CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-29 08:41:22 +00:00
Code Issues Packages Projects Releases Wiki Activity

ggml : use __builtin_amdgcn_sudot4 in __dp4a for gfx11 (#4787)

Browse Source
This commit is contained in:
Konstantin Zhuravlyov
2024-01-07 01:52:42 -05:00
committed by GitHub
parent 67984921a7
commit 63ee677efd
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
ggml-cuda.cu
View File

@@ -183,7 +183,7 @@ static __device__ __forceinline__ int __vsubss4(const int a, const int b) {
static __device__ __forceinline__ int __dp4a(const int a, const int b, int c) {
#if defined(__gfx906__) || defined(__gfx908__) || defined(__gfx90a__) || defined(__gfx1030__)
c = __builtin_amdgcn_sdot4(a, b, c, false);
#elif defined(__gfx1100__)
#elif defined(RDNA3)
c = __builtin_amdgcn_sudot4( true, a, true, b, c, false);
#elif defined(__gfx1010__) || defined(__gfx900__)
int tmp1;