mirrors/looking-glass
2
0
Fork 0
mirror of https://github.com/gnif/LookingGlass.git synced 2024-12-23 06:11:46 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

[host] replaced use of AVX in memcpySSE

Browse source 
There is no need to use AVX instructions as we are not using the
wider registers. Removing their use removes the requirement to run
a AVX capable guest CPU.
This commit is contained in:
Geoffrey McRae 2018-07-10 16:32:50 +10:00
parent b19518a1f8
commit a72ad4e46c
1 changed files with 30 additions and 30 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

60
common/memcpySSE.asm
View file

@ -25,22 +25,22 @@ memcpySSE proc
add r10, r9 add r10, r9
@FullLoop: @FullLoop:
vmovaps xmm0 , xmmword ptr [rdx + 000h] movaps xmm0 , xmmword ptr [rdx + 000h]
vmovaps xmm1 , xmmword ptr [rdx + 010h] movaps xmm1 , xmmword ptr [rdx + 010h]
vmovaps xmm2 , xmmword ptr [rdx + 020h] movaps xmm2 , xmmword ptr [rdx + 020h]
vmovaps xmm3 , xmmword ptr [rdx + 030h] movaps xmm3 , xmmword ptr [rdx + 030h]
vmovaps xmm4 , xmmword ptr [rdx + 040h] movaps xmm4 , xmmword ptr [rdx + 040h]
vmovaps xmm5 , xmmword ptr [rdx + 050h] movaps xmm5 , xmmword ptr [rdx + 050h]
vmovaps xmm6 , xmmword ptr [rdx + 060h] movaps xmm6 , xmmword ptr [rdx + 060h]
vmovaps xmm7 , xmmword ptr [rdx + 070h] movaps xmm7 , xmmword ptr [rdx + 070h]
vmovntdq xmmword ptr [rcx + 000h], xmm0 movntdq xmmword ptr [rcx + 000h], xmm0
vmovntdq xmmword ptr [rcx + 010h], xmm1 movntdq xmmword ptr [rcx + 010h], xmm1
vmovntdq xmmword ptr [rcx + 020h], xmm2 movntdq xmmword ptr [rcx + 020h], xmm2
vmovntdq xmmword ptr [rcx + 030h], xmm3 movntdq xmmword ptr [rcx + 030h], xmm3
vmovntdq xmmword ptr [rcx + 040h], xmm4 movntdq xmmword ptr [rcx + 040h], xmm4
vmovntdq xmmword ptr [rcx + 050h], xmm5 movntdq xmmword ptr [rcx + 050h], xmm5
vmovntdq xmmword ptr [rcx + 060h], xmm6 movntdq xmmword ptr [rcx + 060h], xmm6
vmovntdq xmmword ptr [rcx + 070h], xmm7 movntdq xmmword ptr [rcx + 070h], xmm7
add rdx, 080h add rdx, 080h
add rcx, 080h add rcx, 080h
cmp rcx, r10 cmp rcx, r10
@ -71,20 +71,20 @@ memcpySSE proc
ret ret
@FinalBlocks: @FinalBlocks:
vmovaps xmm6 , xmmword ptr [rdx + 060h] movaps xmm6 , xmmword ptr [rdx + 060h]
vmovntdq xmmword ptr [rcx + 060h], xmm6 movntdq xmmword ptr [rcx + 060h], xmm6
vmovaps xmm5 , xmmword ptr [rdx + 050h] movaps xmm5 , xmmword ptr [rdx + 050h]
vmovntdq xmmword ptr [rcx + 050h], xmm5 movntdq xmmword ptr [rcx + 050h], xmm5
vmovaps xmm4 , xmmword ptr [rdx + 040h] movaps xmm4 , xmmword ptr [rdx + 040h]
vmovntdq xmmword ptr [rcx + 040h], xmm4 movntdq xmmword ptr [rcx + 040h], xmm4
vmovaps xmm3 , xmmword ptr [rdx + 030h] movaps xmm3 , xmmword ptr [rdx + 030h]
vmovntdq xmmword ptr [rcx + 030h], xmm3 movntdq xmmword ptr [rcx + 030h], xmm3
vmovaps xmm2 , xmmword ptr [rdx + 020h] movaps xmm2 , xmmword ptr [rdx + 020h]
vmovntdq xmmword ptr [rcx + 020h], xmm2 movntdq xmmword ptr [rcx + 020h], xmm2
vmovaps xmm1 , xmmword ptr [rdx + 010h] movaps xmm1 , xmmword ptr [rdx + 010h]
vmovntdq xmmword ptr [rcx + 010h], xmm1 movntdq xmmword ptr [rcx + 010h], xmm1
vmovaps xmm0 , xmmword ptr [rdx + 000h] movaps xmm0 , xmmword ptr [rdx + 000h]
vmovntdq xmmword ptr [rcx + 000h], xmm0 movntdq xmmword ptr [rcx + 000h], xmm0
movdqa xmm6 , oword ptr [rsp + 4*8 + 00] movdqa xmm6 , oword ptr [rsp + 4*8 + 00]
movdqa xmm7 , oword ptr [rsp + 4*8 + 16] movdqa xmm7 , oword ptr [rsp + 4*8 + 16]