lib/hash.h

   1 /*
   2  * Copyright (c) 2008, 2009, 2010, 2012, 2013 Nicira, Inc.
   3  *
   4  * Licensed under the Apache License, Version 2.0 (the "License");
   5  * you may not use this file except in compliance with the License.
   6  * You may obtain a copy of the License at:
   7  *
   8  *     http://www.apache.org/licenses/LICENSE-2.0
   9  *
  10  * Unless required by applicable law or agreed to in writing, software
  11  * distributed under the License is distributed on an "AS IS" BASIS,
  12  * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  13  * See the License for the specific language governing permissions and
  14  * limitations under the License.
  15  */
  16 #ifndef HASH_H
  17 #define HASH_H 1
  18
  19 #include <stdbool.h>
  20 #include <stddef.h>
  21 #include <stdint.h>
  22 #include <string.h>
  23 #include "util.h"
  24
  25 #ifdef __cplusplus
  26 extern "C" {
  27 #endif
  28
  29 static inline uint32_t
  30 hash_rot(uint32_t x, int k)
  31 {
  32     return (x << k) | (x >> (32 - k));
  33 }
  34
  35 uint32_t hash_words(const uint32_t data[], size_t n_words, uint32_t basis);
  36 uint32_t hash_bytes(const void *, size_t n_bytes, uint32_t basis);
  37
  38 static inline uint32_t hash_int(uint32_t x, uint32_t basis);
  39 static inline uint32_t hash_2words(uint32_t, uint32_t);
  40 static inline uint32_t hash_uint64(uint64_t);
  41 static inline uint32_t hash_uint64_basis(uint64_t x, uint32_t basis);
  42 uint32_t hash_3words(uint32_t, uint32_t, uint32_t);
  43
  44 static inline uint32_t hash_boolean(bool x, uint32_t basis);
  45 uint32_t hash_double(double, uint32_t basis);
  46
  47 static inline uint32_t hash_pointer(const void *, uint32_t basis);
  48 static inline uint32_t hash_string(const char *, uint32_t basis);
  49
  50 /* Murmurhash by Austin Appleby,
  51  * from http://code.google.com/p/smhasher/source/browse/trunk/MurmurHash3.cpp.
  52  *
  53  * The upstream license there says:
  54  *
  55  * // MurmurHash3 was written by Austin Appleby, and is placed in the public
  56  * // domain. The author hereby disclaims copyright to this source code.
  57  *
  58  * See hash_words() for sample usage. */
  59
  60 static inline uint32_t mhash_add__(uint32_t hash, uint32_t data)
  61 {
  62     data *= 0xcc9e2d51;
  63     data = hash_rot(data, 15);
  64     data *= 0x1b873593;
  65     return hash ^ data;
  66 }
  67
  68 static inline uint32_t mhash_add(uint32_t hash, uint32_t data)
  69 {
  70     hash = mhash_add__(hash, data);
  71     hash = hash_rot(hash, 13);
  72     return hash * 5 + 0xe6546b64;
  73 }
  74
  75 static inline uint32_t mhash_finish(uint32_t hash, size_t n_bytes)
  76 {
  77     hash ^= n_bytes;
  78     hash ^= hash >> 16;
  79     hash *= 0x85ebca6b;
  80     hash ^= hash >> 13;
  81     hash *= 0xc2b2ae35;
  82     hash ^= hash >> 16;
  83     return hash;
  84 }
  85
  86 static inline uint32_t hash_string(const char *s, uint32_t basis)
  87 {
  88     return hash_bytes(s, strlen(s), basis);
  89 }
  90
  91 static inline uint32_t hash_int(uint32_t x, uint32_t basis)
  92 {
  93     return hash_2words(x, basis);
  94 }
  95
  96 /* An attempt at a useful 1-bit hash function.  Has not been analyzed for
  97  * quality. */
  98 static inline uint32_t hash_boolean(bool x, uint32_t basis)
  99 {
 100     const uint32_t P0 = 0xc2b73583;   /* This is hash_int(1, 0). */
 101     const uint32_t P1 = 0xe90f1258;   /* This is hash_int(2, 0). */
 102     return (x ? P0 : P1) ^ hash_rot(basis, 1);
 103 }
 104
 105 static inline uint32_t hash_pointer(const void *p, uint32_t basis)
 106 {
 107     /* Often pointers are hashed simply by casting to integer type, but that
 108      * has pitfalls since the lower bits of a pointer are often all 0 for
 109      * alignment reasons.  It's hard to guess where the entropy really is, so
 110      * we give up here and just use a high-quality hash function.
 111      *
 112      * The double cast suppresses a warning on 64-bit systems about casting to
 113      * an integer to different size.  That's OK in this case, since most of the
 114      * entropy in the pointer is almost certainly in the lower 32 bits. */
 115     return hash_int((uint32_t) (uintptr_t) p, basis);
 116 }
 117
 118 static inline uint32_t hash_2words(uint32_t x, uint32_t y)
 119 {
 120     return mhash_finish(mhash_add(mhash_add(x, 0), y), 8);
 121 }
 122
 123 static inline uint32_t hash_uint64(const uint64_t x)
 124 {
 125     return hash_2words((uint32_t)(x >> 32), (uint32_t)x);
 126 }
 127
 128 static inline uint32_t hash_uint64_basis(const uint64_t x,
 129                                          const uint32_t basis)
 130 {
 131     return hash_3words((uint32_t)(x >> 32), (uint32_t)x, basis);
 132 }
 133 #ifdef __cplusplus
 134 }
 135 #endif
 136
 137 #endif /* hash.h */