Hash map implementation for wmem.

This has two expected uses:
- Many current users of wmem_tree don't actually need the predecessor lookup
  it provides (the lookup_le function family). A hash map provides straight
  insertion and lookup much more efficiently than a wmem_tree when predecessor
  lookup isn't needed.

- Many current users of glib's hash table and hash functions use untrusted data
  for keys, making them vulnerable to algorithmic complexity attacks. Care has
  been taken to make this implementation secure against such attacks, so it
  should be used whenever data is untrusted.

In my benchmarks it is measurably slower than GHashTable, but not excessively
so. Given the additional security it provides this seems like a reasonable
trade-off (and it is still faster than a wmem_tree).

Change-Id: I2d67a0d06029f14c153eaa42d5cfc774aefd9918
Reviewed-on: https://code.wireshark.org/review/1272
Reviewed-by: Evan Huus <eapache@gmail.com>

1 files changed, 284 insertions, 0 deletions

diff --git a/epan/wmem/wmem_map.c b/epan/wmem/wmem_map.c
new file mode 100644
index 0000000000..3413ca8bf0
--- /dev/null
+++ b/epan/wmem/wmem_map.c
@@ -0,0 +1,284 @@
+/* wmem_map.c
+ * Wireshark Memory Manager Hash Map
+ * Copyright 2014, Evan Huus <eapache@gmail.com>
+ *
+ * Wireshark - Network traffic analyzer
+ * By Gerald Combs <gerald@wireshark.org>
+ * Copyright 1998 Gerald Combs
+ *
+ * This program 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 2 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, write to the Free Software Foundation, Inc.,
+ * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
+ */
+
+#include <glib.h>
+
+#include "wmem_core.h"
+#include "wmem_map.h"
+
+static guint32 x; /* Used for universal integer hashing (see the HASH macro) */
+
+/* Used for the wmem_strong_hash() function */
+static guint32 preseed;
+static guint32 postseed;
+
+void
+wmem_init_hashing(void)
+{
+    x = g_random_int();
+
+    preseed  = g_random_int();
+    postseed = g_random_int();
+}
+
+typedef struct _wmem_map_item_t {
+    const void *key;
+    void *value;
+    struct _wmem_map_item_t *next;
+} wmem_map_item_t;
+
+struct _wmem_map_t {
+    guint count; /* number of items stored */
+
+    /* The base-2 logarithm of the actual size of the table. We store this
+     * value for efficiency in hashing, since finding the actual capacity
+     * becomes just a left-shift (see the CAPACITY macro) whereas taking
+     * logarithms is expensive. */
+    guint capacity;
+
+    wmem_map_item_t **table;
+
+    GHashFunc  hash_func;
+    GEqualFunc eql_func;
+
+    wmem_allocator_t *allocator;
+};
+
+/* As per the comment on the 'capacity' member of the wmem_map_t struct, this is
+ * the base-2 logarithm, meaning the actual default capacity is 2^5 = 32 */
+#define WMEM_MAP_DEFAULT_CAPACITY 5
+
+/* Macro for calculating the real capacity of the map by using a left-shift to
+ * do the 2^x operation. */
+#define CAPACITY(MAP) ((guint)(1 << (MAP)->capacity))
+
+/* Efficient universal integer hashing:
+ * https://en.wikipedia.org/wiki/Universal_hashing#Avoiding_modular_arithmetic
+ */
+#define HASH(MAP, KEY) \
+    ((guint32)(((MAP)->hash_func(KEY) * x) >> (32 - (MAP)->capacity)))
+
+wmem_map_t *
+wmem_map_new(wmem_allocator_t *allocator,
+        GHashFunc hash_func, GEqualFunc eql_func)
+{
+    wmem_map_t *map;
+
+    map = wmem_new(allocator, wmem_map_t);
+
+    map->count     = 0;
+    map->capacity  = WMEM_MAP_DEFAULT_CAPACITY;
+    map->table     = wmem_alloc0_array(allocator, wmem_map_item_t*, CAPACITY(map));
+    map->hash_func = hash_func;
+    map->eql_func  = eql_func;
+    map->allocator = allocator;
+
+    return map;
+}
+
+static inline void
+wmem_map_grow(wmem_map_t *map)
+{
+    wmem_map_item_t **old_table, *cur, *nxt;
+    guint             old_cap, i, slot;
+
+    /* store the old table and capacity */
+    old_table = map->table;
+    old_cap   = CAPACITY(map);
+
+    /* double the size (capacity is base-2 logarithm, so this just means
+     * increment it) and allocate new table */
+    map->capacity++;
+    map->table = wmem_alloc0_array(map->allocator, wmem_map_item_t*, CAPACITY(map));
+
+    /* copy all the elements over from the old table */
+    for (i=0; i<old_cap; i++) {
+        cur = old_table[i];
+        while (cur) {
+            nxt              = cur->next;
+            slot             = HASH(map, cur->key);
+            cur->next        = map->table[slot];
+            map->table[slot] = cur;
+            cur              = nxt;
+        }
+    }
+
+    /* free the old table */
+    wmem_free(map->allocator, old_table);
+}
+
+void *
+wmem_map_insert(wmem_map_t *map, const void *key, void *value)
+{
+    wmem_map_item_t **item;
+    void *old_val;
+
+    /* get a pointer to the slot */
+    item = &(map->table[HASH(map, key)]);
+
+    /* check existing items in that slot */
+    while (*item) {
+        if (map->eql_func(key, (*item)->key)) {
+            /* replace and return old value for this key */
+            old_val = (*item)->value;
+            (*item)->value = value;
+            return old_val;
+        }
+        item = &((*item)->next);
+    }
+
+    /* insert new item */
+    (*item) = wmem_new(map->allocator, wmem_map_item_t);
+
+    (*item)->key   = key;
+    (*item)->value = value;
+    (*item)->next  = NULL;
+
+    map->count++;
+
+    /* increase size if we are over-full */
+    if (map->count >= CAPACITY(map)) {
+        wmem_map_grow(map);
+    }
+
+    /* no previous entry, return NULL */
+    return NULL;
+}
+
+void *
+wmem_map_lookup(wmem_map_t *map, const void *key)
+{
+    wmem_map_item_t *item;
+
+    /* find correct slot */
+    item = map->table[HASH(map, key)];
+
+    /* scan list of items in this slot for the correct value */
+    while (item) {
+        if (map->eql_func(key, item->key)) {
+            return item->value;
+        }
+        item = item->next;
+    }
+
+    return NULL;
+}
+
+void *
+wmem_map_remove(wmem_map_t *map, const void *key)
+{
+    wmem_map_item_t **item, *tmp;
+    void *value;
+
+    /* get a pointer to the slot */
+    item = &(map->table[HASH(map, key)]);
+
+    /* check the items in that slot */
+    while (*item) {
+        if (map->eql_func(key, (*item)->key)) {
+            /* found it */
+            tmp     = (*item);
+            value   = tmp->value;
+            (*item) = tmp->next;
+            wmem_free(map->allocator, tmp);
+            map->count--;
+            return value;
+        }
+        item = &((*item)->next);
+    }
+
+    /* didn't find it */
+    return NULL;
+}
+
+/* Borrowed from Perl 5.18. This is based on Bob Jenkin's one-at-a-time
+ * algorithm with some additional randomness seeded in. It is believed to be
+ * generally secure against collision attacks. See
+ * http://blog.booking.com/hardening-perls-hash-function.html
+ */
+guint32
+wmem_strong_hash(const guint8 *buf, const size_t len)
+{
+    const guint8 * const end = (const guint8 *)buf + len;
+    guint32 hash = preseed + len;
+
+    while (buf < end) {
+        hash += (hash << 10);
+        hash ^= (hash >> 6);
+        hash += *buf++;
+    }
+
+    hash += (hash << 10);
+    hash ^= (hash >> 6);
+    hash += ((guint8*)&postseed)[0];
+
+    hash += (hash << 10);
+    hash ^= (hash >> 6);
+    hash += ((guint8*)&postseed)[1];
+
+    hash += (hash << 10);
+    hash ^= (hash >> 6);
+    hash += ((guint8*)&postseed)[2];
+
+    hash += (hash << 10);
+    hash ^= (hash >> 6);
+    hash += ((guint8*)&postseed)[3];
+
+    hash += (hash << 10);
+    hash ^= (hash >> 6);
+
+    hash += (hash << 3);
+    hash ^= (hash >> 11);
+    return (hash + (hash << 15));
+}
+
+guint
+wmem_str_hash(gconstpointer key)
+{
+    return wmem_strong_hash((const guint8 *)key, strlen((const char *)key));
+}
+
+guint
+wmem_int64_hash(gconstpointer key)
+{
+    return wmem_strong_hash((const guint8 *)key, sizeof(guint64));
+}
+
+guint
+wmem_double_hash(gconstpointer key)
+{
+    return wmem_strong_hash((const guint8 *)key, sizeof(double));
+}
+
+/*
+ * Editor modelines  -  http://www.wireshark.org/tools/modelines.html
+ *
+ * Local variables:
+ * c-basic-offset: 4
+ * tab-width: 8
+ * indent-tabs-mode: nil
+ * End:
+ *
+ * vi: set shiftwidth=4 tabstop=8 expandtab:
+ * :indentSize=4:tabSize=8:noTabs=true:
+ */