xref: /freebsd/contrib/llvm-project/llvm/lib/Support/xxhash.cpp (revision 3c4ba5f55438f7afd4f4b0b56f88f2bb505fd6a6)
1 /*
2 *  xxHash - Fast Hash algorithm
3 *  Copyright (C) 2012-2016, Yann Collet
4 *
5 *  BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
6 *
7 *  Redistribution and use in source and binary forms, with or without
8 *  modification, are permitted provided that the following conditions are
9 *  met:
10 *
11 *  * Redistributions of source code must retain the above copyright
12 *  notice, this list of conditions and the following disclaimer.
13 *  * Redistributions in binary form must reproduce the above
14 *  copyright notice, this list of conditions and the following disclaimer
15 *  in the documentation and/or other materials provided with the
16 *  distribution.
17 *
18 *  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
19 *  "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
20 *  LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
21 *  A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
22 *  OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
23 *  SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
24 *  LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
25 *  DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
26 *  THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
27 *  (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
28 *  OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
29 *
30 *  You can contact the author at :
31 *  - xxHash homepage: http://www.xxhash.com
32 *  - xxHash source repository : https://github.com/Cyan4973/xxHash
33 */
34 
35 /* based on revision d2df04efcbef7d7f6886d345861e5dfda4edacc1 Removed
36  * everything but a simple interface for computing XXh64. */
37 
38 #include "llvm/Support/xxhash.h"
39 #include "llvm/Support/Endian.h"
40 
41 #include <stdlib.h>
42 
43 using namespace llvm;
44 using namespace support;
45 
46 static uint64_t rotl64(uint64_t X, size_t R) {
47   return (X << R) | (X >> (64 - R));
48 }
49 
50 static const uint64_t PRIME64_1 = 11400714785074694791ULL;
51 static const uint64_t PRIME64_2 = 14029467366897019727ULL;
52 static const uint64_t PRIME64_3 = 1609587929392839161ULL;
53 static const uint64_t PRIME64_4 = 9650029242287828579ULL;
54 static const uint64_t PRIME64_5 = 2870177450012600261ULL;
55 
56 static uint64_t round(uint64_t Acc, uint64_t Input) {
57   Acc += Input * PRIME64_2;
58   Acc = rotl64(Acc, 31);
59   Acc *= PRIME64_1;
60   return Acc;
61 }
62 
63 static uint64_t mergeRound(uint64_t Acc, uint64_t Val) {
64   Val = round(0, Val);
65   Acc ^= Val;
66   Acc = Acc * PRIME64_1 + PRIME64_4;
67   return Acc;
68 }
69 
70 uint64_t llvm::xxHash64(StringRef Data) {
71   size_t Len = Data.size();
72   uint64_t Seed = 0;
73   const unsigned char *P = Data.bytes_begin();
74   const unsigned char *const BEnd = Data.bytes_end();
75   uint64_t H64;
76 
77   if (Len >= 32) {
78     const unsigned char *const Limit = BEnd - 32;
79     uint64_t V1 = Seed + PRIME64_1 + PRIME64_2;
80     uint64_t V2 = Seed + PRIME64_2;
81     uint64_t V3 = Seed + 0;
82     uint64_t V4 = Seed - PRIME64_1;
83 
84     do {
85       V1 = round(V1, endian::read64le(P));
86       P += 8;
87       V2 = round(V2, endian::read64le(P));
88       P += 8;
89       V3 = round(V3, endian::read64le(P));
90       P += 8;
91       V4 = round(V4, endian::read64le(P));
92       P += 8;
93     } while (P <= Limit);
94 
95     H64 = rotl64(V1, 1) + rotl64(V2, 7) + rotl64(V3, 12) + rotl64(V4, 18);
96     H64 = mergeRound(H64, V1);
97     H64 = mergeRound(H64, V2);
98     H64 = mergeRound(H64, V3);
99     H64 = mergeRound(H64, V4);
100 
101   } else {
102     H64 = Seed + PRIME64_5;
103   }
104 
105   H64 += (uint64_t)Len;
106 
107   while (P + 8 <= BEnd) {
108     uint64_t const K1 = round(0, endian::read64le(P));
109     H64 ^= K1;
110     H64 = rotl64(H64, 27) * PRIME64_1 + PRIME64_4;
111     P += 8;
112   }
113 
114   if (P + 4 <= BEnd) {
115     H64 ^= (uint64_t)(endian::read32le(P)) * PRIME64_1;
116     H64 = rotl64(H64, 23) * PRIME64_2 + PRIME64_3;
117     P += 4;
118   }
119 
120   while (P < BEnd) {
121     H64 ^= (*P) * PRIME64_5;
122     H64 = rotl64(H64, 11) * PRIME64_1;
123     P++;
124   }
125 
126   H64 ^= H64 >> 33;
127   H64 *= PRIME64_2;
128   H64 ^= H64 >> 29;
129   H64 *= PRIME64_3;
130   H64 ^= H64 >> 32;
131 
132   return H64;
133 }
134 
135 uint64_t llvm::xxHash64(ArrayRef<uint8_t> Data) {
136   return xxHash64({(const char *)Data.data(), Data.size()});
137 }
138