blob: faddd389fdf2ab6baedd380d527a2a7773ccacf3 [file] [log] [blame]
; jsimdcpu.asm - SIMD instruction support check
; Copyright 2009 Pierre Ossman <> for Cendio AB
; Copyright (C) 2016, D. R. Commander.
; Based on the x86 SIMD extension for IJG JPEG library
; Copyright (C) 1999-2006, MIYASAKA Masaru.
; For conditions of distribution and use, see copyright notice in
; This file should be assembled with NASM (Netwide Assembler),
; can *not* be assembled with Microsoft's MASM or any compatible
; assembler (including Borland's Turbo Assembler).
; NASM is available from or
; [TAB8]
%include ""
; --------------------------------------------------------------------------
; Check if the CPU supports SIMD instructions
; GLOBAL(unsigned int)
; jpeg_simd_cpu_support(void)
align 32
push ebx
; push ecx ; need not be preserved
; push edx ; need not be preserved
; push esi ; unused
push edi
xor edi, edi ; simd support flag
pop eax
mov edx, eax
xor eax, 1<<21 ; flip ID bit in EFLAGS
push eax
pop eax
xor eax, edx
jz near .return ; CPUID is not supported
; Check for MMX instruction support
xor eax, eax
test eax, eax
jz near .return
xor eax, eax
inc eax
mov eax, edx ; eax = Standard feature flags
test eax, 1<<23 ; bit23:MMX
jz short .no_mmx
or edi, byte JSIMD_MMX
test eax, 1<<25 ; bit25:SSE
jz short .no_sse
or edi, byte JSIMD_SSE
test eax, 1<<26 ; bit26:SSE2
jz short .no_sse2
or edi, byte JSIMD_SSE2
; Check for AVX2 instruction support
mov eax, 7
xor ecx, ecx
mov eax, ebx
test eax, 1<<5 ; bit5:AVX2
jz short .no_avx2
; Check for AVX2 O/S support
mov eax, 1
xor ecx, ecx
test ecx, 1<<27
jz short .no_avx2 ; O/S does not support XSAVE
test ecx, 1<<28
jz short .no_avx2 ; CPU does not support AVX2
xor ecx, ecx
and eax, 6
cmp eax, 6 ; O/S does not manage XMM/YMM state
; using XSAVE
jnz short .no_avx2
or edi, JSIMD_AVX2
; Check for 3DNow! instruction support
mov eax, 0x80000000
cmp eax, 0x80000000
jbe short .return
mov eax, 0x80000001
mov eax, edx ; eax = Extended feature flags
test eax, 1<<31 ; bit31:3DNow!(vendor independent)
jz short .no_3dnow
or edi, byte JSIMD_3DNOW
mov eax, edi
pop edi
; pop esi ; unused
; pop edx ; need not be preserved
; pop ecx ; need not be preserved
pop ebx
; For some reason, the OS X linker does not honor the request to align the
; segment unless we do this.
align 32