google / cpu_features

A cross platform C99 library to get cpu features at runtime.
Apache License 2.0
2.47k stars 262 forks source link
cpu cpu-features cross-platform microarchitecture

cpu_features

A cross-platform C library to retrieve CPU features (such as available instructions) at runtime.

GitHub-CI Status

Linux FreeBSD MacOS Windows
amd64 CMake
Bazel
CMake
Bazel
CMake
Bazel
CMake
Bazel
AArch64 CMake
Bazel
CMake
Bazel
CMake
Bazel
CMake
Bazel
ARM CMake
Bazel
CMake
Bazel
CMake
Bazel
CMake
Bazel
MIPS CMake
Bazel
CMake
Bazel
CMake
Bazel
CMake
Bazel
POWER CMake
Bazel
CMake
Bazel
CMake
Bazel
CMake
Bazel
RISCV CMake
Bazel
CMake
Bazel
CMake
Bazel
CMake
Bazel
LOONGARCH CMake
Bazel
CMake
Bazel
CMake
Bazel
CMake
Bazel
s390x CMake
Bazel
CMake
Bazel
CMake
Bazel
CMake
Bazel

Table of Contents

Design Rationale

Code samples

Note: For C++ code, the library functions are defined in the cpu_features namespace.

Checking features at runtime

Here's a simple example that executes a codepath if the CPU supports both the AES and the SSE4.2 instruction sets:

#include "cpuinfo_x86.h"

// For C++, add `using namespace cpu_features;`
static const X86Features features = GetX86Info().features;

void Compute(void) {
  if (features.aes && features.sse4_2) {
    // Run optimized code.
  } else {
    // Run standard code.
  }
}

Caching for faster evaluation of complex checks

If you wish, you can read all the features at once into a global variable, and then query for the specific features you care about. Below, we store all the ARM features and then check whether AES and NEON are supported.

#include <stdbool.h>
#include "cpuinfo_arm.h"

// For C++, add `using namespace cpu_features;`
static const ArmFeatures features = GetArmInfo().features;
static const bool has_aes_and_neon = features.aes && features.neon;

// use has_aes_and_neon.

This is a good approach to take if you're checking for combinations of features when using a compiler that is slow to extract individual bits from bit-packed structures.

Checking compile time flags

The following code determines whether the compiler was told to use the AVX instruction set (e.g., g++ -mavx) and sets has_avx accordingly.

#include <stdbool.h>
#include "cpuinfo_x86.h"

// For C++, add `using namespace cpu_features;`
static const X86Features features = GetX86Info().features;
static const bool has_avx = CPU_FEATURES_COMPILED_X86_AVX || features.avx;

// use has_avx.

CPU_FEATURES_COMPILED_X86_AVX is set to 1 if the compiler was instructed to use AVX and 0 otherwise, combining compile time and runtime knowledge.

Rejecting poor hardware implementations based on microarchitecture

On x86, the first incarnation of a feature in a microarchitecture might not be the most efficient (e.g. AVX on Sandy Bridge). We provide a function to retrieve the underlying microarchitecture so you can decide whether to use it.

Below, has_fast_avx is set to 1 if the CPU supports the AVX instruction set—but only if it's not Sandy Bridge.

#include <stdbool.h>
#include "cpuinfo_x86.h"

// For C++, add `using namespace cpu_features;`
static const X86Info info = GetX86Info();
static const X86Microarchitecture uarch = GetX86Microarchitecture(&info);
static const bool has_fast_avx = info.features.avx && uarch != INTEL_SNB;

// use has_fast_avx.

This feature is currently available only for x86 microarchitectures.

Running sample code

Building cpu_features (check quickstart below) brings a small executable to test the library.

 % ./build/list_cpu_features
arch            : x86
brand           :        Intel(R) Xeon(R) CPU E5-1650 0 @ 3.20GHz
family          :   6 (0x06)
model           :  45 (0x2D)
stepping        :   7 (0x07)
uarch           : INTEL_SNB
flags           : aes,avx,cx16,smx,sse4_1,sse4_2,ssse3
% ./build/list_cpu_features --json
{"arch":"x86","brand":"       Intel(R) Xeon(R) CPU E5-1650 0 @ 3.20GHz","family":6,"model":45,"stepping":7,"uarch":"INTEL_SNB","flags":["aes","avx","cx16","smx","sse4_1","sse4_2","ssse3"]}

What's supported

x86³ AArch64 ARM MIPS⁴ POWER RISCV Loongarch s390x
Linux yes² yes¹ yes¹ yes¹ yes¹ yes¹ yes¹ yes¹
FreeBSD yes² not yet not yet not yet not yet N/A not yet not yet
MacOs yes² yes⁵ N/A N/A N/A N/A N/A N/A
Windows yes² not yet not yet N/A N/A N/A N/A N/A
Android yes² yes¹ yes¹ yes¹ N/A N/A N/A N/A
iOS N/A not yet not yet N/A N/A N/A N/A N/A
  1. Features revealed from Linux. We gather data from several sources depending on availability:
    • from glibc's getauxval
    • by parsing /proc/self/auxv
    • by parsing /proc/cpuinfo
  2. Features revealed from CPU. features are retrieved by using the cpuid instruction.
  3. Microarchitecture detection. On x86 some features are not always implemented efficiently in hardware (e.g. AVX on Sandybridge). Exposing the microarchitecture allows the client to reject particular microarchitectures.
  4. All flavors of Mips are supported, little and big endian as well as 32/64 bits.
  5. Features revealed from sysctl. features are retrieved by the sysctl instruction.

Android NDK's drop in replacement

cpu_features is now officially supporting Android and offers a drop in replacement of for the NDK's cpu-features.h , see ndk_compat folder for details.

License

The cpu_features library is licensed under the terms of the Apache license. See LICENSE for more information.

Build with CMake

Please check the CMake build instructions.

Quickstart

Community bindings

Links provided here are not affiliated with Google but are kindly provided by the OSS Community.

Send PR to showcase your wrapper here