tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-08-15 04:33:06 -04:00
Code Issues Packages Projects Releases Wiki Activity

cuda : disable unused cudaLaunchHostFunc code (#6078)

Browse Source
This commit is contained in:
slaren
2024-03-15 13:24:03 +01:00
committed by GitHub
parent 46acb36767
commit 3020327f6c
1 changed files with 3 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
ggml-cuda.cu
View File

@@ -11541,6 +11541,7 @@ static void ggml_backend_cuda_event_wait(ggml_backend_t backend, ggml_backend_ev
if (ggml_backend_is_cuda(event->backend)) { if (ggml_backend_is_cuda(event->backend)) {
CUDA_CHECK(cudaStreamWaitEvent(g_cudaStreams[cuda_ctx->device][0], (cudaEvent_t)event->context, 0)); CUDA_CHECK(cudaStreamWaitEvent(g_cudaStreams[cuda_ctx->device][0], (cudaEvent_t)event->context, 0));
} else { } else {
#if 0
// untested // untested
auto wait_fn = [](void * user_data) { auto wait_fn = [](void * user_data) {
ggml_backend_event_t event = (ggml_backend_event_t)user_data; ggml_backend_event_t event = (ggml_backend_event_t)user_data;
@@ -11548,6 +11549,8 @@ static void ggml_backend_cuda_event_wait(ggml_backend_t backend, ggml_backend_ev
}; };
CUDA_CHECK(cudaLaunchHostFunc(g_cudaStreams[cuda_ctx->device][0], wait_fn, event)); CUDA_CHECK(cudaLaunchHostFunc(g_cudaStreams[cuda_ctx->device][0], wait_fn, event));
#endif
GGML_ASSERT(false);
} }
} }