floraachy/OpenBLAS
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Use .p2align instead of .align for compatibility on Sandybridge as well

Browse Source
This commit is contained in:
Martin Kroeker
2018-02-24 19:43:15 +01:00
committed by GitHub
parent 2359c7c1a9
commit 7c1925acec
18 changed files with 32 additions and 32 deletions
Download Patch File Download Diff File Expand all files Collapse all files
kernel/x86_64/zdot_microk_sandy-2.c
+2 -2
View File
@@ -49,7 +49,7 @@ if ( n < 1280 )
"vxorpd %%ymm6, %%ymm6, %%ymm6 \n\t"
"vxorpd %%ymm7, %%ymm7, %%ymm7 \n\t"
".align 16 \n\t"
".p2align 4 \n\t"
"1: \n\t"
"vmovups (%2,%0,8), %%ymm8 \n\t" // 2 * x
"vmovups 32(%2,%0,8), %%ymm9 \n\t" // 2 * x
@@ -137,7 +137,7 @@ if ( n < 1280 )
"vxorpd %%ymm6, %%ymm6, %%ymm6 \n\t"
"vxorpd %%ymm7, %%ymm7, %%ymm7 \n\t"
".align 16 \n\t"
".p2align 4 \n\t"
"1: \n\t"
"prefetcht0 512(%2,%0,8) \n\t"
"vmovups (%2,%0,8), %%ymm8 \n\t" // 2 * x