Forward port r34151 from ruby_1_8_7.

git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_1_8@34379 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

6 files changed, 119 insertions, 19 deletions

diff --git a/ChangeLog b/ChangeLog
index c6b08c3257..409fb81af8 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,29 @@
+Thu Jan 26 16:08:43 2012  URABE Shyouhei  <shyouhei@ruby-lang.org>
+
+	* string.c (rb_str_hash): randomize hash to avoid algorithmic
+	  complexity attacks. CVE-2011-4815
+
+	* st.c (strhash): ditto.
+
+	* string.c (Init_String): initialization of hash_seed to be at the
+	  beginning of the process.
+
+	* st.c (Init_st): ditto.
+
+Thu Jan 26 16:08:43 2012  Tanaka Akira  <akr@fsij.org>
+
+	* inits.c (rb_call_inits): call Init_RandomSeed at first.
+
+	* random.c (seed_initialized): defined.
+	  (fill_random_seed): extracted from random_seed.
+	  (make_seed_value): extracted from random_seed.
+	  (rb_f_rand): initialize random seed at first.
+	  (initial_seed): defined.
+	  (Init_RandomSeed): defined.
+	  (Init_RandomSeed2): defined.
+	  (rb_reset_random_seed): defined.
+	  (Init_Random): call Init_RandomSeed2.
+
 Wed Nov  2 08:16:45 2011  Tanaka Akira  <akr@fsij.org>
 
 	* lib/webrick/utils.rb: fix fcntl call.
diff --git a/inits.c b/inits.c
index 647067f5b4..7c2eae6c48 100644
--- a/inits.c
+++ b/inits.c
@@ -39,6 +39,7 @@ void Init_Precision _((void));
 void Init_sym _((void));
 void Init_process _((void));
 void Init_Random _((void));
+void Init_RandomSeed _((void));
 void Init_Range _((void));
 void Init_Regexp _((void));
 void Init_signal _((void));
@@ -47,10 +48,13 @@ void Init_Struct _((void));
 void Init_Time _((void));
 void Init_var_tables _((void));
 void Init_version _((void));
+void Init_st _((void));
 
 void
 rb_call_inits()
 {
+    Init_RandomSeed();
+    Init_st();
     Init_sym();
     Init_var_tables();
     Init_Object();
diff --git a/random.c b/random.c
index 45a94d0563..75e6dcd851 100644
--- a/random.c
+++ b/random.c
@@ -189,6 +189,7 @@ rb_genrand_real(void)
 #include <fcntl.h>
 #endif
 
+static int seed_initialized = 0;
 static VALUE saved_seed = INT2FIX(0);
 
 static VALUE
@@ -250,27 +251,22 @@ rand_init(vseed)
     return old;
 }
 
-static VALUE
-random_seed()
+#define DEFAULT_SEED_LEN (4 * sizeof(long))
+
+static void
+fill_random_seed(ptr)
+    char *ptr;
 {
     static int n = 0;
+    unsigned long *seed;
     struct timeval tv;
     int fd;
     struct stat statbuf;
+    char *buf = (char*)ptr;
 
-    int seed_len;
-    BDIGIT *digits;
-    unsigned long *seed;
-    NEWOBJ(big, struct RBignum);
-    OBJSETUP(big, rb_cBignum, T_BIGNUM);
-
-    seed_len = 4 * sizeof(long);
-    big->sign = 1;
-    big->len = seed_len / SIZEOF_BDIGITS + 1;
-    digits = big->digits = ALLOC_N(BDIGIT, big->len);
-    seed = (unsigned long *)big->digits;
+    seed = (unsigned long *)buf;
 
-    memset(digits, 0, big->len * SIZEOF_BDIGITS);
+    memset(buf, 0, DEFAULT_SEED_LEN);
 
 #ifdef S_ISCHR
     if ((fd = open("/dev/urandom", O_RDONLY
@@ -282,7 +278,7 @@ random_seed()
 #endif
             )) >= 0) {
         if (fstat(fd, &statbuf) == 0 && S_ISCHR(statbuf.st_mode)) {
-            read(fd, seed, seed_len);
+            read(fd, seed, DEFAULT_SEED_LEN);
         }
         close(fd);
     }
@@ -293,13 +289,37 @@ random_seed()
     seed[1] ^= tv.tv_sec;
     seed[2] ^= getpid() ^ (n++ << 16);
     seed[3] ^= (unsigned long)&seed;
+}
+
+static VALUE
+make_seed_value(char *ptr)
+{
+    BDIGIT *digits;
+    NEWOBJ(big, struct RBignum);
+    OBJSETUP(big, rb_cBignum, T_BIGNUM);
+
+    RBIGNUM_SET_SIGN(big, 1);
+
+    digits = ALLOC_N(char, DEFAULT_SEED_LEN);
+    RBIGNUM(big)->digits = digits;
+    RBIGNUM(big)->len = DEFAULT_SEED_LEN / SIZEOF_BDIGITS;
+
+    MEMCPY(digits, ptr, char, DEFAULT_SEED_LEN);
 
     /* set leading-zero-guard if need. */
-    digits[big->len-1] = digits[big->len-2] <= 1 ? 1 : 0;
+    digits[RBIGNUM_LEN(big)-1] = digits[RBIGNUM_LEN(big)-2] <= 1 ? 1 : 0;
 
     return rb_big_norm((VALUE)big);
 }
 
+static VALUE
+random_seed(void)
+{
+    char buf[DEFAULT_SEED_LEN];
+    fill_random_seed(buf);
+    return make_seed_value(buf);
+}
+
 /*
  *  call-seq:
  *     srand(number=0)    => old_seed
@@ -440,6 +460,9 @@ rb_f_rand(argc, argv, obj)
     long val, max;
 
     rb_scan_args(argc, argv, "01", &vmax);
+    if (!seed_initialized) {
+       rand_init(random_seed());
+    }
     switch (TYPE(vmax)) {
       case T_FLOAT:
 	if (RFLOAT(vmax)->value <= LONG_MAX && RFLOAT(vmax)->value >= LONG_MIN) {
@@ -487,6 +510,8 @@ rb_f_rand(argc, argv, obj)
     return LONG2NUM(val);
 }
 
+static char initial_seed[DEFAULT_SEED_LEN];
+
 void
 rb_reset_random_seed()
 {
@@ -494,9 +519,24 @@ rb_reset_random_seed()
 }
 
 void
+Init_RandomSeed(void)
+{
+    fill_random_seed(initial_seed);
+    init_by_array((unsigned long*)initial_seed, DEFAULT_SEED_LEN/sizeof(unsigned long));
+    seed_initialized = 1;
+}
+
+static void
+Init_RandomSeed2(void)
+{
+    saved_seed = make_seed_value(initial_seed);
+    memset(initial_seed, 0, DEFAULT_SEED_LEN);
+}
+
+void
 Init_Random()
 {
-    rb_reset_random_seed();
+    Init_RandomSeed2();
     rb_define_global_function("srand", rb_f_srand, -1);
     rb_define_global_function("rand", rb_f_rand, -1);
     rb_global_variable(&saved_seed);
diff --git a/st.c b/st.c
index c16c3109a8..21e157aed9 100644
--- a/st.c
+++ b/st.c
@@ -9,6 +9,7 @@
 #include <stdlib.h>
 #endif
 #include <string.h>
+#include <limits.h>
 #include "st.h"
 
 typedef struct st_table_entry st_table_entry;
@@ -521,6 +522,8 @@ st_foreach(table, func, arg)
     return 0;
 }
 
+static unsigned long hash_seed = 0;
+
 static int
 strhash(string)
     register const char *string;
@@ -550,10 +553,11 @@ strhash(string)
 
     return val + (val << 15);
 #else
-    register int val = 0;
+    register unsigned long val = hash_seed;
 
     while ((c = *string++) != '\0') {
 	val = val*997 + c;
+	val = (val << 13) | (val >> (sizeof(st_data_t) * CHAR_BIT - 13));
     }
 
     return val + (val>>5);
@@ -573,3 +577,11 @@ numhash(n)
 {
     return n;
 }
+
+extern unsigned long rb_genrand_int32(void);
+
+void
+Init_st(void)
+{
+    hash_seed = rb_genrand_int32();
+}
diff --git a/string.c b/string.c
index e85bfb88e0..91e481199d 100644
--- a/string.c
+++ b/string.c
@@ -919,13 +919,15 @@ rb_str_concat(str1, str2)
     return str1;
 }
 
+static unsigned long hash_seed;
+
 int
 rb_str_hash(str)
     VALUE str;
 {
     register long len = RSTRING(str)->len;
     register char *p = RSTRING(str)->ptr;
-    register int key = 0;
+    register unsigned long key = hash_seed;
 
 #if defined(HASH_ELFHASH)
     register unsigned int g;
@@ -949,6 +951,7 @@ rb_str_hash(str)
     while (len--) {
 	key = key*65599 + *p;
 	p++;
+	key = (key << 13) | (key >> ((sizeof(unsigned long) * CHAR_BIT) - 13));
     }
     key = key + (key>>5);
 #endif
@@ -5413,4 +5416,6 @@ Init_String()
     rb_define_method(rb_cSymbol, "downcase", sym_downcase, 0);
     rb_define_method(rb_cSymbol, "capitalize", sym_capitalize, 0);
     rb_define_method(rb_cSymbol, "swapcase", sym_swapcase, 0);
+
+    hash_seed = rb_genrand_int32();
 }
diff --git a/test/ruby/test_string.rb b/test/ruby/test_string.rb
index 073aecd6f8..e952233f5b 100644
--- a/test/ruby/test_string.rb
+++ b/test/ruby/test_string.rb
@@ -1,4 +1,5 @@
 require 'test/unit'
+require File.expand_path('envutil', File.dirname(__FILE__))
 
 class TestString < Test::Unit::TestCase
   def check_sum(str, bits=16)
@@ -259,4 +260,16 @@ class TestString < Test::Unit::TestCase
     "abc".end_with?("c")
     assert_nil($&)
   end
+
+  def test_hash_random
+    str = 'abc'
+    a = [str.hash.to_s]
+    cmd = sprintf("%s -e 'print %s.hash'", EnvUtil.rubybin, str.dump)
+    3.times {
+      IO.popen(cmd, "rb") {|o|
+        a << o.read
+      }
+    }
+    assert_not_equal([str.hash.to_s], a.uniq)
+  end
 end