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

Added missing __syncthreads();

Browse source
This commit is contained in:
JohannesGaessler 2023-05-12 21:37:34 +02:00
parent 12fc292ee6
commit 7dc2f57e5e
1 changed files with 1 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

1
ggml-cuda.cu
View file

@ -229,6 +229,7 @@ template <int block_size, int qk, dequantize_kernel_t dequantize_kernel> static
} }
// sum up partial sums and write back result // sum up partial sums and write back result
__syncthreads();
for (int s=block_size/2; s>0; s>>=1) { for (int s=block_size/2; s>0; s>>=1) {
if (tid < s) { if (tid < s) {
tmp[tid] += tmp[tid + s]; tmp[tid] += tmp[tid + s];