CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-31 08:51:55 +00:00
Code Issues Packages Projects Releases Wiki Activity

cuda : remove nrows_x in mul_mat_q_process_tile (#13325)

Browse Source 
Signed-off-by: Xiaodong Ye <xiaodong.ye@mthreads.com>
This commit is contained in:
R0CKSTAR
2025-05-07 15:48:23 +08:00
committed by GitHub
parent 4773d7a02f
commit 1f73301b63
1 changed files with 4 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
ggml/src/ggml-cuda/mmq.cuh
View File

@@ -2522,7 +2522,7 @@ template <ggml_type type, int mmq_x, int nwarps, bool need_check, bool fixup>
static __device__ __forceinline__ void mul_mat_q_process_tile(
const char * __restrict__ x, const int offset_x, const int * __restrict__ y,
const int * __restrict__ ids_dst, float * __restrict__ dst, float * __restrict__ tmp_fixup,
const int nrows_x, const int stride_row_x, const int ncols_y, const int stride_col_dst,
const int stride_row_x, const int ncols_y, const int stride_col_dst,
const int tile_x_max_i, const int tile_y_max_j, const int kb0_start, const int kb0_stop) {
constexpr int qk = ggml_cuda_type_traits<type>::qk;
@@ -2689,7 +2689,7 @@ static __global__ void mul_mat_q(
constexpr bool fixup = false;
mul_mat_q_process_tile<type, mmq_x, nwarps, need_check, fixup>
(x, offset_x, y + offset_y, ids_dst_shared, dst + offset_dst, tmp_fixup, nrows_x, stride_row_x, ncols_y, stride_col_dst,
(x, offset_x, y + offset_y, ids_dst_shared, dst + offset_dst, tmp_fixup, stride_row_x, ncols_y, stride_col_dst,
tile_x_max_i, tile_y_max_j, 0, ncols_x/qk);
return;
}
@@ -2767,7 +2767,7 @@ static __global__ void mul_mat_q(
constexpr bool fixup = false; // All but (potentially) the last iterations write their data to dst rather than the fixup buffer.
mul_mat_q_process_tile<type, mmq_x, nwarps, need_check, fixup>
(x, offset_x, y + offset_y, ids_dst_shared, dst + offset_dst, tmp_fixup, nrows_x, stride_row_x, ncols_y, stride_col_dst,
(x, offset_x, y + offset_y, ids_dst_shared, dst + offset_dst, tmp_fixup, stride_row_x, ncols_y, stride_col_dst,
tile_x_max_i, tile_y_max_j, kb0_start, kb0_stop);
kbc += blocks_per_ne00;
@@ -2834,7 +2834,7 @@ static __global__ void mul_mat_q(
constexpr bool fixup = true; // Last index writes its data to fixup buffer to avoid data races with other blocks.
mul_mat_q_process_tile<type, mmq_x, nwarps, need_check, fixup>
(x, offset_x, y + offset_y, ids_dst_shared, dst + offset_dst, tmp_fixup, nrows_x, stride_row_x, ncols_y, stride_col_dst,
(x, offset_x, y + offset_y, ids_dst_shared, dst + offset_dst, tmp_fixup, stride_row_x, ncols_y, stride_col_dst,
tile_x_max_i, tile_y_max_j, kb0_start, kb0_stop);
}