ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-25 15:24:39 +00:00
Code Issues Packages Projects Releases Wiki Activity

Addressing (Post Merge) code review comments for PR 1845 (#1883)

Browse Source 
* Addressing code review comments.

* Addressing code review comments.

* Reorganized code for better readability.

* add ck_tile gemms for new types in CI

* fix jenkins syntax

* fix script syntax

* Add the test cases back

* Address the review comments

* Address review comments

* clang format

* Solve the merging issues

* Addressed the comments

* clang format

---------

Co-authored-by: illsilin <Illia.Silin@amd.com>
Co-authored-by: ThomasNing <thomas.ning@amd.com>
Co-authored-by: Adam Osewski <19374865+aosewski@users.noreply.github.com>

[ROCm/composable_kernel commit: 66c5f5b0b6]
This commit is contained in:
kylasa
2025-03-06 11:40:30 -08:00
committed by GitHub
parent 710aa99819
commit 7fbcd06a62
32 changed files with 511 additions and 245 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
include/ck_tile/core/arch/generic_memory_space_atomic.hpp
View File

@@ -361,7 +361,7 @@ CK_TILE_DEVICE void atomic_add_g(T* p_dst, const thread_buffer<T, N>& x)
{
if constexpr(N == 2)
{
atomic_add(c_style_pointer_cast<bf16x2_t*>(p_dst), bit_cast<bf16x2_t>(x));
atomic_add(c_style_pointer_cast<bf16x2_t*>(p_dst), x.template get_as<bf16x2_t>()[I0]);
}
else if constexpr(N == 4)
{