potok/opencv
1
0
Fork 0
mirror of https://github.com/opencv/opencv.git synced 2026-07-29 23:33:05 +04:00
Code Packages Releases Wiki Activity

Make sure pagedAttnAVGemmKernel uses no more than the stack

Browse Source 
Divide FAST_GEMM_MAX_STACKBUF by 2 because there are two AutoBuffers.
This commit is contained in:
Vincent Rabaud
2026-06-01 13:16:58 +02:00
parent 61d49e12f6
commit 9937cc22da
2 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
modules/dnn/src/layers/cpu_kernels/fast_gemm_kernels.default.hpp
+1 -1
View File
@@ -13,7 +13,7 @@
#include <opencv2/core/utility.hpp> // parallel_for_
#define FAST_GEMM_STORAGE (1<<20) // 2^20
#define FAST_GEMM_MAX_STACKBUF (1 << 14)
#define FAST_GEMM_MAX_STACKBUF (1 << 13)
#define FAST_GEMM_F32_MC 64
#define FAST_GEMM_F32_NC 240
modules/dnn/src/layers/cpu_kernels/fast_gemm_kernels.simd.hpp
+1 -1
View File
@@ -13,7 +13,7 @@
#include <opencv2/core/utility.hpp> // parallel_for_
#define FAST_GEMM_STORAGE (1<<20) // 2^20
#define FAST_GEMM_MAX_STACKBUF (1 << 14)
#define FAST_GEMM_MAX_STACKBUF (1 << 13)
#if CV_AVX
#define FAST_GEMM_F32_MC 60