cpu.c (3674B)
1 /* 2 * Copyright © 2018, VideoLAN and dav1d authors 3 * Copyright © 2018, Two Orioles, LLC 4 * All rights reserved. 5 * 6 * Redistribution and use in source and binary forms, with or without 7 * modification, are permitted provided that the following conditions are met: 8 * 9 * 1. Redistributions of source code must retain the above copyright notice, this 10 * list of conditions and the following disclaimer. 11 * 12 * 2. Redistributions in binary form must reproduce the above copyright notice, 13 * this list of conditions and the following disclaimer in the documentation 14 * and/or other materials provided with the distribution. 15 * 16 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND 17 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED 18 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE 19 * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR 20 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES 21 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 22 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND 23 * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT 24 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS 25 * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 26 */ 27 28 #include "config.h" 29 30 #include <stdint.h> 31 #include <string.h> 32 33 #include "common/attributes.h" 34 35 #include "src/cpu.h" 36 #include "src/x86/cpu.h" 37 38 typedef struct { 39 uint32_t eax, ebx, edx, ecx; 40 } CpuidRegisters; 41 42 void dav1d_cpu_cpuid(CpuidRegisters *regs, unsigned leaf, unsigned subleaf); 43 uint64_t dav1d_cpu_xgetbv(unsigned xcr); 44 45 #define X(reg, mask) (((reg) & (mask)) == (mask)) 46 47 COLD unsigned dav1d_get_cpu_flags_x86(void) { 48 union { 49 CpuidRegisters r; 50 struct { 51 uint32_t max_leaf; 52 char vendor[12]; 53 }; 54 } cpu; 55 dav1d_cpu_cpuid(&cpu.r, 0, 0); 56 unsigned flags = dav1d_get_default_cpu_flags(); 57 58 if (cpu.max_leaf >= 1) { 59 CpuidRegisters r; 60 dav1d_cpu_cpuid(&r, 1, 0); 61 const unsigned family = ((r.eax >> 8) & 0x0f) + ((r.eax >> 20) & 0xff); 62 63 if (X(r.edx, 0x06008000)) /* CMOV/SSE/SSE2 */ { 64 flags |= DAV1D_X86_CPU_FLAG_SSE2; 65 if (X(r.ecx, 0x00000201)) /* SSE3/SSSE3 */ { 66 flags |= DAV1D_X86_CPU_FLAG_SSSE3; 67 if (X(r.ecx, 0x00080000)) /* SSE4.1 */ 68 flags |= DAV1D_X86_CPU_FLAG_SSE41; 69 } 70 } 71 #if ARCH_X86_64 72 /* We only support >128-bit SIMD on x86-64. */ 73 if (X(r.ecx, 0x18000000)) /* OSXSAVE/AVX */ { 74 const uint64_t xcr0 = dav1d_cpu_xgetbv(0); 75 if (X(xcr0, 0x00000006)) /* XMM/YMM */ { 76 if (cpu.max_leaf >= 7) { 77 dav1d_cpu_cpuid(&r, 7, 0); 78 if (X(r.ebx, 0x00000128)) /* BMI1/BMI2/AVX2 */ { 79 flags |= DAV1D_X86_CPU_FLAG_AVX2; 80 if (X(xcr0, 0x000000e0)) /* ZMM/OPMASK */ { 81 if (X(r.ebx, 0xd0230000) && X(r.ecx, 0x00005f42)) 82 flags |= DAV1D_X86_CPU_FLAG_AVX512ICL; 83 } 84 } 85 } 86 } 87 } 88 #endif 89 if (!memcmp(cpu.vendor, "AuthenticAMD", sizeof(cpu.vendor))) { 90 if ((flags & DAV1D_X86_CPU_FLAG_AVX2) && family <= 0x19) { 91 /* Excavator, Zen, Zen+, Zen 2, Zen 3, Zen 3+, Zen 4 */ 92 flags |= DAV1D_X86_CPU_FLAG_SLOW_GATHER; 93 } 94 } 95 } 96 97 return flags; 98 }