Re: Hash collision vectors in APR?

[Bojan Smojver]

On Fri, 2012-01-13 at 19:57 +1100, Bojan Smojver wrote:
> This probably wouldn't work as expected. If you had /dev/random
> selected for DEV_RANDOM, the if would fail and srand()/rand() bit
> would never get called. No?

Like attached maybe?

-- 
Bojan

Index: tables/apr_hash.c
===================================================================
--- tables/apr_hash.c	(revision 1230868)
+++ tables/apr_hash.c	(working copy)
@@ -18,6 +18,10 @@
 
 #include "apr_general.h"
 #include "apr_pools.h"
+#include "apr_time.h"
+#if APR_HAVE_STDLIB_H
+#include <stdlib.h>     /* for rand, srand */
+#endif
 
 #include "apr_hash.h"
 
@@ -75,7 +79,7 @@
     apr_pool_t          *pool;
     apr_hash_entry_t   **array;
     apr_hash_index_t     iterator;  /* For apr_hash_first(NULL, ...) */
-    unsigned int         count, max;
+    unsigned int         count, max, seed;
     apr_hashfunc_t       hash_func;
     apr_hash_entry_t    *free;  /* List of recycled entries */
 };
@@ -95,13 +99,27 @@
 APR_DECLARE(apr_hash_t *) apr_hash_make(apr_pool_t *pool)
 {
     apr_hash_t *ht;
+    apr_time_t now;
+
     ht = apr_palloc(pool, sizeof(apr_hash_t));
     ht->pool = pool;
     ht->free = NULL;
     ht->count = 0;
     ht->max = INITIAL_MAX;
+#if APR_HAS_RANDOM && defined(DEV_RANDOM)
+    if (!strcmp(DEV_RANDOM, "/dev/random") ||
+        apr_generate_random_bytes(
+            (unsigned char *)&ht->seed, sizeof(ht->seed)) != APR_SUCCESS) {
+#endif
+        now = apr_time_now();
+        srand((unsigned int)((now >> 32) ^ now ^ (apr_uintptr_t)ht));
+        ht->seed = (unsigned int)(rand());
+#if APR_HAS_RANDOM && defined(DEV_RANDOM)
+    }
+#endif
     ht->array = alloc_array(ht, ht->max);
-    ht->hash_func = apr_hashfunc_default;
+    ht->hash_func = NULL;
+
     return ht;
 }
 
@@ -201,10 +219,10 @@
     ht->max = new_max;
 }
 
-APR_DECLARE_NONSTD(unsigned int) apr_hashfunc_default(const char *char_key,
-                                                      apr_ssize_t *klen)
+static unsigned int apr_hashfunc_default_internal(const char *char_key,
+                                                  apr_ssize_t *klen,
+                                                  unsigned int hash)
 {
-    unsigned int hash = 0;
     const unsigned char *key = (const unsigned char *)char_key;
     const unsigned char *p;
     apr_ssize_t i;
@@ -246,7 +264,7 @@
      *
      *                  -- Ralf S. Engelschall <r...@engelschall.com>
      */
-     
+
     if (*klen == APR_HASH_KEY_STRING) {
         for (p = key; *p; p++) {
             hash = hash * 33 + *p;
@@ -262,6 +280,11 @@
     return hash;
 }
 
+APR_DECLARE_NONSTD(unsigned int) apr_hashfunc_default(const char *char_key,
+                                                      apr_ssize_t *klen)
+{
+    return apr_hashfunc_default_internal(char_key, klen, 0);
+}
 
 /*
  * This is where we keep the details of the hash function and control
@@ -280,7 +303,10 @@
     apr_hash_entry_t **hep, *he;
     unsigned int hash;
 
-    hash = ht->hash_func(key, &klen);
+    if (ht->hash_func)
+        hash = ht->hash_func(key, &klen);
+    else
+        hash = apr_hashfunc_default_internal(key, &klen, ht->seed);
 
     /* scan linked list */
     for (hep = &ht->array[hash & ht->max], he = *hep;
@@ -322,6 +348,7 @@
     ht->free = NULL;
     ht->count = orig->count;
     ht->max = orig->max;
+    ht->seed = orig->seed;
     ht->hash_func = orig->hash_func;
     ht->array = (apr_hash_entry_t **)((char *)ht + sizeof(apr_hash_t));