vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

ggml: fix zero division in ‘dne’ calculation in CUDA COUNT_EQUAL operator when ‘ne’ is small (#10213)

Browse source
This commit is contained in:
SXX 2024-11-09 15:35:46 +08:00 • committed by GitHub
parent e89213492d
commit 5b359bb1e3
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 1 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

2
ggml/src/ggml-cuda/count-equal.cu
View file

@ -44,7 +44,7 @@ void ggml_cuda_count_equal(ggml_backend_cuda_context & ctx, ggml_tensor * dst) {
const int64_t ne = ggml_nelements(src0);
GGML_ASSERT(ne < (1 << 30) && "atomicAdd implementation only supports int");
const int64_t dne = GGML_PAD(ne / (4*nsm), CUDA_COUNT_EQUAL_CHUNK_SIZE);
const int64_t dne = GGML_PAD((ne + 4*nsm - 1) / (4*nsm), CUDA_COUNT_EQUAL_CHUNK_SIZE);
CUDA_CHECK(cudaMemsetAsync(dst_d, 0, ggml_nbytes(dst), stream));