tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-07-30 22:23:31 -04:00
Code Issues Packages Projects Releases Wiki Activity

CUDA: fix --split-mode row race condition (#9413)

Browse Source
This commit is contained in:
Johannes Gäßler
2024-09-11 10:22:40 +02:00
committed by GitHub
parent d2b496bff4
commit 5af118efda
2 changed files with 7 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
ggml/src/ggml-cuda/mmq.cuh
View File

@@ -2742,6 +2742,7 @@ struct mmq_args {
int64_t ne00; int64_t ne01; int64_t stride01;
int64_t ne10; int64_t ne11; int64_t stride11;
int64_t ne0;
bool use_stream_k;
};
template<ggml_type type>
@@ -2777,8 +2778,7 @@ static void launch_mul_mat_q(ggml_backend_cuda_context & ctx, const mmq_args & a
const int ntx = (args.ne11 + mmq_x - 1) / mmq_x;
const dim3 block_nums_xy_tiling(nty, ntx, 1);
const bool use_stream_k = cc >= CC_VOLTA && cc < CC_OFFSET_AMD;
if (!use_stream_k) {
if (!args.use_stream_k) {
if (args.ne01 % mmq_y == 0) {
constexpr bool need_check = false;
mul_mat_q<type, mmq_x, MMQ_NWARPS, need_check><<<block_nums_xy_tiling, block_dims, shmem, stream>>>