Restore "Add new hashing and "hash with state" infrastructure"

This reverts commit e6a172f358c0f48c4b744dbd5e9ef6ff0b4ff289,
which was a revert of a3bf60fbb1f05cd2c69d4ff0a2ef99537afdaba7.

Add new hashing and "hash with state" infrastructure

This adds support for three new hash functions: StadtX, Zaphod32 and SBOX,
and reworks some of our hash internals infrastructure to do so.

SBOX is special in that it is designed to be used in conjuction with any
other hash function for hashing short strings very efficiently and very
securely. It features compile time options on how much memory and startup
time are traded off to control the length of keys that SBOX hashes.

This also adds support for caching the hash values of single byte characters
which can be used in conjuction with any other hash, including SBOX, although
SBOX itself is as fast as the lookup cache, so typically you wouldnt use both
at the same time.

This also *removes* support for Jenkins One-At-A-Time. It has served us
well, but it's day is done.

This patch adds three new files: zaphod32_hash.h, stadtx_hash.h,
sbox32_hash.h

1 files changed, 38 insertions, 11 deletions

diff --git a/perl.c b/perl.c
index 658f260997..125afaed51 100644
--- a/perl.c
+++ b/perl.c
@@ -308,27 +308,54 @@ perl_construct(pTHXx)
 #ifdef USE_REENTRANT_API
     Perl_reentrant_init(aTHX);
 #endif
-#if defined(USE_HASH_SEED) || defined(USE_HASH_SEED_EXPLICIT)
-        /* [perl #22371] Algorimic Complexity Attack on Perl 5.6.1, 5.8.0
-         * This MUST be done before any hash stores or fetches take place.
-         * If you set PL_hash_seed (and presumably also PL_hash_seed_set)
-         * yourself, it is your responsibility to provide a good random seed!
-         * You can also define PERL_HASH_SEED in compile time, see hv.h.
-         *
-         * XXX: fix this comment */
     if (PL_hash_seed_set == FALSE) {
+        /* Initialize the hash seed and state at startup. This must be
+         * done very early, before ANY hashes are constructed, and once
+         * setup is fixed for the lifetime of the process.
+         *
+         * If you decide to disable the seeding process you should choose
+         * a suitable seed yourself and define PERL_HASH_SEED to a well chosen
+         * string. See hv_func.h for details.
+         */
+#if defined(USE_HASH_SEED) || defined(USE_HASH_SEED_EXPLICIT)
+        /* get the hash seed from the environment or from an RNG */
         Perl_get_hash_seed(aTHX_ PL_hash_seed);
-        PL_hash_seed_set= TRUE;
-    }
+#else /*if !(defined(USE_HASH_SEED) || defined(USE_HASH_SEED_EXPLICIT)) */
+        /* they want a hard coded seed, check that it is long enough */
+        assert( strlen(PERL_HASH_SEED) >= PERL_HASH_SEED_BYTES );
 #endif /* #if defined(USE_HASH_SEED) || defined(USE_HASH_SEED_EXPLICIT) */
 
+        /* now we use the chosen seed to initialize the state -
+         * in some configurations this may be a relatively speaking
+         * expensive operation, but we only have to do it once at startup */
+        PERL_HASH_SEED_STATE(PERL_HASH_SEED,PL_hash_state);
+
+#ifdef PERL_USE_SINGLE_CHAR_HASH_CACHE
+        /* we can build a special cache for 0/1 byte keys, if people choose
+         * I suspect most of the time it is not worth it */
+        {
+            char str[2]="\0";
+            int i;
+            for (i=0;i<256;i++) {
+                str[0]= i;
+                PERL_HASH_WITH_STATE(PL_hash_state,PL_hash_chars[i],str,1);
+            }
+            PERL_HASH_WITH_STATE(PL_hash_state,PL_hash_chars[256],str,0);
+        }
+#endif
+        /* at this point we have initialezed the hash function, and we can start
+         * constructing hashes */
+        PL_hash_seed_set= TRUE;
+    }
     /* Note that strtab is a rather special HV.  Assumptions are made
        about not iterating on it, and not adding tie magic to it.
        It is properly deallocated in perl_destruct() */
     PL_strtab = newHV();
 
+    /* SHAREKEYS tells us that the hash has its keys shared with PL_strtab,
+     * which is not the case with PL_strtab itself */
     HvSHAREKEYS_off(PL_strtab);			/* mandatory */
-    hv_ksplit(PL_strtab, 512);
+    hv_ksplit(PL_strtab, 1 << 11);
 
     Zero(PL_sv_consts, SV_CONSTS_COUNT, SV*);