From 9f8c20c3e7a1e724a3983a10e7ac485f9c5e4eed Mon Sep 17 00:00:00 2001 From: Geoffrey McRae Date: Tue, 10 Jul 2018 16:38:07 +1000 Subject: [PATCH] [common] no need to use AVX operands here either --- common/memcpySSE.h | 60 +++++++++++++++++++++++----------------------- 1 file changed, 30 insertions(+), 30 deletions(-) diff --git a/common/memcpySSE.h b/common/memcpySSE.h index 61e8b6ca..364b28b5 100644 --- a/common/memcpySSE.h +++ b/common/memcpySSE.h @@ -59,22 +59,22 @@ Place, Suite 330, Boston, MA 02111-1307 USA // perform SIMD block copy "loop_%=: \n\t" - "vmovaps 0x00(%[src]),%%xmm0 \n\t" - "vmovaps 0x10(%[src]),%%xmm1 \n\t" - "vmovaps 0x20(%[src]),%%xmm2 \n\t" - "vmovaps 0x30(%[src]),%%xmm3 \n\t" - "vmovaps 0x40(%[src]),%%xmm4 \n\t" - "vmovaps 0x50(%[src]),%%xmm5 \n\t" - "vmovaps 0x60(%[src]),%%xmm6 \n\t" - "vmovaps 0x70(%[src]),%%xmm7 \n\t" - "vmovntdq %%xmm0 ,0x00(%[dst]) \n\t" - "vmovntdq %%xmm1 ,0x10(%[dst]) \n\t" - "vmovntdq %%xmm2 ,0x20(%[dst]) \n\t" - "vmovntdq %%xmm3 ,0x30(%[dst]) \n\t" - "vmovntdq %%xmm4 ,0x40(%[dst]) \n\t" - "vmovntdq %%xmm5 ,0x50(%[dst]) \n\t" - "vmovntdq %%xmm6 ,0x60(%[dst]) \n\t" - "vmovntdq %%xmm7 ,0x70(%[dst]) \n\t" + "movaps 0x00(%[src]),%%xmm0 \n\t" + "movaps 0x10(%[src]),%%xmm1 \n\t" + "movaps 0x20(%[src]),%%xmm2 \n\t" + "movaps 0x30(%[src]),%%xmm3 \n\t" + "movaps 0x40(%[src]),%%xmm4 \n\t" + "movaps 0x50(%[src]),%%xmm5 \n\t" + "movaps 0x60(%[src]),%%xmm6 \n\t" + "movaps 0x70(%[src]),%%xmm7 \n\t" + "movntdq %%xmm0 ,0x00(%[dst]) \n\t" + "movntdq %%xmm1 ,0x10(%[dst]) \n\t" + "movntdq %%xmm2 ,0x20(%[dst]) \n\t" + "movntdq %%xmm3 ,0x30(%[dst]) \n\t" + "movntdq %%xmm4 ,0x40(%[dst]) \n\t" + "movntdq %%xmm5 ,0x50(%[dst]) \n\t" + "movntdq %%xmm6 ,0x60(%[dst]) \n\t" + "movntdq %%xmm7 ,0x70(%[dst]) \n\t" "add $0x80,%[dst] \n\t" "add $0x80,%[src] \n\t" "cmp %[dst],%[end] \n\t" @@ -94,20 +94,20 @@ Place, Suite 330, Boston, MA 02111-1307 USA "ret \n\t" "BlockTable_%=:\n\t" - "vmovaps 0x60(%[src]),%%xmm6 \n\t" - "vmovntdq %%xmm6 ,0x60(%[dst]) \n\t" - "vmovaps 0x50(%[src]),%%xmm5 \n\t" - "vmovntdq %%xmm5 ,0x50(%[dst]) \n\t" - "vmovaps 0x40(%[src]),%%xmm4 \n\t" - "vmovntdq %%xmm4 ,0x40(%[dst]) \n\t" - "vmovaps 0x30(%[src]),%%xmm3 \n\t" - "vmovntdq %%xmm3 ,0x30(%[dst]) \n\t" - "vmovaps 0x20(%[src]),%%xmm2 \n\t" - "vmovntdq %%xmm2 ,0x20(%[dst]) \n\t" - "vmovaps 0x10(%[src]),%%xmm1 \n\t" - "vmovntdq %%xmm1 ,0x10(%[dst]) \n\t" - "vmovaps 0x00(%[src]),%%xmm0 \n\t" - "vmovntdq %%xmm0 ,0x00(%[dst]) \n\t" + "movaps 0x60(%[src]),%%xmm6 \n\t" + "movntdq %%xmm6 ,0x60(%[dst]) \n\t" + "movaps 0x50(%[src]),%%xmm5 \n\t" + "movntdq %%xmm5 ,0x50(%[dst]) \n\t" + "movaps 0x40(%[src]),%%xmm4 \n\t" + "movntdq %%xmm4 ,0x40(%[dst]) \n\t" + "movaps 0x30(%[src]),%%xmm3 \n\t" + "movntdq %%xmm3 ,0x30(%[dst]) \n\t" + "movaps 0x20(%[src]),%%xmm2 \n\t" + "movntdq %%xmm2 ,0x20(%[dst]) \n\t" + "movaps 0x10(%[src]),%%xmm1 \n\t" + "movntdq %%xmm1 ,0x10(%[dst]) \n\t" + "movaps 0x00(%[src]),%%xmm0 \n\t" + "movntdq %%xmm0 ,0x00(%[dst]) \n\t" "nop\n\t" "nop\n\t"