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

Add SVE implementation for sdot/ddot

Browse Source 
This adds an SVE implementation to sdot/ddot when available, falling back to the previous Advanced SIMD kernel where there's no SVE implementation for the kernel.

All the targets were essentially treating `dot_thunderx2t99.c` as the Advanced SIMD implementation so I've renamed it to better fit with the feature detection.
This commit is contained in:
Chris Sidebottom
2022-11-17 06:49:57 +00:00
parent b6a4ef98b9
commit fd4f52c797
10 changed files with 194 additions and 81 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
kernel/arm64/KERNEL.NEOVERSEN2
View File

@@ -96,8 +96,8 @@ DNRM2KERNEL = dznrm2_thunderx2t99.c
CNRM2KERNEL = scnrm2_thunderx2t99.c
ZNRM2KERNEL = dznrm2_thunderx2t99.c
DDOTKERNEL = dot_thunderx2t99.c
SDOTKERNEL = dot_thunderx2t99.c
DDOTKERNEL = dot.c
SDOTKERNEL = dot.c
CDOTKERNEL = zdot_thunderx2t99.c
ZDOTKERNEL = zdot_thunderx2t99.c
DSDOTKERNEL = dot.S