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

bugfix in sgemv_t_microk_sandy.c

This commit is contained in:
wernsaar 2014-07-20 10:48:41 +02:00
parent c06f9986d4
commit 02eb72ac42
2 changed files with 10 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
kernel/x86_64/KERNEL.SANDYBRIDGE
View File

@ -1,3 +1,11 @@
ifdef OS_WINDOWS
#SGEMVNKERNEL = ../arm/gemv_n.c
SGEMVTKERNEL = ../arm/gemv_t.c
else
#SGEMVNKERNEL = sgemv_n_avx.c
SGEMVTKERNEL = sgemv_t_avx.c
endif
SGEMMKERNEL = sgemm_kernel_16x4_sandy.S
SGEMMINCOPY = ../generic/gemm_ncopy_16.c

3
kernel/x86_64/sgemv_t_microk_sandy.c
View File

@ -80,7 +80,8 @@ static void sgemv_kernel_16( long n, float alpha, float *a, long lda, float *x,
"vhaddps %%xmm12, %%xmm12, %%xmm12\n\t"
"vhaddps %%xmm12, %%xmm12, %%xmm12\n\t"
"vfmaddss (%%rdx), %%xmm12, %%xmm1, %%xmm12\n\t"
"vmulss %%xmm12, %%xmm1, %%xmm12 \n\t"
"vaddss (%%rdx), %%xmm12, %%xmm12\n\t"
"vmovss %%xmm12, (%%rdx) \n\t" // store temp -> y
: