summaryrefslogtreecommitdiffstats
path: root/libmariadb/external/zlib/crc32_simd.c
blob: cc68d4fc3f2606e5ee0e5481ac74be5f3450fb35 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
/* crc32_simd.c
 *
 * Copyright 2017 The Chromium Authors. All rights reserved.
 * Use of this source code is governed by a BSD-style license that can be
 * found in the Chromium source repository LICENSE file.
 */


#include "crc32_simd.h"

#if defined(CRC32_SIMD_SSE42_PCLMUL)

/*
 * crc32_sse42_simd_(): compute the crc32 of the buffer, where the buffer
 * length must be at least 64, and a multiple of 16. Based on:
 *
 * "Fast CRC Computation for Generic Polynomials Using PCLMULQDQ Instruction"
 *  V. Gopal, E. Ozturk, et al., 2009, http://intel.ly/2ySEwL0
 */

#include <emmintrin.h>
#include <smmintrin.h>
#include <wmmintrin.h>

uint32_t ZLIB_INTERNAL crc32_sse42_simd_(  /* SSE4.2+PCLMUL */
    const unsigned char *buf,
    z_size_t len,
    uint32_t crc)
{
    /*
     * Definitions of the bit-reflected domain constants k1,k2,k3, etc and
     * the CRC32+Barrett polynomials given at the end of the paper.
     */
    static const uint64_t zalign(16) k1k2[] = { 0x0154442bd4, 0x01c6e41596 };
    static const uint64_t zalign(16) k3k4[] = { 0x01751997d0, 0x00ccaa009e };
    static const uint64_t zalign(16) k5k0[] = { 0x0163cd6124, 0x0000000000 };
    static const uint64_t zalign(16) poly[] = { 0x01db710641, 0x01f7011641 };

    __m128i x0, x1, x2, x3, x4, x5, x6, x7, x8, y5, y6, y7, y8;

    /*
     * There's at least one block of 64.
     */
    x1 = _mm_loadu_si128((__m128i *)(buf + 0x00));
    x2 = _mm_loadu_si128((__m128i *)(buf + 0x10));
    x3 = _mm_loadu_si128((__m128i *)(buf + 0x20));
    x4 = _mm_loadu_si128((__m128i *)(buf + 0x30));

    x1 = _mm_xor_si128(x1, _mm_cvtsi32_si128(crc));

    x0 = _mm_load_si128((__m128i *)k1k2);

    buf += 64;
    len -= 64;

    /*
     * Parallel fold blocks of 64, if any.
     */
    while (len >= 64)
    {
        x5 = _mm_clmulepi64_si128(x1, x0, 0x00);
        x6 = _mm_clmulepi64_si128(x2, x0, 0x00);
        x7 = _mm_clmulepi64_si128(x3, x0, 0x00);
        x8 = _mm_clmulepi64_si128(x4, x0, 0x00);

        x1 = _mm_clmulepi64_si128(x1, x0, 0x11);
        x2 = _mm_clmulepi64_si128(x2, x0, 0x11);
        x3 = _mm_clmulepi64_si128(x3, x0, 0x11);
        x4 = _mm_clmulepi64_si128(x4, x0, 0x11);

        y5 = _mm_loadu_si128((__m128i *)(buf + 0x00));
        y6 = _mm_loadu_si128((__m128i *)(buf + 0x10));
        y7 = _mm_loadu_si128((__m128i *)(buf + 0x20));
        y8 = _mm_loadu_si128((__m128i *)(buf + 0x30));

        x1 = _mm_xor_si128(x1, x5);
        x2 = _mm_xor_si128(x2, x6);
        x3 = _mm_xor_si128(x3, x7);
        x4 = _mm_xor_si128(x4, x8);

        x1 = _mm_xor_si128(x1, y5);
        x2 = _mm_xor_si128(x2, y6);
        x3 = _mm_xor_si128(x3, y7);
        x4 = _mm_xor_si128(x4, y8);

        buf += 64;
        len -= 64;
    }

    /*
     * Fold into 128-bits.
     */
    x0 = _mm_load_si128((__m128i *)k3k4);

    x5 = _mm_clmulepi64_si128(x1, x0, 0x00);
    x1 = _mm_clmulepi64_si128(x1, x0, 0x11);
    x1 = _mm_xor_si128(x1, x2);
    x1 = _mm_xor_si128(x1, x5);

    x5 = _mm_clmulepi64_si128(x1, x0, 0x00);
    x1 = _mm_clmulepi64_si128(x1, x0, 0x11);
    x1 = _mm_xor_si128(x1, x3);
    x1 = _mm_xor_si128(x1, x5);

    x5 = _mm_clmulepi64_si128(x1, x0, 0x00);
    x1 = _mm_clmulepi64_si128(x1, x0, 0x11);
    x1 = _mm_xor_si128(x1, x4);
    x1 = _mm_xor_si128(x1, x5);

    /*
     * Single fold blocks of 16, if any.
     */
    while (len >= 16)
    {
        x2 = _mm_loadu_si128((__m128i *)buf);

        x5 = _mm_clmulepi64_si128(x1, x0, 0x00);
        x1 = _mm_clmulepi64_si128(x1, x0, 0x11);
        x1 = _mm_xor_si128(x1, x2);
        x1 = _mm_xor_si128(x1, x5);

        buf += 16;
        len -= 16;
    }

    /*
     * Fold 128-bits to 64-bits.
     */
    x2 = _mm_clmulepi64_si128(x1, x0, 0x10);
    x3 = _mm_setr_epi32(~0, 0, ~0, 0);
    x1 = _mm_srli_si128(x1, 8);
    x1 = _mm_xor_si128(x1, x2);

    x0 = _mm_loadl_epi64((__m128i*)k5k0);

    x2 = _mm_srli_si128(x1, 4);
    x1 = _mm_and_si128(x1, x3);
    x1 = _mm_clmulepi64_si128(x1, x0, 0x00);
    x1 = _mm_xor_si128(x1, x2);

    /*
     * Barret reduce to 32-bits.
     */
    x0 = _mm_load_si128((__m128i*)poly);

    x2 = _mm_and_si128(x1, x3);
    x2 = _mm_clmulepi64_si128(x2, x0, 0x10);
    x2 = _mm_and_si128(x2, x3);
    x2 = _mm_clmulepi64_si128(x2, x0, 0x00);
    x1 = _mm_xor_si128(x1, x2);

    /*
     * Return the crc32.
     */
    return _mm_extract_epi32(x1, 1);
}

#elif defined(CRC32_ARMV8_CRC32)

/* CRC32 checksums using ARMv8-a crypto instructions.
 *
 * TODO: implement a version using the PMULL instruction.
 */

#if defined(__clang__)
/* CRC32 intrinsics are #ifdef'ed out of arm_acle.h unless we build with an
 * armv8 target, which is incompatible with ThinLTO optimizations on Android.
 * (Namely, mixing and matching different module-level targets makes ThinLTO
 * warn, and Android defaults to armv7-a. This restriction does not apply to
 * function-level `target`s, however.)
 *
 * Since we only need four crc intrinsics, and since clang's implementation of
 * those are just wrappers around compiler builtins, it's simplest to #define
 * those builtins directly. If this #define list grows too much (or we depend on
 * an intrinsic that isn't a trivial wrapper), we may have to find a better way
 * to go about this.
 *
 * NOTE: clang currently complains that "'+soft-float-abi' is not a recognized
 * feature for this target (ignoring feature)." This appears to be a harmless
 * bug in clang.
 */
#define __crc32b __builtin_arm_crc32b
#define __crc32d __builtin_arm_crc32d
#define __crc32w __builtin_arm_crc32w
#define __crc32cw __builtin_arm_crc32cw

#if defined(__aarch64__)
#define TARGET_ARMV8_WITH_CRC __attribute__((target("crc")))
#else  // !defined(__aarch64__)
#define TARGET_ARMV8_WITH_CRC __attribute__((target("armv8-a,crc")))
#endif  // defined(__aarch64__)

#elif defined(__GNUC__)
/* For GCC, we are setting CRC extensions at module level, so ThinLTO is not
 * allowed. We can just include arm_acle.h.
 */
#include <arm_acle.h>
#define TARGET_ARMV8_WITH_CRC
#else  // !defined(__GNUC__) && !defined(_aarch64__)
#error ARM CRC32 SIMD extensions only supported for Clang and GCC
#endif

TARGET_ARMV8_WITH_CRC
uint32_t ZLIB_INTERNAL armv8_crc32_little(unsigned long crc,
                                          const unsigned char *buf,
                                          z_size_t len)
{
    uint32_t c = (uint32_t) ~crc;

    while (len && ((uintptr_t)buf & 7)) {
        c = __crc32b(c, *buf++);
        --len;
    }

    const uint64_t *buf8 = (const uint64_t *)buf;

    while (len >= 64) {
        c = __crc32d(c, *buf8++);
        c = __crc32d(c, *buf8++);
        c = __crc32d(c, *buf8++);
        c = __crc32d(c, *buf8++);

        c = __crc32d(c, *buf8++);
        c = __crc32d(c, *buf8++);
        c = __crc32d(c, *buf8++);
        c = __crc32d(c, *buf8++);
        len -= 64;
    }

    while (len >= 8) {
        c = __crc32d(c, *buf8++);
        len -= 8;
    }

    buf = (const unsigned char *)buf8;

    while (len--) {
        c = __crc32b(c, *buf++);
    }

    return ~c;
}

#endif