ -465,17 +536,19 @@ TEST_BEGIN(test_arenas_lg_dirty_mult)
 	ssize_t lg_dirty_mult, orig_lg_dirty_mult, prev_lg_dirty_mult;
 	size_t sz = sizeof(ssize_t);
 
-	assert_d_eq(mallctl("arenas.lg_dirty_mult", &orig_lg_dirty_mult, &sz,
-	    NULL, 0), 0, "Unexpected mallctl() failure");
+	test_skip_if(opt_purge != purge_mode_ratio);
+
+	assert_d_eq(mallctl("arenas.lg_dirty_mult", (void *)&orig_lg_dirty_mult,
+	    &sz, NULL, 0), 0, "Unexpected mallctl() failure");
 
 	lg_dirty_mult = -2;
 	assert_d_eq(mallctl("arenas.lg_dirty_mult", NULL, NULL,
-	    &lg_dirty_mult, sizeof(ssize_t)), EFAULT,
+	    (void *)&lg_dirty_mult, sizeof(ssize_t)), EFAULT,
 	    "Unexpected mallctl() success");
 
 	lg_dirty_mult = (sizeof(size_t) << 3);
 	assert_d_eq(mallctl("arenas.lg_dirty_mult", NULL, NULL,
-	    &lg_dirty_mult, sizeof(ssize_t)), EFAULT,
+	    (void *)&lg_dirty_mult, sizeof(ssize_t)), EFAULT,
 	    "Unexpected mallctl() success");
 
 	for (prev_lg_dirty_mult = orig_lg_dirty_mult, lg_dirty_mult = -1;
@@ -483,23 +556,56 @@ TEST_BEGIN(test_arenas_lg_dirty_mult)
 	    lg_dirty_mult, lg_dirty_mult++) {
 		ssize_t old_lg_dirty_mult;
 
-		assert_d_eq(mallctl("arenas.lg_dirty_mult", &old_lg_dirty_mult,
-		    &sz, &lg_dirty_mult, sizeof(ssize_t)), 0,
-		    "Unexpected mallctl() failure");
+		assert_d_eq(mallctl("arenas.lg_dirty_mult",
+		    (void *)&old_lg_dirty_mult, &sz, (void *)&lg_dirty_mult,
+		    sizeof(ssize_t)), 0, "Unexpected mallctl() failure");
 		assert_zd_eq(old_lg_dirty_mult, prev_lg_dirty_mult,
 		    "Unexpected old arenas.lg_dirty_mult");
 	}
 }
 TEST_END
 
+TEST_BEGIN(test_arenas_decay_time)
+{
+	ssize_t decay_time, orig_decay_time, prev_decay_time;
+	size_t sz = sizeof(ssize_t);
+
+	test_skip_if(opt_purge != purge_mode_decay);
+
+	assert_d_eq(mallctl("arenas.decay_time", (void *)&orig_decay_time, &sz,
+	    NULL, 0), 0, "Unexpected mallctl() failure");
+
+	decay_time = -2;
+	assert_d_eq(mallctl("arenas.decay_time", NULL, NULL,
+	    (void *)&decay_time, sizeof(ssize_t)), EFAULT,
+	    "Unexpected mallctl() success");
+
+	decay_time = 0x7fffffff;
+	assert_d_eq(mallctl("arenas.decay_time", NULL, NULL,
+	    (void *)&decay_time, sizeof(ssize_t)), 0,
+	    "Expected mallctl() failure");
+
+	for (prev_decay_time = decay_time, decay_time = -1;
+	    decay_time < 20; prev_decay_time = decay_time, decay_time++) {
+		ssize_t old_decay_time;
+
+		assert_d_eq(mallctl("arenas.decay_time",
+		    (void *)&old_decay_time, &sz, (void *)&decay_time,
+		    sizeof(ssize_t)), 0, "Unexpected mallctl() failure");
+		assert_zd_eq(old_decay_time, prev_decay_time,
+		    "Unexpected old arenas.decay_time");
+	}
+}
+TEST_END
+
 TEST_BEGIN(test_arenas_constants)
 {
 
 #define	TEST_ARENAS_CONSTANT(t, name, expected) do {			\
 	t name;								\
 	size_t sz = sizeof(t);						\
-	assert_d_eq(mallctl("arenas."#name, &name, &sz, NULL, 0), 0,	\
-	    "Unexpected mallctl() failure");				\
+	assert_d_eq(mallctl("arenas."#name, (void *)&name, &sz, NULL,	\
+	    0), 0, "Unexpected mallctl() failure");			\
 	assert_zu_eq(name, expected, "Incorrect "#name" size");		\
 } while (0)
 
@@ -519,8 +625,8 @@ TEST_BEGIN(test_arenas_bin_constants)
 #define	TEST_ARENAS_BIN_CONSTANT(t, name, expected) do {		\
 	t name;								\
 	size_t sz = sizeof(t);						\
-	assert_d_eq(mallctl("arenas.bin.0."#name, &name, &sz, NULL, 0),	\
-	    0, "Unexpected mallctl() failure");				\
+	assert_d_eq(mallctl("arenas.bin.0."#name, (void *)&name, &sz,	\
+	    NULL, 0), 0, "Unexpected mallctl() failure");		\
 	assert_zu_eq(name, expected, "Incorrect "#name" size");		\
 } while (0)
 
@@ -538,8 +644,8 @@ TEST_BEGIN(test_arenas_lrun_constants)
 #define	TEST_ARENAS_LRUN_CONSTANT(t, name, expected) do {		\
 	t name;								\
 	size_t sz = sizeof(t);						\
-	assert_d_eq(mallctl("arenas.lrun.0."#name, &name, &sz, NULL,	\
-	    0), 0, "Unexpected mallctl() failure");			\
+	assert_d_eq(mallctl("arenas.lrun.0."#name, (void *)&name, &sz,	\
+	    NULL, 0), 0, "Unexpected mallctl() failure");		\
 	assert_zu_eq(name, expected, "Incorrect "#name" size");		\
 } while (0)
 
@@ -555,8 +661,8 @@ TEST_BEGIN(test_arenas_hchunk_constants)
 #define	TEST_ARENAS_HCHUNK_CONSTANT(t, name, expected) do {		\
 	t name;								\
 	size_t sz = sizeof(t);						\
-	assert_d_eq(mallctl("arenas.hchunk.0."#name, &name, &sz, NULL,	\
-	    0), 0, "Unexpected mallctl() failure");			\
+	assert_d_eq(mallctl("arenas.hchunk.0."#name, (void *)&name,	\
+	    &sz, NULL, 0), 0, "Unexpected mallctl() failure");		\
 	assert_zu_eq(name, expected, "Incorrect "#name" size");		\
 } while (0)
 
@@ -571,12 +677,12 @@ TEST_BEGIN(test_arenas_extend)
 	unsigned narenas_before, arena, narenas_after;
 	size_t sz = sizeof(unsigned);
 
-	assert_d_eq(mallctl("arenas.narenas", &narenas_before, &sz, NULL, 0), 0,
-	    "Unexpected mallctl() failure");
-	assert_d_eq(mallctl("arenas.extend", &arena, &sz, NULL, 0), 0,
-	    "Unexpected mallctl() failure");
-	assert_d_eq(mallctl("arenas.narenas", &narenas_after, &sz, NULL, 0), 0,
+	assert_d_eq(mallctl("arenas.narenas", (void *)&narenas_before, &sz,
+	    NULL, 0), 0, "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("arenas.extend", (void *)&arena, &sz, NULL, 0), 0,
 	    "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("arenas.narenas", (void *)&narenas_after, &sz, NULL,
+	    0), 0, "Unexpected mallctl() failure");
 
 	assert_u_eq(narenas_before+1, narenas_after,
 	    "Unexpected number of arenas before versus after extension");
@@ -590,12 +696,14 @@ TEST_BEGIN(test_stats_arenas)
 #define	TEST_STATS_ARENAS(t, name) do {					\
 	t name;								\
 	size_t sz = sizeof(t);						\
-	assert_d_eq(mallctl("stats.arenas.0."#name, &name, &sz, NULL,	\
-	    0), 0, "Unexpected mallctl() failure");			\
+	assert_d_eq(mallctl("stats.arenas.0."#name, (void *)&name, &sz,	\
+	    NULL, 0), 0, "Unexpected mallctl() failure");		\
 } while (0)
 
-	TEST_STATS_ARENAS(const char *, dss);
 	TEST_STATS_ARENAS(unsigned, nthreads);
+	TEST_STATS_ARENAS(const char *, dss);
+	TEST_STATS_ARENAS(ssize_t, lg_dirty_mult);
+	TEST_STATS_ARENAS(ssize_t, decay_time);
 	TEST_STATS_ARENAS(size_t, pactive);
 	TEST_STATS_ARENAS(size_t, pdirty);
 
@@ -620,10 +728,13 @@ main(void)
 	    test_tcache,
 	    test_thread_arena,
 	    test_arena_i_lg_dirty_mult,
+	    test_arena_i_decay_time,
 	    test_arena_i_purge,
+	    test_arena_i_decay,
 	    test_arena_i_dss,
 	    test_arenas_initialized,
 	    test_arenas_lg_dirty_mult,
+	    test_arenas_decay_time,
 	    test_arenas_constants,
 	    test_arenas_bin_constants,
 	    test_arenas_lrun_constants,
diff --git a/deps/jemalloc/test/unit/math.c b/deps/jemalloc/test/unit/math.c
index ebec77a62..adb72bed9 100644
--- a/deps/jemalloc/test/unit/math.c
+++ b/deps/jemalloc/test/unit/math.c
@@ -5,6 +5,10 @@
 
 #include <float.h>
 
+#ifdef __PGI
+#undef INFINITY
+#endif
+
 #ifndef INFINITY
 #define	INFINITY (DBL_MAX + DBL_MAX)
 #endif
diff --git a/deps/jemalloc/test/unit/nstime.c b/deps/jemalloc/test/unit/nstime.c
new file mode 100644
index 000000000..0368bc26e
--- /dev/null
+++ b/deps/jemalloc/test/unit/nstime.c
@@ -0,0 +1,227 @@
+#include "test/jemalloc_test.h"
+
+#define	BILLION	UINT64_C(1000000000)
+
+TEST_BEGIN(test_nstime_init)
+{
+	nstime_t nst;
+
+	nstime_init(&nst, 42000000043);
+	assert_u64_eq(nstime_ns(&nst), 42000000043, "ns incorrectly read");
+	assert_u64_eq(nstime_sec(&nst), 42, "sec incorrectly read");
+	assert_u64_eq(nstime_nsec(&nst), 43, "nsec incorrectly read");
+}
+TEST_END
+
+TEST_BEGIN(test_nstime_init2)
+{
+	nstime_t nst;
+
+	nstime_init2(&nst, 42, 43);
+	assert_u64_eq(nstime_sec(&nst), 42, "sec incorrectly read");
+	assert_u64_eq(nstime_nsec(&nst), 43, "nsec incorrectly read");
+}
+TEST_END
+
+TEST_BEGIN(test_nstime_copy)
+{
+	nstime_t nsta, nstb;
+
+	nstime_init2(&nsta, 42, 43);
+	nstime_init(&nstb, 0);
+	nstime_copy(&nstb, &nsta);
+	assert_u64_eq(nstime_sec(&nstb), 42, "sec incorrectly copied");
+	assert_u64_eq(nstime_nsec(&nstb), 43, "nsec incorrectly copied");
+}
+TEST_END
+
+TEST_BEGIN(test_nstime_compare)
+{
+	nstime_t nsta, nstb;
+
+	nstime_init2(&nsta, 42, 43);
+	nstime_copy(&nstb, &nsta);
+	assert_d_eq(nstime_compare(&nsta, &nstb), 0, "Times should be equal");
+	assert_d_eq(nstime_compare(&nstb, &nsta), 0, "Times should be equal");
+
+	nstime_init2(&nstb, 42, 42);
+	assert_d_eq(nstime_compare(&nsta, &nstb), 1,
+	    "nsta should be greater than nstb");
+	assert_d_eq(nstime_compare(&nstb, &nsta), -1,
+	    "nstb should be less than nsta");
+
+	nstime_init2(&nstb, 42, 44);
+	assert_d_eq(nstime_compare(&nsta, &nstb), -1,
+	    "nsta should be less than nstb");
+	assert_d_eq(nstime_compare(&nstb, &nsta), 1,
+	    "nstb should be greater than nsta");
+
+	nstime_init2(&nstb, 41, BILLION - 1);
+	assert_d_eq(nstime_compare(&nsta, &nstb), 1,
+	    "nsta should be greater than nstb");
+	assert_d_eq(nstime_compare(&nstb, &nsta), -1,
+	    "nstb should be less than nsta");
+
+	nstime_init2(&nstb, 43, 0);
+	assert_d_eq(nstime_compare(&nsta, &nstb), -1,
+	    "nsta should be less than nstb");
+	assert_d_eq(nstime_compare(&nstb, &nsta), 1,
+	    "nstb should be greater than nsta");
+}
+TEST_END
+
+TEST_BEGIN(test_nstime_add)
+{
+	nstime_t nsta, nstb;
+
+	nstime_init2(&nsta, 42, 43);
+	nstime_copy(&nstb, &nsta);
+	nstime_add(&nsta, &nstb);
+	nstime_init2(&nstb, 84, 86);
+	assert_d_eq(nstime_compare(&nsta, &nstb), 0,
+	    "Incorrect addition result");
+
+	nstime_init2(&nsta, 42, BILLION - 1);
+	nstime_copy(&nstb, &nsta);
+	nstime_add(&nsta, &nstb);
+	nstime_init2(&nstb, 85, BILLION - 2);
+	assert_d_eq(nstime_compare(&nsta, &nstb), 0,
+	    "Incorrect addition result");
+}
+TEST_END
+
+TEST_BEGIN(test_nstime_subtract)
+{
+	nstime_t nsta, nstb;
+
+	nstime_init2(&nsta, 42, 43);
+	nstime_copy(&nstb, &nsta);
+	nstime_subtract(&nsta, &nstb);
+	nstime_init(&nstb, 0);
+	assert_d_eq(nstime_compare(&nsta, &nstb), 0,
+	    "Incorrect subtraction result");
+
+	nstime_init2(&nsta, 42, 43);
+	nstime_init2(&nstb, 41, 44);
+	nstime_subtract(&nsta, &nstb);
+	nstime_init2(&nstb, 0, BILLION - 1);
+	assert_d_eq(nstime_compare(&nsta, &nstb), 0,
+	    "Incorrect subtraction result");
+}
+TEST_END
+
+TEST_BEGIN(test_nstime_imultiply)
+{
+	nstime_t nsta, nstb;
+
+	nstime_init2(&nsta, 42, 43);
+	nstime_imultiply(&nsta, 10);
+	nstime_init2(&nstb, 420, 430);
+	assert_d_eq(nstime_compare(&nsta, &nstb), 0,
+	    "Incorrect multiplication result");
+
+	nstime_init2(&nsta, 42, 666666666);
+	nstime_imultiply(&nsta, 3);
+	nstime_init2(&nstb, 127, 999999998);
+	assert_d_eq(nstime_compare(&nsta, &nstb), 0,
+	    "Incorrect multiplication result");
+}
+TEST_END
+
+TEST_BEGIN(test_nstime_idivide)
+{
+	nstime_t nsta, nstb;
+
+	nstime_init2(&nsta, 42, 43);
+	nstime_copy(&nstb, &nsta);
+	nstime_imultiply(&nsta, 10);
+	nstime_idivide(&nsta, 10);
+	assert_d_eq(nstime_compare(&nsta, &nstb), 0,
+	    "Incorrect division result");
+
+	nstime_init2(&nsta, 42, 666666666);
+	nstime_copy(&nstb, &nsta);
+	nstime_imultiply(&nsta, 3);
+	nstime_idivide(&nsta, 3);
+	assert_d_eq(nstime_compare(&nsta, &nstb), 0,
+	    "Incorrect division result");
+}
+TEST_END
+
+TEST_BEGIN(test_nstime_divide)
+{
+	nstime_t nsta, nstb, nstc;
+
+	nstime_init2(&nsta, 42, 43);
+	nstime_copy(&nstb, &nsta);
+	nstime_imultiply(&nsta, 10);
+	assert_u64_eq(nstime_divide(&nsta, &nstb), 10,
+	    "Incorrect division result");
+
+	nstime_init2(&nsta, 42, 43);
+	nstime_copy(&nstb, &nsta);
+	nstime_imultiply(&nsta, 10);
+	nstime_init(&nstc, 1);
+	nstime_add(&nsta, &nstc);
+	assert_u64_eq(nstime_divide(&nsta, &nstb), 10,
+	    "Incorrect division result");
+
+	nstime_init2(&nsta, 42, 43);
+	nstime_copy(&nstb, &nsta);
+	nstime_imultiply(&nsta, 10);
+	nstime_init(&nstc, 1);
+	nstime_subtract(&nsta, &nstc);
+	assert_u64_eq(nstime_divide(&nsta, &nstb), 9,
+	    "Incorrect division result");
+}
+TEST_END
+
+TEST_BEGIN(test_nstime_monotonic)
+{
+
+	nstime_monotonic();
+}
+TEST_END
+
+TEST_BEGIN(test_nstime_update)
+{
+	nstime_t nst;
+
+	nstime_init(&nst, 0);
+
+	assert_false(nstime_update(&nst), "Basic time update failed.");
+
+	/* Only Rip Van Winkle sleeps this long. */
+	{
+		nstime_t addend;
+		nstime_init2(&addend, 631152000, 0);
+		nstime_add(&nst, &addend);
+	}
+	{
+		nstime_t nst0;
+		nstime_copy(&nst0, &nst);
+		assert_true(nstime_update(&nst),
+		    "Update should detect time roll-back.");
+		assert_d_eq(nstime_compare(&nst, &nst0), 0,
+		    "Time should not have been modified");
+	}
+}
+TEST_END
+
+int
+main(void)
+{
+
+	return (test(
+	    test_nstime_init,
+	    test_nstime_init2,
+	    test_nstime_copy,
+	    test_nstime_compare,
+	    test_nstime_add,
+	    test_nstime_subtract,
+	    test_nstime_imultiply,
+	    test_nstime_idivide,
+	    test_nstime_divide,
+	    test_nstime_monotonic,
+	    test_nstime_update));
+}
diff --git a/deps/jemalloc/test/unit/pack.c b/deps/jemalloc/test/unit/pack.c
new file mode 100644
index 000000000..0b6ffcd21
--- /dev/null
+++ b/deps/jemalloc/test/unit/pack.c
@@ -0,0 +1,206 @@
+#include "test/jemalloc_test.h"
+
+const char *malloc_conf =
+    /* Use smallest possible chunk size. */
+    "lg_chunk:0"
+    /* Immediately purge to minimize fragmentation. */
+    ",lg_dirty_mult:-1"
+    ",decay_time:-1"
+    ;
+
+/*
+ * Size class that is a divisor of the page size, ideally 4+ regions per run.
+ */
+#if LG_PAGE <= 14
+#define	SZ	(ZU(1) << (LG_PAGE - 2))
+#else
+#define	SZ	4096
+#endif
+
+/*
+ * Number of chunks to consume at high water mark.  Should be at least 2 so that
+ * if mmap()ed memory grows downward, downward growth of mmap()ed memory is
+ * tested.
+ */
+#define	NCHUNKS	8
+
+static unsigned
+binind_compute(void)
+{
+	size_t sz;
+	unsigned nbins, i;
+
+	sz = sizeof(nbins);
+	assert_d_eq(mallctl("arenas.nbins", (void *)&nbins, &sz, NULL, 0), 0,
+	    "Unexpected mallctl failure");
+
+	for (i = 0; i < nbins; i++) {
+		size_t mib[4];
+		size_t miblen = sizeof(mib)/sizeof(size_t);
+		size_t size;
+
+		assert_d_eq(mallctlnametomib("arenas.bin.0.size", mib,
+		    &miblen), 0, "Unexpected mallctlnametomb failure");
+		mib[2] = (size_t)i;
+
+		sz = sizeof(size);
+		assert_d_eq(mallctlbymib(mib, miblen, (void *)&size, &sz, NULL,
+		    0), 0, "Unexpected mallctlbymib failure");
+		if (size == SZ)
+			return (i);
+	}
+
+	test_fail("Unable to compute nregs_per_run");
+	return (0);
+}
+
+static size_t
+nregs_per_run_compute(void)
+{
+	uint32_t nregs;
+	size_t sz;
+	unsigned binind = binind_compute();
+	size_t mib[4];
+	size_t miblen = sizeof(mib)/sizeof(size_t);
+
+	assert_d_eq(mallctlnametomib("arenas.bin.0.nregs", mib, &miblen), 0,
+	    "Unexpected mallctlnametomb failure");
+	mib[2] = (size_t)binind;
+	sz = sizeof(nregs);
+	assert_d_eq(mallctlbymib(mib, miblen, (void *)&nregs, &sz, NULL,
+	    0), 0, "Unexpected mallctlbymib failure");
+	return (nregs);
+}
+
+static size_t
+npages_per_run_compute(void)
+{
+	size_t sz;
+	unsigned binind = binind_compute();
+	size_t mib[4];
+	size_t miblen = sizeof(mib)/sizeof(size_t);
+	size_t run_size;
+
+	assert_d_eq(mallctlnametomib("arenas.bin.0.run_size", mib, &miblen), 0,
+	    "Unexpected mallctlnametomb failure");
+	mib[2] = (size_t)binind;
+	sz = sizeof(run_size);
+	assert_d_eq(mallctlbymib(mib, miblen, (void *)&run_size, &sz, NULL,
+	    0), 0, "Unexpected mallctlbymib failure");
+	return (run_size >> LG_PAGE);
+}
+
+static size_t
+npages_per_chunk_compute(void)
+{
+
+	return ((chunksize >> LG_PAGE) - map_bias);
+}
+
+static size_t
+nruns_per_chunk_compute(void)
+{
+
+	return (npages_per_chunk_compute() / npages_per_run_compute());
+}
+
+static unsigned
+arenas_extend_mallctl(void)
+{
+	unsigned arena_ind;
+	size_t sz;
+
+	sz = sizeof(arena_ind);
+	assert_d_eq(mallctl("arenas.extend", (void *)&arena_ind, &sz, NULL, 0),
+	    0, "Error in arenas.extend");
+
+	return (arena_ind);
+}
+
+static void
+arena_reset_mallctl(unsigned arena_ind)
+{
+	size_t mib[3];
+	size_t miblen = sizeof(mib)/sizeof(size_t);
+
+	assert_d_eq(mallctlnametomib("arena.0.reset", mib, &miblen), 0,
+	    "Unexpected mallctlnametomib() failure");
+	mib[1] = (size_t)arena_ind;
+	assert_d_eq(mallctlbymib(mib, miblen, NULL, NULL, NULL, 0), 0,
+	    "Unexpected mallctlbymib() failure");
+}
+
+TEST_BEGIN(test_pack)
+{
+	unsigned arena_ind = arenas_extend_mallctl();
+	size_t nregs_per_run = nregs_per_run_compute();
+	size_t nruns_per_chunk = nruns_per_chunk_compute();
+	size_t nruns = nruns_per_chunk * NCHUNKS;
+	size_t nregs = nregs_per_run * nruns;
+	VARIABLE_ARRAY(void *, ptrs, nregs);
+	size_t i, j, offset;
+
+	/* Fill matrix. */
+	for (i = offset = 0; i < nruns; i++) {
+		for (j = 0; j < nregs_per_run; j++) {
+			void *p = mallocx(SZ, MALLOCX_ARENA(arena_ind) |
+			    MALLOCX_TCACHE_NONE);
+			assert_ptr_not_null(p,
+			    "Unexpected mallocx(%zu, MALLOCX_ARENA(%u) |"
+			    " MALLOCX_TCACHE_NONE) failure, run=%zu, reg=%zu",
+			    SZ, arena_ind, i, j);
+			ptrs[(i * nregs_per_run) + j] = p;
+		}
+	}
+
+	/*
+	 * Free all but one region of each run, but rotate which region is
+	 * preserved, so that subsequent allocations exercise the within-run
+	 * layout policy.
+	 */
+	offset = 0;
+	for (i = offset = 0;
+	    i < nruns;
+	    i++, offset = (offset + 1) % nregs_per_run) {
+		for (j = 0; j < nregs_per_run; j++) {
+			void *p = ptrs[(i * nregs_per_run) + j];
+			if (offset == j)
+				continue;
+			dallocx(p, MALLOCX_ARENA(arena_ind) |
+			    MALLOCX_TCACHE_NONE);
+		}
+	}
+
+	/*
+	 * Logically refill matrix, skipping preserved regions and verifying
+	 * that the matrix is unmodified.
+	 */
+	offset = 0;
+	for (i = offset = 0;
+	    i < nruns;
+	    i++, offset = (offset + 1) % nregs_per_run) {
+		for (j = 0; j < nregs_per_run; j++) {
+			void *p;
+
+			if (offset == j)
+				continue;
+			p = mallocx(SZ, MALLOCX_ARENA(arena_ind) |
+			    MALLOCX_TCACHE_NONE);
+			assert_ptr_eq(p, ptrs[(i * nregs_per_run) + j],
+			    "Unexpected refill discrepancy, run=%zu, reg=%zu\n",
+			    i, j);
+		}
+	}
+
+	/* Clean up. */
+	arena_reset_mallctl(arena_ind);
+}
+TEST_END
+
+int
+main(void)
+{
+
+	return (test(
+	    test_pack));
+}
diff --git a/deps/jemalloc/test/unit/pages.c b/deps/jemalloc/test/unit/pages.c
new file mode 100644
index 000000000..d31a35e68
--- /dev/null
+++ b/deps/jemalloc/test/unit/pages.c
@@ -0,0 +1,27 @@
+#include "test/jemalloc_test.h"
+
+TEST_BEGIN(test_pages_huge)
+{
+	bool commit;
+	void *pages;
+
+	commit = true;
+	pages = pages_map(NULL, PAGE, &commit);
+	assert_ptr_not_null(pages, "Unexpected pages_map() error");
+
+	assert_false(pages_huge(pages, PAGE),
+	    "Unexpected pages_huge() result");
+	assert_false(pages_nohuge(pages, PAGE),
+	    "Unexpected pages_nohuge() result");
+
+	pages_unmap(pages, PAGE);
+}
+TEST_END
+
+int
+main(void)
+{
+
+	return (test(
+	    test_pages_huge));
+}
diff --git a/deps/jemalloc/test/unit/ph.c b/deps/jemalloc/test/unit/ph.c
new file mode 100644
index 000000000..da442f07e
--- /dev/null
+++ b/deps/jemalloc/test/unit/ph.c
@@ -0,0 +1,290 @@
+#include "test/jemalloc_test.h"
+
+typedef struct node_s node_t;
+
+struct node_s {
+#define	NODE_MAGIC 0x9823af7e
+	uint32_t magic;
+	phn(node_t) link;
+	uint64_t key;
+};
+
+static int
+node_cmp(const node_t *a, const node_t *b)
+{
+	int ret;
+
+	ret = (a->key > b->key) - (a->key < b->key);
+	if (ret == 0) {
+		/*
+		 * Duplicates are not allowed in the heap, so force an
+		 * arbitrary ordering for non-identical items with equal keys.
+		 */
+		ret = (((uintptr_t)a) > ((uintptr_t)b))
+		    - (((uintptr_t)a) < ((uintptr_t)b));
+	}
+	return (ret);
+}
+
+static int
+node_cmp_magic(const node_t *a, const node_t *b) {
+
+	assert_u32_eq(a->magic, NODE_MAGIC, "Bad magic");
+	assert_u32_eq(b->magic, NODE_MAGIC, "Bad magic");
+
+	return (node_cmp(a, b));
+}
+
+typedef ph(node_t) heap_t;
+ph_gen(static, heap_, heap_t, node_t, link, node_cmp_magic);
+
+static void
+node_print(const node_t *node, unsigned depth)
+{
+	unsigned i;
+	node_t *leftmost_child, *sibling;
+
+	for (i = 0; i < depth; i++)
+		malloc_printf("\t");
+	malloc_printf("%2"FMTu64"\n", node->key);
+
+	leftmost_child = phn_lchild_get(node_t, link, node);
+	if (leftmost_child == NULL)
+		return;
+	node_print(leftmost_child, depth + 1);
+
+	for (sibling = phn_next_get(node_t, link, leftmost_child); sibling !=
+	    NULL; sibling = phn_next_get(node_t, link, sibling)) {
+		node_print(sibling, depth + 1);
+	}
+}
+
+static void
+heap_print(const heap_t *heap)
+{
+	node_t *auxelm;
+
+	malloc_printf("vvv heap %p vvv\n", heap);
+	if (heap->ph_root == NULL)
+		goto label_return;
+
+	node_print(heap->ph_root, 0);
+
+	for (auxelm = phn_next_get(node_t, link, heap->ph_root); auxelm != NULL;
+	    auxelm = phn_next_get(node_t, link, auxelm)) {
+		assert_ptr_eq(phn_next_get(node_t, link, phn_prev_get(node_t,
+		    link, auxelm)), auxelm,
+		    "auxelm's prev doesn't link to auxelm");
+		node_print(auxelm, 0);
+	}
+
+label_return:
+	malloc_printf("^^^ heap %p ^^^\n", heap);
+}
+
+static unsigned
+node_validate(const node_t *node, const node_t *parent)
+{
+	unsigned nnodes = 1;
+	node_t *leftmost_child, *sibling;
+
+	if (parent != NULL) {
+		assert_d_ge(node_cmp_magic(node, parent), 0,
+		    "Child is less than parent");
+	}
+
+	leftmost_child = phn_lchild_get(node_t, link, node);
+	if (leftmost_child == NULL)
+		return (nnodes);
+	assert_ptr_eq((void *)phn_prev_get(node_t, link, leftmost_child),
+	    (void *)node, "Leftmost child does not link to node");
+	nnodes += node_validate(leftmost_child, node);
+
+	for (sibling = phn_next_get(node_t, link, leftmost_child); sibling !=
+	    NULL; sibling = phn_next_get(node_t, link, sibling)) {
+		assert_ptr_eq(phn_next_get(node_t, link, phn_prev_get(node_t,
+		    link, sibling)), sibling,
+		    "sibling's prev doesn't link to sibling");
+		nnodes += node_validate(sibling, node);
+	}
+	return (nnodes);
+}
+
+static unsigned
+heap_validate(const heap_t *heap)
+{
+	unsigned nnodes = 0;
+	node_t *auxelm;
+
+	if (heap->ph_root == NULL)
+		goto label_return;
+
+	nnodes += node_validate(heap->ph_root, NULL);
+
+	for (auxelm = phn_next_get(node_t, link, heap->ph_root); auxelm != NULL;
+	    auxelm = phn_next_get(node_t, link, auxelm)) {
+		assert_ptr_eq(phn_next_get(node_t, link, phn_prev_get(node_t,
+		    link, auxelm)), auxelm,
+		    "auxelm's prev doesn't link to auxelm");
+		nnodes += node_validate(auxelm, NULL);
+	}
+
+label_return:
+	if (false)
+		heap_print(heap);
+	return (nnodes);
+}
+
+TEST_BEGIN(test_ph_empty)
+{
+	heap_t heap;
+
+	heap_new(&heap);
+	assert_true(heap_empty(&heap), "Heap should be empty");
+	assert_ptr_null(heap_first(&heap), "Unexpected node");
+}
+TEST_END
+
+static void
+node_remove(heap_t *heap, node_t *node)
+{
+
+	heap_remove(heap, node);
+
+	node->magic = 0;
+}
+
+static node_t *
+node_remove_first(heap_t *heap)
+{
+	node_t *node = heap_remove_first(heap);
+	node->magic = 0;
+	return (node);
+}
+
+TEST_BEGIN(test_ph_random)
+{
+#define	NNODES 25
+#define	NBAGS 250
+#define	SEED 42
+	sfmt_t *sfmt;
+	uint64_t bag[NNODES];
+	heap_t heap;
+	node_t nodes[NNODES];
+	unsigned i, j, k;
+
+	sfmt = init_gen_rand(SEED);
+	for (i = 0; i < NBAGS; i++) {
+		switch (i) {
+		case 0:
+			/* Insert in order. */
+			for (j = 0; j < NNODES; j++)
+				bag[j] = j;
+			break;
+		case 1:
+			/* Insert in reverse order. */
+			for (j = 0; j < NNODES; j++)
+				bag[j] = NNODES - j - 1;
+			break;
+		default:
+			for (j = 0; j < NNODES; j++)
+				bag[j] = gen_rand64_range(sfmt, NNODES);
+		}
+
+		for (j = 1; j <= NNODES; j++) {
+			/* Initialize heap and nodes. */
+			heap_new(&heap);
+			assert_u_eq(heap_validate(&heap), 0,
+			    "Incorrect node count");
+			for (k = 0; k < j; k++) {
+				nodes[k].magic = NODE_MAGIC;
+				nodes[k].key = bag[k];
+			}
+
+			/* Insert nodes. */
+			for (k = 0; k < j; k++) {
+				heap_insert(&heap, &nodes[k]);
+				if (i % 13 == 12) {
+					/* Trigger merging. */
+					assert_ptr_not_null(heap_first(&heap),
+					    "Heap should not be empty");
+				}
+				assert_u_eq(heap_validate(&heap), k + 1,
+				    "Incorrect node count");
+			}
+
+			assert_false(heap_empty(&heap),
+			    "Heap should not be empty");
+
+			/* Remove nodes. */
+			switch (i % 4) {
+			case 0:
+				for (k = 0; k < j; k++) {
+					assert_u_eq(heap_validate(&heap), j - k,
+					    "Incorrect node count");
+					node_remove(&heap, &nodes[k]);
+					assert_u_eq(heap_validate(&heap), j - k
+					    - 1, "Incorrect node count");
+				}
+				break;
+			case 1:
+				for (k = j; k > 0; k--) {
+					node_remove(&heap, &nodes[k-1]);
+					assert_u_eq(heap_validate(&heap), k - 1,
+					    "Incorrect node count");
+				}
+				break;
+			case 2: {
+				node_t *prev = NULL;
+				for (k = 0; k < j; k++) {
+					node_t *node = node_remove_first(&heap);
+					assert_u_eq(heap_validate(&heap), j - k
+					    - 1, "Incorrect node count");
+					if (prev != NULL) {
+						assert_d_ge(node_cmp(node,
+						    prev), 0,
+						    "Bad removal order");
+					}
+					prev = node;
+				}
+				break;
+			} case 3: {
+				node_t *prev = NULL;
+				for (k = 0; k < j; k++) {
+					node_t *node = heap_first(&heap);
+					assert_u_eq(heap_validate(&heap), j - k,
+					    "Incorrect node count");
+					if (prev != NULL) {
+						assert_d_ge(node_cmp(node,
+						    prev), 0,
+						    "Bad removal order");
+					}
+					node_remove(&heap, node);
+					assert_u_eq(heap_validate(&heap), j - k
+					    - 1, "Incorrect node count");
+					prev = node;
+				}
+				break;
+			} default:
+				not_reached();
+			}
+
+			assert_ptr_null(heap_first(&heap),
+			    "Heap should be empty");
+			assert_true(heap_empty(&heap), "Heap should be empty");
+		}
+	}
+	fini_gen_rand(sfmt);
+#undef NNODES
+#undef SEED
+}
+TEST_END
+
+int
+main(void)
+{
+
+	return (test(
+	    test_ph_empty,
+	    test_ph_random));
+}
diff --git a/deps/jemalloc/test/unit/prng.c b/deps/jemalloc/test/unit/prng.c
new file mode 100644
index 000000000..80c9d733f
--- /dev/null
+++ b/deps/jemalloc/test/unit/prng.c
@@ -0,0 +1,263 @@
+#include "test/jemalloc_test.h"
+
+static void
+test_prng_lg_range_u32(bool atomic)
+{
+	uint32_t sa, sb, ra, rb;
+	unsigned lg_range;
+
+	sa = 42;
+	ra = prng_lg_range_u32(&sa, 32, atomic);
+	sa = 42;
+	rb = prng_lg_range_u32(&sa, 32, atomic);
+	assert_u32_eq(ra, rb,
+	    "Repeated generation should produce repeated results");
+
+	sb = 42;
+	rb = prng_lg_range_u32(&sb, 32, atomic);
+	assert_u32_eq(ra, rb,
+	    "Equivalent generation should produce equivalent results");
+
+	sa = 42;
+	ra = prng_lg_range_u32(&sa, 32, atomic);
+	rb = prng_lg_range_u32(&sa, 32, atomic);
+	assert_u32_ne(ra, rb,
+	    "Full-width results must not immediately repeat");
+
+	sa = 42;
+	ra = prng_lg_range_u32(&sa, 32, atomic);
+	for (lg_range = 31; lg_range > 0; lg_range--) {
+		sb = 42;
+		rb = prng_lg_range_u32(&sb, lg_range, atomic);
+		assert_u32_eq((rb & (UINT32_C(0xffffffff) << lg_range)),
+		    0, "High order bits should be 0, lg_range=%u", lg_range);
+		assert_u32_eq(rb, (ra >> (32 - lg_range)),
+		    "Expected high order bits of full-width result, "
+		    "lg_range=%u", lg_range);
+	}
+}
+
+static void
+test_prng_lg_range_u64(void)
+{
+	uint64_t sa, sb, ra, rb;
+	unsigned lg_range;
+
+	sa = 42;
+	ra = prng_lg_range_u64(&sa, 64);
+	sa = 42;
+	rb = prng_lg_range_u64(&sa, 64);
+	assert_u64_eq(ra, rb,
+	    "Repeated generation should produce repeated results");
+
+	sb = 42;
+	rb = prng_lg_range_u64(&sb, 64);
+	assert_u64_eq(ra, rb,
+	    "Equivalent generation should produce equivalent results");
+
+	sa = 42;
+	ra = prng_lg_range_u64(&sa, 64);
+	rb = prng_lg_range_u64(&sa, 64);
+	assert_u64_ne(ra, rb,
+	    "Full-width results must not immediately repeat");
+
+	sa = 42;
+	ra = prng_lg_range_u64(&sa, 64);
+	for (lg_range = 63; lg_range > 0; lg_range--) {
+		sb = 42;
+		rb = prng_lg_range_u64(&sb, lg_range);
+		assert_u64_eq((rb & (UINT64_C(0xffffffffffffffff) << lg_range)),
+		    0, "High order bits should be 0, lg_range=%u", lg_range);
+		assert_u64_eq(rb, (ra >> (64 - lg_range)),
+		    "Expected high order bits of full-width result, "
+		    "lg_range=%u", lg_range);
+	}
+}
+
+static void
+test_prng_lg_range_zu(bool atomic)
+{
+	size_t sa, sb, ra, rb;
+	unsigned lg_range;
+
+	sa = 42;
+	ra = prng_lg_range_zu(&sa, ZU(1) << (3 + LG_SIZEOF_PTR), atomic);
+	sa = 42;
+	rb = prng_lg_range_zu(&sa, ZU(1) << (3 + LG_SIZEOF_PTR), atomic);
+	assert_zu_eq(ra, rb,
+	    "Repeated generation should produce repeated results");
+
+	sb = 42;
+	rb = prng_lg_range_zu(&sb, ZU(1) << (3 + LG_SIZEOF_PTR), atomic);
+	assert_zu_eq(ra, rb,
+	    "Equivalent generation should produce equivalent results");
+
+	sa = 42;
+	ra = prng_lg_range_zu(&sa, ZU(1) << (3 + LG_SIZEOF_PTR), atomic);
+	rb = prng_lg_range_zu(&sa, ZU(1) << (3 + LG_SIZEOF_PTR), atomic);
+	assert_zu_ne(ra, rb,
+	    "Full-width results must not immediately repeat");
+
+	sa = 42;
+	ra = prng_lg_range_zu(&sa, ZU(1) << (3 + LG_SIZEOF_PTR), atomic);
+	for (lg_range = (ZU(1) << (3 + LG_SIZEOF_PTR)) - 1; lg_range > 0;
+	    lg_range--) {
+		sb = 42;
+		rb = prng_lg_range_zu(&sb, lg_range, atomic);
+		assert_zu_eq((rb & (SIZE_T_MAX << lg_range)),
+		    0, "High order bits should be 0, lg_range=%u", lg_range);
+		assert_zu_eq(rb, (ra >> ((ZU(1) << (3 + LG_SIZEOF_PTR)) -
+		    lg_range)), "Expected high order bits of full-width "
+		    "result, lg_range=%u", lg_range);
+	}
+}
+
+TEST_BEGIN(test_prng_lg_range_u32_nonatomic)
+{
+
+	test_prng_lg_range_u32(false);
+}
+TEST_END
+
+TEST_BEGIN(test_prng_lg_range_u32_atomic)
+{
+
+	test_prng_lg_range_u32(true);
+}
+TEST_END
+
+TEST_BEGIN(test_prng_lg_range_u64_nonatomic)
+{
+
+	test_prng_lg_range_u64();
+}
+TEST_END
+
+TEST_BEGIN(test_prng_lg_range_zu_nonatomic)
+{
+
+	test_prng_lg_range_zu(false);
+}
+TEST_END
+
+TEST_BEGIN(test_prng_lg_range_zu_atomic)
+{
+
+	test_prng_lg_range_zu(true);
+}
+TEST_END
+
+static void
+test_prng_range_u32(bool atomic)
+{
+	uint32_t range;
+#define	MAX_RANGE	10000000
+#define	RANGE_STEP	97
+#define	NREPS		10
+
+	for (range = 2; range < MAX_RANGE; range += RANGE_STEP) {
+		uint32_t s;
+		unsigned rep;
+
+		s = range;
+		for (rep = 0; rep < NREPS; rep++) {
+			uint32_t r = prng_range_u32(&s, range, atomic);
+
+			assert_u32_lt(r, range, "Out of range");
+		}
+	}
+}
+
+static void
+test_prng_range_u64(void)
+{
+	uint64_t range;
+#define	MAX_RANGE	10000000
+#define	RANGE_STEP	97
+#define	NREPS		10
+
+	for (range = 2; range < MAX_RANGE; range += RANGE_STEP) {
+		uint64_t s;
+		unsigned rep;
+
+		s = range;
+		for (rep = 0; rep < NREPS; rep++) {
+			uint64_t r = prng_range_u64(&s, range);
+
+			assert_u64_lt(r, range, "Out of range");
+		}
+	}
+}
+
+static void
+test_prng_range_zu(bool atomic)
+{
+	size_t range;
+#define	MAX_RANGE	10000000
+#define	RANGE_STEP	97
+#define	NREPS		10
+
+	for (range = 2; range < MAX_RANGE; range += RANGE_STEP) {
+		size_t s;
+		unsigned rep;
+
+		s = range;
+		for (rep = 0; rep < NREPS; rep++) {
+			size_t r = prng_range_zu(&s, range, atomic);
+
+			assert_zu_lt(r, range, "Out of range");
+		}
+	}
+}
+
+TEST_BEGIN(test_prng_range_u32_nonatomic)
+{
+
+	test_prng_range_u32(false);
+}
+TEST_END
+
+TEST_BEGIN(test_prng_range_u32_atomic)
+{
+
+	test_prng_range_u32(true);
+}
+TEST_END
+
+TEST_BEGIN(test_prng_range_u64_nonatomic)
+{
+
+	test_prng_range_u64();
+}
+TEST_END
+
+TEST_BEGIN(test_prng_range_zu_nonatomic)
+{
+
+	test_prng_range_zu(false);
+}
+TEST_END
+
+TEST_BEGIN(test_prng_range_zu_atomic)
+{
+
+	test_prng_range_zu(true);
+}
+TEST_END
+
+int
+main(void)
+{
+
+	return (test(
+	    test_prng_lg_range_u32_nonatomic,
+	    test_prng_lg_range_u32_atomic,
+	    test_prng_lg_range_u64_nonatomic,
+	    test_prng_lg_range_zu_nonatomic,
+	    test_prng_lg_range_zu_atomic,
+	    test_prng_range_u32_nonatomic,
+	    test_prng_range_u32_atomic,
+	    test_prng_range_u64_nonatomic,
+	    test_prng_range_zu_nonatomic,
+	    test_prng_range_zu_atomic));
+}
diff --git a/deps/jemalloc/test/unit/prof_accum.c b/deps/jemalloc/test/unit/prof_accum.c
old mode 100644
new mode 100755
index fd229e0fd..d941b5bc6
--- a/deps/jemalloc/test/unit/prof_accum.c
+++ b/deps/jemalloc/test/unit/prof_accum.c
@@ -68,8 +68,9 @@ TEST_BEGIN(test_idump)
 	test_skip_if(!config_prof);
 
 	active = true;
-	assert_d_eq(mallctl("prof.active", NULL, NULL, &active, sizeof(active)),
-	    0, "Unexpected mallctl failure while activating profiling");
+	assert_d_eq(mallctl("prof.active", NULL, NULL, (void *)&active,
+	    sizeof(active)), 0,
+	    "Unexpected mallctl failure while activating profiling");
 
 	prof_dump_open = prof_dump_open_intercept;
 
diff --git a/deps/jemalloc/test/unit/prof_active.c b/deps/jemalloc/test/unit/prof_active.c
old mode 100644
new mode 100755
index 814909572..d00943a4c
--- a/deps/jemalloc/test/unit/prof_active.c
+++ b/deps/jemalloc/test/unit/prof_active.c
@@ -12,7 +12,7 @@ mallctl_bool_get(const char *name, bool expected, const char *func, int line)
 	size_t sz;
 
 	sz = sizeof(old);
-	assert_d_eq(mallctl(name, &old, &sz, NULL, 0), 0,
+	assert_d_eq(mallctl(name, (void *)&old, &sz, NULL, 0), 0,
 	    "%s():%d: Unexpected mallctl failure reading %s", func, line, name);
 	assert_b_eq(old, expected, "%s():%d: Unexpected %s value", func, line,
 	    name);
@@ -26,7 +26,8 @@ mallctl_bool_set(const char *name, bool old_expected, bool val_new,
 	size_t sz;
 
 	sz = sizeof(old);
-	assert_d_eq(mallctl(name, &old, &sz, &val_new, sizeof(val_new)), 0,
+	assert_d_eq(mallctl(name, (void *)&old, &sz, (void *)&val_new,
+	    sizeof(val_new)), 0,
 	    "%s():%d: Unexpected mallctl failure reading/writing %s", func,
 	    line, name);
 	assert_b_eq(old, old_expected, "%s():%d: Unexpected %s value", func,
diff --git a/deps/jemalloc/test/unit/prof_gdump.c b/deps/jemalloc/test/unit/prof_gdump.c
old mode 100644
new mode 100755
index a0e6ee921..996cb6704
--- a/deps/jemalloc/test/unit/prof_gdump.c
+++ b/deps/jemalloc/test/unit/prof_gdump.c
@@ -28,8 +28,9 @@ TEST_BEGIN(test_gdump)
 	test_skip_if(!config_prof);
 
 	active = true;
-	assert_d_eq(mallctl("prof.active", NULL, NULL, &active, sizeof(active)),
-	    0, "Unexpected mallctl failure while activating profiling");
+	assert_d_eq(mallctl("prof.active", NULL, NULL, (void *)&active,
+	    sizeof(active)), 0,
+	    "Unexpected mallctl failure while activating profiling");
 
 	prof_dump_open = prof_dump_open_intercept;
 
@@ -45,8 +46,8 @@ TEST_BEGIN(test_gdump)
 
 	gdump = false;
 	sz = sizeof(gdump_old);
-	assert_d_eq(mallctl("prof.gdump", &gdump_old, &sz, &gdump,
-	    sizeof(gdump)), 0,
+	assert_d_eq(mallctl("prof.gdump", (void *)&gdump_old, &sz,
+	    (void *)&gdump, sizeof(gdump)), 0,
 	    "Unexpected mallctl failure while disabling prof.gdump");
 	assert(gdump_old);
 	did_prof_dump_open = false;
@@ -56,8 +57,8 @@ TEST_BEGIN(test_gdump)
 
 	gdump = true;
 	sz = sizeof(gdump_old);
-	assert_d_eq(mallctl("prof.gdump", &gdump_old, &sz, &gdump,
-	    sizeof(gdump)), 0,
+	assert_d_eq(mallctl("prof.gdump", (void *)&gdump_old, &sz,
+	    (void *)&gdump, sizeof(gdump)), 0,
 	    "Unexpected mallctl failure while enabling prof.gdump");
 	assert(!gdump_old);
 	did_prof_dump_open = false;
diff --git a/deps/jemalloc/test/unit/prof_idump.c b/deps/jemalloc/test/unit/prof_idump.c
old mode 100644
new mode 100755
index bdea53ecd..16c6462de
--- a/deps/jemalloc/test/unit/prof_idump.c
+++ b/deps/jemalloc/test/unit/prof_idump.c
@@ -29,8 +29,9 @@ TEST_BEGIN(test_idump)
 	test_skip_if(!config_prof);
 
 	active = true;
-	assert_d_eq(mallctl("prof.active", NULL, NULL, &active, sizeof(active)),
-	    0, "Unexpected mallctl failure while activating profiling");
+	assert_d_eq(mallctl("prof.active", NULL, NULL, (void *)&active,
+	    sizeof(active)), 0,
+	    "Unexpected mallctl failure while activating profiling");
 
 	prof_dump_open = prof_dump_open_intercept;
 
diff --git a/deps/jemalloc/test/unit/prof_reset.c b/deps/jemalloc/test/unit/prof_reset.c
old mode 100644
new mode 100755
index 69983e5e5..59d70796a
--- a/deps/jemalloc/test/unit/prof_reset.c
+++ b/deps/jemalloc/test/unit/prof_reset.c
@@ -20,8 +20,8 @@ static void
 set_prof_active(bool active)
 {
 
-	assert_d_eq(mallctl("prof.active", NULL, NULL, &active, sizeof(active)),
-	    0, "Unexpected mallctl failure");
+	assert_d_eq(mallctl("prof.active", NULL, NULL, (void *)&active,
+	    sizeof(active)), 0, "Unexpected mallctl failure");
 }
 
 static size_t
@@ -30,7 +30,8 @@ get_lg_prof_sample(void)
 	size_t lg_prof_sample;
 	size_t sz = sizeof(size_t);
 
-	assert_d_eq(mallctl("prof.lg_sample", &lg_prof_sample, &sz, NULL, 0), 0,
+	assert_d_eq(mallctl("prof.lg_sample", (void *)&lg_prof_sample, &sz,
+	    NULL, 0), 0,
 	    "Unexpected mallctl failure while reading profiling sample rate");
 	return (lg_prof_sample);
 }
@@ -39,7 +40,7 @@ static void
 do_prof_reset(size_t lg_prof_sample)
 {
 	assert_d_eq(mallctl("prof.reset", NULL, NULL,
-	    &lg_prof_sample, sizeof(size_t)), 0,
+	    (void *)&lg_prof_sample, sizeof(size_t)), 0,
 	    "Unexpected mallctl failure while resetting profile data");
 	assert_zu_eq(lg_prof_sample, get_lg_prof_sample(),
 	    "Expected profile sample rate change");
@@ -54,8 +55,8 @@ TEST_BEGIN(test_prof_reset_basic)
 	test_skip_if(!config_prof);
 
 	sz = sizeof(size_t);
-	assert_d_eq(mallctl("opt.lg_prof_sample", &lg_prof_sample_orig, &sz,
-	    NULL, 0), 0,
+	assert_d_eq(mallctl("opt.lg_prof_sample", (void *)&lg_prof_sample_orig,
+	    &sz, NULL, 0), 0,
 	    "Unexpected mallctl failure while reading profiling sample rate");
 	assert_zu_eq(lg_prof_sample_orig, 0,
 	    "Unexpected profiling sample rate");
@@ -94,7 +95,8 @@ TEST_END
 bool prof_dump_header_intercepted = false;
 prof_cnt_t cnt_all_copy = {0, 0, 0, 0};
 static bool
-prof_dump_header_intercept(bool propagate_err, const prof_cnt_t *cnt_all)
+prof_dump_header_intercept(tsdn_t *tsdn, bool propagate_err,
+    const prof_cnt_t *cnt_all)
 {
 
 	prof_dump_header_intercepted = true;
diff --git a/deps/jemalloc/test/unit/prof_thread_name.c b/deps/jemalloc/test/unit/prof_thread_name.c
old mode 100644
new mode 100755
index f501158d7..9ec549776
--- a/deps/jemalloc/test/unit/prof_thread_name.c
+++ b/deps/jemalloc/test/unit/prof_thread_name.c
@@ -12,8 +12,9 @@ mallctl_thread_name_get_impl(const char *thread_name_expected, const char *func,
 	size_t sz;
 
 	sz = sizeof(thread_name_old);
-	assert_d_eq(mallctl("thread.prof.name", &thread_name_old, &sz, NULL, 0),
-	    0, "%s():%d: Unexpected mallctl failure reading thread.prof.name",
+	assert_d_eq(mallctl("thread.prof.name", (void *)&thread_name_old, &sz,
+	    NULL, 0), 0,
+	    "%s():%d: Unexpected mallctl failure reading thread.prof.name",
 	    func, line);
 	assert_str_eq(thread_name_old, thread_name_expected,
 	    "%s():%d: Unexpected thread.prof.name value", func, line);
@@ -26,8 +27,8 @@ mallctl_thread_name_set_impl(const char *thread_name, const char *func,
     int line)
 {
 
-	assert_d_eq(mallctl("thread.prof.name", NULL, NULL, &thread_name,
-	    sizeof(thread_name)), 0,
+	assert_d_eq(mallctl("thread.prof.name", NULL, NULL,
+	    (void *)&thread_name, sizeof(thread_name)), 0,
 	    "%s():%d: Unexpected mallctl failure reading thread.prof.name",
 	    func, line);
 	mallctl_thread_name_get_impl(thread_name, func, line);
@@ -46,15 +47,15 @@ TEST_BEGIN(test_prof_thread_name_validation)
 
 	/* NULL input shouldn't be allowed. */
 	thread_name = NULL;
-	assert_d_eq(mallctl("thread.prof.name", NULL, NULL, &thread_name,
-	    sizeof(thread_name)), EFAULT,
+	assert_d_eq(mallctl("thread.prof.name", NULL, NULL,
+	    (void *)&thread_name, sizeof(thread_name)), EFAULT,
 	    "Unexpected mallctl result writing \"%s\" to thread.prof.name",
 	    thread_name);
 
 	/* '\n' shouldn't be allowed. */
 	thread_name = "hi\nthere";
-	assert_d_eq(mallctl("thread.prof.name", NULL, NULL, &thread_name,
-	    sizeof(thread_name)), EFAULT,
+	assert_d_eq(mallctl("thread.prof.name", NULL, NULL,
+	    (void *)&thread_name, sizeof(thread_name)), EFAULT,
 	    "Unexpected mallctl result writing \"%s\" to thread.prof.name",
 	    thread_name);
 
@@ -64,8 +65,9 @@ TEST_BEGIN(test_prof_thread_name_validation)
 		size_t sz;
 
 		sz = sizeof(thread_name_old);
-		assert_d_eq(mallctl("thread.prof.name", &thread_name_old, &sz,
-		    &thread_name, sizeof(thread_name)), EPERM,
+		assert_d_eq(mallctl("thread.prof.name",
+		    (void *)&thread_name_old, &sz, (void *)&thread_name,
+		    sizeof(thread_name)), EPERM,
 		    "Unexpected mallctl result writing \"%s\" to "
 		    "thread.prof.name", thread_name);
 	}
diff --git a/deps/jemalloc/test/unit/rb.c b/deps/jemalloc/test/unit/rb.c
index b38eb0e33..cf3d3a783 100644
--- a/deps/jemalloc/test/unit/rb.c
+++ b/deps/jemalloc/test/unit/rb.c
@@ -3,7 +3,7 @@
 #define	rbtn_black_height(a_type, a_field, a_rbt, r_height) do {	\
     a_type *rbp_bh_t;							\
     for (rbp_bh_t = (a_rbt)->rbt_root, (r_height) = 0;			\
-      rbp_bh_t != &(a_rbt)->rbt_nil;					\
+	 rbp_bh_t != NULL;						\
       rbp_bh_t = rbtn_left_get(a_type, a_field, rbp_bh_t)) {		\
 	if (!rbtn_red_get(a_type, a_field, rbp_bh_t)) {			\
 	    (r_height)++;						\
@@ -21,7 +21,7 @@ struct node_s {
 };
 
 static int
-node_cmp(node_t *a, node_t *b) {
+node_cmp(const node_t *a, const node_t *b) {
 	int ret;
 
 	assert_u32_eq(a->magic, NODE_MAGIC, "Bad magic");
@@ -68,38 +68,43 @@ TEST_BEGIN(test_rb_empty)
 TEST_END
 
 static unsigned
-tree_recurse(node_t *node, unsigned black_height, unsigned black_depth,
-    node_t *nil)
+tree_recurse(node_t *node, unsigned black_height, unsigned black_depth)
 {
 	unsigned ret = 0;
-	node_t *left_node = rbtn_left_get(node_t, link, node);
-	node_t *right_node = rbtn_right_get(node_t, link, node);
+	node_t *left_node;
+	node_t *right_node;
+
+	if (node == NULL)
+		return (ret);
+
+	left_node = rbtn_left_get(node_t, link, node);
+	right_node = rbtn_right_get(node_t, link, node);
 
 	if (!rbtn_red_get(node_t, link, node))
 		black_depth++;
 
 	/* Red nodes must be interleaved with black nodes. */
 	if (rbtn_red_get(node_t, link, node)) {
-		assert_false(rbtn_red_get(node_t, link, left_node),
-		    "Node should be black");
-		assert_false(rbtn_red_get(node_t, link, right_node),
-		    "Node should be black");
+		if (left_node != NULL)
+			assert_false(rbtn_red_get(node_t, link, left_node),
+				"Node should be black");
+		if (right_node != NULL)
+			assert_false(rbtn_red_get(node_t, link, right_node),
+			    "Node should be black");
 	}
 
-	if (node == nil)
-		return (ret);
 	/* Self. */
 	assert_u32_eq(node->magic, NODE_MAGIC, "Bad magic");
 
 	/* Left subtree. */
-	if (left_node != nil)
-		ret += tree_recurse(left_node, black_height, black_depth, nil);
+	if (left_node != NULL)
+		ret += tree_recurse(left_node, black_height, black_depth);
 	else
 		ret += (black_depth != black_height);
 
 	/* Right subtree. */
-	if (right_node != nil)
-		ret += tree_recurse(right_node, black_height, black_depth, nil);
+	if (right_node != NULL)
+		ret += tree_recurse(right_node, black_height, black_depth);
 	else
 		ret += (black_depth != black_height);
 
@@ -181,8 +186,7 @@ node_remove(tree_t *tree, node_t *node, unsigned nnodes)
 	node->magic = 0;
 
 	rbtn_black_height(node_t, link, tree, black_height);
-	imbalances = tree_recurse(tree->rbt_root, black_height, 0,
-	    &(tree->rbt_nil));
+	imbalances = tree_recurse(tree->rbt_root, black_height, 0);
 	assert_u_eq(imbalances, 0, "Tree is unbalanced");
 	assert_u_eq(tree_iterate(tree), nnodes-1,
 	    "Unexpected node iteration count");
@@ -212,6 +216,15 @@ remove_reverse_iterate_cb(tree_t *tree, node_t *node, void *data)
 	return (ret);
 }
 
+static void
+destroy_cb(node_t *node, void *data)
+{
+	unsigned *nnodes = (unsigned *)data;
+
+	assert_u_gt(*nnodes, 0, "Destruction removed too many nodes");
+	(*nnodes)--;
+}
+
 TEST_BEGIN(test_rb_random)
 {
 #define	NNODES 25
@@ -244,7 +257,6 @@ TEST_BEGIN(test_rb_random)
 		for (j = 1; j <= NNODES; j++) {
 			/* Initialize tree and nodes. */
 			tree_new(&tree);
-			tree.rbt_nil.magic = 0;
 			for (k = 0; k < j; k++) {
 				nodes[k].magic = NODE_MAGIC;
 				nodes[k].key = bag[k];
@@ -257,7 +269,7 @@ TEST_BEGIN(test_rb_random)
 				rbtn_black_height(node_t, link, &tree,
 				    black_height);
 				imbalances = tree_recurse(tree.rbt_root,
-				    black_height, 0, &(tree.rbt_nil));
+				    black_height, 0);
 				assert_u_eq(imbalances, 0,
 				    "Tree is unbalanced");
 
@@ -278,7 +290,7 @@ TEST_BEGIN(test_rb_random)
 			}
 
 			/* Remove nodes. */
-			switch (i % 4) {
+			switch (i % 5) {
 			case 0:
 				for (k = 0; k < j; k++)
 					node_remove(&tree, &nodes[k], j - k);
@@ -314,6 +326,12 @@ TEST_BEGIN(test_rb_random)
 				assert_u_eq(nnodes, 0,
 				    "Removal terminated early");
 				break;
+			} case 4: {
+				unsigned nnodes = j;
+				tree_destroy(&tree, destroy_cb, &nnodes);
+				assert_u_eq(nnodes, 0,
+				    "Destruction terminated early");
+				break;
 			} default:
 				not_reached();
 			}
diff --git a/deps/jemalloc/test/unit/run_quantize.c b/deps/jemalloc/test/unit/run_quantize.c
new file mode 100644
index 000000000..089176f39
--- /dev/null
+++ b/deps/jemalloc/test/unit/run_quantize.c
@@ -0,0 +1,149 @@
+#include "test/jemalloc_test.h"
+
+TEST_BEGIN(test_small_run_size)
+{
+	unsigned nbins, i;
+	size_t sz, run_size;
+	size_t mib[4];
+	size_t miblen = sizeof(mib) / sizeof(size_t);
+
+	/*
+	 * Iterate over all small size classes, get their run sizes, and verify
+	 * that the quantized size is the same as the run size.
+	 */
+
+	sz = sizeof(unsigned);
+	assert_d_eq(mallctl("arenas.nbins", (void *)&nbins, &sz, NULL, 0), 0,
+	    "Unexpected mallctl failure");
+
+	assert_d_eq(mallctlnametomib("arenas.bin.0.run_size", mib, &miblen), 0,
+	    "Unexpected mallctlnametomib failure");
+	for (i = 0; i < nbins; i++) {
+		mib[2] = i;
+		sz = sizeof(size_t);
+		assert_d_eq(mallctlbymib(mib, miblen, (void *)&run_size, &sz,
+		    NULL, 0), 0, "Unexpected mallctlbymib failure");
+		assert_zu_eq(run_size, run_quantize_floor(run_size),
+		    "Small run quantization should be a no-op (run_size=%zu)",
+		    run_size);
+		assert_zu_eq(run_size, run_quantize_ceil(run_size),
+		    "Small run quantization should be a no-op (run_size=%zu)",
+		    run_size);
+	}
+}
+TEST_END
+
+TEST_BEGIN(test_large_run_size)
+{
+	bool cache_oblivious;
+	unsigned nlruns, i;
+	size_t sz, run_size_prev, ceil_prev;
+	size_t mib[4];
+	size_t miblen = sizeof(mib) / sizeof(size_t);
+
+	/*
+	 * Iterate over all large size classes, get their run sizes, and verify
+	 * that the quantized size is the same as the run size.
+	 */
+
+	sz = sizeof(bool);
+	assert_d_eq(mallctl("config.cache_oblivious", (void *)&cache_oblivious,
+	    &sz, NULL, 0), 0, "Unexpected mallctl failure");
+
+	sz = sizeof(unsigned);
+	assert_d_eq(mallctl("arenas.nlruns", (void *)&nlruns, &sz, NULL, 0), 0,
+	    "Unexpected mallctl failure");
+
+	assert_d_eq(mallctlnametomib("arenas.lrun.0.size", mib, &miblen), 0,
+	    "Unexpected mallctlnametomib failure");
+	for (i = 0; i < nlruns; i++) {
+		size_t lrun_size, run_size, floor, ceil;
+
+		mib[2] = i;
+		sz = sizeof(size_t);
+		assert_d_eq(mallctlbymib(mib, miblen, (void *)&lrun_size, &sz,
+		    NULL, 0), 0, "Unexpected mallctlbymib failure");
+		run_size = cache_oblivious ? lrun_size + PAGE : lrun_size;
+		floor = run_quantize_floor(run_size);
+		ceil = run_quantize_ceil(run_size);
+
+		assert_zu_eq(run_size, floor,
+		    "Large run quantization should be a no-op for precise "
+		    "size (lrun_size=%zu, run_size=%zu)", lrun_size, run_size);
+		assert_zu_eq(run_size, ceil,
+		    "Large run quantization should be a no-op for precise "
+		    "size (lrun_size=%zu, run_size=%zu)", lrun_size, run_size);
+
+		if (i > 0) {
+			assert_zu_eq(run_size_prev, run_quantize_floor(run_size
+			    - PAGE), "Floor should be a precise size");
+			if (run_size_prev < ceil_prev) {
+				assert_zu_eq(ceil_prev, run_size,
+				    "Ceiling should be a precise size "
+				    "(run_size_prev=%zu, ceil_prev=%zu, "
+				    "run_size=%zu)", run_size_prev, ceil_prev,
+				    run_size);
+			}
+		}
+		run_size_prev = floor;
+		ceil_prev = run_quantize_ceil(run_size + PAGE);
+	}
+}
+TEST_END
+
+TEST_BEGIN(test_monotonic)
+{
+	unsigned nbins, nlruns, i;
+	size_t sz, floor_prev, ceil_prev;
+
+	/*
+	 * Iterate over all run sizes and verify that
+	 * run_quantize_{floor,ceil}() are monotonic.
+	 */
+
+	sz = sizeof(unsigned);
+	assert_d_eq(mallctl("arenas.nbins", (void *)&nbins, &sz, NULL, 0), 0,
+	    "Unexpected mallctl failure");
+
+	sz = sizeof(unsigned);
+	assert_d_eq(mallctl("arenas.nlruns", (void *)&nlruns, &sz, NULL, 0), 0,
+	    "Unexpected mallctl failure");
+
+	floor_prev = 0;
+	ceil_prev = 0;
+	for (i = 1; i <= chunksize >> LG_PAGE; i++) {
+		size_t run_size, floor, ceil;
+
+		run_size = i << LG_PAGE;
+		floor = run_quantize_floor(run_size);
+		ceil = run_quantize_ceil(run_size);
+
+		assert_zu_le(floor, run_size,
+		    "Floor should be <= (floor=%zu, run_size=%zu, ceil=%zu)",
+		    floor, run_size, ceil);
+		assert_zu_ge(ceil, run_size,
+		    "Ceiling should be >= (floor=%zu, run_size=%zu, ceil=%zu)",
+		    floor, run_size, ceil);
+
+		assert_zu_le(floor_prev, floor, "Floor should be monotonic "
+		    "(floor_prev=%zu, floor=%zu, run_size=%zu, ceil=%zu)",
+		    floor_prev, floor, run_size, ceil);
+		assert_zu_le(ceil_prev, ceil, "Ceiling should be monotonic "
+		    "(floor=%zu, run_size=%zu, ceil_prev=%zu, ceil=%zu)",
+		    floor, run_size, ceil_prev, ceil);
+
+		floor_prev = floor;
+		ceil_prev = ceil;
+	}
+}
+TEST_END
+
+int
+main(void)
+{
+
+	return (test(
+	    test_small_run_size,
+	    test_large_run_size,
+	    test_monotonic));
+}
diff --git a/deps/jemalloc/test/unit/size_classes.c b/deps/jemalloc/test/unit/size_classes.c
old mode 100644
new mode 100755
index d3aaebd77..81cc60617
--- a/deps/jemalloc/test/unit/size_classes.c
+++ b/deps/jemalloc/test/unit/size_classes.c
@@ -8,8 +8,8 @@ get_max_size_class(void)
 	size_t sz, miblen, max_size_class;
 
 	sz = sizeof(unsigned);
-	assert_d_eq(mallctl("arenas.nhchunks", &nhchunks, &sz, NULL, 0), 0,
-	    "Unexpected mallctl() error");
+	assert_d_eq(mallctl("arenas.nhchunks", (void *)&nhchunks, &sz, NULL, 0),
+	    0, "Unexpected mallctl() error");
 
 	miblen = sizeof(mib) / sizeof(size_t);
 	assert_d_eq(mallctlnametomib("arenas.hchunk.0.size", mib, &miblen), 0,
@@ -17,8 +17,8 @@ get_max_size_class(void)
 	mib[2] = nhchunks - 1;
 
 	sz = sizeof(size_t);
-	assert_d_eq(mallctlbymib(mib, miblen, &max_size_class, &sz, NULL, 0), 0,
-	    "Unexpected mallctlbymib() error");
+	assert_d_eq(mallctlbymib(mib, miblen, (void *)&max_size_class, &sz,
+	    NULL, 0), 0, "Unexpected mallctlbymib() error");
 
 	return (max_size_class);
 }
@@ -80,10 +80,105 @@ TEST_BEGIN(test_size_classes)
 }
 TEST_END
 
+TEST_BEGIN(test_psize_classes)
+{
+	size_t size_class, max_size_class;
+	pszind_t pind, max_pind;
+
+	max_size_class = get_max_size_class();
+	max_pind = psz2ind(max_size_class);
+
+	for (pind = 0, size_class = pind2sz(pind); pind < max_pind ||
+	    size_class < max_size_class; pind++, size_class =
+	    pind2sz(pind)) {
+		assert_true(pind < max_pind,
+		    "Loop conditionals should be equivalent; pind=%u, "
+		    "size_class=%zu (%#zx)", pind, size_class, size_class);
+		assert_true(size_class < max_size_class,
+		    "Loop conditionals should be equivalent; pind=%u, "
+		    "size_class=%zu (%#zx)", pind, size_class, size_class);
+
+		assert_u_eq(pind, psz2ind(size_class),
+		    "psz2ind() does not reverse pind2sz(): pind=%u -->"
+		    " size_class=%zu --> pind=%u --> size_class=%zu", pind,
+		    size_class, psz2ind(size_class),
+		    pind2sz(psz2ind(size_class)));
+		assert_zu_eq(size_class, pind2sz(psz2ind(size_class)),
+		    "pind2sz() does not reverse psz2ind(): pind=%u -->"
+		    " size_class=%zu --> pind=%u --> size_class=%zu", pind,
+		    size_class, psz2ind(size_class),
+		    pind2sz(psz2ind(size_class)));
+
+		assert_u_eq(pind+1, psz2ind(size_class+1),
+		    "Next size_class does not round up properly");
+
+		assert_zu_eq(size_class, (pind > 0) ?
+		    psz2u(pind2sz(pind-1)+1) : psz2u(1),
+		    "psz2u() does not round up to size class");
+		assert_zu_eq(size_class, psz2u(size_class-1),
+		    "psz2u() does not round up to size class");
+		assert_zu_eq(size_class, psz2u(size_class),
+		    "psz2u() does not compute same size class");
+		assert_zu_eq(psz2u(size_class+1), pind2sz(pind+1),
+		    "psz2u() does not round up to next size class");
+	}
+
+	assert_u_eq(pind, psz2ind(pind2sz(pind)),
+	    "psz2ind() does not reverse pind2sz()");
+	assert_zu_eq(max_size_class, pind2sz(psz2ind(max_size_class)),
+	    "pind2sz() does not reverse psz2ind()");
+
+	assert_zu_eq(size_class, psz2u(pind2sz(pind-1)+1),
+	    "psz2u() does not round up to size class");
+	assert_zu_eq(size_class, psz2u(size_class-1),
+	    "psz2u() does not round up to size class");
+	assert_zu_eq(size_class, psz2u(size_class),
+	    "psz2u() does not compute same size class");
+}
+TEST_END
+
+TEST_BEGIN(test_overflow)
+{
+	size_t max_size_class;
+
+	max_size_class = get_max_size_class();
+
+	assert_u_eq(size2index(max_size_class+1), NSIZES,
+	    "size2index() should return NSIZES on overflow");
+	assert_u_eq(size2index(ZU(PTRDIFF_MAX)+1), NSIZES,
+	    "size2index() should return NSIZES on overflow");
+	assert_u_eq(size2index(SIZE_T_MAX), NSIZES,
+	    "size2index() should return NSIZES on overflow");
+
+	assert_zu_eq(s2u(max_size_class+1), 0,
+	    "s2u() should return 0 for unsupported size");
+	assert_zu_eq(s2u(ZU(PTRDIFF_MAX)+1), 0,
+	    "s2u() should return 0 for unsupported size");
+	assert_zu_eq(s2u(SIZE_T_MAX), 0,
+	    "s2u() should return 0 on overflow");
+
+	assert_u_eq(psz2ind(max_size_class+1), NPSIZES,
+	    "psz2ind() should return NPSIZES on overflow");
+	assert_u_eq(psz2ind(ZU(PTRDIFF_MAX)+1), NPSIZES,
+	    "psz2ind() should return NPSIZES on overflow");
+	assert_u_eq(psz2ind(SIZE_T_MAX), NPSIZES,
+	    "psz2ind() should return NPSIZES on overflow");
+
+	assert_zu_eq(psz2u(max_size_class+1), 0,
+	    "psz2u() should return 0 for unsupported size");
+	assert_zu_eq(psz2u(ZU(PTRDIFF_MAX)+1), 0,
+	    "psz2u() should return 0 for unsupported size");
+	assert_zu_eq(psz2u(SIZE_T_MAX), 0,
+	    "psz2u() should return 0 on overflow");
+}
+TEST_END
+
 int
 main(void)
 {
 
 	return (test(
-	    test_size_classes));
+	    test_size_classes,
+	    test_psize_classes,
+	    test_overflow));
 }
diff --git a/deps/jemalloc/test/unit/smoothstep.c b/deps/jemalloc/test/unit/smoothstep.c
new file mode 100644
index 000000000..4cfb21343
--- /dev/null
+++ b/deps/jemalloc/test/unit/smoothstep.c
@@ -0,0 +1,106 @@
+#include "test/jemalloc_test.h"
+
+static const uint64_t smoothstep_tab[] = {
+#define	STEP(step, h, x, y) \
+	h,
+	SMOOTHSTEP
+#undef STEP
+};
+
+TEST_BEGIN(test_smoothstep_integral)
+{
+	uint64_t sum, min, max;
+	unsigned i;
+
+	/*
+	 * The integral of smoothstep in the [0..1] range equals 1/2.  Verify
+	 * that the fixed point representation's integral is no more than
+	 * rounding error distant from 1/2.  Regarding rounding, each table
+	 * element is rounded down to the nearest fixed point value, so the
+	 * integral may be off by as much as SMOOTHSTEP_NSTEPS ulps.
+	 */
+	sum = 0;
+	for (i = 0; i < SMOOTHSTEP_NSTEPS; i++)
+		sum += smoothstep_tab[i];
+
+	max = (KQU(1) << (SMOOTHSTEP_BFP-1)) * (SMOOTHSTEP_NSTEPS+1);
+	min = max - SMOOTHSTEP_NSTEPS;
+
+	assert_u64_ge(sum, min,
+	    "Integral too small, even accounting for truncation");
+	assert_u64_le(sum, max, "Integral exceeds 1/2");
+	if (false) {
+		malloc_printf("%"FMTu64" ulps under 1/2 (limit %d)\n",
+		    max - sum, SMOOTHSTEP_NSTEPS);
+	}
+}
+TEST_END
+
+TEST_BEGIN(test_smoothstep_monotonic)
+{
+	uint64_t prev_h;
+	unsigned i;
+
+	/*
+	 * The smoothstep function is monotonic in [0..1], i.e. its slope is
+	 * non-negative.  In practice we want to parametrize table generation
+	 * such that piecewise slope is greater than zero, but do not require
+	 * that here.
+	 */
+	prev_h = 0;
+	for (i = 0; i < SMOOTHSTEP_NSTEPS; i++) {
+		uint64_t h = smoothstep_tab[i];
+		assert_u64_ge(h, prev_h, "Piecewise non-monotonic, i=%u", i);
+		prev_h = h;
+	}
+	assert_u64_eq(smoothstep_tab[SMOOTHSTEP_NSTEPS-1],
+	    (KQU(1) << SMOOTHSTEP_BFP), "Last step must equal 1");
+}
+TEST_END
+
+TEST_BEGIN(test_smoothstep_slope)
+{
+	uint64_t prev_h, prev_delta;
+	unsigned i;
+
+	/*
+	 * The smoothstep slope strictly increases until x=0.5, and then
+	 * strictly decreases until x=1.0.  Verify the slightly weaker
+	 * requirement of monotonicity, so that inadequate table precision does
+	 * not cause false test failures.
+	 */
+	prev_h = 0;
+	prev_delta = 0;
+	for (i = 0; i < SMOOTHSTEP_NSTEPS / 2 + SMOOTHSTEP_NSTEPS % 2; i++) {
+		uint64_t h = smoothstep_tab[i];
+		uint64_t delta = h - prev_h;
+		assert_u64_ge(delta, prev_delta,
+		    "Slope must monotonically increase in 0.0 <= x <= 0.5, "
+		    "i=%u", i);
+		prev_h = h;
+		prev_delta = delta;
+	}
+
+	prev_h = KQU(1) << SMOOTHSTEP_BFP;
+	prev_delta = 0;
+	for (i = SMOOTHSTEP_NSTEPS-1; i >= SMOOTHSTEP_NSTEPS / 2; i--) {
+		uint64_t h = smoothstep_tab[i];
+		uint64_t delta = prev_h - h;
+		assert_u64_ge(delta, prev_delta,
+		    "Slope must monotonically decrease in 0.5 <= x <= 1.0, "
+		    "i=%u", i);
+		prev_h = h;
+		prev_delta = delta;
+	}
+}
+TEST_END
+
+int
+main(void)
+{
+
+	return (test(
+	    test_smoothstep_integral,
+	    test_smoothstep_monotonic,
+	    test_smoothstep_slope));
+}
diff --git a/deps/jemalloc/test/unit/stats.c b/deps/jemalloc/test/unit/stats.c
old mode 100644
new mode 100755
index 8e4bc631e..315717dfb
--- a/deps/jemalloc/test/unit/stats.c
+++ b/deps/jemalloc/test/unit/stats.c
@@ -7,18 +7,18 @@ TEST_BEGIN(test_stats_summary)
 	int expected = config_stats ? 0 : ENOENT;
 
 	sz = sizeof(cactive);
-	assert_d_eq(mallctl("stats.cactive", &cactive, &sz, NULL, 0), expected,
-	    "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.cactive", (void *)&cactive, &sz, NULL, 0),
+	    expected, "Unexpected mallctl() result");
 
 	sz = sizeof(size_t);
-	assert_d_eq(mallctl("stats.allocated", &allocated, &sz, NULL, 0),
+	assert_d_eq(mallctl("stats.allocated", (void *)&allocated, &sz, NULL,
+	    0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.active", (void *)&active, &sz, NULL, 0),
 	    expected, "Unexpected mallctl() result");
-	assert_d_eq(mallctl("stats.active", &active, &sz, NULL, 0), expected,
-	    "Unexpected mallctl() result");
-	assert_d_eq(mallctl("stats.resident", &resident, &sz, NULL, 0),
+	assert_d_eq(mallctl("stats.resident", (void *)&resident, &sz, NULL, 0),
+	    expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.mapped", (void *)&mapped, &sz, NULL, 0),
 	    expected, "Unexpected mallctl() result");
-	assert_d_eq(mallctl("stats.mapped", &mapped, &sz, NULL, 0), expected,
-	    "Unexpected mallctl() result");
 
 	if (config_stats) {
 		assert_zu_le(active, *cactive,
@@ -45,19 +45,19 @@ TEST_BEGIN(test_stats_huge)
 	p = mallocx(large_maxclass+1, 0);
 	assert_ptr_not_null(p, "Unexpected mallocx() failure");
 
-	assert_d_eq(mallctl("epoch", NULL, NULL, &epoch, sizeof(epoch)), 0,
-	    "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, sizeof(epoch)),
+	    0, "Unexpected mallctl() failure");
 
 	sz = sizeof(size_t);
-	assert_d_eq(mallctl("stats.arenas.0.huge.allocated", &allocated, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.huge.allocated", (void *)&allocated,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
 	sz = sizeof(uint64_t);
-	assert_d_eq(mallctl("stats.arenas.0.huge.nmalloc", &nmalloc, &sz, NULL,
-	    0), expected, "Unexpected mallctl() result");
-	assert_d_eq(mallctl("stats.arenas.0.huge.ndalloc", &ndalloc, &sz, NULL,
-	    0), expected, "Unexpected mallctl() result");
-	assert_d_eq(mallctl("stats.arenas.0.huge.nrequests", &nrequests, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.huge.nmalloc", (void *)&nmalloc,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.huge.ndalloc", (void *)&ndalloc,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.huge.nrequests", (void *)&nrequests,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
 
 	if (config_stats) {
 		assert_zu_gt(allocated, 0,
@@ -83,8 +83,8 @@ TEST_BEGIN(test_stats_arenas_summary)
 	uint64_t npurge, nmadvise, purged;
 
 	arena = 0;
-	assert_d_eq(mallctl("thread.arena", NULL, NULL, &arena, sizeof(arena)),
-	    0, "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("thread.arena", NULL, NULL, (void *)&arena,
+	    sizeof(arena)), 0, "Unexpected mallctl() failure");
 
 	little = mallocx(SMALL_MAXCLASS, 0);
 	assert_ptr_not_null(little, "Unexpected mallocx() failure");
@@ -93,22 +93,26 @@ TEST_BEGIN(test_stats_arenas_summary)
 	huge = mallocx(chunksize, 0);
 	assert_ptr_not_null(huge, "Unexpected mallocx() failure");
 
+	dallocx(little, 0);
+	dallocx(large, 0);
+	dallocx(huge, 0);
+
 	assert_d_eq(mallctl("arena.0.purge", NULL, NULL, NULL, 0), 0,
 	    "Unexpected mallctl() failure");
 
-	assert_d_eq(mallctl("epoch", NULL, NULL, &epoch, sizeof(epoch)), 0,
-	    "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, sizeof(epoch)),
+	    0, "Unexpected mallctl() failure");
 
 	sz = sizeof(size_t);
-	assert_d_eq(mallctl("stats.arenas.0.mapped", &mapped, &sz, NULL, 0),
-	    expected, "Unexepected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.mapped", (void *)&mapped, &sz, NULL,
+	    0), expected, "Unexepected mallctl() result");
 	sz = sizeof(uint64_t);
-	assert_d_eq(mallctl("stats.arenas.0.npurge", &npurge, &sz, NULL, 0),
-	    expected, "Unexepected mallctl() result");
-	assert_d_eq(mallctl("stats.arenas.0.nmadvise", &nmadvise, &sz, NULL, 0),
-	    expected, "Unexepected mallctl() result");
-	assert_d_eq(mallctl("stats.arenas.0.purged", &purged, &sz, NULL, 0),
-	    expected, "Unexepected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.npurge", (void *)&npurge, &sz, NULL,
+	    0), expected, "Unexepected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.nmadvise", (void *)&nmadvise, &sz,
+	    NULL, 0), expected, "Unexepected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.purged", (void *)&purged, &sz, NULL,
+	    0), expected, "Unexepected mallctl() result");
 
 	if (config_stats) {
 		assert_u64_gt(npurge, 0,
@@ -116,10 +120,6 @@ TEST_BEGIN(test_stats_arenas_summary)
 		assert_u64_le(nmadvise, purged,
 		    "nmadvise should be no greater than purged");
 	}
-
-	dallocx(little, 0);
-	dallocx(large, 0);
-	dallocx(huge, 0);
 }
 TEST_END
 
@@ -150,8 +150,8 @@ TEST_BEGIN(test_stats_arenas_small)
 	no_lazy_lock(); /* Lazy locking would dodge tcache testing. */
 
 	arena = 0;
-	assert_d_eq(mallctl("thread.arena", NULL, NULL, &arena, sizeof(arena)),
-	    0, "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("thread.arena", NULL, NULL, (void *)&arena,
+	    sizeof(arena)), 0, "Unexpected mallctl() failure");
 
 	p = mallocx(SMALL_MAXCLASS, 0);
 	assert_ptr_not_null(p, "Unexpected mallocx() failure");
@@ -159,19 +159,21 @@ TEST_BEGIN(test_stats_arenas_small)
 	assert_d_eq(mallctl("thread.tcache.flush", NULL, NULL, NULL, 0),
 	    config_tcache ? 0 : ENOENT, "Unexpected mallctl() result");
 
-	assert_d_eq(mallctl("epoch", NULL, NULL, &epoch, sizeof(epoch)), 0,
-	    "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, sizeof(epoch)),
+	    0, "Unexpected mallctl() failure");
 
 	sz = sizeof(size_t);
-	assert_d_eq(mallctl("stats.arenas.0.small.allocated", &allocated, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.small.allocated",
+	    (void *)&allocated, &sz, NULL, 0), expected,
+	    "Unexpected mallctl() result");
 	sz = sizeof(uint64_t);
-	assert_d_eq(mallctl("stats.arenas.0.small.nmalloc", &nmalloc, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
-	assert_d_eq(mallctl("stats.arenas.0.small.ndalloc", &ndalloc, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
-	assert_d_eq(mallctl("stats.arenas.0.small.nrequests", &nrequests, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.small.nmalloc", (void *)&nmalloc,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.small.ndalloc", (void *)&ndalloc,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.small.nrequests",
+	    (void *)&nrequests, &sz, NULL, 0), expected,
+	    "Unexpected mallctl() result");
 
 	if (config_stats) {
 		assert_zu_gt(allocated, 0,
@@ -197,34 +199,36 @@ TEST_BEGIN(test_stats_arenas_large)
 	int expected = config_stats ? 0 : ENOENT;
 
 	arena = 0;
-	assert_d_eq(mallctl("thread.arena", NULL, NULL, &arena, sizeof(arena)),
-	    0, "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("thread.arena", NULL, NULL, (void *)&arena,
+	    sizeof(arena)), 0, "Unexpected mallctl() failure");
 
 	p = mallocx(large_maxclass, 0);
 	assert_ptr_not_null(p, "Unexpected mallocx() failure");
 
-	assert_d_eq(mallctl("epoch", NULL, NULL, &epoch, sizeof(epoch)), 0,
-	    "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, sizeof(epoch)),
+	    0, "Unexpected mallctl() failure");
 
 	sz = sizeof(size_t);
-	assert_d_eq(mallctl("stats.arenas.0.large.allocated", &allocated, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.large.allocated",
+	    (void *)&allocated, &sz, NULL, 0), expected,
+	    "Unexpected mallctl() result");
 	sz = sizeof(uint64_t);
-	assert_d_eq(mallctl("stats.arenas.0.large.nmalloc", &nmalloc, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
-	assert_d_eq(mallctl("stats.arenas.0.large.ndalloc", &ndalloc, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
-	assert_d_eq(mallctl("stats.arenas.0.large.nrequests", &nrequests, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.large.nmalloc", (void *)&nmalloc,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.large.ndalloc", (void *)&ndalloc,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.large.nrequests",
+	    (void *)&nrequests, &sz, NULL, 0), expected,
+	    "Unexpected mallctl() result");
 
 	if (config_stats) {
 		assert_zu_gt(allocated, 0,
 		    "allocated should be greater than zero");
-		assert_zu_gt(nmalloc, 0,
+		assert_u64_gt(nmalloc, 0,
 		    "nmalloc should be greater than zero");
-		assert_zu_ge(nmalloc, ndalloc,
+		assert_u64_ge(nmalloc, ndalloc,
 		    "nmalloc should be at least as large as ndalloc");
-		assert_zu_gt(nrequests, 0,
+		assert_u64_gt(nrequests, 0,
 		    "nrequests should be greater than zero");
 	}
 
@@ -241,30 +245,30 @@ TEST_BEGIN(test_stats_arenas_huge)
 	int expected = config_stats ? 0 : ENOENT;
 
 	arena = 0;
-	assert_d_eq(mallctl("thread.arena", NULL, NULL, &arena, sizeof(arena)),
-	    0, "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("thread.arena", NULL, NULL, (void *)&arena,
+	    sizeof(arena)), 0, "Unexpected mallctl() failure");
 
 	p = mallocx(chunksize, 0);
 	assert_ptr_not_null(p, "Unexpected mallocx() failure");
 
-	assert_d_eq(mallctl("epoch", NULL, NULL, &epoch, sizeof(epoch)), 0,
-	    "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, sizeof(epoch)),
+	    0, "Unexpected mallctl() failure");
 
 	sz = sizeof(size_t);
-	assert_d_eq(mallctl("stats.arenas.0.huge.allocated", &allocated, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.huge.allocated", (void *)&allocated,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
 	sz = sizeof(uint64_t);
-	assert_d_eq(mallctl("stats.arenas.0.huge.nmalloc", &nmalloc, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
-	assert_d_eq(mallctl("stats.arenas.0.huge.ndalloc", &ndalloc, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.huge.nmalloc", (void *)&nmalloc,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.huge.ndalloc", (void *)&ndalloc,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
 
 	if (config_stats) {
 		assert_zu_gt(allocated, 0,
 		    "allocated should be greater than zero");
-		assert_zu_gt(nmalloc, 0,
+		assert_u64_gt(nmalloc, 0,
 		    "nmalloc should be greater than zero");
-		assert_zu_ge(nmalloc, ndalloc,
+		assert_u64_ge(nmalloc, ndalloc,
 		    "nmalloc should be at least as large as ndalloc");
 	}
 
@@ -282,8 +286,8 @@ TEST_BEGIN(test_stats_arenas_bins)
 	int expected = config_stats ? 0 : ENOENT;
 
 	arena = 0;
-	assert_d_eq(mallctl("thread.arena", NULL, NULL, &arena, sizeof(arena)),
-	    0, "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("thread.arena", NULL, NULL, (void *)&arena,
+	    sizeof(arena)), 0, "Unexpected mallctl() failure");
 
 	p = mallocx(arena_bin_info[0].reg_size, 0);
 	assert_ptr_not_null(p, "Unexpected mallocx() failure");
@@ -291,35 +295,36 @@ TEST_BEGIN(test_stats_arenas_bins)
 	assert_d_eq(mallctl("thread.tcache.flush", NULL, NULL, NULL, 0),
 	    config_tcache ? 0 : ENOENT, "Unexpected mallctl() result");
 
-	assert_d_eq(mallctl("epoch", NULL, NULL, &epoch, sizeof(epoch)), 0,
-	    "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, sizeof(epoch)),
+	    0, "Unexpected mallctl() failure");
 
 	sz = sizeof(uint64_t);
-	assert_d_eq(mallctl("stats.arenas.0.bins.0.nmalloc", &nmalloc, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
-	assert_d_eq(mallctl("stats.arenas.0.bins.0.ndalloc", &ndalloc, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
-	assert_d_eq(mallctl("stats.arenas.0.bins.0.nrequests", &nrequests, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.bins.0.nmalloc", (void *)&nmalloc,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.bins.0.ndalloc", (void *)&ndalloc,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.bins.0.nrequests",
+	    (void *)&nrequests, &sz, NULL, 0), expected,
+	    "Unexpected mallctl() result");
 	sz = sizeof(size_t);
-	assert_d_eq(mallctl("stats.arenas.0.bins.0.curregs", &curregs, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.bins.0.curregs", (void *)&curregs,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
 
 	sz = sizeof(uint64_t);
-	assert_d_eq(mallctl("stats.arenas.0.bins.0.nfills", &nfills, &sz,
-	    NULL, 0), config_tcache ? expected : ENOENT,
+	assert_d_eq(mallctl("stats.arenas.0.bins.0.nfills", (void *)&nfills,
+	    &sz, NULL, 0), config_tcache ? expected : ENOENT,
 	    "Unexpected mallctl() result");
-	assert_d_eq(mallctl("stats.arenas.0.bins.0.nflushes", &nflushes, &sz,
-	    NULL, 0), config_tcache ? expected : ENOENT,
+	assert_d_eq(mallctl("stats.arenas.0.bins.0.nflushes", (void *)&nflushes,
+	    &sz, NULL, 0), config_tcache ? expected : ENOENT,
 	    "Unexpected mallctl() result");
 
-	assert_d_eq(mallctl("stats.arenas.0.bins.0.nruns", &nruns, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
-	assert_d_eq(mallctl("stats.arenas.0.bins.0.nreruns", &nreruns, &sz,
+	assert_d_eq(mallctl("stats.arenas.0.bins.0.nruns", (void *)&nruns, &sz,
 	    NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.bins.0.nreruns", (void *)&nreruns,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
 	sz = sizeof(size_t);
-	assert_d_eq(mallctl("stats.arenas.0.bins.0.curruns", &curruns, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.bins.0.curruns", (void *)&curruns,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
 
 	if (config_stats) {
 		assert_u64_gt(nmalloc, 0,
@@ -355,25 +360,26 @@ TEST_BEGIN(test_stats_arenas_lruns)
 	int expected = config_stats ? 0 : ENOENT;
 
 	arena = 0;
-	assert_d_eq(mallctl("thread.arena", NULL, NULL, &arena, sizeof(arena)),
-	    0, "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("thread.arena", NULL, NULL, (void *)&arena,
+	    sizeof(arena)), 0, "Unexpected mallctl() failure");
 
 	p = mallocx(LARGE_MINCLASS, 0);
 	assert_ptr_not_null(p, "Unexpected mallocx() failure");
 
-	assert_d_eq(mallctl("epoch", NULL, NULL, &epoch, sizeof(epoch)), 0,
-	    "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, sizeof(epoch)),
+	    0, "Unexpected mallctl() failure");
 
 	sz = sizeof(uint64_t);
-	assert_d_eq(mallctl("stats.arenas.0.lruns.0.nmalloc", &nmalloc, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
-	assert_d_eq(mallctl("stats.arenas.0.lruns.0.ndalloc", &ndalloc, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
-	assert_d_eq(mallctl("stats.arenas.0.lruns.0.nrequests", &nrequests, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.lruns.0.nmalloc", (void *)&nmalloc,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.lruns.0.ndalloc", (void *)&ndalloc,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.lruns.0.nrequests",
+	    (void *)&nrequests, &sz, NULL, 0), expected,
+	    "Unexpected mallctl() result");
 	sz = sizeof(size_t);
-	assert_d_eq(mallctl("stats.arenas.0.lruns.0.curruns", &curruns, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.lruns.0.curruns", (void *)&curruns,
+	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
 
 	if (config_stats) {
 		assert_u64_gt(nmalloc, 0,
@@ -399,23 +405,26 @@ TEST_BEGIN(test_stats_arenas_hchunks)
 	int expected = config_stats ? 0 : ENOENT;
 
 	arena = 0;
-	assert_d_eq(mallctl("thread.arena", NULL, NULL, &arena, sizeof(arena)),
-	    0, "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("thread.arena", NULL, NULL, (void *)&arena,
+	    sizeof(arena)), 0, "Unexpected mallctl() failure");
 
 	p = mallocx(chunksize, 0);
 	assert_ptr_not_null(p, "Unexpected mallocx() failure");
 
-	assert_d_eq(mallctl("epoch", NULL, NULL, &epoch, sizeof(epoch)), 0,
-	    "Unexpected mallctl() failure");
+	assert_d_eq(mallctl("epoch", NULL, NULL, (void *)&epoch, sizeof(epoch)),
+	    0, "Unexpected mallctl() failure");
 
 	sz = sizeof(uint64_t);
-	assert_d_eq(mallctl("stats.arenas.0.hchunks.0.nmalloc", &nmalloc, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
-	assert_d_eq(mallctl("stats.arenas.0.hchunks.0.ndalloc", &ndalloc, &sz,
-	    NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.hchunks.0.nmalloc",
+	    (void *)&nmalloc, &sz, NULL, 0), expected,
+	    "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.hchunks.0.ndalloc",
+	    (void *)&ndalloc, &sz, NULL, 0), expected,
+	    "Unexpected mallctl() result");
 	sz = sizeof(size_t);
-	assert_d_eq(mallctl("stats.arenas.0.hchunks.0.curhchunks", &curhchunks,
-	    &sz, NULL, 0), expected, "Unexpected mallctl() result");
+	assert_d_eq(mallctl("stats.arenas.0.hchunks.0.curhchunks",
+	    (void *)&curhchunks, &sz, NULL, 0), expected,
+	    "Unexpected mallctl() result");
 
 	if (config_stats) {
 		assert_u64_gt(nmalloc, 0,
diff --git a/deps/jemalloc/test/unit/ticker.c b/deps/jemalloc/test/unit/ticker.c
new file mode 100644
index 000000000..e737020ab
--- /dev/null
+++ b/deps/jemalloc/test/unit/ticker.c
@@ -0,0 +1,76 @@
+#include "test/jemalloc_test.h"
+
+TEST_BEGIN(test_ticker_tick)
+{
+#define	NREPS 2
+#define	NTICKS 3
+	ticker_t ticker;
+	int32_t i, j;
+
+	ticker_init(&ticker, NTICKS);
+	for (i = 0; i < NREPS; i++) {
+		for (j = 0; j < NTICKS; j++) {
+			assert_u_eq(ticker_read(&ticker), NTICKS - j,
+			    "Unexpected ticker value (i=%d, j=%d)", i, j);
+			assert_false(ticker_tick(&ticker),
+			    "Unexpected ticker fire (i=%d, j=%d)", i, j);
+		}
+		assert_u32_eq(ticker_read(&ticker), 0,
+		    "Expected ticker depletion");
+		assert_true(ticker_tick(&ticker),
+		    "Expected ticker fire (i=%d)", i);
+		assert_u32_eq(ticker_read(&ticker), NTICKS,
+		    "Expected ticker reset");
+	}
+#undef NTICKS
+}
+TEST_END
+
+TEST_BEGIN(test_ticker_ticks)
+{
+#define	NTICKS 3
+	ticker_t ticker;
+
+	ticker_init(&ticker, NTICKS);
+
+	assert_u_eq(ticker_read(&ticker), NTICKS, "Unexpected ticker value");
+	assert_false(ticker_ticks(&ticker, NTICKS), "Unexpected ticker fire");
+	assert_u_eq(ticker_read(&ticker), 0, "Unexpected ticker value");
+	assert_true(ticker_ticks(&ticker, NTICKS), "Expected ticker fire");
+	assert_u_eq(ticker_read(&ticker), NTICKS, "Unexpected ticker value");
+
+	assert_true(ticker_ticks(&ticker, NTICKS + 1), "Expected ticker fire");
+	assert_u_eq(ticker_read(&ticker), NTICKS, "Unexpected ticker value");
+#undef NTICKS
+}
+TEST_END
+
+TEST_BEGIN(test_ticker_copy)
+{
+#define	NTICKS 3
+	ticker_t ta, tb;
+
+	ticker_init(&ta, NTICKS);
+	ticker_copy(&tb, &ta);
+	assert_u_eq(ticker_read(&tb), NTICKS, "Unexpected ticker value");
+	assert_true(ticker_ticks(&tb, NTICKS + 1), "Expected ticker fire");
+	assert_u_eq(ticker_read(&tb), NTICKS, "Unexpected ticker value");
+
+	ticker_tick(&ta);
+	ticker_copy(&tb, &ta);
+	assert_u_eq(ticker_read(&tb), NTICKS - 1, "Unexpected ticker value");
+	assert_true(ticker_ticks(&tb, NTICKS), "Expected ticker fire");
+	assert_u_eq(ticker_read(&tb), NTICKS, "Unexpected ticker value");
+#undef NTICKS
+}
+TEST_END
+
+int
+main(void)
+{
+
+	return (test(
+	    test_ticker_tick,
+	    test_ticker_ticks,
+	    test_ticker_copy));
+}
diff --git a/deps/jemalloc/test/unit/tsd.c b/deps/jemalloc/test/unit/tsd.c
index 8be787fda..d5f96ac36 100644
--- a/deps/jemalloc/test/unit/tsd.c
+++ b/deps/jemalloc/test/unit/tsd.c
@@ -58,18 +58,18 @@ thd_start(void *arg)
 	data_t d = (data_t)(uintptr_t)arg;
 	void *p;
 
-	assert_x_eq(*data_tsd_get(), DATA_INIT,
+	assert_x_eq(*data_tsd_get(true), DATA_INIT,
 	    "Initial tsd get should return initialization value");
 
 	p = malloc(1);
 	assert_ptr_not_null(p, "Unexpected malloc() failure");
 
 	data_tsd_set(&d);
-	assert_x_eq(*data_tsd_get(), d,
+	assert_x_eq(*data_tsd_get(true), d,
 	    "After tsd set, tsd get should return value that was set");
 
 	d = 0;
-	assert_x_eq(*data_tsd_get(), (data_t)(uintptr_t)arg,
+	assert_x_eq(*data_tsd_get(true), (data_t)(uintptr_t)arg,
 	    "Resetting local data should have no effect on tsd");
 
 	free(p);
@@ -79,7 +79,7 @@ thd_start(void *arg)
 TEST_BEGIN(test_tsd_main_thread)
 {
 
-	thd_start((void *) 0xa5f3e329);
+	thd_start((void *)(uintptr_t)0xa5f3e329);
 }
 TEST_END
 
@@ -99,6 +99,11 @@ int
 main(void)
 {
 
+	/* Core tsd bootstrapping must happen prior to data_tsd_boot(). */
+	if (nallocx(1, 0) == 0) {
+		malloc_printf("Initialization error");
+		return (test_status_fail);
+	}
 	data_tsd_boot();
 
 	return (test(
diff --git a/deps/jemalloc/test/unit/util.c b/deps/jemalloc/test/unit/util.c
index 8ab39a458..b1f9abd9b 100644
--- a/deps/jemalloc/test/unit/util.c
+++ b/deps/jemalloc/test/unit/util.c
@@ -1,33 +1,54 @@
 #include "test/jemalloc_test.h"
 
-TEST_BEGIN(test_pow2_ceil)
+#define	TEST_POW2_CEIL(t, suf, pri) do {				\
+	unsigned i, pow2;						\
+	t x;								\
+									\
+	assert_##suf##_eq(pow2_ceil_##suf(0), 0, "Unexpected result");	\
+									\
+	for (i = 0; i < sizeof(t) * 8; i++) {				\
+		assert_##suf##_eq(pow2_ceil_##suf(((t)1) << i), ((t)1)	\
+		    << i, "Unexpected result");				\
+	}								\
+									\
+	for (i = 2; i < sizeof(t) * 8; i++) {				\
+		assert_##suf##_eq(pow2_ceil_##suf((((t)1) << i) - 1),	\
+		    ((t)1) << i, "Unexpected result");			\
+	}								\
+									\
+	for (i = 0; i < sizeof(t) * 8 - 1; i++) {			\
+		assert_##suf##_eq(pow2_ceil_##suf((((t)1) << i) + 1),	\
+		    ((t)1) << (i+1), "Unexpected result");		\
+	}								\
+									\
+	for (pow2 = 1; pow2 < 25; pow2++) {				\
+		for (x = (((t)1) << (pow2-1)) + 1; x <= ((t)1) << pow2;	\
+		    x++) {						\
+			assert_##suf##_eq(pow2_ceil_##suf(x),		\
+			    ((t)1) << pow2,				\
+			    "Unexpected result, x=%"pri, x);		\
+		}							\
+	}								\
+} while (0)
+
+TEST_BEGIN(test_pow2_ceil_u64)
 {
-	unsigned i, pow2;
-	size_t x;
 
-	assert_zu_eq(pow2_ceil(0), 0, "Unexpected result");
+	TEST_POW2_CEIL(uint64_t, u64, FMTu64);
+}
+TEST_END
 
-	for (i = 0; i < sizeof(size_t) * 8; i++) {
-		assert_zu_eq(pow2_ceil(ZU(1) << i), ZU(1) << i,
-		    "Unexpected result");
-	}
+TEST_BEGIN(test_pow2_ceil_u32)
+{
 
-	for (i = 2; i < sizeof(size_t) * 8; i++) {
-		assert_zu_eq(pow2_ceil((ZU(1) << i) - 1), ZU(1) << i,
-		    "Unexpected result");
-	}
+	TEST_POW2_CEIL(uint32_t, u32, FMTu32);
+}
+TEST_END
 
-	for (i = 0; i < sizeof(size_t) * 8 - 1; i++) {
-		assert_zu_eq(pow2_ceil((ZU(1) << i) + 1), ZU(1) << (i+1),
-		    "Unexpected result");
-	}
+TEST_BEGIN(test_pow2_ceil_zu)
+{
 
-	for (pow2 = 1; pow2 < 25; pow2++) {
-		for (x = (ZU(1) << (pow2-1)) + 1; x <= ZU(1) << pow2; x++) {
-			assert_zu_eq(pow2_ceil(x), ZU(1) << pow2,
-			    "Unexpected result, x=%zu", x);
-		}
-	}
+	TEST_POW2_CEIL(size_t, zu, "zu");
 }
 TEST_END
 
@@ -54,6 +75,7 @@ TEST_BEGIN(test_malloc_strtoumax)
 	};
 #define	ERR(e)		e, #e
 #define	KUMAX(x)	((uintmax_t)x##ULL)
+#define	KSMAX(x)	((uintmax_t)(intmax_t)x##LL)
 	struct test_s tests[] = {
 		{"0",		"0",	-1,	ERR(EINVAL),	UINTMAX_MAX},
 		{"0",		"0",	1,	ERR(EINVAL),	UINTMAX_MAX},
@@ -66,13 +88,13 @@ TEST_BEGIN(test_malloc_strtoumax)
 
 		{"42",		"",	0,	ERR(0),		KUMAX(42)},
 		{"+42",		"",	0,	ERR(0),		KUMAX(42)},
-		{"-42",		"",	0,	ERR(0),		KUMAX(-42)},
+		{"-42",		"",	0,	ERR(0),		KSMAX(-42)},
 		{"042",		"",	0,	ERR(0),		KUMAX(042)},
 		{"+042",	"",	0,	ERR(0),		KUMAX(042)},
-		{"-042",	"",	0,	ERR(0),		KUMAX(-042)},
+		{"-042",	"",	0,	ERR(0),		KSMAX(-042)},
 		{"0x42",	"",	0,	ERR(0),		KUMAX(0x42)},
 		{"+0x42",	"",	0,	ERR(0),		KUMAX(0x42)},
-		{"-0x42",	"",	0,	ERR(0),		KUMAX(-0x42)},
+		{"-0x42",	"",	0,	ERR(0),		KSMAX(-0x42)},
 
 		{"0",		"",	0,	ERR(0),		KUMAX(0)},
 		{"1",		"",	0,	ERR(0),		KUMAX(1)},
@@ -109,6 +131,7 @@ TEST_BEGIN(test_malloc_strtoumax)
 	};
 #undef ERR
 #undef KUMAX
+#undef KSMAX
 	unsigned i;
 
 	for (i = 0; i < sizeof(tests)/sizeof(struct test_s); i++) {
@@ -139,14 +162,14 @@ TEST_BEGIN(test_malloc_snprintf_truncated)
 {
 #define	BUFLEN	15
 	char buf[BUFLEN];
-	int result;
+	size_t result;
 	size_t len;
-#define TEST(expected_str_untruncated, ...) do {			\
+#define	TEST(expected_str_untruncated, ...) do {			\
 	result = malloc_snprintf(buf, len, __VA_ARGS__);		\
 	assert_d_eq(strncmp(buf, expected_str_untruncated, len-1), 0,	\
 	    "Unexpected string inequality (\"%s\" vs \"%s\")",		\
-	    buf, expected_str_untruncated);		\
-	assert_d_eq(result, strlen(expected_str_untruncated),		\
+	    buf, expected_str_untruncated);				\
+	assert_zu_eq(result, strlen(expected_str_untruncated),		\
 	    "Unexpected result");					\
 } while (0)
 
@@ -172,11 +195,11 @@ TEST_BEGIN(test_malloc_snprintf)
 {
 #define	BUFLEN	128
 	char buf[BUFLEN];
-	int result;
+	size_t result;
 #define	TEST(expected_str, ...) do {					\
 	result = malloc_snprintf(buf, sizeof(buf), __VA_ARGS__);	\
 	assert_str_eq(buf, expected_str, "Unexpected output");		\
-	assert_d_eq(result, strlen(expected_str), "Unexpected result");	\
+	assert_zu_eq(result, strlen(expected_str), "Unexpected result");\
 } while (0)
 
 	TEST("hello", "hello");
@@ -286,7 +309,9 @@ main(void)
 {
 
 	return (test(
-	    test_pow2_ceil,
+	    test_pow2_ceil_u64,
+	    test_pow2_ceil_u32,
+	    test_pow2_ceil_zu,
 	    test_malloc_strtoumax_no_endptr,
 	    test_malloc_strtoumax,
 	    test_malloc_snprintf_truncated,
diff --git a/deps/jemalloc/test/unit/witness.c b/deps/jemalloc/test/unit/witness.c
new file mode 100644
index 000000000..ed172753c
--- /dev/null
+++ b/deps/jemalloc/test/unit/witness.c
@@ -0,0 +1,278 @@
+#include "test/jemalloc_test.h"
+
+static witness_lock_error_t *witness_lock_error_orig;
+static witness_owner_error_t *witness_owner_error_orig;
+static witness_not_owner_error_t *witness_not_owner_error_orig;
+static witness_lockless_error_t *witness_lockless_error_orig;
+
+static bool saw_lock_error;
+static bool saw_owner_error;
+static bool saw_not_owner_error;
+static bool saw_lockless_error;
+
+static void
+witness_lock_error_intercept(const witness_list_t *witnesses,
+    const witness_t *witness)
+{
+
+	saw_lock_error = true;
+}
+
+static void
+witness_owner_error_intercept(const witness_t *witness)
+{
+
+	saw_owner_error = true;
+}
+
+static void
+witness_not_owner_error_intercept(const witness_t *witness)
+{
+
+	saw_not_owner_error = true;
+}
+
+static void
+witness_lockless_error_intercept(const witness_list_t *witnesses)
+{
+
+	saw_lockless_error = true;
+}
+
+static int
+witness_comp(const witness_t *a, const witness_t *b)
+{
+
+	assert_u_eq(a->rank, b->rank, "Witnesses should have equal rank");
+
+	return (strcmp(a->name, b->name));
+}
+
+static int
+witness_comp_reverse(const witness_t *a, const witness_t *b)
+{
+
+	assert_u_eq(a->rank, b->rank, "Witnesses should have equal rank");
+
+	return (-strcmp(a->name, b->name));
+}
+
+TEST_BEGIN(test_witness)
+{
+	witness_t a, b;
+	tsdn_t *tsdn;
+
+	test_skip_if(!config_debug);
+
+	tsdn = tsdn_fetch();
+
+	witness_assert_lockless(tsdn);
+
+	witness_init(&a, "a", 1, NULL);
+	witness_assert_not_owner(tsdn, &a);
+	witness_lock(tsdn, &a);
+	witness_assert_owner(tsdn, &a);
+
+	witness_init(&b, "b", 2, NULL);
+	witness_assert_not_owner(tsdn, &b);
+	witness_lock(tsdn, &b);
+	witness_assert_owner(tsdn, &b);
+
+	witness_unlock(tsdn, &a);
+	witness_unlock(tsdn, &b);
+
+	witness_assert_lockless(tsdn);
+}
+TEST_END
+
+TEST_BEGIN(test_witness_comp)
+{
+	witness_t a, b, c, d;
+	tsdn_t *tsdn;
+
+	test_skip_if(!config_debug);
+
+	tsdn = tsdn_fetch();
+
+	witness_assert_lockless(tsdn);
+
+	witness_init(&a, "a", 1, witness_comp);
+	witness_assert_not_owner(tsdn, &a);
+	witness_lock(tsdn, &a);
+	witness_assert_owner(tsdn, &a);
+
+	witness_init(&b, "b", 1, witness_comp);
+	witness_assert_not_owner(tsdn, &b);
+	witness_lock(tsdn, &b);
+	witness_assert_owner(tsdn, &b);
+	witness_unlock(tsdn, &b);
+
+	witness_lock_error_orig = witness_lock_error;
+	witness_lock_error = witness_lock_error_intercept;
+	saw_lock_error = false;
+
+	witness_init(&c, "c", 1, witness_comp_reverse);
+	witness_assert_not_owner(tsdn, &c);
+	assert_false(saw_lock_error, "Unexpected witness lock error");
+	witness_lock(tsdn, &c);
+	assert_true(saw_lock_error, "Expected witness lock error");
+	witness_unlock(tsdn, &c);
+
+	saw_lock_error = false;
+
+	witness_init(&d, "d", 1, NULL);
+	witness_assert_not_owner(tsdn, &d);
+	assert_false(saw_lock_error, "Unexpected witness lock error");
+	witness_lock(tsdn, &d);
+	assert_true(saw_lock_error, "Expected witness lock error");
+	witness_unlock(tsdn, &d);
+
+	witness_unlock(tsdn, &a);
+
+	witness_assert_lockless(tsdn);
+
+	witness_lock_error = witness_lock_error_orig;
+}
+TEST_END
+
+TEST_BEGIN(test_witness_reversal)
+{
+	witness_t a, b;
+	tsdn_t *tsdn;
+
+	test_skip_if(!config_debug);
+
+	witness_lock_error_orig = witness_lock_error;
+	witness_lock_error = witness_lock_error_intercept;
+	saw_lock_error = false;
+
+	tsdn = tsdn_fetch();
+
+	witness_assert_lockless(tsdn);
+
+	witness_init(&a, "a", 1, NULL);
+	witness_init(&b, "b", 2, NULL);
+
+	witness_lock(tsdn, &b);
+	assert_false(saw_lock_error, "Unexpected witness lock error");
+	witness_lock(tsdn, &a);
+	assert_true(saw_lock_error, "Expected witness lock error");
+
+	witness_unlock(tsdn, &a);
+	witness_unlock(tsdn, &b);
+
+	witness_assert_lockless(tsdn);
+
+	witness_lock_error = witness_lock_error_orig;
+}
+TEST_END
+
+TEST_BEGIN(test_witness_recursive)
+{
+	witness_t a;
+	tsdn_t *tsdn;
+
+	test_skip_if(!config_debug);
+
+	witness_not_owner_error_orig = witness_not_owner_error;
+	witness_not_owner_error = witness_not_owner_error_intercept;
+	saw_not_owner_error = false;
+
+	witness_lock_error_orig = witness_lock_error;
+	witness_lock_error = witness_lock_error_intercept;
+	saw_lock_error = false;
+
+	tsdn = tsdn_fetch();
+
+	witness_assert_lockless(tsdn);
+
+	witness_init(&a, "a", 1, NULL);
+
+	witness_lock(tsdn, &a);
+	assert_false(saw_lock_error, "Unexpected witness lock error");
+	assert_false(saw_not_owner_error, "Unexpected witness not owner error");
+	witness_lock(tsdn, &a);
+	assert_true(saw_lock_error, "Expected witness lock error");
+	assert_true(saw_not_owner_error, "Expected witness not owner error");
+
+	witness_unlock(tsdn, &a);
+
+	witness_assert_lockless(tsdn);
+
+	witness_owner_error = witness_owner_error_orig;
+	witness_lock_error = witness_lock_error_orig;
+
+}
+TEST_END
+
+TEST_BEGIN(test_witness_unlock_not_owned)
+{
+	witness_t a;
+	tsdn_t *tsdn;
+
+	test_skip_if(!config_debug);
+
+	witness_owner_error_orig = witness_owner_error;
+	witness_owner_error = witness_owner_error_intercept;
+	saw_owner_error = false;
+
+	tsdn = tsdn_fetch();
+
+	witness_assert_lockless(tsdn);
+
+	witness_init(&a, "a", 1, NULL);
+
+	assert_false(saw_owner_error, "Unexpected owner error");
+	witness_unlock(tsdn, &a);
+	assert_true(saw_owner_error, "Expected owner error");
+
+	witness_assert_lockless(tsdn);
+
+	witness_owner_error = witness_owner_error_orig;
+}
+TEST_END
+
+TEST_BEGIN(test_witness_lockful)
+{
+	witness_t a;
+	tsdn_t *tsdn;
+
+	test_skip_if(!config_debug);
+
+	witness_lockless_error_orig = witness_lockless_error;
+	witness_lockless_error = witness_lockless_error_intercept;
+	saw_lockless_error = false;
+
+	tsdn = tsdn_fetch();
+
+	witness_assert_lockless(tsdn);
+
+	witness_init(&a, "a", 1, NULL);
+
+	assert_false(saw_lockless_error, "Unexpected lockless error");
+	witness_assert_lockless(tsdn);
+
+	witness_lock(tsdn, &a);
+	witness_assert_lockless(tsdn);
+	assert_true(saw_lockless_error, "Expected lockless error");
+
+	witness_unlock(tsdn, &a);
+
+	witness_assert_lockless(tsdn);
+
+	witness_lockless_error = witness_lockless_error_orig;
+}
+TEST_END
+
+int
+main(void)
+{
+
+	return (test(
+	    test_witness,
+	    test_witness_comp,
+	    test_witness_reversal,
+	    test_witness_recursive,
+	    test_witness_unlock_not_owned,
+	    test_witness_lockful));
+}
diff --git a/deps/jemalloc/test/unit/zero.c b/deps/jemalloc/test/unit/zero.c
index 93afc2b87..30ebe37a4 100644
--- a/deps/jemalloc/test/unit/zero.c
+++ b/deps/jemalloc/test/unit/zero.c
@@ -8,39 +8,41 @@ const char *malloc_conf =
 static void
 test_zero(size_t sz_min, size_t sz_max)
 {
-	char *s;
+	uint8_t *s;
 	size_t sz_prev, sz, i;
+#define	MAGIC	((uint8_t)0x61)
 
 	sz_prev = 0;
-	s = (char *)mallocx(sz_min, 0);
+	s = (uint8_t *)mallocx(sz_min, 0);
 	assert_ptr_not_null((void *)s, "Unexpected mallocx() failure");
 
 	for (sz = sallocx(s, 0); sz <= sz_max;
 	    sz_prev = sz, sz = sallocx(s, 0)) {
 		if (sz_prev > 0) {
-			assert_c_eq(s[0], 'a',
+			assert_u_eq(s[0], MAGIC,
 			    "Previously allocated byte %zu/%zu is corrupted",
 			    ZU(0), sz_prev);
-			assert_c_eq(s[sz_prev-1], 'a',
+			assert_u_eq(s[sz_prev-1], MAGIC,
 			    "Previously allocated byte %zu/%zu is corrupted",
 			    sz_prev-1, sz_prev);
 		}
 
 		for (i = sz_prev; i < sz; i++) {
-			assert_c_eq(s[i], 0x0,
+			assert_u_eq(s[i], 0x0,
 			    "Newly allocated byte %zu/%zu isn't zero-filled",
 			    i, sz);
-			s[i] = 'a';
+			s[i] = MAGIC;
 		}
 
 		if (xallocx(s, sz+1, 0, 0) == sz) {
-			s = (char *)rallocx(s, sz+1, 0);
+			s = (uint8_t *)rallocx(s, sz+1, 0);
 			assert_ptr_not_null((void *)s,
 			    "Unexpected rallocx() failure");
 		}
 	}
 
 	dallocx(s, 0);
+#undef MAGIC
 }
 
 TEST_BEGIN(test_zero_small)
-- 
cgit v1.2.1


From 3a7410a8a69dca0ff5d4cd4e48abf974d2e53b9f Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Fri, 27 Jan 2017 16:54:20 +0100
Subject: ziplist: better comments, some refactoring.

---
 src/ziplist.c | 350 +++++++++++++++++++++++++++++++++++++++++-----------------
 1 file changed, 250 insertions(+), 100 deletions(-)

diff --git a/src/ziplist.c b/src/ziplist.c
index f270cdbf9..81d23ca38 100644
--- a/src/ziplist.c
+++ b/src/ziplist.c
@@ -9,78 +9,149 @@
  * ----------------------------------------------------------------------------
  *
  * ZIPLIST OVERALL LAYOUT
+ * ======================
  *
  * The general layout of the ziplist is as follows:
  *
  * <zlbytes> <zltail> <zllen> <entry> <entry> ... <entry> <zlend>
  *
- * All fields are stored in little endian.
+ * NOTE: all fields are stored in little endian, if not specified otherwise.
  *
  * <uint32_t zlbytes> is an unsigned integer to hold the number of bytes that
- * the ziplist occupies. This value needs to be stored to be able to resize the
- * entire structure without the need to traverse it first.
+ * the ziplist occupies, including the four bytes of the zlbytes field itself.
+ * This value needs to be stored to be able to resize the entire structure
+ * without the need to traverse it first.
  *
  * <uint32_t zltail> is the offset to the last entry in the list. This allows
  * a pop operation on the far side of the list without the need for full
  * traversal.
  *
- * <uint16_t zllen> is the number of entries. When this value is larger
- * than 2^16-2, we need to traverse the entire list to know how many items it
- * holds.
+ * <uint16_t zllen> is the number of entries. When there are more than
+ * 2^16-2 entires, this value is set to 2^16-1 and we need to traverse the
+ * entire list to know how many items it holds.
  *
- * <uint8_t zlend> is a single byte special value, equal to 255, which
- * indicates the end of the list.
+ * <uint8_t zlend> is a special entry representing the end of the ziplist.
+ * Is encoded as a single byte equal to 255. No other normal entry starts
+ * with a byte set to the value of 255.
  *
  * ZIPLIST ENTRIES
+ * ===============
  *
- * Every entry in the ziplist is prefixed by a header that contains two pieces
+ * Every entry in the ziplist is prefixed by metadata that contains two pieces
  * of information. First, the length of the previous entry is stored to be
- * able to traverse the list from back to front. Second, the encoding with an
- * optional string length of the entry itself is stored.
+ * able to traverse the list from back to front. Second, the entry encoding is
+ * provided. It represents the entry type, integer or string, and in the case
+ * of strings it also represents the length of the string payload.
+ * So a complete entry is stored like this:
  *
- * The length of the previous entry is encoded in the following way:
- * If this length is smaller than 254 bytes, it will only consume a single
- * byte that takes the length as value. When the length is greater than or
- * equal to 254, it will consume 5 bytes. The first byte is set to 254 to
- * indicate a larger value is following. The remaining 4 bytes take the
- * length of the previous entry as value.
+ * <prevlen> <encoding> <entry-data>
  *
- * The other header field of the entry itself depends on the contents of the
- * entry. When the entry is a string, the first 2 bits of this header will hold
- * the type of encoding used to store the length of the string, followed by the
- * actual length of the string. When the entry is an integer the first 2 bits
- * are both set to 1. The following 2 bits are used to specify what kind of
- * integer will be stored after this header. An overview of the different
- * types and encodings is as follows:
+ * Sometimes the encoding represents the entry itself, like for small integers
+ * as we'll see later. In such a case the <entry-data> part is missing, and we
+ * could have just:
+ *
+ * <prevlen> <encoding>
+ *
+ * The length of the previous entry, <prevlen>, is encoded in the following way:
+ * If this length is smaller than 255 bytes, it will only consume a single
+ * byte representing the length as an unsinged 8 bit integer. When the length
+ * is greater than or equal to 255, it will consume 5 bytes. The first byte is
+ * set to 255 (FF) to indicate a larger value is following. The remaining 4
+ * bytes take the length of the previous entry as value.
+ *
+ * So practically an entry is encoded in the following way:
+ *
+ * <prevlen from 0 to 254> <encoding> <entry>
+ *
+ * Or alternatively if the previous entry length is greater than 254 bytes
+ * the following encoding is used:
+ *
+ * 0xFF <4 bytes unsigned little endian prevlen> <encoding> <entry>
+ *
+ * The encoding field of the entry depends on the content of the
+ * entry. When the entry is a string, the first 2 bits of the encoding first
+ * byte will hold the type of encoding used to store the length of the string,
+ * followed by the actual length of the string. When the entry is an integer
+ * the first 2 bits are both set to 1. The following 2 bits are used to specify
+ * what kind of integer will be stored after this header. An overview of the
+ * different types and encodings is as follows. The first byte is always enough
+ * to determine the kind of entry.
  *
  * |00pppppp| - 1 byte
  *      String value with length less than or equal to 63 bytes (6 bits).
+ *      "pppppp" represents the unsigned 6 bit length.
  * |01pppppp|qqqqqqqq| - 2 bytes
  *      String value with length less than or equal to 16383 bytes (14 bits).
- * |10______|qqqqqqqq|rrrrrrrr|ssssssss|tttttttt| - 5 bytes
+ *      IMPORTANT: The 14 bit number is stored in big endian.
+ * |10000000|qqqqqqqq|rrrrrrrr|ssssssss|tttttttt| - 5 bytes
  *      String value with length greater than or equal to 16384 bytes.
- * |11000000| - 1 byte
+ *      Only the 4 bytes following the first byte represents the length
+ *      up to 32^2-1. The 6 lower bits of the first byte are not used and
+ *      are set to zero.
+ *      IMPORTANT: The 32 bit number is stored in big endian.
+ * |11000000| - 3 bytes
  *      Integer encoded as int16_t (2 bytes).
- * |11010000| - 1 byte
+ * |11010000| - 5 bytes
  *      Integer encoded as int32_t (4 bytes).
- * |11100000| - 1 byte
+ * |11100000| - 9 bytes
  *      Integer encoded as int64_t (8 bytes).
- * |11110000| - 1 byte
+ * |11110000| - 4 bytes
  *      Integer encoded as 24 bit signed (3 bytes).
- * |11111110| - 1 byte
+ * |11111110| - 2 bytes
  *      Integer encoded as 8 bit signed (1 byte).
  * |1111xxxx| - (with xxxx between 0000 and 1101) immediate 4 bit integer.
  *      Unsigned integer from 0 to 12. The encoded value is actually from
  *      1 to 13 because 0000 and 1111 can not be used, so 1 should be
  *      subtracted from the encoded 4 bit value to obtain the right value.
- * |11111111| - End of ziplist.
+ * |11111111| - End of ziplist special entry.
+ *
+ * Like for the ziplist header, all the integers are represented in little
+ * endian byte order, even when this code is compiled in big endian systems.
+ *
+ * EXAMPLES OF ACTUAL ZIPLISTS
+ * ===========================
+ *
+ * The following is a ziplist containing the two elements representing
+ * the strings "2" and "5". It is composed of 15 bytes, that we visually
+ * split into sections:
+ *
+ *  [0f 00 00 00] [0c 00 00 00] [02 00] [00 f3] [02 f6] [ff]
+ *        |             |          |       |       |     |
+ *     zlbytes        zltail    entries   "2"     "5"   end
  *
- * All the integers are represented in little endian byte order.
+ * The first 4 bytes represent the number 15, that is the number of bytes
+ * the whole ziplist is composed of. The second 4 bytes are the offset
+ * at which the last ziplist entry is found, that is 12, in fact the
+ * last entry, that is "5", is at offset 12 inside the ziplist.
+ * The next 16 bit integer represents the number of elements inside the
+ * ziplist, its value is 2 since there are just two elements inside.
+ * Finally "00 f3" is the first entry representing the number 2. It is
+ * composed of the previous entry length, which is zero because this is
+ * our first entry, and the byte F3 which corresponds to the encoding
+ * |1111xxxx| with xxxx between 0001 and 1101. We need to remove the "F"
+ * higher order bits 1111, and subtract 1 from the "3", so the entry value
+ * is "2". The next entry has a prevlen of 02, since the first entry is
+ * composed of exactly two bytes. The entry itself, F6, is encoded exactly
+ * like the first entry, and 6-1 = 5, so the value of the entry is 5.
+ * Finally the special entry FF signals the end of the ziplist.
+ *
+ * Adding another element to the above string with the value "Hello World"
+ * allows us to show how the ziplist encodes small strings. We'll just show
+ * the hex dump of the entry itself. Imagine the bytes as following the
+ * entry that stores "5" in the ziplist above:
+ *
+ * [02] [0b] [48 65 6c 6c 6f 20 57 6f 72 6c 64]
+ *
+ * The first byte, 02, is the length of the previous entry. The next
+ * byte represents the encoding in the pattern |00pppppp| that means
+ * that the entry is a string of length <pppppp>, so 0B means that
+ * an 11 bytes string follows. From the third byte (48) to the last (64)
+ * there are just the ASCII characters for "Hello World".
  *
  * ----------------------------------------------------------------------------
  *
  * Copyright (c) 2009-2012, Pieter Noordhuis <pcnoordhuis at gmail dot com>
- * Copyright (c) 2009-2012, Salvatore Sanfilippo <antirez at gmail dot com>
+ * Copyright (c) 2009-2017, Salvatore Sanfilippo <antirez at gmail dot com>
  * All rights reserved.
  *
  * Redistribution and use in source and binary forms, with or without
@@ -119,8 +190,13 @@
 #include "endianconv.h"
 #include "redisassert.h"
 
-#define ZIP_END 255
-#define ZIP_BIGLEN 254
+#define ZIP_END 255         /* Special "end of ziplist" entry. */
+#define ZIP_BIG_PREVLEN 254 /* Max number of bytes of the previous entry, for
+                               the "prevlen" field prefixing each entry, to be
+                               represented with just a single byte. Otherwise
+                               it is represented as FF AA BB CC DD, where
+                               AA BB CC DD are a 4 bytes unsigned integer
+                               representing the previous entry len. */
 
 /* Different encoding/length possibilities */
 #define ZIP_STR_MASK 0xc0
@@ -133,41 +209,83 @@
 #define ZIP_INT_64B (0xc0 | 2<<4)
 #define ZIP_INT_24B (0xc0 | 3<<4)
 #define ZIP_INT_8B 0xfe
-/* 4 bit integer immediate encoding */
-#define ZIP_INT_IMM_MASK 0x0f
+
+/* 4 bit integer immediate encoding |1111xxxx| with xxxx between
+ * 0001 and 1101. */
+#define ZIP_INT_IMM_MASK 0x0f   /* Mask to extract the 4 bits value. To add
+                                   one is needed to reconstruct the value. */
 #define ZIP_INT_IMM_MIN 0xf1    /* 11110001 */
 #define ZIP_INT_IMM_MAX 0xfd    /* 11111101 */
-#define ZIP_INT_IMM_VAL(v) (v & ZIP_INT_IMM_MASK)
 
 #define INT24_MAX 0x7fffff
 #define INT24_MIN (-INT24_MAX - 1)
 
-/* Macro to determine type */
+/* Macro to determine if the entry is a string. String entries never start
+ * with "11" as most significant bits of the first byte. */
 #define ZIP_IS_STR(enc) (((enc) & ZIP_STR_MASK) < ZIP_STR_MASK)
 
-/* Utility macros */
+/* Utility macros.*/
+
+/* Return total bytes a ziplist is composed of. */
 #define ZIPLIST_BYTES(zl)       (*((uint32_t*)(zl)))
+
+/* Return the offset of the last item inside the ziplist. */
 #define ZIPLIST_TAIL_OFFSET(zl) (*((uint32_t*)((zl)+sizeof(uint32_t))))
+
+/* Return the length of a ziplist, or UINT16_MAX if the length cannot be
+ * determined without scanning the whole ziplist. */
 #define ZIPLIST_LENGTH(zl)      (*((uint16_t*)((zl)+sizeof(uint32_t)*2)))
+
+/* The size of a ziplist header: two 32 bit integers for the total
+ * bytes count and last item offset. One 16 bit integer for the number
+ * of items field. */
 #define ZIPLIST_HEADER_SIZE     (sizeof(uint32_t)*2+sizeof(uint16_t))
+
+/* Size of the "end of ziplist" entry. Just one byte. */
 #define ZIPLIST_END_SIZE        (sizeof(uint8_t))
+
+/* Return the pointer to the first entry of a ziplist. */
 #define ZIPLIST_ENTRY_HEAD(zl)  ((zl)+ZIPLIST_HEADER_SIZE)
+
+/* Return the pointer to the last entry of a ziplist, using the
+ * last entry offset inside the ziplist header. */
 #define ZIPLIST_ENTRY_TAIL(zl)  ((zl)+intrev32ifbe(ZIPLIST_TAIL_OFFSET(zl)))
+
+/* Return the pointer to the last byte of a ziplist, which is, the
+ * end of ziplist FF entry. */
 #define ZIPLIST_ENTRY_END(zl)   ((zl)+intrev32ifbe(ZIPLIST_BYTES(zl))-1)
 
-/* We know a positive increment can only be 1 because entries can only be
- * pushed one at a time. */
+/* Increment the number of items field in the ziplist header. Note that this
+ * macro should never overflow the unsigned 16 bit integer, since entires are
+ * always pushed one at a time. When UINT16_MAX is reached we want the count
+ * to stay there to signal that a full scan is needed to get the number of
+ * items inside the ziplist. */
 #define ZIPLIST_INCR_LENGTH(zl,incr) { \
     if (ZIPLIST_LENGTH(zl) < UINT16_MAX) \
         ZIPLIST_LENGTH(zl) = intrev16ifbe(intrev16ifbe(ZIPLIST_LENGTH(zl))+incr); \
 }
 
+/* We use this function to receive information about a ziplist entry.
+ * Note that this is not how the data is actually encoded, is just what we
+ * get filled by a function in order to operate more easily. */
 typedef struct zlentry {
-    unsigned int prevrawlensize, prevrawlen;
-    unsigned int lensize, len;
-    unsigned int headersize;
-    unsigned char encoding;
-    unsigned char *p;
+    unsigned int prevrawlensize; /* Bytes used to encode the previos entry len*/
+    unsigned int prevrawlen;     /* Previous entry len. */
+    unsigned int lensize;        /* Bytes used to encode this entry type/len.
+                                    For example strings have a 1, 2 or 5 bytes
+                                    header. Integers always use a single byte.*/
+    unsigned int len;            /* Bytes used to represent the actual entry.
+                                    For strings this is just the string length
+                                    while for integers it is 1, 2, 3, 4, 8 or
+                                    0 (for 4 bit immediate) depending on the
+                                    number range. */
+    unsigned int headersize;     /* prevrawlensize + lensize. */
+    unsigned char encoding;      /* Set to ZIP_STR_* or ZIP_INT_* depending on
+                                    the entry encoding. However for 4 bits
+                                    immediate integers this can assume a range
+                                    of values and must be range-checked. */
+    unsigned char *p;            /* Pointer to the very start of the entry, that
+                                    is, this points to prev-entry-len field. */
 } zlentry;
 
 #define ZIPLIST_ENTRY_ZERO(zle) { \
@@ -178,15 +296,13 @@ typedef struct zlentry {
 }
 
 /* Extract the encoding from the byte pointed by 'ptr' and set it into
- * 'encoding'. */
+ * 'encoding' field of the zlentry structure. */
 #define ZIP_ENTRY_ENCODING(ptr, encoding) do {  \
     (encoding) = (ptr[0]); \
     if ((encoding) < ZIP_STR_MASK) (encoding) &= ZIP_STR_MASK; \
 } while(0)
 
-void ziplistRepr(unsigned char *zl);
-
-/* Return bytes needed to store integer encoded by 'encoding' */
+/* Return bytes needed to store integer encoded by 'encoding'. */
 unsigned int zipIntSize(unsigned char encoding) {
     switch(encoding) {
     case ZIP_INT_8B:  return 1;
@@ -194,15 +310,26 @@ unsigned int zipIntSize(unsigned char encoding) {
     case ZIP_INT_24B: return 3;
     case ZIP_INT_32B: return 4;
     case ZIP_INT_64B: return 8;
-    default: return 0; /* 4 bit immediate */
     }
-    assert(NULL);
+    if (encoding >= ZIP_INT_IMM_MIN && encoding <= ZIP_INT_IMM_MAX)
+        return 0; /* 4 bit immediate */
+    panic("Invalid integer encoding 0x%02X", encoding);
     return 0;
 }
 
-/* Encode the length 'rawlen' writing it in 'p'. If p is NULL it just returns
- * the amount of bytes required to encode such a length. */
-unsigned int zipEncodeLength(unsigned char *p, unsigned char encoding, unsigned int rawlen) {
+/* Write the encoidng header of the entry in 'p'. If p is NULL it just returns
+ * the amount of bytes required to encode such a length. Arguments:
+ *
+ * 'encoding' is the encoding we are using for the entry. It could be
+ * ZIP_INT_* or ZIP_STR_* or between ZIP_INT_IMM_MIN and ZIP_INT_IMM_MAX
+ * for single-byte small immediate integers.
+ *
+ * 'rawlen' is only used for ZIP_STR_* encodings and is the length of the
+ * srting that this entry represents.
+ *
+ * The function returns the number of bytes used by the encoding/length
+ * header stored in 'p'. */
+unsigned int zipStoreEntryEncoding(unsigned char *p, unsigned char encoding, unsigned int rawlen) {
     unsigned char len = 1, buf[5];
 
     if (ZIP_IS_STR(encoding)) {
@@ -231,15 +358,16 @@ unsigned int zipEncodeLength(unsigned char *p, unsigned char encoding, unsigned
         buf[0] = encoding;
     }
 
-    /* Store this length at p */
+    /* Store this length at p. */
     memcpy(p,buf,len);
     return len;
 }
 
-/* Decode the length encoded in 'ptr'. The 'encoding' variable will hold the
- * entries encoding, the 'lensize' variable will hold the number of bytes
- * required to encode the entries length, and the 'len' variable will hold the
- * entries length. */
+/* Decode the entry encoding type and data length (string length for strings,
+ * number of bytes used for the integer for integer entries) encoded in 'ptr'.
+ * The 'encoding' variable will hold the entry encoding, the 'lensize'
+ * variable will hold the number of bytes required to encode the entry
+ * length, and the 'len' variable will hold the entry length. */
 #define ZIP_DECODE_LENGTH(ptr, encoding, lensize, len) do {                    \
     ZIP_ENTRY_ENCODING((ptr), (encoding));                                     \
     if ((encoding) < ZIP_STR_MASK) {                                           \
@@ -256,7 +384,7 @@ unsigned int zipEncodeLength(unsigned char *p, unsigned char encoding, unsigned
                     ((ptr)[3] <<  8) |                                         \
                     ((ptr)[4]);                                                \
         } else {                                                               \
-            assert(NULL);                                                      \
+            panic("Invalid string encoding 0x%02X", (encoding));               \
         }                                                                      \
     } else {                                                                   \
         (lensize) = 1;                                                         \
@@ -264,45 +392,49 @@ unsigned int zipEncodeLength(unsigned char *p, unsigned char encoding, unsigned
     }                                                                          \
 } while(0);
 
+/* Encode the length of the previous entry and write it to "p". This only
+ * uses the larger encoding (required in __ziplistCascadeUpdate). */
+int zipStorePrevEntryLengthLarge(unsigned char *p, unsigned int len) {
+    if (p != NULL) {
+        p[0] = ZIP_BIG_PREVLEN;
+        memcpy(p+1,&len,sizeof(len));
+        memrev32ifbe(p+1);
+    }
+    return 1+sizeof(len);
+}
+
 /* Encode the length of the previous entry and write it to "p". Return the
  * number of bytes needed to encode this length if "p" is NULL. */
-unsigned int zipPrevEncodeLength(unsigned char *p, unsigned int len) {
+unsigned int zipStorePrevEntryLength(unsigned char *p, unsigned int len) {
     if (p == NULL) {
-        return (len < ZIP_BIGLEN) ? 1 : sizeof(len)+1;
+        return (len < ZIP_BIG_PREVLEN) ? 1 : sizeof(len)+1;
     } else {
-        if (len < ZIP_BIGLEN) {
+        if (len < ZIP_BIG_PREVLEN) {
             p[0] = len;
             return 1;
         } else {
-            p[0] = ZIP_BIGLEN;
-            memcpy(p+1,&len,sizeof(len));
-            memrev32ifbe(p+1);
-            return 1+sizeof(len);
+            return zipStorePrevEntryLengthLarge(p,len);
         }
     }
 }
 
-/* Encode the length of the previous entry and write it to "p". This only
- * uses the larger encoding (required in __ziplistCascadeUpdate). */
-void zipPrevEncodeLengthForceLarge(unsigned char *p, unsigned int len) {
-    if (p == NULL) return;
-    p[0] = ZIP_BIGLEN;
-    memcpy(p+1,&len,sizeof(len));
-    memrev32ifbe(p+1);
-}
-
-/* Decode the number of bytes required to store the length of the previous
- * element, from the perspective of the entry pointed to by 'ptr'. */
+/* Return the number of bytes used to encode the length of the previous
+ * entry. The length is returned by setting the var 'prevlensize'. */
 #define ZIP_DECODE_PREVLENSIZE(ptr, prevlensize) do {                          \
-    if ((ptr)[0] < ZIP_BIGLEN) {                                               \
+    if ((ptr)[0] < ZIP_BIG_PREVLEN) {                                          \
         (prevlensize) = 1;                                                     \
     } else {                                                                   \
         (prevlensize) = 5;                                                     \
     }                                                                          \
 } while(0);
 
-/* Decode the length of the previous element, from the perspective of the entry
- * pointed to by 'ptr'. */
+/* Return the length of the previous element, and the number of bytes that
+ * are used in order to encode the previous element length.
+ * 'ptr' must point to the prevlen prefix of an entry (that encodes the
+ * length of the previos entry in order to navigate the elements backward).
+ * The length of the previous entry is stored in 'prevlen', the number of
+ * bytes needed to encode the previous entry length are stored in
+ * 'prevlensize'. */
 #define ZIP_DECODE_PREVLEN(ptr, prevlensize, prevlen) do {                     \
     ZIP_DECODE_PREVLENSIZE(ptr, prevlensize);                                  \
     if ((prevlensize) == 1) {                                                  \
@@ -314,12 +446,25 @@ void zipPrevEncodeLengthForceLarge(unsigned char *p, unsigned int len) {
     }                                                                          \
 } while(0);
 
-/* Return the difference in number of bytes needed to store the length of the
- * previous element 'len', in the entry pointed to by 'p'. */
+/* Given a pointer 'p' to the prevlen info that prefixes an entry, this
+ * function returns the difference in number of bytes needed to encode
+ * the prevlen if the previous entry changes of size.
+ *
+ * So if A is the number of bytes used right now to encode the 'prevlen'
+ * field.
+ *
+ * And B is the number of bytes that are needed in order to encode the
+ * 'prevlen' if the previous element will be updated to one of size 'len'.
+ *
+ * Then the function returns B - A
+ *
+ * So the function returns a positive number if more space is needed,
+ * a negative number if less space is needed, or zero if the same space
+ * is needed. */
 int zipPrevLenByteDiff(unsigned char *p, unsigned int len) {
     unsigned int prevlensize;
     ZIP_DECODE_PREVLENSIZE(p, prevlensize);
-    return zipPrevEncodeLength(NULL, len) - prevlensize;
+    return zipStorePrevEntryLength(NULL, len) - prevlensize;
 }
 
 /* Return the total number of bytes used by the entry pointed to by 'p'. */
@@ -456,8 +601,8 @@ unsigned char *ziplistResize(unsigned char *zl, unsigned int len) {
  * causes a realloc and memmove). However, encoding the prevlen may require
  * that this entry is grown as well. This effect may cascade throughout
  * the ziplist when there are consecutive entries with a size close to
- * ZIP_BIGLEN, so we need to check that the prevlen can be encoded in every
- * consecutive entry.
+ * ZIP_BIG_PREVLEN, so we need to check that the prevlen can be encoded in
+ * every consecutive entry.
  *
  * Note that this effect can also happen in reverse, where the bytes required
  * to encode the prevlen field can shrink. This effect is deliberately ignored,
@@ -477,7 +622,7 @@ unsigned char *__ziplistCascadeUpdate(unsigned char *zl, unsigned char *p) {
     while (p[0] != ZIP_END) {
         zipEntry(p, &cur);
         rawlen = cur.headersize + cur.len;
-        rawlensize = zipPrevEncodeLength(NULL,rawlen);
+        rawlensize = zipStorePrevEntryLength(NULL,rawlen);
 
         /* Abort if there is no next entry. */
         if (p[rawlen] == ZIP_END) break;
@@ -508,7 +653,7 @@ unsigned char *__ziplistCascadeUpdate(unsigned char *zl, unsigned char *p) {
             memmove(np+rawlensize,
                 np+next.prevrawlensize,
                 curlen-noffset-next.prevrawlensize-1);
-            zipPrevEncodeLength(np,rawlen);
+            zipStorePrevEntryLength(np,rawlen);
 
             /* Advance the cursor */
             p += rawlen;
@@ -517,9 +662,9 @@ unsigned char *__ziplistCascadeUpdate(unsigned char *zl, unsigned char *p) {
             if (next.prevrawlensize > rawlensize) {
                 /* This would result in shrinking, which we want to avoid.
                  * So, set "rawlen" in the available bytes. */
-                zipPrevEncodeLengthForceLarge(p+rawlen,rawlen);
+                zipStorePrevEntryLengthLarge(p+rawlen,rawlen);
             } else {
-                zipPrevEncodeLength(p+rawlen,rawlen);
+                zipStorePrevEntryLength(p+rawlen,rawlen);
             }
 
             /* Stop here, as the raw length of "next" has not changed. */
@@ -542,7 +687,7 @@ unsigned char *__ziplistDelete(unsigned char *zl, unsigned char *p, unsigned int
         deleted++;
     }
 
-    totlen = p-first.p;
+    totlen = p-first.p; /* Bytes taken by the element(s) to delete. */
     if (totlen > 0) {
         if (p[0] != ZIP_END) {
             /* Storing `prevrawlen` in this entry may increase or decrease the
@@ -550,8 +695,13 @@ unsigned char *__ziplistDelete(unsigned char *zl, unsigned char *p, unsigned int
              * There always is room to store this, because it was previously
              * stored by an entry that is now being deleted. */
             nextdiff = zipPrevLenByteDiff(p,first.prevrawlen);
+
+            /* Note that there is always space when p jumps backward: if
+             * the new previous entry is large, one of the deleted elements
+             * had a 5 bytes prevlen header, so there is for sure at least
+             * 5 bytes free and we need just 4. */
             p -= nextdiff;
-            zipPrevEncodeLength(p,first.prevrawlen);
+            zipStorePrevEntryLength(p,first.prevrawlen);
 
             /* Update offset for tail */
             ZIPLIST_TAIL_OFFSET(zl) =
@@ -616,14 +766,14 @@ unsigned char *__ziplistInsert(unsigned char *zl, unsigned char *p, unsigned cha
         /* 'encoding' is set to the appropriate integer encoding */
         reqlen = zipIntSize(encoding);
     } else {
-        /* 'encoding' is untouched, however zipEncodeLength will use the
+        /* 'encoding' is untouched, however zipStoreEntryEncoding will use the
          * string length to figure out how to encode it. */
         reqlen = slen;
     }
     /* We need space for both the length of the previous entry and
      * the length of the payload. */
-    reqlen += zipPrevEncodeLength(NULL,prevlen);
-    reqlen += zipEncodeLength(NULL,encoding,slen);
+    reqlen += zipStorePrevEntryLength(NULL,prevlen);
+    reqlen += zipStoreEntryEncoding(NULL,encoding,slen);
 
     /* When the insert position is not equal to the tail, we need to
      * make sure that the next entry can hold this entry's length in
@@ -641,7 +791,7 @@ unsigned char *__ziplistInsert(unsigned char *zl, unsigned char *p, unsigned cha
         memmove(p+reqlen,p-nextdiff,curlen-offset-1+nextdiff);
 
         /* Encode this entry's raw length in the next entry. */
-        zipPrevEncodeLength(p+reqlen,reqlen);
+        zipStorePrevEntryLength(p+reqlen,reqlen);
 
         /* Update offset for tail */
         ZIPLIST_TAIL_OFFSET(zl) =
@@ -669,8 +819,8 @@ unsigned char *__ziplistInsert(unsigned char *zl, unsigned char *p, unsigned cha
     }
 
     /* Write the entry */
-    p += zipPrevEncodeLength(p,prevlen);
-    p += zipEncodeLength(p,encoding,slen);
+    p += zipStorePrevEntryLength(p,prevlen);
+    p += zipStoreEntryEncoding(p,encoding,slen);
     if (ZIP_IS_STR(encoding)) {
         memcpy(p,s,slen);
     } else {
-- 
cgit v1.2.1


From c495d095ae495ea5253443ee4562aaa30681a854 Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Mon, 30 Jan 2017 16:34:08 +0100
Subject: Ziplist: insertion bug under particular conditions fixed.

Ziplists had a bug that was discovered while investigating a different
issue, resulting in a corrupted ziplist representation, and a likely
segmentation foult and/or data corruption of the last element of the
ziplist, once the ziplist is accessed again.

The bug happens when a specific set of insertions / deletions is
performed so that an entry is encoded to have a "prevlen" field (the
length of the previous entry) of 5 bytes but with a count that could be
encoded in a "prevlen" field of a since byte. This could happen when the
"cascading update" process called by ziplistInsert()/ziplistDelete() in
certain contitious forces the prevlen to be bigger than necessary in
order to avoid too much data moving around.

Once such an entry is generated, inserting a very small entry
immediately before it will result in a resizing of the ziplist for a
count smaller than the current ziplist length (which is a violation,
inserting code expects the ziplist to get bigger actually). So an FF
byte is inserted in a misplaced position. Moreover a realloc() is
performed with a count smaller than the ziplist current length so the
final bytes could be trashed as well.

SECURITY IMPLICATIONS:

Currently it looks like an attacker can only crash a Redis server by
providing specifically choosen commands. However a FF byte is written
and there are other memory operations that depend on a wrong count, so
even if it is not immediately apparent how to mount an attack in order
to execute code remotely, it is not impossible at all that this could be
done. Attacks always get better... and we did not spent enough time in
order to think how to exploit this issue, but security researchers
or malicious attackers could.
---
 src/ziplist.c | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

diff --git a/src/ziplist.c b/src/ziplist.c
index 81d23ca38..0da1b0ca6 100644
--- a/src/ziplist.c
+++ b/src/ziplist.c
@@ -778,7 +778,12 @@ unsigned char *__ziplistInsert(unsigned char *zl, unsigned char *p, unsigned cha
     /* When the insert position is not equal to the tail, we need to
      * make sure that the next entry can hold this entry's length in
      * its prevlen field. */
+    int forcelarge = 0;
     nextdiff = (p[0] != ZIP_END) ? zipPrevLenByteDiff(p,reqlen) : 0;
+    if (nextdiff == -4 && reqlen < 4) {
+        nextdiff = 0;
+        forcelarge = 1;
+    }
 
     /* Store offset because a realloc may change the address of zl. */
     offset = p-zl;
@@ -791,7 +796,10 @@ unsigned char *__ziplistInsert(unsigned char *zl, unsigned char *p, unsigned cha
         memmove(p+reqlen,p-nextdiff,curlen-offset-1+nextdiff);
 
         /* Encode this entry's raw length in the next entry. */
-        zipStorePrevEntryLength(p+reqlen,reqlen);
+        if (forcelarge)
+            zipStorePrevEntryLength(p+reqlen,reqlen);
+        else
+            zipStorePrevEntryLengthLarge(p+reqlen,reqlen);
 
         /* Update offset for tail */
         ZIPLIST_TAIL_OFFSET(zl) =
-- 
cgit v1.2.1


From 0dbfb1d154b0df28bbdb16a59ae7342d4a3f9281 Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Wed, 1 Feb 2017 17:01:31 +0100
Subject: Fix ziplist fix...

---
 src/ziplist.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/ziplist.c b/src/ziplist.c
index 0da1b0ca6..e407937ff 100644
--- a/src/ziplist.c
+++ b/src/ziplist.c
@@ -797,9 +797,9 @@ unsigned char *__ziplistInsert(unsigned char *zl, unsigned char *p, unsigned cha
 
         /* Encode this entry's raw length in the next entry. */
         if (forcelarge)
-            zipStorePrevEntryLength(p+reqlen,reqlen);
-        else
             zipStorePrevEntryLengthLarge(p+reqlen,reqlen);
+        else
+            zipStorePrevEntryLength(p+reqlen,reqlen);
 
         /* Update offset for tail */
         ZIPLIST_TAIL_OFFSET(zl) =
-- 
cgit v1.2.1


From f917e0da4cda8a0e6cd3242180b268706cdd2dd2 Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Thu, 9 Feb 2017 09:58:38 +0100
Subject: Fix MIGRATE closing of cached socket on error.

After investigating issue #3796, it was discovered that MIGRATE
could call migrateCloseSocket() after the original MIGRATE c->argv
was already rewritten as a DEL operation. As a result the host/port
passed to migrateCloseSocket() could be anything, often a NULL pointer
that gets deferenced crashing the server.

Now the socket is closed at an earlier time when there is a socket
error in a later stage where no retry will be performed, before we
rewrite the argument vector. Moreover a check was added so that later,
in the socket_err label, there is no further attempt at closing the
socket if the argument was rewritten.

This fix should resolve the bug reported in #3796.
---
 src/cluster.c | 29 +++++++++++++++++++++++------
 1 file changed, 23 insertions(+), 6 deletions(-)

diff --git a/src/cluster.c b/src/cluster.c
index 6e3fc8b00..4d7b0502d 100644
--- a/src/cluster.c
+++ b/src/cluster.c
@@ -4756,6 +4756,7 @@ void migrateCommand(client *c) {
     rio cmd, payload;
     int may_retry = 1;
     int write_error = 0;
+    int argv_rewritten = 0;
 
     /* To support the KEYS option we need the following additional state. */
     int first_key = 3; /* Argument index of the first key. */
@@ -4939,12 +4940,20 @@ try_again:
         goto socket_err; /* A retry is guaranteed because of tested conditions.*/
     }
 
+    /* On socket errors, close the migration socket now that we still have
+     * the original host/port in the ARGV. Later the original command may be
+     * rewritten to DEL and will be too later. */
+    if (socket_error) migrateCloseSocket(c->argv[1],c->argv[2]);
+
     if (!copy) {
-        /* Translate MIGRATE as DEL for replication/AOF. */
+        /* Translate MIGRATE as DEL for replication/AOF. Note that we do
+         * this only for the keys for which we received an acknowledgement
+         * from the receiving Redis server, by using the del_idx index. */
         if (del_idx > 1) {
             newargv[0] = createStringObject("DEL",3);
             /* Note that the following call takes ownership of newargv. */
             replaceClientCommandVector(c,del_idx,newargv);
+            argv_rewritten = 1;
         } else {
             /* No key transfer acknowledged, no need to rewrite as DEL. */
             zfree(newargv);
@@ -4953,8 +4962,8 @@ try_again:
     }
 
     /* If we are here and a socket error happened, we don't want to retry.
-     * Just signal the problem to the client, but only do it if we don't
-     * already queued a different error reported by the destination server. */
+     * Just signal the problem to the client, but only do it if we did not
+     * already queue a different error reported by the destination server. */
     if (!error_from_target && socket_error) {
         may_retry = 0;
         goto socket_err;
@@ -4962,7 +4971,11 @@ try_again:
 
     if (!error_from_target) {
         /* Success! Update the last_dbid in migrateCachedSocket, so that we can
-         * avoid SELECT the next time if the target DB is the same. Reply +OK. */
+         * avoid SELECT the next time if the target DB is the same. Reply +OK.
+         *
+         * Note: If we reached this point, even if socket_error is true
+         * still the SELECT command succeeded (otherwise the code jumps to
+         * socket_err label. */
         cs->last_dbid = dbid;
         addReply(c,shared.ok);
     } else {
@@ -4972,7 +4985,6 @@ try_again:
 
     sdsfree(cmd.io.buffer.ptr);
     zfree(ov); zfree(kv); zfree(newargv);
-    if (socket_error) migrateCloseSocket(c->argv[1],c->argv[2]);
     return;
 
 /* On socket errors we try to close the cached socket and try again.
@@ -4982,7 +4994,12 @@ socket_err:
     /* Cleanup we want to perform in both the retry and no retry case.
      * Note: Closing the migrate socket will also force SELECT next time. */
     sdsfree(cmd.io.buffer.ptr);
-    migrateCloseSocket(c->argv[1],c->argv[2]);
+
+    /* If the command was rewritten as DEL and there was a socket error,
+     * we already closed the socket earlier. While migrateCloseSocket()
+     * is idempotent, the host/port arguments are now gone, so don't do it
+     * again. */
+    if (!argv_rewritten) migrateCloseSocket(c->argv[1],c->argv[2]);
     zfree(newargv);
     newargv = NULL; /* This will get reallocated on retry. */
 
-- 
cgit v1.2.1


From 1e272a6b52d663e0b4db8f42162c4461405b7f84 Mon Sep 17 00:00:00 2001
From: Salvatore Sanfilippo <antirez@gmail.com>
Date: Sun, 19 Feb 2017 14:01:58 +0000
Subject: ARM: Fix 64 bit unaligned access in MurmurHash64A().

---
 src/config.h      | 6 ++++++
 src/hyperloglog.c | 4 ++++
 2 files changed, 10 insertions(+)

diff --git a/src/config.h b/src/config.h
index 9fd53626e..354f8f5e5 100644
--- a/src/config.h
+++ b/src/config.h
@@ -206,4 +206,10 @@ void setproctitle(const char *fmt, ...);
 #endif
 #endif
 
+/* Make sure we can test for ARM just checking for __arm__, since sometimes
+ * __arm is defined but __arm__ is not. */
+#if defined(__arm) && !defined(__arm__)
+#define __arm__
+#endif
+
 #endif
diff --git a/src/hyperloglog.c b/src/hyperloglog.c
index 0800bf59d..7de5786f9 100644
--- a/src/hyperloglog.c
+++ b/src/hyperloglog.c
@@ -401,7 +401,11 @@ uint64_t MurmurHash64A (const void * key, int len, unsigned int seed) {
         uint64_t k;
 
 #if (BYTE_ORDER == LITTLE_ENDIAN)
+	#ifdef __arm__
+	memcpy(&k,data,sizeof(uint64_t));
+	#else
         k = *((uint64_t*)data);
+	#endif
 #else
         k = (uint64_t) data[0];
         k |= (uint64_t) data[1] << 8;
-- 
cgit v1.2.1


From 72d6d64771ce52519c240fb122dbb35f989b4669 Mon Sep 17 00:00:00 2001
From: Salvatore Sanfilippo <antirez@gmail.com>
Date: Sun, 19 Feb 2017 14:59:39 +0000
Subject: ARM: Avoid memcpy() in MurmurHash64A() if we are using 64 bit ARM.

However note that in architectures supporting 64 bit unaligned
accesses memcpy(...,...,8) is likely translated to a simple
word memory movement anyway.
---
 src/Makefile      | 1 +
 src/config.h      | 3 +++
 src/hyperloglog.c | 2 +-
 3 files changed, 5 insertions(+), 1 deletion(-)

diff --git a/src/Makefile b/src/Makefile
index 3f445f40f..f211eb84b 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -14,6 +14,7 @@
 
 release_hdr := $(shell sh -c './mkreleasehdr.sh')
 uname_S := $(shell sh -c 'uname -s 2>/dev/null || echo not')
+uname_M := $(shell sh -c 'uname -m 2>/dev/null || echo not')
 OPTIMIZATION?=-O2
 DEPENDENCY_TARGETS=hiredis linenoise lua
 NODEPS:=clean distclean
diff --git a/src/config.h b/src/config.h
index 354f8f5e5..1005dcc2a 100644
--- a/src/config.h
+++ b/src/config.h
@@ -211,5 +211,8 @@ void setproctitle(const char *fmt, ...);
 #if defined(__arm) && !defined(__arm__)
 #define __arm__
 #endif
+#if defined (__aarch64__) && !defined(__arm64__)
+#define __arm64__
+#endif
 
 #endif
diff --git a/src/hyperloglog.c b/src/hyperloglog.c
index 7de5786f9..b8a63a73f 100644
--- a/src/hyperloglog.c
+++ b/src/hyperloglog.c
@@ -401,7 +401,7 @@ uint64_t MurmurHash64A (const void * key, int len, unsigned int seed) {
         uint64_t k;
 
 #if (BYTE_ORDER == LITTLE_ENDIAN)
-	#ifdef __arm__
+	#if defined(__arm__) && !defined(__arm64__)
 	memcpy(&k,data,sizeof(uint64_t));
 	#else
         k = *((uint64_t*)data);
-- 
cgit v1.2.1


From 4e9cf4cc7ed4b732fc4bb592f19ceb41d132954e Mon Sep 17 00:00:00 2001
From: Salvatore Sanfilippo <antirez@gmail.com>
Date: Sun, 19 Feb 2017 15:02:37 +0000
Subject: ARM: Use libc malloc by default.

I'm not sure how much test Jemalloc gets on ARM, moreover
compiling Redis with Jemalloc support in not very powerful
devices, like most ARMs people will build Redis on, is extremely
slow. It is possible to enable Jemalloc build anyway if needed
by using "make MALLOC=jemalloc".
---
 src/Makefile | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/src/Makefile b/src/Makefile
index f211eb84b..83f5c6d79 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -28,11 +28,14 @@ PREFIX?=/usr/local
 INSTALL_BIN=$(PREFIX)/bin
 INSTALL=install
 
-# Default allocator
+# Default allocator defaults to Jemalloc if it's not an ARM
+MALLOC=libc
+ifneq ($(uname_M),armv6l)
+ifneq ($(uname_M),armv7l)
 ifeq ($(uname_S),Linux)
 	MALLOC=jemalloc
-else
-	MALLOC=libc
+endif
+endif
 endif
 
 # Backwards compatibility for selecting an allocator
-- 
cgit v1.2.1


From 7329cc39818a05c168e7d1e791afb03c089f1933 Mon Sep 17 00:00:00 2001
From: Salvatore Sanfilippo <antirez@gmail.com>
Date: Sun, 19 Feb 2017 15:07:08 +0000
Subject: ARM: Avoid fast path for BITOP.

GCC will produce certain unaligned multi load-store instructions
that will be trapped by the Linux kernel since ARM v6 cannot
handle them with unaligned addresses. Better to use the slower
but safer implementation instead of generating the exception which
should be anyway very slow.
---
 src/bitops.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/src/bitops.c b/src/bitops.c
index 46eee22c3..7ab72633c 100644
--- a/src/bitops.c
+++ b/src/bitops.c
@@ -654,8 +654,11 @@ void bitopCommand(client *c) {
 
         /* Fast path: as far as we have data for all the input bitmaps we
          * can take a fast path that performs much better than the
-         * vanilla algorithm. */
+         * vanilla algorithm. On ARM we skip the fast path since it will
+         * result in GCC compiling the code using multiple-words load/store
+         * operations that are not supported even in ARM >= v6. */
         j = 0;
+        #ifndef __arm__
         if (minlen >= sizeof(unsigned long)*4 && numkeys <= 16) {
             unsigned long *lp[16];
             unsigned long *lres = (unsigned long*) res;
@@ -716,6 +719,7 @@ void bitopCommand(client *c) {
                 }
             }
         }
+        #endif
 
         /* j is set to the next byte to process by the previous loop. */
         for (; j < maxlen; j++) {
-- 
cgit v1.2.1


From 76d87f47c7075a04dfd759fc5bd102dfad005dcc Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Mon, 20 Feb 2017 10:18:41 +0100
Subject: Don't leak file descriptor on syncWithMaster().

Close #3804.
---
 src/replication.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/src/replication.c b/src/replication.c
index df2e23f3a..bde37bc43 100644
--- a/src/replication.c
+++ b/src/replication.c
@@ -1568,7 +1568,7 @@ int slaveTryPartialResynchronization(int fd, int read_reply) {
  * establish a connection with the master. */
 void syncWithMaster(aeEventLoop *el, int fd, void *privdata, int mask) {
     char tmpfile[256], *err = NULL;
-    int dfd, maxtries = 5;
+    int dfd = -1, maxtries = 5;
     int sockerr = 0, psync_result;
     socklen_t errlen = sizeof(sockerr);
     UNUSED(el);
@@ -1832,6 +1832,7 @@ void syncWithMaster(aeEventLoop *el, int fd, void *privdata, int mask) {
 
 error:
     aeDeleteFileEvent(server.el,fd,AE_READABLE|AE_WRITABLE);
+    if (dfd != -1) close(dfd);
     close(fd);
     server.repl_transfer_s = -1;
     server.repl_state = REPL_STATE_CONNECT;
-- 
cgit v1.2.1


From 9b05aafb50348838f45bfddcd689e7d8d1d3c950 Mon Sep 17 00:00:00 2001
From: "John.Koepi" <john.koepi@gmail.com>
Date: Sat, 23 Jul 2016 16:31:56 +0200
Subject: fix #2883, #2857 pipe fds leak when fork() failed on bg aof rw

---
 src/aof.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/src/aof.c b/src/aof.c
index 07d8561da..f73da1e25 100644
--- a/src/aof.c
+++ b/src/aof.c
@@ -1357,6 +1357,7 @@ int rewriteAppendOnlyFileBackground(void) {
             serverLog(LL_WARNING,
                 "Can't rewrite append only file in background: fork: %s",
                 strerror(errno));
+            aofClosePipes();
             return C_ERR;
         }
         serverLog(LL_NOTICE,
-- 
cgit v1.2.1


From adeed29a99dcd0efdbfe4dbd5da74e7b01966c67 Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Mon, 20 Feb 2017 16:09:54 +0100
Subject: Use SipHash hash function to mitigate HashDos attempts.

This change attempts to switch to an hash function which mitigates
the effects of the HashDoS attack (denial of service attack trying
to force data structures to worst case behavior) while at the same time
providing Redis with an hash function that does not expect the input
data to be word aligned, a condition no longer true now that sds.c
strings have a varialbe length header.

Note that it is possible sometimes that even using an hash function
for which collisions cannot be generated without knowing the seed,
special implementation details or the exposure of the seed in an
indirect way (for example the ability to add elements to a Set and
check the return in which Redis returns them with SMEMBERS) may
make the attacker's life simpler in the process of trying to guess
the correct seed, however the next step would be to switch to a
log(N) data structure when too many items in a single bucket are
detected: this seems like an overkill in the case of Redis.

SPEED REGRESION TESTS:

In order to verify that switching from MurmurHash to SipHash had
no impact on speed, a set of benchmarks involving fast insertion
of 5 million of keys were performed.

The result shows Redis with SipHash in high pipelining conditions
to be about 4% slower compared to using the previous hash function.
However this could partially be related to the fact that the current
implementation does not attempt to hash whole words at a time but
reads single bytes, in order to have an output which is endian-netural
and at the same time working on systems where unaligned memory accesses
are a problem.

Further X86 specific optimizations should be tested, the function
may easily get at the same level of MurMurHash2 if a few optimizations
are performed.
---
 src/Makefile   |   2 +-
 src/config.c   |   2 +-
 src/debug.c    |   2 -
 src/dict.c     |  75 +++----------
 src/dict.h     |  10 +-
 src/latency.c  |   2 +-
 src/module.c   |   2 +-
 src/sentinel.c |   2 +-
 src/server.c   |  14 +--
 src/server.h   |   2 +-
 src/siphash.c  | 328 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 src/t_zset.c   |   2 +-
 12 files changed, 361 insertions(+), 82 deletions(-)
 create mode 100644 src/siphash.c

diff --git a/src/Makefile b/src/Makefile
index 3f445f40f..5f7536e83 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -128,7 +128,7 @@ endif
 
 REDIS_SERVER_NAME=redis-server
 REDIS_SENTINEL_NAME=redis-sentinel
-REDIS_SERVER_OBJ=adlist.o quicklist.o ae.o anet.o dict.o server.o sds.o zmalloc.o lzf_c.o lzf_d.o pqsort.o zipmap.o sha1.o ziplist.o release.o networking.o util.o object.o db.o replication.o rdb.o t_string.o t_list.o t_set.o t_zset.o t_hash.o config.o aof.o pubsub.o multi.o debug.o sort.o intset.o syncio.o cluster.o crc16.o endianconv.o slowlog.o scripting.o bio.o rio.o rand.o memtest.o crc64.o bitops.o sentinel.o notify.o setproctitle.o blocked.o hyperloglog.o latency.o sparkline.o redis-check-rdb.o geo.o lazyfree.o module.o evict.o expire.o geohash.o geohash_helper.o childinfo.o defrag.o
+REDIS_SERVER_OBJ=adlist.o quicklist.o ae.o anet.o dict.o server.o sds.o zmalloc.o lzf_c.o lzf_d.o pqsort.o zipmap.o sha1.o ziplist.o release.o networking.o util.o object.o db.o replication.o rdb.o t_string.o t_list.o t_set.o t_zset.o t_hash.o config.o aof.o pubsub.o multi.o debug.o sort.o intset.o syncio.o cluster.o crc16.o endianconv.o slowlog.o scripting.o bio.o rio.o rand.o memtest.o crc64.o bitops.o sentinel.o notify.o setproctitle.o blocked.o hyperloglog.o latency.o sparkline.o redis-check-rdb.o geo.o lazyfree.o module.o evict.o expire.o geohash.o geohash_helper.o childinfo.o defrag.o siphash.o
 REDIS_CLI_NAME=redis-cli
 REDIS_CLI_OBJ=anet.o adlist.o redis-cli.o zmalloc.o release.o anet.o ae.o crc64.o
 REDIS_BENCHMARK_NAME=redis-benchmark
diff --git a/src/config.c b/src/config.c
index 83651877c..900274f65 100644
--- a/src/config.c
+++ b/src/config.c
@@ -1423,7 +1423,7 @@ void configGetCommand(client *c) {
 /* We use the following dictionary type to store where a configuration
  * option is mentioned in the old configuration file, so it's
  * like "maxmemory" -> list of line numbers (first line is zero). */
-unsigned int dictSdsCaseHash(const void *key);
+uint64_t dictSdsCaseHash(const void *key);
 int dictSdsKeyCaseCompare(void *privdata, const void *key1, const void *key2);
 void dictSdsDestructor(void *privdata, void *val);
 void dictListDestructor(void *privdata, void *val);
diff --git a/src/debug.c b/src/debug.c
index a6bc62dc8..a4caa49f2 100644
--- a/src/debug.c
+++ b/src/debug.c
@@ -1029,8 +1029,6 @@ void sigsegvHandler(int sig, siginfo_t *info, void *secret) {
     /* Log INFO and CLIENT LIST */
     serverLogRaw(LL_WARNING|LL_RAW, "\n------ INFO OUTPUT ------\n");
     infostring = genRedisInfoString("all");
-    infostring = sdscatprintf(infostring, "hash_init_value: %u\n",
-        dictGetHashFunctionSeed());
     serverLogRaw(LL_WARNING|LL_RAW, infostring);
     serverLogRaw(LL_WARNING|LL_RAW, "\n------ CLIENT LIST OUTPUT ------\n");
     clients = getAllClientsInfoString();
diff --git a/src/dict.c b/src/dict.c
index 59aef7724..8ce735961 100644
--- a/src/dict.c
+++ b/src/dict.c
@@ -37,11 +37,11 @@
 
 #include <stdio.h>
 #include <stdlib.h>
+#include <stdint.h>
 #include <string.h>
 #include <stdarg.h>
 #include <limits.h>
 #include <sys/time.h>
-#include <ctype.h>
 
 #include "dict.h"
 #include "zmalloc.h"
@@ -71,77 +71,28 @@ static int _dictInit(dict *ht, dictType *type, void *privDataPtr);
 
 /* -------------------------- hash functions -------------------------------- */
 
-static uint32_t dict_hash_function_seed = 5381;
+static uint8_t dict_hash_function_seed[16];
 
-void dictSetHashFunctionSeed(uint32_t seed) {
-    dict_hash_function_seed = seed;
+void dictSetHashFunctionSeed(uint8_t *seed) {
+    memcpy(dict_hash_function_seed,seed,sizeof(dict_hash_function_seed));
 }
 
-uint32_t dictGetHashFunctionSeed(void) {
+uint8_t *dictGetHashFunctionSeed(void) {
     return dict_hash_function_seed;
 }
 
-/* MurmurHash2, by Austin Appleby
- * Note - This code makes a few assumptions about how your machine behaves -
- * 1. We can read a 4-byte value from any address without crashing
- * 2. sizeof(int) == 4
- *
- * And it has a few limitations -
- *
- * 1. It will not work incrementally.
- * 2. It will not produce the same results on little-endian and big-endian
- *    machines.
- */
-unsigned int dictGenHashFunction(const void *key, int len) {
-    /* 'm' and 'r' are mixing constants generated offline.
-     They're not really 'magic', they just happen to work well.  */
-    uint32_t seed = dict_hash_function_seed;
-    const uint32_t m = 0x5bd1e995;
-    const int r = 24;
-
-    /* Initialize the hash to a 'random' value */
-    uint32_t h = seed ^ len;
-
-    /* Mix 4 bytes at a time into the hash */
-    const unsigned char *data = (const unsigned char *)key;
-
-    while(len >= 4) {
-        uint32_t k = *(uint32_t*)data;
-
-        k *= m;
-        k ^= k >> r;
-        k *= m;
+/* The default hashing function uses SipHash implementation
+ * in siphash.c. */
 
-        h *= m;
-        h ^= k;
+uint64_t siphash(const uint8_t *in, const size_t inlen, const uint8_t *k);
+uint64_t siphash_nocase(const uint8_t *in, const size_t inlen, const uint8_t *k);
 
-        data += 4;
-        len -= 4;
-    }
-
-    /* Handle the last few bytes of the input array  */
-    switch(len) {
-    case 3: h ^= data[2] << 16;
-    case 2: h ^= data[1] << 8;
-    case 1: h ^= data[0]; h *= m;
-    };
-
-    /* Do a few final mixes of the hash to ensure the last few
-     * bytes are well-incorporated. */
-    h ^= h >> 13;
-    h *= m;
-    h ^= h >> 15;
-
-    return (unsigned int)h;
+uint64_t dictGenHashFunction(const void *key, int len) {
+    return siphash(key,len,dict_hash_function_seed);
 }
 
-/* And a case insensitive hash function (based on djb hash) */
-unsigned int dictGenCaseHashFunction(const unsigned char *buf, int len) {
-    unsigned int hash = (unsigned int)dict_hash_function_seed;
-
-    while (len--)
-        hash = ((hash << 5) + hash) + (tolower(*buf++)); /* hash * 33 + c */
-    return hash;
+uint64_t dictGenCaseHashFunction(const unsigned char *buf, int len) {
+    return siphash_nocase(buf,len,dict_hash_function_seed);
 }
 
 /* ----------------------------- API implementation ------------------------- */
diff --git a/src/dict.h b/src/dict.h
index 60a423a2c..bf316a00f 100644
--- a/src/dict.h
+++ b/src/dict.h
@@ -56,7 +56,7 @@ typedef struct dictEntry {
 } dictEntry;
 
 typedef struct dictType {
-    unsigned int (*hashFunction)(const void *key);
+    uint64_t (*hashFunction)(const void *key);
     void *(*keyDup)(void *privdata, const void *key);
     void *(*valDup)(void *privdata, const void *obj);
     int (*keyCompare)(void *privdata, const void *key1, const void *key2);
@@ -168,15 +168,15 @@ void dictReleaseIterator(dictIterator *iter);
 dictEntry *dictGetRandomKey(dict *d);
 unsigned int dictGetSomeKeys(dict *d, dictEntry **des, unsigned int count);
 void dictGetStats(char *buf, size_t bufsize, dict *d);
-unsigned int dictGenHashFunction(const void *key, int len);
-unsigned int dictGenCaseHashFunction(const unsigned char *buf, int len);
+uint64_t dictGenHashFunction(const void *key, int len);
+uint64_t dictGenCaseHashFunction(const unsigned char *buf, int len);
 void dictEmpty(dict *d, void(callback)(void*));
 void dictEnableResize(void);
 void dictDisableResize(void);
 int dictRehash(dict *d, int n);
 int dictRehashMilliseconds(dict *d, int ms);
-void dictSetHashFunctionSeed(unsigned int initval);
-unsigned int dictGetHashFunctionSeed(void);
+void dictSetHashFunctionSeed(uint8_t *seed);
+uint8_t *dictGetHashFunctionSeed(void);
 unsigned long dictScan(dict *d, unsigned long v, dictScanFunction *fn, dictScanBucketFunction *bucketfn, void *privdata);
 unsigned int dictGetHash(dict *d, const void *key);
 dictEntry **dictFindEntryRefByPtrAndHash(dict *d, const void *oldptr, unsigned int hash);
diff --git a/src/latency.c b/src/latency.c
index 53e0ec7be..9e9f1f13a 100644
--- a/src/latency.c
+++ b/src/latency.c
@@ -41,7 +41,7 @@ int dictStringKeyCompare(void *privdata, const void *key1, const void *key2) {
     return strcmp(key1,key2) == 0;
 }
 
-unsigned int dictStringHash(const void *key) {
+uint64_t dictStringHash(const void *key) {
     return dictGenHashFunction(key, strlen(key));
 }
 
diff --git a/src/module.c b/src/module.c
index 1fbc5094f..3b90eae4a 100644
--- a/src/module.c
+++ b/src/module.c
@@ -3264,7 +3264,7 @@ void *RM_GetBlockedClientPrivateData(RedisModuleCtx *ctx) {
 /* server.moduleapi dictionary type. Only uses plain C strings since
  * this gets queries from modules. */
 
-unsigned int dictCStringKeyHash(const void *key) {
+uint64_t dictCStringKeyHash(const void *key) {
     return dictGenHashFunction((unsigned char*)key, strlen((char*)key));
 }
 
diff --git a/src/sentinel.c b/src/sentinel.c
index 1f47dd337..6c6a3a0cd 100644
--- a/src/sentinel.c
+++ b/src/sentinel.c
@@ -379,7 +379,7 @@ void sentinelSimFailureCrash(void);
 
 /* ========================= Dictionary types =============================== */
 
-unsigned int dictSdsHash(const void *key);
+uint64_t dictSdsHash(const void *key);
 int dictSdsKeyCompare(void *privdata, const void *key1, const void *key2);
 void releaseSentinelRedisInstance(sentinelRedisInstance *ri);
 
diff --git a/src/server.c b/src/server.c
index 8bf6510de..0494a4e75 100644
--- a/src/server.c
+++ b/src/server.c
@@ -482,16 +482,16 @@ int dictObjKeyCompare(void *privdata, const void *key1,
     return dictSdsKeyCompare(privdata,o1->ptr,o2->ptr);
 }
 
-unsigned int dictObjHash(const void *key) {
+uint64_t dictObjHash(const void *key) {
     const robj *o = key;
     return dictGenHashFunction(o->ptr, sdslen((sds)o->ptr));
 }
 
-unsigned int dictSdsHash(const void *key) {
+uint64_t dictSdsHash(const void *key) {
     return dictGenHashFunction((unsigned char*)key, sdslen((char*)key));
 }
 
-unsigned int dictSdsCaseHash(const void *key) {
+uint64_t dictSdsCaseHash(const void *key) {
     return dictGenCaseHashFunction((unsigned char*)key, sdslen((char*)key));
 }
 
@@ -513,7 +513,7 @@ int dictEncObjKeyCompare(void *privdata, const void *key1,
     return cmp;
 }
 
-unsigned int dictEncObjHash(const void *key) {
+uint64_t dictEncObjHash(const void *key) {
     robj *o = (robj*) key;
 
     if (sdsEncodedObject(o)) {
@@ -526,7 +526,7 @@ unsigned int dictEncObjHash(const void *key) {
             len = ll2string(buf,32,(long)o->ptr);
             return dictGenHashFunction((unsigned char*)buf, len);
         } else {
-            unsigned int hash;
+            uint64_t hash;
 
             o = getDecodedObject(o);
             hash = dictGenHashFunction(o->ptr, sdslen((sds)o->ptr));
@@ -3639,7 +3639,9 @@ int main(int argc, char **argv) {
     zmalloc_set_oom_handler(redisOutOfMemoryHandler);
     srand(time(NULL)^getpid());
     gettimeofday(&tv,NULL);
-    dictSetHashFunctionSeed(tv.tv_sec^tv.tv_usec^getpid());
+    char hashseed[16];
+    getRandomHexChars(hashseed,sizeof(hashseed));
+    dictSetHashFunctionSeed((uint8_t*)hashseed);
     server.sentinel_mode = checkForSentinelMode(argc,argv);
     initServerConfig();
     moduleInitModulesSystem();
diff --git a/src/server.h b/src/server.h
index 30d8be849..75ff384cd 100644
--- a/src/server.h
+++ b/src/server.h
@@ -1765,7 +1765,7 @@ unsigned long LFUGetTimeInMinutes(void);
 uint8_t LFULogIncr(uint8_t value);
 
 /* Keys hashing / comparison functions for dict.c hash tables. */
-unsigned int dictSdsHash(const void *key);
+uint64_t dictSdsHash(const void *key);
 int dictSdsKeyCompare(void *privdata, const void *key1, const void *key2);
 void dictSdsDestructor(void *privdata, void *val);
 
diff --git a/src/siphash.c b/src/siphash.c
new file mode 100644
index 000000000..04e571fd7
--- /dev/null
+++ b/src/siphash.c
@@ -0,0 +1,328 @@
+/*
+   SipHash reference C implementation
+
+   Copyright (c) 2012-2016 Jean-Philippe Aumasson
+   <jeanphilippe.aumasson@gmail.com>
+   Copyright (c) 2012-2014 Daniel J. Bernstein <djb@cr.yp.to>
+   Copyright (c) 2017 Salvatore Sanfilippo <antirez@gmail.com>
+
+   To the extent possible under law, the author(s) have dedicated all copyright
+   and related and neighboring rights to this software to the public domain
+   worldwide. This software is distributed without any warranty.
+
+   You should have received a copy of the CC0 Public Domain Dedication along
+   with this software. If not, see
+   <http://creativecommons.org/publicdomain/zero/1.0/>.
+
+   ----------------------------------------------------------------------------
+
+   This version was modified by Salvatore Sanfilippo <antirez@gmail.com>
+   in the following ways:
+
+   1. Hard-code 2-4 rounds in the hope the compiler can optimize it more
+      in this raw from. Anyway we always want the standard 2-4 variant.
+   2. Modify the prototype and implementation so that the function directly
+      returns an uint64_t value, the hash itself, instead of receiving an
+      output buffer. This also means that the output size is set to 8 bytes
+      and the 16 bytes output code handling was removed.
+   3. Provide a case insensitive variant to be used when hashing strings that
+      must be considered identical by the hash table regardless of the case.
+      If we don't have directly a case insensitive hash function, we need to
+      perform a text transformation in some temporary buffer, which is costly.
+   4. Remove debugging code.
+   5. Modified the original test.c file to be a stand-alone function testing
+      the function in the new form (returing an uint64_t) using just the
+      relevant test vector.
+ */
+#include <assert.h>
+#include <stdint.h>
+#include <stdio.h>
+#include <string.h>
+#include <ctype.h>
+
+#define ROTL(x, b) (uint64_t)(((x) << (b)) | ((x) >> (64 - (b))))
+
+#define U32TO8_LE(p, v)                                                        \
+    (p)[0] = (uint8_t)((v));                                                   \
+    (p)[1] = (uint8_t)((v) >> 8);                                              \
+    (p)[2] = (uint8_t)((v) >> 16);                                             \
+    (p)[3] = (uint8_t)((v) >> 24);
+
+#define U64TO8_LE(p, v)                                                        \
+    U32TO8_LE((p), (uint32_t)((v)));                                           \
+    U32TO8_LE((p) + 4, (uint32_t)((v) >> 32));
+
+#define U8TO64_LE(p)                                                           \
+    (((uint64_t)((p)[0])) | ((uint64_t)((p)[1]) << 8) |                        \
+     ((uint64_t)((p)[2]) << 16) | ((uint64_t)((p)[3]) << 24) |                 \
+     ((uint64_t)((p)[4]) << 32) | ((uint64_t)((p)[5]) << 40) |                 \
+     ((uint64_t)((p)[6]) << 48) | ((uint64_t)((p)[7]) << 56))
+
+#define U8TO64_LE_NOCASE(p)                                                    \
+    (((uint64_t)(tolower((p)[0]))) |                                           \
+     ((uint64_t)(tolower((p)[1])) << 8) |                                      \
+     ((uint64_t)(tolower((p)[2])) << 16) |                                     \
+     ((uint64_t)(tolower((p)[3])) << 24) |                                     \
+     ((uint64_t)(tolower((p)[4])) << 32) |                                              \
+     ((uint64_t)(tolower((p)[5])) << 40) |                                              \
+     ((uint64_t)(tolower((p)[6])) << 48) |                                              \
+     ((uint64_t)(tolower((p)[7])) << 56))
+
+#define SIPROUND                                                               \
+    do {                                                                       \
+        v0 += v1;                                                              \
+        v1 = ROTL(v1, 13);                                                     \
+        v1 ^= v0;                                                              \
+        v0 = ROTL(v0, 32);                                                     \
+        v2 += v3;                                                              \
+        v3 = ROTL(v3, 16);                                                     \
+        v3 ^= v2;                                                              \
+        v0 += v3;                                                              \
+        v3 = ROTL(v3, 21);                                                     \
+        v3 ^= v0;                                                              \
+        v2 += v1;                                                              \
+        v1 = ROTL(v1, 17);                                                     \
+        v1 ^= v2;                                                              \
+        v2 = ROTL(v2, 32);                                                     \
+    } while (0)
+
+uint64_t siphash(const uint8_t *in, const size_t inlen, const uint8_t *k) {
+    uint64_t hash;
+    uint8_t *out = (uint8_t*) &hash;
+    uint64_t v0 = 0x736f6d6570736575ULL;
+    uint64_t v1 = 0x646f72616e646f6dULL;
+    uint64_t v2 = 0x6c7967656e657261ULL;
+    uint64_t v3 = 0x7465646279746573ULL;
+    uint64_t k0 = U8TO64_LE(k);
+    uint64_t k1 = U8TO64_LE(k + 8);
+    uint64_t m;
+    const uint8_t *end = in + inlen - (inlen % sizeof(uint64_t));
+    const int left = inlen & 7;
+    uint64_t b = ((uint64_t)inlen) << 56;
+    v3 ^= k1;
+    v2 ^= k0;
+    v1 ^= k1;
+    v0 ^= k0;
+
+    for (; in != end; in += 8) {
+        m = U8TO64_LE(in);
+        v3 ^= m;
+
+        SIPROUND;
+        SIPROUND;
+
+        v0 ^= m;
+    }
+
+    switch (left) {
+    case 7: b |= ((uint64_t)in[6]) << 48;
+    case 6: b |= ((uint64_t)in[5]) << 40;
+    case 5: b |= ((uint64_t)in[4]) << 32;
+    case 4: b |= ((uint64_t)in[3]) << 24;
+    case 3: b |= ((uint64_t)in[2]) << 16;
+    case 2: b |= ((uint64_t)in[1]) << 8;
+    case 1: b |= ((uint64_t)in[0]); break;
+    case 0: break;
+    }
+
+    v3 ^= b;
+
+    SIPROUND;
+    SIPROUND;
+
+    v0 ^= b;
+    v2 ^= 0xff;
+
+    SIPROUND;
+    SIPROUND;
+    SIPROUND;
+    SIPROUND;
+
+    b = v0 ^ v1 ^ v2 ^ v3;
+    U64TO8_LE(out, b);
+
+    return hash;
+}
+
+uint64_t siphash_nocase(const uint8_t *in, const size_t inlen, const uint8_t *k)
+{
+    uint64_t hash;
+    uint8_t *out = (uint8_t*) &hash;
+    uint64_t v0 = 0x736f6d6570736575ULL;
+    uint64_t v1 = 0x646f72616e646f6dULL;
+    uint64_t v2 = 0x6c7967656e657261ULL;
+    uint64_t v3 = 0x7465646279746573ULL;
+    uint64_t k0 = U8TO64_LE(k);
+    uint64_t k1 = U8TO64_LE(k + 8);
+    uint64_t m;
+    const uint8_t *end = in + inlen - (inlen % sizeof(uint64_t));
+    const int left = inlen & 7;
+    uint64_t b = ((uint64_t)inlen) << 56;
+    v3 ^= k1;
+    v2 ^= k0;
+    v1 ^= k1;
+    v0 ^= k0;
+
+    for (; in != end; in += 8) {
+        m = U8TO64_LE_NOCASE(in);
+        v3 ^= m;
+
+        SIPROUND;
+        SIPROUND;
+
+        v0 ^= m;
+    }
+
+    switch (left) {
+    case 7: b |= ((uint64_t)tolower(in[6])) << 48;
+    case 6: b |= ((uint64_t)tolower(in[5])) << 40;
+    case 5: b |= ((uint64_t)tolower(in[4])) << 32;
+    case 4: b |= ((uint64_t)tolower(in[3])) << 24;
+    case 3: b |= ((uint64_t)tolower(in[2])) << 16;
+    case 2: b |= ((uint64_t)tolower(in[1])) << 8;
+    case 1: b |= ((uint64_t)tolower(in[0])); break;
+    case 0: break;
+    }
+
+    v3 ^= b;
+
+    SIPROUND;
+    SIPROUND;
+
+    v0 ^= b;
+    v2 ^= 0xff;
+
+    SIPROUND;
+    SIPROUND;
+    SIPROUND;
+    SIPROUND;
+
+    b = v0 ^ v1 ^ v2 ^ v3;
+    U64TO8_LE(out, b);
+
+    return hash;
+}
+
+
+/* --------------------------------- TEST ------------------------------------ */
+
+#ifdef SIPHASH_TEST
+
+const uint8_t vectors_sip64[64][8] = {
+    { 0x31, 0x0e, 0x0e, 0xdd, 0x47, 0xdb, 0x6f, 0x72, },
+    { 0xfd, 0x67, 0xdc, 0x93, 0xc5, 0x39, 0xf8, 0x74, },
+    { 0x5a, 0x4f, 0xa9, 0xd9, 0x09, 0x80, 0x6c, 0x0d, },
+    { 0x2d, 0x7e, 0xfb, 0xd7, 0x96, 0x66, 0x67, 0x85, },
+    { 0xb7, 0x87, 0x71, 0x27, 0xe0, 0x94, 0x27, 0xcf, },
+    { 0x8d, 0xa6, 0x99, 0xcd, 0x64, 0x55, 0x76, 0x18, },
+    { 0xce, 0xe3, 0xfe, 0x58, 0x6e, 0x46, 0xc9, 0xcb, },
+    { 0x37, 0xd1, 0x01, 0x8b, 0xf5, 0x00, 0x02, 0xab, },
+    { 0x62, 0x24, 0x93, 0x9a, 0x79, 0xf5, 0xf5, 0x93, },
+    { 0xb0, 0xe4, 0xa9, 0x0b, 0xdf, 0x82, 0x00, 0x9e, },
+    { 0xf3, 0xb9, 0xdd, 0x94, 0xc5, 0xbb, 0x5d, 0x7a, },
+    { 0xa7, 0xad, 0x6b, 0x22, 0x46, 0x2f, 0xb3, 0xf4, },
+    { 0xfb, 0xe5, 0x0e, 0x86, 0xbc, 0x8f, 0x1e, 0x75, },
+    { 0x90, 0x3d, 0x84, 0xc0, 0x27, 0x56, 0xea, 0x14, },
+    { 0xee, 0xf2, 0x7a, 0x8e, 0x90, 0xca, 0x23, 0xf7, },
+    { 0xe5, 0x45, 0xbe, 0x49, 0x61, 0xca, 0x29, 0xa1, },
+    { 0xdb, 0x9b, 0xc2, 0x57, 0x7f, 0xcc, 0x2a, 0x3f, },
+    { 0x94, 0x47, 0xbe, 0x2c, 0xf5, 0xe9, 0x9a, 0x69, },
+    { 0x9c, 0xd3, 0x8d, 0x96, 0xf0, 0xb3, 0xc1, 0x4b, },
+    { 0xbd, 0x61, 0x79, 0xa7, 0x1d, 0xc9, 0x6d, 0xbb, },
+    { 0x98, 0xee, 0xa2, 0x1a, 0xf2, 0x5c, 0xd6, 0xbe, },
+    { 0xc7, 0x67, 0x3b, 0x2e, 0xb0, 0xcb, 0xf2, 0xd0, },
+    { 0x88, 0x3e, 0xa3, 0xe3, 0x95, 0x67, 0x53, 0x93, },
+    { 0xc8, 0xce, 0x5c, 0xcd, 0x8c, 0x03, 0x0c, 0xa8, },
+    { 0x94, 0xaf, 0x49, 0xf6, 0xc6, 0x50, 0xad, 0xb8, },
+    { 0xea, 0xb8, 0x85, 0x8a, 0xde, 0x92, 0xe1, 0xbc, },
+    { 0xf3, 0x15, 0xbb, 0x5b, 0xb8, 0x35, 0xd8, 0x17, },
+    { 0xad, 0xcf, 0x6b, 0x07, 0x63, 0x61, 0x2e, 0x2f, },
+    { 0xa5, 0xc9, 0x1d, 0xa7, 0xac, 0xaa, 0x4d, 0xde, },
+    { 0x71, 0x65, 0x95, 0x87, 0x66, 0x50, 0xa2, 0xa6, },
+    { 0x28, 0xef, 0x49, 0x5c, 0x53, 0xa3, 0x87, 0xad, },
+    { 0x42, 0xc3, 0x41, 0xd8, 0xfa, 0x92, 0xd8, 0x32, },
+    { 0xce, 0x7c, 0xf2, 0x72, 0x2f, 0x51, 0x27, 0x71, },
+    { 0xe3, 0x78, 0x59, 0xf9, 0x46, 0x23, 0xf3, 0xa7, },
+    { 0x38, 0x12, 0x05, 0xbb, 0x1a, 0xb0, 0xe0, 0x12, },
+    { 0xae, 0x97, 0xa1, 0x0f, 0xd4, 0x34, 0xe0, 0x15, },
+    { 0xb4, 0xa3, 0x15, 0x08, 0xbe, 0xff, 0x4d, 0x31, },
+    { 0x81, 0x39, 0x62, 0x29, 0xf0, 0x90, 0x79, 0x02, },
+    { 0x4d, 0x0c, 0xf4, 0x9e, 0xe5, 0xd4, 0xdc, 0xca, },
+    { 0x5c, 0x73, 0x33, 0x6a, 0x76, 0xd8, 0xbf, 0x9a, },
+    { 0xd0, 0xa7, 0x04, 0x53, 0x6b, 0xa9, 0x3e, 0x0e, },
+    { 0x92, 0x59, 0x58, 0xfc, 0xd6, 0x42, 0x0c, 0xad, },
+    { 0xa9, 0x15, 0xc2, 0x9b, 0xc8, 0x06, 0x73, 0x18, },
+    { 0x95, 0x2b, 0x79, 0xf3, 0xbc, 0x0a, 0xa6, 0xd4, },
+    { 0xf2, 0x1d, 0xf2, 0xe4, 0x1d, 0x45, 0x35, 0xf9, },
+    { 0x87, 0x57, 0x75, 0x19, 0x04, 0x8f, 0x53, 0xa9, },
+    { 0x10, 0xa5, 0x6c, 0xf5, 0xdf, 0xcd, 0x9a, 0xdb, },
+    { 0xeb, 0x75, 0x09, 0x5c, 0xcd, 0x98, 0x6c, 0xd0, },
+    { 0x51, 0xa9, 0xcb, 0x9e, 0xcb, 0xa3, 0x12, 0xe6, },
+    { 0x96, 0xaf, 0xad, 0xfc, 0x2c, 0xe6, 0x66, 0xc7, },
+    { 0x72, 0xfe, 0x52, 0x97, 0x5a, 0x43, 0x64, 0xee, },
+    { 0x5a, 0x16, 0x45, 0xb2, 0x76, 0xd5, 0x92, 0xa1, },
+    { 0xb2, 0x74, 0xcb, 0x8e, 0xbf, 0x87, 0x87, 0x0a, },
+    { 0x6f, 0x9b, 0xb4, 0x20, 0x3d, 0xe7, 0xb3, 0x81, },
+    { 0xea, 0xec, 0xb2, 0xa3, 0x0b, 0x22, 0xa8, 0x7f, },
+    { 0x99, 0x24, 0xa4, 0x3c, 0xc1, 0x31, 0x57, 0x24, },
+    { 0xbd, 0x83, 0x8d, 0x3a, 0xaf, 0xbf, 0x8d, 0xb7, },
+    { 0x0b, 0x1a, 0x2a, 0x32, 0x65, 0xd5, 0x1a, 0xea, },
+    { 0x13, 0x50, 0x79, 0xa3, 0x23, 0x1c, 0xe6, 0x60, },
+    { 0x93, 0x2b, 0x28, 0x46, 0xe4, 0xd7, 0x06, 0x66, },
+    { 0xe1, 0x91, 0x5f, 0x5c, 0xb1, 0xec, 0xa4, 0x6c, },
+    { 0xf3, 0x25, 0x96, 0x5c, 0xa1, 0x6d, 0x62, 0x9f, },
+    { 0x57, 0x5f, 0xf2, 0x8e, 0x60, 0x38, 0x1b, 0xe5, },
+    { 0x72, 0x45, 0x06, 0xeb, 0x4c, 0x32, 0x8a, 0x95, },
+};
+
+
+/* Test siphash using a test vector. Returns 0 if the function passed
+ * all the tests, otherwise 1 is returned. */
+int siphash_test(void) {
+    uint8_t in[64], k[16];
+    int i;
+    int fails = 0;
+
+    for (i = 0; i < 16; ++i)
+        k[i] = i;
+
+    for (i = 0; i < 64; ++i) {
+        in[i] = i;
+        uint64_t hash = siphash(in, i, k);
+        const uint8_t *v = NULL;
+        v = (uint8_t *)vectors_sip64;
+        if (memcmp(&hash, v + (i * 8), 8)) {
+            /* printf("fail for %d bytes\n", i); */
+            fails++;
+        }
+    }
+
+    /* Run a few basic tests with the case insensitive version. */
+    uint64_t h1, h2;
+    h1 = siphash((uint8_t*)"hello world",11,(uint8_t*)"1234567812345678");
+    h2 = siphash_nocase((uint8_t*)"hello world",11,(uint8_t*)"1234567812345678");
+    if (h1 != h2) fails++;
+
+    h1 = siphash((uint8_t*)"hello world",11,(uint8_t*)"1234567812345678");
+    h2 = siphash_nocase((uint8_t*)"HELLO world",11,(uint8_t*)"1234567812345678");
+    if (h1 != h2) fails++;
+
+    h1 = siphash((uint8_t*)"HELLO world",11,(uint8_t*)"1234567812345678");
+    h2 = siphash_nocase((uint8_t*)"HELLO world",11,(uint8_t*)"1234567812345678");
+    if (h1 == h2) fails++;
+
+    if (!fails) return 0;
+    return 1;
+}
+
+int main(void) {
+    if (siphash_test() == 0) {
+        printf("SipHash test: OK\n");
+        return 0;
+    } else {
+        printf("SipHash test: FAILED\n");
+        return 1;
+    }
+}
+
+#endif
diff --git a/src/t_zset.c b/src/t_zset.c
index d36fa30ae..f7f4c6eb2 100644
--- a/src/t_zset.c
+++ b/src/t_zset.c
@@ -2110,7 +2110,7 @@ inline static void zunionInterAggregate(double *target, double val, int aggregat
     }
 }
 
-unsigned int dictSdsHash(const void *key);
+uint64_t dictSdsHash(const void *key);
 int dictSdsKeyCompare(void *privdata, const void *key1, const void *key2);
 
 dictType setAccumulatorDictType = {
-- 
cgit v1.2.1


From 05ea8c612238bc8b9a8eda2427d6cf8a3afb4ac5 Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Mon, 20 Feb 2017 17:32:46 +0100
Subject: SipHash x86 optimizations.

---
 src/siphash.c | 25 +++++++++++++++++++++++--
 1 file changed, 23 insertions(+), 2 deletions(-)

diff --git a/src/siphash.c b/src/siphash.c
index 04e571fd7..3bcada728 100644
--- a/src/siphash.c
+++ b/src/siphash.c
@@ -40,6 +40,13 @@
 #include <string.h>
 #include <ctype.h>
 
+/* Test of the CPU is Little Endian and supports not aligned accesses.
+ * Two interesting conditions to speedup the function that happen to be
+ * in most of x86 servers. */
+#if defined(__X86_64__) || defined(__x86_64__) || defined (__i386__)
+#define UNALIGNED_LE_CPU
+#endif
+
 #define ROTL(x, b) (uint64_t)(((x) << (b)) | ((x) >> (64 - (b))))
 
 #define U32TO8_LE(p, v)                                                        \
@@ -52,11 +59,15 @@
     U32TO8_LE((p), (uint32_t)((v)));                                           \
     U32TO8_LE((p) + 4, (uint32_t)((v) >> 32));
 
+#ifdef UNALIGNED_LE_CPU
+#define U8TO64_LE(p) (*((uint64_t*)(p)))
+#else
 #define U8TO64_LE(p)                                                           \
     (((uint64_t)((p)[0])) | ((uint64_t)((p)[1]) << 8) |                        \
      ((uint64_t)((p)[2]) << 16) | ((uint64_t)((p)[3]) << 24) |                 \
      ((uint64_t)((p)[4]) << 32) | ((uint64_t)((p)[5]) << 40) |                 \
      ((uint64_t)((p)[6]) << 48) | ((uint64_t)((p)[7]) << 56))
+#endif
 
 #define U8TO64_LE_NOCASE(p)                                                    \
     (((uint64_t)(tolower((p)[0]))) |                                           \
@@ -87,8 +98,10 @@
     } while (0)
 
 uint64_t siphash(const uint8_t *in, const size_t inlen, const uint8_t *k) {
+#ifndef UNALIGNED_LE_CPU
     uint64_t hash;
     uint8_t *out = (uint8_t*) &hash;
+#endif
     uint64_t v0 = 0x736f6d6570736575ULL;
     uint64_t v1 = 0x646f72616e646f6dULL;
     uint64_t v2 = 0x6c7967656e657261ULL;
@@ -139,15 +152,20 @@ uint64_t siphash(const uint8_t *in, const size_t inlen, const uint8_t *k) {
     SIPROUND;
 
     b = v0 ^ v1 ^ v2 ^ v3;
+#ifndef UNALIGNED_LE_CPU
     U64TO8_LE(out, b);
-
     return hash;
+#else
+    return b;
+#endif
 }
 
 uint64_t siphash_nocase(const uint8_t *in, const size_t inlen, const uint8_t *k)
 {
+#ifndef UNALIGNED_LE_CPU
     uint64_t hash;
     uint8_t *out = (uint8_t*) &hash;
+#endif
     uint64_t v0 = 0x736f6d6570736575ULL;
     uint64_t v1 = 0x646f72616e646f6dULL;
     uint64_t v2 = 0x6c7967656e657261ULL;
@@ -198,9 +216,12 @@ uint64_t siphash_nocase(const uint8_t *in, const size_t inlen, const uint8_t *k)
     SIPROUND;
 
     b = v0 ^ v1 ^ v2 ^ v3;
+#ifndef UNALIGNED_LE_CPU
     U64TO8_LE(out, b);
-
     return hash;
+#else
+    return b;
+#endif
 }
 
 
-- 
cgit v1.2.1


From 84fa8230e5a75c8ff4eed6afa56d7a6babdb8fb9 Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Mon, 20 Feb 2017 17:39:44 +0100
Subject: Use locale agnostic tolower() in dict.c hash function.

---
 src/Makefile  |  2 +-
 src/dict.c    |  2 +-
 src/siphash.c | 40 +++++++++++++++++++++++++---------------
 3 files changed, 27 insertions(+), 17 deletions(-)

diff --git a/src/Makefile b/src/Makefile
index 5f7536e83..cb7afa023 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -204,7 +204,7 @@ $(REDIS_CHECK_AOF_NAME): $(REDIS_CHECK_AOF_OBJ)
 	$(REDIS_LD) -o $@ $^ $(FINAL_LIBS)
 
 dict-benchmark: dict.c zmalloc.c sds.c
-	$(REDIS_CC) $(FINAL_CFLAGS) dict.c zmalloc.c sds.c -D DICT_BENCHMARK_MAIN -o dict-benchmark
+	$(REDIS_CC) $(FINAL_CFLAGS) dict.c zmalloc.c sds.c siphash.c -D DICT_BENCHMARK_MAIN -o dict-benchmark
 
 # Because the jemalloc.h header is generated as a part of the jemalloc build,
 # building it should complete before building any other object. Instead of
diff --git a/src/dict.c b/src/dict.c
index 8ce735961..69fb3b8f8 100644
--- a/src/dict.c
+++ b/src/dict.c
@@ -1109,7 +1109,7 @@ void dictGetStats(char *buf, size_t bufsize, dict *d) {
 
 #include "sds.h"
 
-unsigned int hashCallback(const void *key) {
+uint64_t hashCallback(const void *key) {
     return dictGenHashFunction((unsigned char*)key, sdslen((char*)key));
 }
 
diff --git a/src/siphash.c b/src/siphash.c
index 3bcada728..7219d4b88 100644
--- a/src/siphash.c
+++ b/src/siphash.c
@@ -40,6 +40,16 @@
 #include <string.h>
 #include <ctype.h>
 
+/* Fast tolower() alike function that does not care about locale
+ * but just returns a-z insetad of A-Z. */
+int siptlw(int c) {
+    if (c >= 'A' && c <= 'Z') {
+        return c+('a'-'A');
+    } else {
+        return c;
+    }
+}
+
 /* Test of the CPU is Little Endian and supports not aligned accesses.
  * Two interesting conditions to speedup the function that happen to be
  * in most of x86 servers. */
@@ -70,14 +80,14 @@
 #endif
 
 #define U8TO64_LE_NOCASE(p)                                                    \
-    (((uint64_t)(tolower((p)[0]))) |                                           \
-     ((uint64_t)(tolower((p)[1])) << 8) |                                      \
-     ((uint64_t)(tolower((p)[2])) << 16) |                                     \
-     ((uint64_t)(tolower((p)[3])) << 24) |                                     \
-     ((uint64_t)(tolower((p)[4])) << 32) |                                              \
-     ((uint64_t)(tolower((p)[5])) << 40) |                                              \
-     ((uint64_t)(tolower((p)[6])) << 48) |                                              \
-     ((uint64_t)(tolower((p)[7])) << 56))
+    (((uint64_t)(siptlw((p)[0]))) |                                           \
+     ((uint64_t)(siptlw((p)[1])) << 8) |                                      \
+     ((uint64_t)(siptlw((p)[2])) << 16) |                                     \
+     ((uint64_t)(siptlw((p)[3])) << 24) |                                     \
+     ((uint64_t)(siptlw((p)[4])) << 32) |                                              \
+     ((uint64_t)(siptlw((p)[5])) << 40) |                                              \
+     ((uint64_t)(siptlw((p)[6])) << 48) |                                              \
+     ((uint64_t)(siptlw((p)[7])) << 56))
 
 #define SIPROUND                                                               \
     do {                                                                       \
@@ -192,13 +202,13 @@ uint64_t siphash_nocase(const uint8_t *in, const size_t inlen, const uint8_t *k)
     }
 
     switch (left) {
-    case 7: b |= ((uint64_t)tolower(in[6])) << 48;
-    case 6: b |= ((uint64_t)tolower(in[5])) << 40;
-    case 5: b |= ((uint64_t)tolower(in[4])) << 32;
-    case 4: b |= ((uint64_t)tolower(in[3])) << 24;
-    case 3: b |= ((uint64_t)tolower(in[2])) << 16;
-    case 2: b |= ((uint64_t)tolower(in[1])) << 8;
-    case 1: b |= ((uint64_t)tolower(in[0])); break;
+    case 7: b |= ((uint64_t)siptlw(in[6])) << 48;
+    case 6: b |= ((uint64_t)siptlw(in[5])) << 40;
+    case 5: b |= ((uint64_t)siptlw(in[4])) << 32;
+    case 4: b |= ((uint64_t)siptlw(in[3])) << 24;
+    case 3: b |= ((uint64_t)siptlw(in[2])) << 16;
+    case 2: b |= ((uint64_t)siptlw(in[1])) << 8;
+    case 1: b |= ((uint64_t)siptlw(in[0])); break;
     case 0: break;
     }
 
-- 
cgit v1.2.1


From cd90389b30f745ef1ed04fd579adbd914189a10a Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Tue, 21 Feb 2017 12:25:29 +0100
Subject: freeMemoryIfNeeded(): improve code and lazyfree handling.

1. Refactor memory overhead computation into a function.
2. Every 10 keys evicted, check if memory usage already reached
   the target value directly, since we otherwise don't count all
   the memory reclaimed by the background thread right now.
---
 src/evict.c | 61 ++++++++++++++++++++++++++++++++++++++++++-------------------
 1 file changed, 42 insertions(+), 19 deletions(-)

diff --git a/src/evict.c b/src/evict.c
index 802997ce8..62753c5a7 100644
--- a/src/evict.c
+++ b/src/evict.c
@@ -336,20 +336,13 @@ unsigned long LFUDecrAndReturn(robj *o) {
  * server when there is data to add in order to make space if needed.
  * --------------------------------------------------------------------------*/
 
-int freeMemoryIfNeeded(void) {
-    size_t mem_reported, mem_used, mem_tofree, mem_freed;
+/* We don't want to count AOF buffers and slaves output buffers as
+ * used memory: the eviction should use mostly data size. This function
+ * returns the sum of AOF and slaves buffer. */
+size_t freeMemoryGetNotCountedMemory(void) {
+    size_t overhead = 0;
     int slaves = listLength(server.slaves);
-    mstime_t latency, eviction_latency;
-    long long delta;
-
-    /* Check if we are over the memory usage limit. If we are not, no need
-     * to subtract the slaves output buffers. We can just return ASAP. */
-    mem_reported = zmalloc_used_memory();
-    if (mem_reported <= server.maxmemory) return C_OK;
 
-    /* Remove the size of slaves output buffers and AOF buffer from the
-     * count of used memory. */
-    mem_used = mem_reported;
     if (slaves) {
         listIter li;
         listNode *ln;
@@ -357,17 +350,31 @@ int freeMemoryIfNeeded(void) {
         listRewind(server.slaves,&li);
         while((ln = listNext(&li))) {
             client *slave = listNodeValue(ln);
-            unsigned long obuf_bytes = getClientOutputBufferMemoryUsage(slave);
-            if (obuf_bytes > mem_used)
-                mem_used = 0;
-            else
-                mem_used -= obuf_bytes;
+            overhead += getClientOutputBufferMemoryUsage(slave);
         }
     }
     if (server.aof_state != AOF_OFF) {
-        mem_used -= sdslen(server.aof_buf);
-        mem_used -= aofRewriteBufferSize();
+        overhead += sdslen(server.aof_buf)+aofRewriteBufferSize();
     }
+    return overhead;
+}
+
+int freeMemoryIfNeeded(void) {
+    size_t mem_reported, mem_used, mem_tofree, mem_freed;
+    mstime_t latency, eviction_latency;
+    long long delta;
+    int slaves = listLength(server.slaves);
+
+    /* Check if we are over the memory usage limit. If we are not, no need
+     * to subtract the slaves output buffers. We can just return ASAP. */
+    mem_reported = zmalloc_used_memory();
+    if (mem_reported <= server.maxmemory) return C_OK;
+
+    /* Remove the size of slaves output buffers and AOF buffer from the
+     * count of used memory. */
+    mem_used = mem_reported;
+    size_t overhead = freeMemoryGetNotCountedMemory();
+    mem_used = (mem_used > overhead) ? mem_used-overhead : 0;
 
     /* Check if we are still over the memory limit. */
     if (mem_used <= server.maxmemory) return C_OK;
@@ -498,6 +505,22 @@ int freeMemoryIfNeeded(void) {
              * deliver data to the slaves fast enough, so we force the
              * transmission here inside the loop. */
             if (slaves) flushSlavesOutputBuffers();
+
+            /* Normally our stop condition is the ability to release
+             * a fixed, pre-computed amount of memory. However when we
+             * are deleting objects in another thread, it's better to
+             * check, from time to time, if we already reached our target
+             * memory, since the "mem_freed" amount is computed only
+             * across the dbAsyncDelete() call, while the thread can
+             * release the memory all the time. */
+            if (server.lazyfree_lazy_eviction && !(keys_freed % 16)) {
+                overhead = freeMemoryGetNotCountedMemory();
+                mem_used = zmalloc_used_memory();
+                mem_used = (mem_used > overhead) ? mem_used-overhead : 0;
+                if (mem_used <= server.maxmemory) {
+                    mem_freed = mem_tofree;
+                }
+            }
         }
 
         if (!keys_freed) {
-- 
cgit v1.2.1


From 0285c2714b6f1f4517d2ac3bc34177f874205a8b Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Tue, 21 Feb 2017 17:07:28 +0100
Subject: SipHash 2-4 -> SipHash 1-2.

For performance reasons we use a reduced rounds variant of
SipHash. This should still provide enough protection and the
effects in the hash table distribution are non existing.
If some real world attack on SipHash 1-2 will be found we can
trivially switch to something more secure. Anyway it is a
big step forward from Murmurhash, for which it is trivial to
generate *seed independent* colliding keys... The speed
penatly introduced by SipHash 2-4, around 4%, was a too big
price to pay compared to the effectiveness of the HashDoS
attack against SipHash 1-2, and considering so far in the
Redis history, no such an incident ever happened even while
using trivially to collide hash functions.
---
 src/siphash.c | 29 +++++++++++++++--------------
 1 file changed, 15 insertions(+), 14 deletions(-)

diff --git a/src/siphash.c b/src/siphash.c
index 7219d4b88..6c41fe6b6 100644
--- a/src/siphash.c
+++ b/src/siphash.c
@@ -19,18 +19,23 @@
    This version was modified by Salvatore Sanfilippo <antirez@gmail.com>
    in the following ways:
 
-   1. Hard-code 2-4 rounds in the hope the compiler can optimize it more
+   1. We use SipHash 1-2. This is not believed to be as strong as the
+      suggested 2-4 variant, but AFAIK there are not trivial attacks
+      against this reduced-rounds version, and it runs at the same speed
+      as Murmurhash2 that we used previously, why the 2-4 variant slowed
+      down Redis by a 4% figure more or less.
+   2. Hard-code rounds in the hope the compiler can optimize it more
       in this raw from. Anyway we always want the standard 2-4 variant.
-   2. Modify the prototype and implementation so that the function directly
+   3. Modify the prototype and implementation so that the function directly
       returns an uint64_t value, the hash itself, instead of receiving an
       output buffer. This also means that the output size is set to 8 bytes
       and the 16 bytes output code handling was removed.
-   3. Provide a case insensitive variant to be used when hashing strings that
+   4. Provide a case insensitive variant to be used when hashing strings that
       must be considered identical by the hash table regardless of the case.
       If we don't have directly a case insensitive hash function, we need to
       perform a text transformation in some temporary buffer, which is costly.
-   4. Remove debugging code.
-   5. Modified the original test.c file to be a stand-alone function testing
+   5. Remove debugging code.
+   6. Modified the original test.c file to be a stand-alone function testing
       the function in the new form (returing an uint64_t) using just the
       relevant test vector.
  */
@@ -131,7 +136,6 @@ uint64_t siphash(const uint8_t *in, const size_t inlen, const uint8_t *k) {
         m = U8TO64_LE(in);
         v3 ^= m;
 
-        SIPROUND;
         SIPROUND;
 
         v0 ^= m;
@@ -150,14 +154,11 @@ uint64_t siphash(const uint8_t *in, const size_t inlen, const uint8_t *k) {
 
     v3 ^= b;
 
-    SIPROUND;
     SIPROUND;
 
     v0 ^= b;
     v2 ^= 0xff;
 
-    SIPROUND;
-    SIPROUND;
     SIPROUND;
     SIPROUND;
 
@@ -195,7 +196,6 @@ uint64_t siphash_nocase(const uint8_t *in, const size_t inlen, const uint8_t *k)
         m = U8TO64_LE_NOCASE(in);
         v3 ^= m;
 
-        SIPROUND;
         SIPROUND;
 
         v0 ^= m;
@@ -214,14 +214,11 @@ uint64_t siphash_nocase(const uint8_t *in, const size_t inlen, const uint8_t *k)
 
     v3 ^= b;
 
-    SIPROUND;
     SIPROUND;
 
     v0 ^= b;
     v2 ^= 0xff;
 
-    SIPROUND;
-    SIPROUND;
     SIPROUND;
     SIPROUND;
 
@@ -308,7 +305,11 @@ const uint8_t vectors_sip64[64][8] = {
 
 
 /* Test siphash using a test vector. Returns 0 if the function passed
- * all the tests, otherwise 1 is returned. */
+ * all the tests, otherwise 1 is returned.
+ *
+ * IMPORTANT: The test vector is for SipHash 2-4. Before running
+ * the test revert back the siphash() function to 2-4 rounds since
+ * now it uses 1-2 rounds. */
 int siphash_test(void) {
     uint8_t in[64], k[16];
     int i;
-- 
cgit v1.2.1


From 7c8ddab4f88a76413641b3dc49b0c42ac8ee3b2e Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Wed, 22 Feb 2017 12:00:09 +0100
Subject: Test: fix conditional execution of HINCRBYFLOAT representation test.

---
 tests/unit/type/hash.tcl | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/unit/type/hash.tcl b/tests/unit/type/hash.tcl
index 02560ede2..d2c679d32 100644
--- a/tests/unit/type/hash.tcl
+++ b/tests/unit/type/hash.tcl
@@ -525,7 +525,7 @@ start_server {tags {"hash"}} {
     # 1.23 cannot be represented correctly with 64 bit doubles, so we skip
     # the test, since we are only testing pretty printing here and is not
     # a bug if the program outputs things like 1.299999...
-    if {!$::valgrind || ![string match *x86_64* [exec uname -a]]} {
+    if {!$::valgrind && [string match *x86_64* [exec uname -a]]} {
         test {Test HINCRBYFLOAT for correct float representation (issue #2846)} {
             r del myhash
             assert {[r hincrbyfloat myhash float 1.23] eq {1.23}}
-- 
cgit v1.2.1


From 2b36706a48759ac6ed991ab1f8257de122425a75 Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Wed, 22 Feb 2017 12:27:01 +0100
Subject: Test: replication-psync, wait more to detect write load.

Slow systems like the original Raspberry PI need more time
than 5 seconds to start the script and detect writes.
After fixing the Raspberry PI can pass the unit without issues.
---
 tests/integration/replication-psync.tcl | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/integration/replication-psync.tcl b/tests/integration/replication-psync.tcl
index 3a41ceb21..da1e9cf5b 100644
--- a/tests/integration/replication-psync.tcl
+++ b/tests/integration/replication-psync.tcl
@@ -47,7 +47,7 @@ proc test_psync {descr duration backlog_size backlog_ttl delay cond diskless rec
 
             # Check that the background clients are actually writing.
             test {Detect write load to master} {
-                wait_for_condition 50 100 {
+                wait_for_condition 50 1000 {
                     [$master dbsize] > 100
                 } else {
                     fail "Can't detect write load from background clients."
-- 
cgit v1.2.1


From 95883313b5a405916fabed34c3af290d6072c817 Mon Sep 17 00:00:00 2001
From: antirez <antirez@gmail.com>
Date: Wed, 22 Feb 2017 13:08:21 +0100
Subject: Solaris fixes about tail usage and atomic vars.

Testing with Solaris C compiler (SunOS 5.11 11.2 sun4v sparc sun4v)
there were issues compiling due to atomicvar.h and running the
tests also failed because of "tail" usage not conform with Solaris
tail implementation. This commit fixes both the issues.
---
 src/atomicvar.h           | 2 +-
 tests/integration/aof.tcl | 6 +++---
 tests/integration/rdb.tcl | 4 ++--
 tests/unit/aofrw.tcl      | 2 +-
 4 files changed, 7 insertions(+), 7 deletions(-)

diff --git a/src/atomicvar.h b/src/atomicvar.h
index 4aa8fa173..c522db3e9 100644
--- a/src/atomicvar.h
+++ b/src/atomicvar.h
@@ -51,7 +51,7 @@
 #ifndef __ATOMIC_VAR_H
 #define __ATOMIC_VAR_H
 
-#if defined(__ATOMIC_RELAXED) && (!defined(__clang__) || !defined(__APPLE__) || __apple_build_version__ > 4210057)
+#if defined(__ATOMIC_RELAXED) && !defined(__sun) && (!defined(__clang__) || !defined(__APPLE__) || __apple_build_version__ > 4210057)
 /* Implementation using __atomic macros. */
 
 #define atomicIncr(var,count,mutex) __atomic_add_fetch(&var,(count),__ATOMIC_RELAXED)
diff --git a/tests/integration/aof.tcl b/tests/integration/aof.tcl
index 832f996e1..e397faeeb 100644
--- a/tests/integration/aof.tcl
+++ b/tests/integration/aof.tcl
@@ -88,7 +88,7 @@ tags {"aof"} {
             set pattern "*Bad file format reading the append only file*"
             set retry 10
             while {$retry} {
-                set result [exec tail -n1 < [dict get $srv stdout]]
+                set result [exec tail -1 < [dict get $srv stdout]]
                 if {[string match $pattern $result]} {
                     break
                 }
@@ -113,7 +113,7 @@ tags {"aof"} {
             set pattern "*Unexpected end of file reading the append only file*"
             set retry 10
             while {$retry} {
-                set result [exec tail -n1 < [dict get $srv stdout]]
+                set result [exec tail -1 < [dict get $srv stdout]]
                 if {[string match $pattern $result]} {
                     break
                 }
@@ -137,7 +137,7 @@ tags {"aof"} {
             set pattern "*Unexpected end of file reading the append only file*"
             set retry 10
             while {$retry} {
-                set result [exec tail -n1 < [dict get $srv stdout]]
+                set result [exec tail -1 < [dict get $srv stdout]]
                 if {[string match $pattern $result]} {
                     break
                 }
diff --git a/tests/integration/rdb.tcl b/tests/integration/rdb.tcl
index 2ed47cc58..66aad4cc7 100644
--- a/tests/integration/rdb.tcl
+++ b/tests/integration/rdb.tcl
@@ -66,7 +66,7 @@ if {!$isroot} {
         test {Server should not start if RDB file can't be open} {
             wait_for_condition 50 100 {
                 [string match {*Fatal error loading*} \
-                    [exec tail -n1 < [dict get $srv stdout]]]
+                    [exec tail -1 < [dict get $srv stdout]]]
             } else {
                 fail "Server started even if RDB was unreadable!"
             }
@@ -90,7 +90,7 @@ start_server_and_kill_it [list "dir" $server_path] {
     test {Server should not start if RDB is corrupted} {
         wait_for_condition 50 100 {
             [string match {*CRC error*} \
-                [exec tail -n10 < [dict get $srv stdout]]]
+                [exec tail -10 < [dict get $srv stdout]]]
         } else {
             fail "Server started even if RDB was corrupted!"
         }
diff --git a/tests/unit/aofrw.tcl b/tests/unit/aofrw.tcl
index c5430eedc..dff7588ff 100644
--- a/tests/unit/aofrw.tcl
+++ b/tests/unit/aofrw.tcl
@@ -73,7 +73,7 @@ start_server {tags {"aofrw"}} {
         r config set appendonly no
         r exec
         wait_for_condition 50 100 {
-            [string match {*Killing*AOF*child*} [exec tail -n5 < [srv 0 stdout]]]
+            [string match {*Killing*AOF*child*} [exec tail -5 < [srv 0 stdout]]]
         } else {
             fail "Can't find 'Killing AOF child' into recent logs"
         }
-- 
cgit v1.2.1


From d7826823c05ad82f710da74c3d5d402898187888 Mon Sep 17 00:00:00 2001
From: Salvatore Sanfilippo <antirez@gmail.com>
Date: Thu, 23 Feb 2017 22:38:44 +0800
Subject: Fix BITPOS unaligned memory access.

---
 src/bitops.c | 21 ++++++++++++++-------
 1 file changed, 14 insertions(+), 7 deletions(-)

diff --git a/src/bitops.c b/src/bitops.c
index 7ab72633c..d5b96ce1e 100644
--- a/src/bitops.c
+++ b/src/bitops.c
@@ -104,6 +104,7 @@ long redisBitpos(void *s, unsigned long count, int bit) {
     unsigned long skipval, word = 0, one;
     long pos = 0; /* Position of bit, to return to the caller. */
     unsigned long j;
+    int found;
 
     /* Process whole words first, seeking for first word that is not
      * all ones or all zeros respectively if we are lookig for zeros
@@ -117,21 +118,27 @@ long redisBitpos(void *s, unsigned long count, int bit) {
     /* Skip initial bits not aligned to sizeof(unsigned long) byte by byte. */
     skipval = bit ? 0 : UCHAR_MAX;
     c = (unsigned char*) s;
+    found = 0;
     while((unsigned long)c & (sizeof(*l)-1) && count) {
-        if (*c != skipval) break;
+        if (*c != skipval) {
+            found = 1;
+            break;
+        }
         c++;
         count--;
         pos += 8;
     }
 
     /* Skip bits with full word step. */
-    skipval = bit ? 0 : ULONG_MAX;
     l = (unsigned long*) c;
-    while (count >= sizeof(*l)) {
-        if (*l != skipval) break;
-        l++;
-        count -= sizeof(*l);
-        pos += sizeof(*l)*8;
+    if (!found) {
+        skipval = bit ? 0 : ULONG_MAX;
+        while (count >= sizeof(*l)) {
+            if (*l != skipval) break;
+            l++;
+            count -= sizeof(*l);
+            pos += sizeof(*l)*8;
+        }
     }
 
     /* Load bytes into "word" considering the first byte as the most significant
-- 
cgit v1.2.1


From b3391fd853d55b3da04ede024adcc6bf017c78f1 Mon Sep 17 00:00:00 2001
From: Salvatore Sanfilippo <antirez@gmail.com>
Date: Thu, 23 Feb 2017 22:39:44 +0800
Subject: Use ARM unaligned accesses ifdefs for SPARC as well.

---
 src/bitops.c      | 2 +-
 src/config.h      | 9 +++++++++
 src/hyperloglog.c | 2 +-
 3 files changed, 11 insertions(+), 2 deletions(-)

diff --git a/src/bitops.c b/src/bitops.c
index d5b96ce1e..43450fca3 100644
--- a/src/bitops.c
+++ b/src/bitops.c
@@ -665,7 +665,7 @@ void bitopCommand(client *c) {
          * result in GCC compiling the code using multiple-words load/store
          * operations that are not supported even in ARM >= v6. */
         j = 0;
-        #ifndef __arm__
+        #ifndef USE_ALIGNED_ACCESS
         if (minlen >= sizeof(unsigned long)*4 && numkeys <= 16) {
             unsigned long *lp[16];
             unsigned long *lres = (unsigned long*) res;
diff --git a/src/config.h b/src/config.h
index 1005dcc2a..c23f1c789 100644
--- a/src/config.h
+++ b/src/config.h
@@ -215,4 +215,13 @@ void setproctitle(const char *fmt, ...);
 #define __arm64__
 #endif
 
+/* Make sure we can test for SPARC just checking for __sparc__. */
+#if defined(__sparc) && !defined(__sparc__)
+#define __sparc__
+#endif
+
+#if defined(__sparc__) || defined(__arm__)
+#define USE_ALIGNED_ACCESS
+#endif
+
 #endif
diff --git a/src/hyperloglog.c b/src/hyperloglog.c
index b8a63a73f..49516f824 100644
--- a/src/hyperloglog.c
+++ b/src/hyperloglog.c
@@ -401,7 +401,7 @@ uint64_t MurmurHash64A (const void * key, int len, unsigned int seed) {
         uint64_t k;
 
 #if (BYTE_ORDER == LITTLE_ENDIAN)
-	#if defined(__arm__) && !defined(__arm64__)
+	#ifdef USE_ALIGNED_ACCESS
 	memcpy(&k,data,sizeof(uint64_t));
 	#else
         k = *((uint64_t*)data);
-- 
cgit v1.2.1


From 9cc83d2ad9d5ed0e6a396a7e26814ac2f2445d1c Mon Sep 17 00:00:00 2001
From: Salvatore Sanfilippo <antirez@gmail.com>
Date: Fri, 24 Feb 2017 00:00:13 +0800
Subject: Makefile: fix building with Solaris C compiler, 64 bit.

---
 src/Makefile | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/src/Makefile b/src/Makefile
index 8cf6ba6f8..5d2b2b395 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -65,6 +65,13 @@ DEBUG=-g -ggdb
 
 ifeq ($(uname_S),SunOS)
 	# SunOS
+        ifneq ($(@@),32bit)
+		CFLAGS+= -m64
+		LDFLAGS+= -m64
+	endif
+	DEBUG=-g
+	DEBUG_FLAGS=-g
+	export CFLAGS LDFLAGS DEBUG DEBUG_FLAGS
 	INSTALL=cp -pf
 	FINAL_CFLAGS+= -D__EXTENSIONS__ -D_XPG6
 	FINAL_LIBS+= -ldl -lnsl -lsocket -lresolv -lpthread -lrt
-- 
cgit v1.2.1