Geek Answers Handbook

What is the minimum version of OS X for use with AVX / AVX2?

I have an image drawing routine that has been compiled multiple times for SSE, SSE2, SSE3, SSE4.1, SSE4.2, AVX and AVX2. My program dynamically sends one of these binary options by checking the CPUID flags.

On Windows, I will check the version of Windows and disable sending AVX / AVX2 if the OS does not support them. (For example, only Windows 7 Service Pack 1 (SP1) or later supports AVX / AVX2.)

I want to do the same on Mac OS X, but I'm not sure which version of OS X supports AVX / AVX2.

Note that I want to know this is the minimum version of OS X for use with AVX / AVX2. Not machine models capable of AVX / AVX2.

+1
sse avx avx2 macos
source share
2 answers

To discover the functions of the instruction set, there are two source files that I refer to:

Both of these files will tell you how to detect SSE via AVX2, as well as XOP, FMA3, FMA4, if your OS supports AVX and other functions.

I use Agner code (one source file for MSVC, GCC, Clang, ICC), so let's take a look at this first.

Below are the relevant code snippets from instrset_detect.cpp for AVX detection:

 iset = 0; // default value int abcd[4] = {0,0,0,0}; // cpuid results cpuid(abcd, 0); // call cpuid function 0 //.... iset = 6; // 6: SSE4.2 supported if ((abcd[2] & (1 << 27)) == 0) return iset; // no OSXSAVE if ((xgetbv(0) & 6) != 6) return iset; // AVX not enabled in OS if ((abcd[2] & (1 << 28)) == 0) return iset; // no AVX iset = 7; // 7: AVX supported

with xgetbv defined as

 // Define interface to xgetbv instruction static inline int64_t xgetbv (int ctr) { #if (defined (_MSC_FULL_VER) && _MSC_FULL_VER >= 160040000) || (defined (__INTEL_COMPILER) && __INTEL_COMPILER >= 1200) // Microsoft or Intel compiler supporting _xgetbv intrinsic return _xgetbv(ctr); // intrinsic function for XGETBV #elif defined(__GNUC__) // use inline assembly, Gnu/AT&T syntax uint32_t a, d; __asm("xgetbv" : "=a"(a),"=d"(d) : "c"(ctr) : ); return a | (uint64_t(d) << 32); #else // #elif defined (_WIN32) // other compiler. try inline assembly with masm/intel/MS syntax //see the source file }

I did not enable the cpuid function (see source code), and I removed the non-GCC inline assembly from xgetbv to make the answer shorter.

Here is detect_OS_AVX() from Mystical cpu_x86.cpp for detecting AVX:

 bool cpu_x86::detect_OS_AVX(){ // Copied from: http://stackoverflow.com/a/22521619/922184 bool avxSupported = false; int cpuInfo[4]; cpuid(cpuInfo, 1); bool osUsesXSAVE_XRSTORE = (cpuInfo[2] & (1 << 27)) != 0; bool cpuAVXSuport = (cpuInfo[2] & (1 << 28)) != 0; if (osUsesXSAVE_XRSTORE && cpuAVXSuport) { uint64_t xcrFeatureMask = xgetbv(_XCR_XFEATURE_ENABLED_MASK); avxSupported = (xcrFeatureMask & 0x6) == 0x6; } return avxSupported; }

Mystic seems to have come up with this solution from this answer .

Please note that both source files have basically the same thing: check bit 27 of OSXSAVE, check bit AVX 28 of CPUID, check result with xgetbv .

+4
source share

For AVX, the answer is pretty simple:

You need at least OS X 10.6.7

Please note that support for 10J3250 and 10J4138 will support it.

For AVX2, which will be 10.8.4 build 12E3067 or 12E4022

+1
source share

More articles:


All Articles