ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-14 18:17:44 +00:00
Code Issues Packages Projects Releases Wiki Activity

bug fix

Browse Source 
[ROCm/composable_kernel commit: d707993933]
This commit is contained in:
Chao Liu
2019-09-15 20:57:07 -05:00
parent d89b9c2e08
commit ff99bc281e
3 changed files with 5 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
composable_kernel/include/kernel_algorithm/gridwise_convolution_implicit_gemm_v4r1_nchw_kcyx_nkhw_padded_lds_double_buffer.hpp
View File

@@ -112,7 +112,7 @@ struct GridwiseConvolutionImplicitGemm_v4r1_nchw_kcyx_nkhw_padded_lds_double_buf
"be violated");
// divide block work by [K, B]
static_assert(K % KPerBlock == 0 && B % BPerBlock == 0 && E % EPerBlock == 0,
static_assert(K % KPerBlock == 0 && B % BPerBlock == 0 && E % (2 * EPerBlock) == 0,
"wrong! cannot divide work evenly among block");
constexpr index_t KBlockWork = K / KPerBlock;

2
composable_kernel/include/kernel_algorithm/gridwise_convolution_implicit_gemm_v4r4_nchw_kcyx_nkhw_padded_lds_double_buffer.hpp
View File

@@ -96,7 +96,7 @@ struct GridwiseConvolutionImplicitGemm_v4r4_nchw_kcyx_nkhw_padded_lds_double_buf
"be violated");
// divide block work by [K, B]
static_assert(K % KPerBlock == 0 && B % BPerBlock == 0 && E % EPerBlock == 0,
static_assert(K % KPerBlock == 0 && B % BPerBlock == 0 && E % (2 * EPerBlock) == 0,
"wrong! cannot divide work evenly among block");
constexpr index_t KBlockWork = K / KPerBlock;