Replace hashing function for string keys implementation with xxhash
[ghc.git] / rts / xxhash.h
1 /*
2 xxHash - Extremely Fast Hash algorithm
3 Header File
4 Copyright (C) 2012-2016, Yann Collet.
5
6 BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
7
8 Redistribution and use in source and binary forms, with or without
9 modification, are permitted provided that the following conditions are
10 met:
11
12 * Redistributions of source code must retain the above copyright
13 notice, this list of conditions and the following disclaimer.
14 * Redistributions in binary form must reproduce the above
15 copyright notice, this list of conditions and the following disclaimer
16 in the documentation and/or other materials provided with the
17 distribution.
18
19 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
20 "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
21 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
22 A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
23 OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
24 SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
25 LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
26 DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
27 THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
28 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
29 OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
30
31 You can contact the author at :
32 - xxHash source repository : https://github.com/Cyan4973/xxHash
33 */
34
35 /* Notice extracted from xxHash homepage :
36
37 xxHash is an extremely fast Hash algorithm, running at RAM speed limits.
38 It also successfully passes all tests from the SMHasher suite.
39
40 Comparison (single thread, Windows Seven 32 bits, using SMHasher on a Core 2 Duo @3GHz)
41
42 Name Speed Q.Score Author
43 xxHash 5.4 GB/s 10
44 CrapWow 3.2 GB/s 2 Andrew
45 MumurHash 3a 2.7 GB/s 10 Austin Appleby
46 SpookyHash 2.0 GB/s 10 Bob Jenkins
47 SBox 1.4 GB/s 9 Bret Mulvey
48 Lookup3 1.2 GB/s 9 Bob Jenkins
49 SuperFastHash 1.2 GB/s 1 Paul Hsieh
50 CityHash64 1.05 GB/s 10 Pike & Alakuijala
51 FNV 0.55 GB/s 5 Fowler, Noll, Vo
52 CRC32 0.43 GB/s 9
53 MD5-32 0.33 GB/s 10 Ronald L. Rivest
54 SHA1-32 0.28 GB/s 10
55
56 Q.Score is a measure of quality of the hash function.
57 It depends on successfully passing SMHasher test set.
58 10 is a perfect score.
59
60 A 64-bits version, named XXH64, is available since r35.
61 It offers much better speed, but for 64-bits applications only.
62 Name Speed on 64 bits Speed on 32 bits
63 XXH64 13.8 GB/s 1.9 GB/s
64 XXH32 6.8 GB/s 6.0 GB/s
65 */
66
67 #ifndef XXHASH_H_5627135585666179
68 #define XXHASH_H_5627135585666179 1
69
70 #if defined (__cplusplus)
71 extern "C" {
72 #endif
73
74
75 /* ****************************
76 * Definitions
77 ******************************/
78 #include <stddef.h> /* size_t */
79 typedef enum { XXH_OK=0, XXH_ERROR } XXH_errorcode;
80
81
82 /* ****************************
83 * API modifier
84 ******************************/
85 /** XXH_PRIVATE_API
86 * This is useful to include xxhash functions in `static` mode
87 * in order to inline them, and remove their symbol from the public list.
88 * Methodology :
89 * #define XXH_PRIVATE_API
90 * #include "xxhash.h"
91 * `xxhash.c` is automatically included.
92 * It's not useful to compile and link it as a separate module.
93 */
94 #ifdef XXH_PRIVATE_API
95 # ifndef XXH_STATIC_LINKING_ONLY
96 # define XXH_STATIC_LINKING_ONLY
97 # endif
98 # if defined(__GNUC__)
99 # define XXH_PUBLIC_API static __inline __attribute__((unused))
100 # elif defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */)
101 # define XXH_PUBLIC_API static inline
102 # elif defined(_MSC_VER)
103 # define XXH_PUBLIC_API static __inline
104 # else
105 # define XXH_PUBLIC_API static /* this version may generate warnings for unused static functions; disable the relevant warning */
106 # endif
107 #else
108 # define XXH_PUBLIC_API /* do nothing */
109 #endif /* XXH_PRIVATE_API */
110
111 /*!XXH_NAMESPACE, aka Namespace Emulation :
112
113 If you want to include _and expose_ xxHash functions from within your own library,
114 but also want to avoid symbol collisions with other libraries which may also include xxHash,
115
116 you can use XXH_NAMESPACE, to automatically prefix any public symbol from xxhash library
117 with the value of XXH_NAMESPACE (therefore, avoid NULL and numeric values).
118
119 Note that no change is required within the calling program as long as it includes `xxhash.h` :
120 regular symbol name will be automatically translated by this header.
121 */
122 #ifdef XXH_NAMESPACE
123 # define XXH_CAT(A,B) A##B
124 # define XXH_NAME2(A,B) XXH_CAT(A,B)
125 # define XXH_versionNumber XXH_NAME2(XXH_NAMESPACE, XXH_versionNumber)
126 # define XXH32 XXH_NAME2(XXH_NAMESPACE, XXH32)
127 # define XXH32_createState XXH_NAME2(XXH_NAMESPACE, XXH32_createState)
128 # define XXH32_freeState XXH_NAME2(XXH_NAMESPACE, XXH32_freeState)
129 # define XXH32_reset XXH_NAME2(XXH_NAMESPACE, XXH32_reset)
130 # define XXH32_update XXH_NAME2(XXH_NAMESPACE, XXH32_update)
131 # define XXH32_digest XXH_NAME2(XXH_NAMESPACE, XXH32_digest)
132 # define XXH32_copyState XXH_NAME2(XXH_NAMESPACE, XXH32_copyState)
133 # define XXH32_canonicalFromHash XXH_NAME2(XXH_NAMESPACE, XXH32_canonicalFromHash)
134 # define XXH32_hashFromCanonical XXH_NAME2(XXH_NAMESPACE, XXH32_hashFromCanonical)
135 # define XXH64 XXH_NAME2(XXH_NAMESPACE, XXH64)
136 # define XXH64_createState XXH_NAME2(XXH_NAMESPACE, XXH64_createState)
137 # define XXH64_freeState XXH_NAME2(XXH_NAMESPACE, XXH64_freeState)
138 # define XXH64_reset XXH_NAME2(XXH_NAMESPACE, XXH64_reset)
139 # define XXH64_update XXH_NAME2(XXH_NAMESPACE, XXH64_update)
140 # define XXH64_digest XXH_NAME2(XXH_NAMESPACE, XXH64_digest)
141 # define XXH64_copyState XXH_NAME2(XXH_NAMESPACE, XXH64_copyState)
142 # define XXH64_canonicalFromHash XXH_NAME2(XXH_NAMESPACE, XXH64_canonicalFromHash)
143 # define XXH64_hashFromCanonical XXH_NAME2(XXH_NAMESPACE, XXH64_hashFromCanonical)
144 #endif
145
146
147 /* *************************************
148 * Version
149 ***************************************/
150 #define XXH_VERSION_MAJOR 0
151 #define XXH_VERSION_MINOR 6
152 #define XXH_VERSION_RELEASE 2
153 #define XXH_VERSION_NUMBER (XXH_VERSION_MAJOR *100*100 + XXH_VERSION_MINOR *100 + XXH_VERSION_RELEASE)
154 XXH_PUBLIC_API unsigned XXH_versionNumber (void);
155
156
157 /*-**********************************************************************
158 * 32-bits hash
159 ************************************************************************/
160 typedef unsigned int XXH32_hash_t;
161
162 /*! XXH32() :
163 Calculate the 32-bits hash of sequence "length" bytes stored at memory address "input".
164 The memory between input & input+length must be valid (allocated and read-accessible).
165 "seed" can be used to alter the result predictably.
166 Speed on Core 2 Duo @ 3 GHz (single thread, SMHasher benchmark) : 5.4 GB/s */
167 XXH_PUBLIC_API XXH32_hash_t XXH32 (const void* input, size_t length, unsigned int seed);
168
169 /*====== Streaming ======*/
170 typedef struct XXH32_state_s XXH32_state_t; /* incomplete type */
171 XXH_PUBLIC_API XXH32_state_t* XXH32_createState(void);
172 XXH_PUBLIC_API XXH_errorcode XXH32_freeState(XXH32_state_t* statePtr);
173 XXH_PUBLIC_API void XXH32_copyState(XXH32_state_t* dst_state, const XXH32_state_t* src_state);
174
175 XXH_PUBLIC_API XXH_errorcode XXH32_reset (XXH32_state_t* statePtr, unsigned int seed);
176 XXH_PUBLIC_API XXH_errorcode XXH32_update (XXH32_state_t* statePtr, const void* input, size_t length);
177 XXH_PUBLIC_API XXH32_hash_t XXH32_digest (const XXH32_state_t* statePtr);
178
179 /*
180 These functions generate the xxHash of an input provided in multiple segments.
181 Note that, for small input, they are slower than single-call functions, due to state management.
182 For small input, prefer `XXH32()` and `XXH64()` .
183
184 XXH state must first be allocated, using XXH*_createState() .
185
186 Start a new hash by initializing state with a seed, using XXH*_reset().
187
188 Then, feed the hash state by calling XXH*_update() as many times as necessary.
189 Obviously, input must be allocated and read accessible.
190 The function returns an error code, with 0 meaning OK, and any other value meaning there is an error.
191
192 Finally, a hash value can be produced anytime, by using XXH*_digest().
193 This function returns the nn-bits hash as an int or long long.
194
195 It's still possible to continue inserting input into the hash state after a digest,
196 and generate some new hashes later on, by calling again XXH*_digest().
197
198 When done, free XXH state space if it was allocated dynamically.
199 */
200
201 /*====== Canonical representation ======*/
202
203 typedef struct { unsigned char digest[4]; } XXH32_canonical_t;
204 XXH_PUBLIC_API void XXH32_canonicalFromHash(XXH32_canonical_t* dst, XXH32_hash_t hash);
205 XXH_PUBLIC_API XXH32_hash_t XXH32_hashFromCanonical(const XXH32_canonical_t* src);
206
207 /* Default result type for XXH functions are primitive unsigned 32 and 64 bits.
208 * The canonical representation uses human-readable write convention, aka big-endian (large digits first).
209 * These functions allow transformation of hash result into and from its canonical format.
210 * This way, hash values can be written into a file / memory, and remain comparable on different systems and programs.
211 */
212
213
214 #ifndef XXH_NO_LONG_LONG
215 /*-**********************************************************************
216 * 64-bits hash
217 ************************************************************************/
218 typedef unsigned long long XXH64_hash_t;
219
220 /*! XXH64() :
221 Calculate the 64-bits hash of sequence of length "len" stored at memory address "input".
222 "seed" can be used to alter the result predictably.
223 This function runs faster on 64-bits systems, but slower on 32-bits systems (see benchmark).
224 */
225 XXH_PUBLIC_API XXH64_hash_t XXH64 (const void* input, size_t length, unsigned long long seed);
226
227 /*====== Streaming ======*/
228 typedef struct XXH64_state_s XXH64_state_t; /* incomplete type */
229 XXH_PUBLIC_API XXH64_state_t* XXH64_createState(void);
230 XXH_PUBLIC_API XXH_errorcode XXH64_freeState(XXH64_state_t* statePtr);
231 XXH_PUBLIC_API void XXH64_copyState(XXH64_state_t* dst_state, const XXH64_state_t* src_state);
232
233 XXH_PUBLIC_API XXH_errorcode XXH64_reset (XXH64_state_t* statePtr, unsigned long long seed);
234 XXH_PUBLIC_API XXH_errorcode XXH64_update (XXH64_state_t* statePtr, const void* input, size_t length);
235 XXH_PUBLIC_API XXH64_hash_t XXH64_digest (const XXH64_state_t* statePtr);
236
237 /*====== Canonical representation ======*/
238 typedef struct { unsigned char digest[8]; } XXH64_canonical_t;
239 XXH_PUBLIC_API void XXH64_canonicalFromHash(XXH64_canonical_t* dst, XXH64_hash_t hash);
240 XXH_PUBLIC_API XXH64_hash_t XXH64_hashFromCanonical(const XXH64_canonical_t* src);
241 #endif /* XXH_NO_LONG_LONG */
242
243
244 #ifdef XXH_STATIC_LINKING_ONLY
245
246 /* ================================================================================================
247 This section contains definitions which are not guaranteed to remain stable.
248 They may change in future versions, becoming incompatible with a different version of the library.
249 They shall only be used with static linking.
250 Never use these definitions in association with dynamic linking !
251 =================================================================================================== */
252
253 /* These definitions are only meant to make possible
254 static allocation of XXH state, on stack or in a struct for example.
255 Never use members directly. */
256
257 struct XXH32_state_s {
258 unsigned total_len_32;
259 unsigned large_len;
260 unsigned v1;
261 unsigned v2;
262 unsigned v3;
263 unsigned v4;
264 unsigned mem32[4]; /* buffer defined as U32 for alignment */
265 unsigned memsize;
266 unsigned reserved; /* never read nor write, will be removed in a future version */
267 }; /* typedef'd to XXH32_state_t */
268
269 #ifndef XXH_NO_LONG_LONG /* remove 64-bits support */
270 struct XXH64_state_s {
271 unsigned long long total_len;
272 unsigned long long v1;
273 unsigned long long v2;
274 unsigned long long v3;
275 unsigned long long v4;
276 unsigned long long mem64[4]; /* buffer defined as U64 for alignment */
277 unsigned memsize;
278 unsigned reserved[2]; /* never read nor write, will be removed in a future version */
279 }; /* typedef'd to XXH64_state_t */
280 #endif
281
282 #ifdef XXH_PRIVATE_API
283 # include "xxhash.c" /* include xxhash function bodies as `static`, for inlining */
284 #endif
285
286 #endif /* XXH_STATIC_LINKING_ONLY */
287
288
289 #if defined (__cplusplus)
290 }
291 #endif
292
293 #endif /* XXHASH_H_5627135585666179 */