skia2/src/core/SkCpu.h
Mike Klein b437351d88 add _skx stages
This just makes sure all the plumbing is in place to use the Skylake
Xeon subset of AVX-512 instructions.  So far,

  - no Windows
  - no lowp
  - nothing explicitly making use of AVX-512 registers or instructions

This initial pass should run essentially identically to the _hsw AVX2
code we've been using previously.  Clang _does_ use AVX-512-only
instructions to implement some of the higher-level concepts we've coded,
but it's really a pretty subtle difference.

Next steps will bump N from 8 to 16 and start threading through an
AVX-512-friendly mask instead of tail.  I'll also want to take a harder
look at how we do blending like if_then_else()... the default codegen
here doesn't really take advantage of AVX-512 the way I'd like here.

CQ_INCLUDE_TRYBOTS=skia.primary:Test-Debian9-Clang-GCE-CPU-AVX512-x86_64-Debug

Change-Id: I6c9442488a449ea4770617bb22b2669859cc92e2
Reviewed-on: https://skia-review.googlesource.com/54062
Commit-Queue: Mike Klein <mtklein@chromium.org>
Reviewed-by: Herb Derby <herb@google.com>
2017-10-02 22:05:49 +00:00

114 lines
3.0 KiB
C++

/*
* Copyright 2016 Google Inc.
*
* Use of this source code is governed by a BSD-style license that can be
* found in the LICENSE file.
*/
#ifndef SkCpu_DEFINED
#define SkCpu_DEFINED
#include "SkTypes.h"
struct SkCpu {
enum {
SSE1 = 1 << 0,
SSE2 = 1 << 1,
SSE3 = 1 << 2,
SSSE3 = 1 << 3,
SSE41 = 1 << 4,
SSE42 = 1 << 5,
AVX = 1 << 6,
F16C = 1 << 7,
FMA = 1 << 8,
AVX2 = 1 << 9,
BMI1 = 1 << 10,
BMI2 = 1 << 11,
// Handy alias for all the cool Haswell+ instructions.
HSW = AVX2 | BMI1 | BMI2 | F16C | FMA,
AVX512F = 1 << 12,
AVX512DQ = 1 << 13,
AVX512IFMA = 1 << 14,
AVX512PF = 1 << 15,
AVX512ER = 1 << 16,
AVX512CD = 1 << 17,
AVX512BW = 1 << 18,
AVX512VL = 1 << 19,
// Handy alias for all the cool Skylake Xeon+ instructions.
SKX = AVX512F | AVX512DQ | AVX512CD | AVX512BW | AVX512VL,
};
enum {
NEON = 1 << 0,
NEON_FMA = 1 << 1,
VFP_FP16 = 1 << 2,
CRC32 = 1 << 3,
};
static void CacheRuntimeFeatures();
static bool Supports(uint32_t);
private:
static uint32_t gCachedFeatures;
};
inline bool SkCpu::Supports(uint32_t mask) {
uint32_t features = gCachedFeatures;
// If we mask in compile-time known lower limits, the compiler can
// often compile away this entire function.
#if SK_CPU_X86
#if SK_CPU_SSE_LEVEL >= SK_CPU_SSE_LEVEL_SSE1
features |= SSE1;
#endif
#if SK_CPU_SSE_LEVEL >= SK_CPU_SSE_LEVEL_SSE2
features |= SSE2;
#endif
#if SK_CPU_SSE_LEVEL >= SK_CPU_SSE_LEVEL_SSE3
features |= SSE3;
#endif
#if SK_CPU_SSE_LEVEL >= SK_CPU_SSE_LEVEL_SSSE3
features |= SSSE3;
#endif
#if SK_CPU_SSE_LEVEL >= SK_CPU_SSE_LEVEL_SSE41
features |= SSE41;
#endif
#if SK_CPU_SSE_LEVEL >= SK_CPU_SSE_LEVEL_SSE42
features |= SSE42;
#endif
#if SK_CPU_SSE_LEVEL >= SK_CPU_SSE_LEVEL_AVX
features |= AVX;
#endif
// F16C goes here if we add SK_CPU_SSE_LEVEL_F16C
#if SK_CPU_SSE_LEVEL >= SK_CPU_SSE_LEVEL_AVX2
features |= AVX2;
#endif
// FMA doesn't fit neatly into this total ordering.
// It's available on Haswell+ just like AVX2, but it's technically a different bit.
// TODO: circle back on this if we find ourselves limited by lack of compile-time FMA
#if defined(SK_CPU_LIMIT_SSE41)
features &= (SkCpu::SSE1 | SkCpu::SSE2 | SkCpu::SSE3 | SkCpu::SSSE3 | SkCpu::SSE41);
#elif defined(SK_CPU_LIMIT_SSE2)
features &= (SkCpu::SSE1 | SkCpu::SSE2);
#endif
#else
#if defined(SK_ARM_HAS_NEON)
features |= NEON;
#endif
#if defined(SK_CPU_ARM64)
features |= NEON|NEON_FMA|VFP_FP16;
#endif
#if defined(SK_ARM_HAS_CRC32)
features |= CRC32;
#endif
#endif
return (features & mask) == mask;
}
#endif//SkCpu_DEFINED