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

Add __restrict__ to dequantize_mul_mat kernels

Browse source
This commit is contained in:
JohannesGaessler 2023-07-08 22:53:43 +02:00
parent 2d55023143
commit b90c80bdbf
1 changed files with 2 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

3
ggml-cuda.cu
View file

@ -1485,7 +1485,8 @@ static __global__ void dequantize_mul_mat_vec(const void * vx, const dfloat * y,
template <dequantize_1_kernel_t dequantize_kernel>
static __global__ void dequantize_mul_mat(
const void * vx, const float * y, float * dst, const int ncols_x, const int nrows_x, const int ncols_y, const int nrows_dst) {
const void * __restrict__ vx, const float * __restrict__ y, float * __restrict__ dst,
const int ncols_x, const int nrows_x, const int ncols_y, const int nrows_dst) {
const int nrows_y = ncols_x;
const int ncols_dst = ncols_y;