From a05ce3ad9d34b1085430a12b84984dbdacc3a6c8 Mon Sep 17 00:00:00 2001 From: Justine Tunney Date: Wed, 29 May 2024 10:13:37 -0700 Subject: [PATCH] Support avx512f + vpclmulqdq crc32() acceleration Cosmo's _Cz_crc32() function now goes 73 GiB/s on Threadripper. This will significantly improve the performance of the PKZIP file format. This algorithm is also used by apelink, to create deterministic ids. --- test/libc/str/crc32z_test.c | 354 +++++++++++++++++- third_party/zlib/BUILD.mk | 23 +- third_party/zlib/cpu_features.internal.h | 2 +- third_party/zlib/crc32.c | 5 +- .../zlib/{crc32_simd.c => crc32_simd.inc} | 0 third_party/zlib/crc32_simd_avx512.c | 3 + third_party/zlib/crc32_simd_neon.c | 3 + third_party/zlib/crc32_simd_sse42.c | 3 + 8 files changed, 385 insertions(+), 8 deletions(-) rename third_party/zlib/{crc32_simd.c => crc32_simd.inc} (100%) create mode 100644 third_party/zlib/crc32_simd_avx512.c create mode 100644 third_party/zlib/crc32_simd_neon.c create mode 100644 third_party/zlib/crc32_simd_sse42.c diff --git a/test/libc/str/crc32z_test.c b/test/libc/str/crc32z_test.c index 7099a5a11..92f329ace 100644 --- a/test/libc/str/crc32z_test.c +++ b/test/libc/str/crc32z_test.c @@ -17,17 +17,21 @@ │ PERFORMANCE OF THIS SOFTWARE. │ ╚─────────────────────────────────────────────────────────────────────────────*/ #include "libc/assert.h" +#include "libc/calls/struct/timespec.h" #include "libc/dce.h" #include "libc/intrin/asan.internal.h" #include "libc/mem/gc.h" #include "libc/mem/mem.h" #include "libc/nexgen32e/crc32.h" #include "libc/nexgen32e/x86feature.h" +#include "libc/stdio/rand.h" #include "libc/stdio/stdio.h" #include "libc/str/str.h" -#include "libc/testlib/ezbench.h" #include "libc/testlib/hyperion.h" #include "libc/testlib/testlib.h" +#include "third_party/aarch64/arm_acle.internal.h" +#include "third_party/aarch64/arm_neon.internal.h" +#include "third_party/intel/immintrin.internal.h" #include "third_party/zlib/zlib.h" #define FANATICS "Fanatics" @@ -38,6 +42,330 @@ static const char hyperion[] = "From forth the loftiest fashion of his sleep / " "..."; +unsigned crc32_reference(unsigned crc, const void *data, size_t size) { + static unsigned tab[256]; + if (!tab[255]) { + for (int d = 0; d < 256; ++d) { + unsigned r = d; + for (int i = 0; i < 8; ++i) + r = r >> 1 ^ (r & 1 ? 0xedb88320u : 0); + tab[d] = r; + } + } + const unsigned char *p = (const unsigned char *)data; + const unsigned char *pe = p + size; + crc = ~crc; + while (p < pe) + crc = crc >> 8 ^ tab[(crc & 255) ^ *p++]; + crc = ~crc; + return crc; +} + +// Calculates CRC32 at 73 GiB/s on AMD Ryzen Threadripper PRO 7995WX. +// "Fast CRC Computation for Generic Polynomials Using PCLMULQDQ Instruction" +// V. Gopal, E. Ozturk, et al., 2009, http://intel.ly/2ySEwL0 +uint32_t crc32_uber_alles(uint32_t crc, const void *data, size_t len) { + const uint8_t *buf = (const uint8_t *)data; + crc = ~crc; +#if defined(__AVX512F__) && defined(__VPCLMULQDQ__) && defined(__PCLMUL__) + if (len >= 256) { + __attribute__((__aligned__(64))) static const uint64_t k1k2[] = { + 0x011542778a, 0x01322d1430, 0x011542778a, 0x01322d1430, + 0x011542778a, 0x01322d1430, 0x011542778a, 0x01322d1430, + }; + __attribute__((__aligned__(64))) static const uint64_t k3k4[] = { + 0x0154442bd4, 0x01c6e41596, 0x0154442bd4, 0x01c6e41596, + 0x0154442bd4, 0x01c6e41596, 0x0154442bd4, 0x01c6e41596, + }; + __attribute__((__aligned__(64))) static const uint64_t k5k6[] = { + 0x01751997d0, + 0x00ccaa009e, + }; + __attribute__((__aligned__(64))) static const uint64_t k7k8[] = { + 0x0163cd6124, + 0x0000000000, + }; + __attribute__((__aligned__(64))) static const uint64_t poly[] = { + 0x01db710641, + 0x01f7011641, + }; + __m512i x0, x1, x2, x3, x4, x5, x6, x7, x8, y5, y6, y7, y8; + __m128i a0, a1, a2, a3; + x1 = _mm512_loadu_si512((__m512i *)(buf + 0x00)); + x2 = _mm512_loadu_si512((__m512i *)(buf + 0x40)); + x3 = _mm512_loadu_si512((__m512i *)(buf + 0x80)); + x4 = _mm512_loadu_si512((__m512i *)(buf + 0xC0)); + x1 = _mm512_xor_si512(x1, _mm512_castsi128_si512(_mm_cvtsi32_si128(crc))); + x0 = _mm512_load_si512((__m512i *)k1k2); + buf += 256; + len -= 256; + while (len >= 256) { + x5 = _mm512_clmulepi64_epi128(x1, x0, 0x00); + x6 = _mm512_clmulepi64_epi128(x2, x0, 0x00); + x7 = _mm512_clmulepi64_epi128(x3, x0, 0x00); + x8 = _mm512_clmulepi64_epi128(x4, x0, 0x00); + x1 = _mm512_clmulepi64_epi128(x1, x0, 0x11); + x2 = _mm512_clmulepi64_epi128(x2, x0, 0x11); + x3 = _mm512_clmulepi64_epi128(x3, x0, 0x11); + x4 = _mm512_clmulepi64_epi128(x4, x0, 0x11); + y5 = _mm512_loadu_si512((__m512i *)(buf + 0x00)); + y6 = _mm512_loadu_si512((__m512i *)(buf + 0x40)); + y7 = _mm512_loadu_si512((__m512i *)(buf + 0x80)); + y8 = _mm512_loadu_si512((__m512i *)(buf + 0xC0)); + x1 = _mm512_xor_si512(x1, x5); + x2 = _mm512_xor_si512(x2, x6); + x3 = _mm512_xor_si512(x3, x7); + x4 = _mm512_xor_si512(x4, x8); + x1 = _mm512_xor_si512(x1, y5); + x2 = _mm512_xor_si512(x2, y6); + x3 = _mm512_xor_si512(x3, y7); + x4 = _mm512_xor_si512(x4, y8); + buf += 256; + len -= 256; + } + x0 = _mm512_load_si512((__m512i *)k3k4); + x5 = _mm512_clmulepi64_epi128(x1, x0, 0x00); + x1 = _mm512_clmulepi64_epi128(x1, x0, 0x11); + x1 = _mm512_xor_si512(x1, x2); + x1 = _mm512_xor_si512(x1, x5); + x5 = _mm512_clmulepi64_epi128(x1, x0, 0x00); + x1 = _mm512_clmulepi64_epi128(x1, x0, 0x11); + x1 = _mm512_xor_si512(x1, x3); + x1 = _mm512_xor_si512(x1, x5); + x5 = _mm512_clmulepi64_epi128(x1, x0, 0x00); + x1 = _mm512_clmulepi64_epi128(x1, x0, 0x11); + x1 = _mm512_xor_si512(x1, x4); + x1 = _mm512_xor_si512(x1, x5); + while (len >= 64) { + x2 = _mm512_loadu_si512((__m512i *)buf); + x5 = _mm512_clmulepi64_epi128(x1, x0, 0x00); + x1 = _mm512_clmulepi64_epi128(x1, x0, 0x11); + x1 = _mm512_xor_si512(x1, x2); + x1 = _mm512_xor_si512(x1, x5); + buf += 64; + len -= 64; + } + a0 = _mm_load_si128((__m128i *)k5k6); + a1 = _mm512_extracti32x4_epi32(x1, 0); + a2 = _mm512_extracti32x4_epi32(x1, 1); + a3 = _mm_clmulepi64_si128(a1, a0, 0x00); + a1 = _mm_clmulepi64_si128(a1, a0, 0x11); + a1 = _mm_xor_si128(a1, a3); + a1 = _mm_xor_si128(a1, a2); + a2 = _mm512_extracti32x4_epi32(x1, 2); + a3 = _mm_clmulepi64_si128(a1, a0, 0x00); + a1 = _mm_clmulepi64_si128(a1, a0, 0x11); + a1 = _mm_xor_si128(a1, a3); + a1 = _mm_xor_si128(a1, a2); + a2 = _mm512_extracti32x4_epi32(x1, 3); + a3 = _mm_clmulepi64_si128(a1, a0, 0x00); + a1 = _mm_clmulepi64_si128(a1, a0, 0x11); + a1 = _mm_xor_si128(a1, a3); + a1 = _mm_xor_si128(a1, a2); + a2 = _mm_clmulepi64_si128(a1, a0, 0x10); + a3 = _mm_setr_epi32(~0, 0, ~0, 0); + a1 = _mm_srli_si128(a1, 8); + a1 = _mm_xor_si128(a1, a2); + a0 = _mm_loadl_epi64((__m128i *)k7k8); + a2 = _mm_srli_si128(a1, 4); + a1 = _mm_and_si128(a1, a3); + a1 = _mm_clmulepi64_si128(a1, a0, 0x00); + a1 = _mm_xor_si128(a1, a2); + a0 = _mm_load_si128((__m128i *)poly); + a2 = _mm_and_si128(a1, a3); + a2 = _mm_clmulepi64_si128(a2, a0, 0x10); + a2 = _mm_and_si128(a2, a3); + a2 = _mm_clmulepi64_si128(a2, a0, 0x00); + a1 = _mm_xor_si128(a1, a2); + crc = _mm_extract_epi32(a1, 1); + } +#endif +#if defined(__SSE4_2__) && defined(__PCLMUL__) + if (len >= 64) { + __attribute__((__aligned__(16))) static const uint64_t k1k2[] = { + 0x0154442bd4, + 0x01c6e41596, + }; + __attribute__((__aligned__(16))) static const uint64_t k3k4[] = { + 0x01751997d0, + 0x00ccaa009e, + }; + __attribute__((__aligned__(16))) static const uint64_t k5k0[] = { + 0x0163cd6124, + 0x0000000000, + }; + __attribute__((__aligned__(16))) static const uint64_t poly[] = { + 0x01db710641, + 0x01f7011641, + }; + __m128i x0, x1, x2, x3, x4, x5, x6, x7, x8, y5, y6, y7, y8; + 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; + 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; + } + 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); + 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; + } + 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); + 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); + crc = _mm_extract_epi32(x1, 1); + } +#endif + static uint32_t tab[256]; + if (!tab[255]) { + // generates table for byte-wise crc calculation on the polynomial + // x^32+x^26+x^23+x^22+x^16+x^12+x^11+x^10+x^8+x^7+x^5+x^4+x^2+x+1 + uint32_t polynomial = 0xedb88320; // bits are reversed + for (int d = 0; d < 256; ++d) { + uint32_t r = d; + for (int i = 0; i < 8; ++i) + r = r >> 1 ^ (r & 1 ? polynomial : 0); + tab[d] = r; + } + } + for (size_t i = 0; i < len; ++i) + crc = crc >> 8 ^ tab[(crc & 255) ^ buf[i]]; + return ~crc; +} + +uint32_t crc32_bloated(uint32_t crc, const void *buf, size_t buf_len) { + const uint8_t *ptr = (const uint8_t *)buf; + static const uint32_t s_crc_table[256] = { + 0x00000000, 0x77073096, 0xEE0E612C, 0x990951BA, 0x076DC419, 0x706AF48F, + 0xE963A535, 0x9E6495A3, 0x0EDB8832, 0x79DCB8A4, 0xE0D5E91E, 0x97D2D988, + 0x09B64C2B, 0x7EB17CBD, 0xE7B82D07, 0x90BF1D91, 0x1DB71064, 0x6AB020F2, + 0xF3B97148, 0x84BE41DE, 0x1ADAD47D, 0x6DDDE4EB, 0xF4D4B551, 0x83D385C7, + 0x136C9856, 0x646BA8C0, 0xFD62F97A, 0x8A65C9EC, 0x14015C4F, 0x63066CD9, + 0xFA0F3D63, 0x8D080DF5, 0x3B6E20C8, 0x4C69105E, 0xD56041E4, 0xA2677172, + 0x3C03E4D1, 0x4B04D447, 0xD20D85FD, 0xA50AB56B, 0x35B5A8FA, 0x42B2986C, + 0xDBBBC9D6, 0xACBCF940, 0x32D86CE3, 0x45DF5C75, 0xDCD60DCF, 0xABD13D59, + 0x26D930AC, 0x51DE003A, 0xC8D75180, 0xBFD06116, 0x21B4F4B5, 0x56B3C423, + 0xCFBA9599, 0xB8BDA50F, 0x2802B89E, 0x5F058808, 0xC60CD9B2, 0xB10BE924, + 0x2F6F7C87, 0x58684C11, 0xC1611DAB, 0xB6662D3D, 0x76DC4190, 0x01DB7106, + 0x98D220BC, 0xEFD5102A, 0x71B18589, 0x06B6B51F, 0x9FBFE4A5, 0xE8B8D433, + 0x7807C9A2, 0x0F00F934, 0x9609A88E, 0xE10E9818, 0x7F6A0DBB, 0x086D3D2D, + 0x91646C97, 0xE6635C01, 0x6B6B51F4, 0x1C6C6162, 0x856530D8, 0xF262004E, + 0x6C0695ED, 0x1B01A57B, 0x8208F4C1, 0xF50FC457, 0x65B0D9C6, 0x12B7E950, + 0x8BBEB8EA, 0xFCB9887C, 0x62DD1DDF, 0x15DA2D49, 0x8CD37CF3, 0xFBD44C65, + 0x4DB26158, 0x3AB551CE, 0xA3BC0074, 0xD4BB30E2, 0x4ADFA541, 0x3DD895D7, + 0xA4D1C46D, 0xD3D6F4FB, 0x4369E96A, 0x346ED9FC, 0xAD678846, 0xDA60B8D0, + 0x44042D73, 0x33031DE5, 0xAA0A4C5F, 0xDD0D7CC9, 0x5005713C, 0x270241AA, + 0xBE0B1010, 0xC90C2086, 0x5768B525, 0x206F85B3, 0xB966D409, 0xCE61E49F, + 0x5EDEF90E, 0x29D9C998, 0xB0D09822, 0xC7D7A8B4, 0x59B33D17, 0x2EB40D81, + 0xB7BD5C3B, 0xC0BA6CAD, 0xEDB88320, 0x9ABFB3B6, 0x03B6E20C, 0x74B1D29A, + 0xEAD54739, 0x9DD277AF, 0x04DB2615, 0x73DC1683, 0xE3630B12, 0x94643B84, + 0x0D6D6A3E, 0x7A6A5AA8, 0xE40ECF0B, 0x9309FF9D, 0x0A00AE27, 0x7D079EB1, + 0xF00F9344, 0x8708A3D2, 0x1E01F268, 0x6906C2FE, 0xF762575D, 0x806567CB, + 0x196C3671, 0x6E6B06E7, 0xFED41B76, 0x89D32BE0, 0x10DA7A5A, 0x67DD4ACC, + 0xF9B9DF6F, 0x8EBEEFF9, 0x17B7BE43, 0x60B08ED5, 0xD6D6A3E8, 0xA1D1937E, + 0x38D8C2C4, 0x4FDFF252, 0xD1BB67F1, 0xA6BC5767, 0x3FB506DD, 0x48B2364B, + 0xD80D2BDA, 0xAF0A1B4C, 0x36034AF6, 0x41047A60, 0xDF60EFC3, 0xA867DF55, + 0x316E8EEF, 0x4669BE79, 0xCB61B38C, 0xBC66831A, 0x256FD2A0, 0x5268E236, + 0xCC0C7795, 0xBB0B4703, 0x220216B9, 0x5505262F, 0xC5BA3BBE, 0xB2BD0B28, + 0x2BB45A92, 0x5CB36A04, 0xC2D7FFA7, 0xB5D0CF31, 0x2CD99E8B, 0x5BDEAE1D, + 0x9B64C2B0, 0xEC63F226, 0x756AA39C, 0x026D930A, 0x9C0906A9, 0xEB0E363F, + 0x72076785, 0x05005713, 0x95BF4A82, 0xE2B87A14, 0x7BB12BAE, 0x0CB61B38, + 0x92D28E9B, 0xE5D5BE0D, 0x7CDCEFB7, 0x0BDBDF21, 0x86D3D2D4, 0xF1D4E242, + 0x68DDB3F8, 0x1FDA836E, 0x81BE16CD, 0xF6B9265B, 0x6FB077E1, 0x18B74777, + 0x88085AE6, 0xFF0F6A70, 0x66063BCA, 0x11010B5C, 0x8F659EFF, 0xF862AE69, + 0x616BFFD3, 0x166CCF45, 0xA00AE278, 0xD70DD2EE, 0x4E048354, 0x3903B3C2, + 0xA7672661, 0xD06016F7, 0x4969474D, 0x3E6E77DB, 0xAED16A4A, 0xD9D65ADC, + 0x40DF0B66, 0x37D83BF0, 0xA9BCAE53, 0xDEBB9EC5, 0x47B2CF7F, 0x30B5FFE9, + 0xBDBDF21C, 0xCABAC28A, 0x53B39330, 0x24B4A3A6, 0xBAD03605, 0xCDD70693, + 0x54DE5729, 0x23D967BF, 0xB3667A2E, 0xC4614AB8, 0x5D681B02, 0x2A6F2B94, + 0xB40BBE37, 0xC30C8EA1, 0x5A05DF1B, 0x2D02EF8D}; + uint32_t crc32 = (uint32_t)crc ^ 0xFFFFFFFF; + const uint8_t *pByte_buf = (const uint8_t *)ptr; + while (buf_len >= 4) { + crc32 = (crc32 >> 8) ^ s_crc_table[(crc32 ^ pByte_buf[0]) & 0xFF]; + crc32 = (crc32 >> 8) ^ s_crc_table[(crc32 ^ pByte_buf[1]) & 0xFF]; + crc32 = (crc32 >> 8) ^ s_crc_table[(crc32 ^ pByte_buf[2]) & 0xFF]; + crc32 = (crc32 >> 8) ^ s_crc_table[(crc32 ^ pByte_buf[3]) & 0xFF]; + pByte_buf += 4; + buf_len -= 4; + } + while (buf_len) { + crc32 = (crc32 >> 8) ^ s_crc_table[(crc32 ^ pByte_buf[0]) & 0xFF]; + ++pByte_buf; + --buf_len; + } + return ~crc32; +} + +TEST(crc32_z, fuzz) { + char buf[512]; + for (int n = 0; n < 512; ++n) { + for (int i = 0; i < n; ++i) + buf[i] = rand(); + for (int i = 0; i < 10; ++i) { + int x = lemur64(); + ASSERT_EQ(crc32_bloated(x, buf, n), crc32_z(x, buf, n)); + ASSERT_EQ(crc32_reference(x, buf, n), crc32_z(x, buf, n)); + ASSERT_EQ(crc32_uber_alles(x, buf, n), crc32_z(x, buf, n)); + } + } +} + TEST(crc32_z, test) { char *p; p = gc(strdup(kHyperion)); @@ -54,7 +382,27 @@ TEST(crc32_z, test) { EXPECT_EQ(0xe9ded8e6, crc32_z(0, p, kHyperionSize)); } +#define BENCHMARK_GBPS(ITERATIONS, BYTES_PER_RUN, CODE) \ + do { \ + struct timespec start = timespec_real(); \ + for (int i = 0; i < ITERATIONS; ++i) { \ + CODE; \ + } \ + long bytes = BYTES_PER_RUN * ITERATIONS; \ + struct timespec elapsed = timespec_sub(timespec_real(), start); \ + long nanos_elapsed = timespec_tonanos(elapsed); \ + double seconds_elapsed = nanos_elapsed * 1e-9; \ + double bytes_per_second = bytes / seconds_elapsed; \ + double gib_per_second = bytes_per_second / (1024 * 1024 * 1024); \ + printf("%12g GiB/s %s\n", gib_per_second, #CODE); \ + } while (0) + BENCH(crc32_z, bench) { - EZBENCH_N("crc32_z", kHyperionSize, crc32_z(0, kHyperion, kHyperionSize)); - EZBENCH_N("crc32_z", kHyperionSize, crc32_z(0, kHyperion, kHyperionSize)); + volatile unsigned vv = 0; + BENCHMARK_GBPS(100, kHyperionSize, + vv += crc32_z(0, kHyperion, kHyperionSize)); + BENCHMARK_GBPS(100, kHyperionSize, + vv += crc32_bloated(0, kHyperion, kHyperionSize)); + BENCHMARK_GBPS(100, kHyperionSize, + vv += crc32_uber_alles(0, kHyperion, kHyperionSize)); } diff --git a/third_party/zlib/BUILD.mk b/third_party/zlib/BUILD.mk index fb6390bcf..fa383a8b2 100644 --- a/third_party/zlib/BUILD.mk +++ b/third_party/zlib/BUILD.mk @@ -38,16 +38,31 @@ $(THIRD_PARTY_ZLIB_A).pkg: \ ifeq ($(ARCH), x86_64) o/$(MODE)/third_party/zlib/adler32_simd.o: private \ TARGET_ARCH += \ + -O3 \ -mssse3 o/$(MODE)/third_party/zlib/crc_folding.o \ -o/$(MODE)/third_party/zlib/crc32_simd.o: private \ +o/$(MODE)/third_party/zlib/crc32_simd_sse42.o: private \ TARGET_ARCH += \ + -O3 \ -msse4.2 \ - -mpclmul + -mpclmul \ + -UCRC32_SIMD_AVX512_PCLMUL \ + -DCRC32_SIMD_SSE42_PCLMUL \ + -DBUILD_SSE42 +o/$(MODE)/third_party/zlib/crc32_simd_avx512.o: private \ + TARGET_ARCH += \ + -O3 \ + -mpclmul \ + -mavx512f \ + -mvpclmulqdq \ + -UCRC32_SIMD_SSE42_PCLMUL \ + -DCRC32_SIMD_AVX512_PCLMUL \ + -DBUILD_AVX512 $(THIRD_PARTY_ZLIB_A_OBJS): private \ CPPFLAGS += \ -DADLER32_SIMD_SSSE3 \ -DCRC32_SIMD_SSE42_PCLMUL \ + -DCRC32_SIMD_AVX512_PCLMUL \ -DDEFLATE_SLIDE_HASH_SSE2 \ -DINFLATE_CHUNK_SIMD_SSE2 \ -DINFLATE_CHUNK_READ_64LE @@ -55,8 +70,10 @@ endif ifeq ($(ARCH), aarch64) o/$(MODE)/third_party/zlib/deflate.o \ -o/$(MODE)/third_party/zlib/crc32_simd.o: private \ +o/$(MODE)/third_party/zlib/crc32_simd_neon.o: private \ TARGET_ARCH += \ + -O3 \ + -DBUILD_NEON \ -march=armv8-a+aes+crc $(THIRD_PARTY_ZLIB_A_OBJS): private \ CPPFLAGS += \ diff --git a/third_party/zlib/cpu_features.internal.h b/third_party/zlib/cpu_features.internal.h index 48da0e7b4..f4a24393d 100644 --- a/third_party/zlib/cpu_features.internal.h +++ b/third_party/zlib/cpu_features.internal.h @@ -16,7 +16,7 @@ COSMOPOLITAN_C_START_ #define x86_cpu_enable_sse2 X86_HAVE(SSE2) #define x86_cpu_enable_ssse3 X86_HAVE(SSSE3) #define x86_cpu_enable_simd (X86_HAVE(SSE4_2) && X86_HAVE(PCLMUL)) -#define x86_cpu_enable_avx512 X86_HAVE(AVX512F) +#define x86_cpu_enable_avx512 (X86_HAVE(AVX512F) && X86_HAVE(PCLMUL) && X86_HAVE(VPCLMULQDQ)) #define cpu_check_features() ((void)0) #elif defined(__aarch64__) diff --git a/third_party/zlib/crc32.c b/third_party/zlib/crc32.c index a5879c78c..b1e15b804 100644 --- a/third_party/zlib/crc32.c +++ b/third_party/zlib/crc32.c @@ -780,6 +780,7 @@ uint32_t ZEXPORT crc32_z(crc, buf_, len) } #endif +#if defined(__x86_64__) #if defined(CRC32_SIMD_AVX512_PCLMUL) if (x86_cpu_enable_avx512 && len >= Z_CRC32_AVX512_MINIMUM_LENGTH) { /* crc32 64-byte chunks */ @@ -792,7 +793,8 @@ uint32_t ZEXPORT crc32_z(crc, buf_, len) /* Fall into the default crc32 for the remaining data. */ buf += chunk_size; } -#elif defined(CRC32_SIMD_SSE42_PCLMUL) +#endif +#if defined(CRC32_SIMD_SSE42_PCLMUL) if (x86_cpu_enable_simd && len >= Z_CRC32_SSE42_MINIMUM_LENGTH) { /* crc32 16-byte chunks */ z_size_t chunk_size = len & ~Z_CRC32_SSE42_CHUNKSIZE_MASK; @@ -804,6 +806,7 @@ uint32_t ZEXPORT crc32_z(crc, buf_, len) /* Fall into the default crc32 for the remaining data. */ buf += chunk_size; } +#endif #elif defined(CRC32_ARMV8_CRC32) if (arm_cpu_enable_crc32) { #if defined(__aarch64__) diff --git a/third_party/zlib/crc32_simd.c b/third_party/zlib/crc32_simd.inc similarity index 100% rename from third_party/zlib/crc32_simd.c rename to third_party/zlib/crc32_simd.inc diff --git a/third_party/zlib/crc32_simd_avx512.c b/third_party/zlib/crc32_simd_avx512.c new file mode 100644 index 000000000..d911af038 --- /dev/null +++ b/third_party/zlib/crc32_simd_avx512.c @@ -0,0 +1,3 @@ +#ifdef BUILD_AVX512 +#include "third_party/zlib/crc32_simd.inc" +#endif diff --git a/third_party/zlib/crc32_simd_neon.c b/third_party/zlib/crc32_simd_neon.c new file mode 100644 index 000000000..441fc6126 --- /dev/null +++ b/third_party/zlib/crc32_simd_neon.c @@ -0,0 +1,3 @@ +#ifdef BUILD_NEON +#include "third_party/zlib/crc32_simd.inc" +#endif diff --git a/third_party/zlib/crc32_simd_sse42.c b/third_party/zlib/crc32_simd_sse42.c new file mode 100644 index 000000000..070207352 --- /dev/null +++ b/third_party/zlib/crc32_simd_sse42.c @@ -0,0 +1,3 @@ +#ifdef BUILD_SSE42 +#include "third_party/zlib/crc32_simd.inc" +#endif