mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-14 10:09:41 +00:00

Files

Ville Pietilä 6e0454216d [CK] Disable compilation of problematic bwd weight conv instances for gfx90a (#6343 )

## Motivation

Due to compiler version update, there are test failures in the test
suite `test_grouped_convnd_bwd_weight` when running on `gfx90a`. There
are four failing tests for FP16/BF16 that arise from a single kernel
instance. As the problem is in the current `develop` branch, the test
failures are blocking any PR merges into `develop`. An example of a
failed CI runs is here:
[http://micimaster.amd.com/blue/organizations/jenkins/rocm-libraries-folder%2FComposable%20Kernel/detail/develop/558/pipeline/](http://micimaster.amd.com/blue/organizations/jenkins/rocm-libraries-folder%2FComposable%20Kernel/detail/develop/558/pipeline/).
The underlying compiler problem is potentially the same as described in
#6342 as tests are passing for clang compiler version 20.0 and failing
for clang compiler version 22.0.

## Technical Details

This PR disables the compilation of the problematic bwd weight conv
instance for `gfx90a` by adding a new CMake flag `CK_USE_GFX90A` that
allows us to detect when we are compiling for `gfx90a`. Using the new
CMake flag, compilation of instance
`DeviceGroupedConvBwdWeight_Xdl_CShuffleV3<64, 128, 32, 32, Default, 8,
4, 1, 8, 8, 8, 8, 1, 1, 2>` is disabled for `gfx90a`.

Co-authored-by: Ville Pietilä <>

2026-04-13 13:40:27 +02:00

algorithm

[CK][CK Tile] Improve access for merged groups and remove modulo from xor (#5334 )

2026-03-20 15:45:45 +00:00

arch

[CK Tile] Add Tile Distribution Encoding Calculator (#5515 )

2026-04-13 08:00:31 +00:00

container

CK: Remove 41 commented-out dead code blocks (~200 lines) (#6302 )

2026-04-10 11:17:11 -04:00

numeric

CK: Remove 41 commented-out dead code blocks (~200 lines) (#6302 )

2026-04-10 11:17:11 -04:00

tensor

CK: Remove 41 commented-out dead code blocks (~200 lines) (#6302 )

2026-04-10 11:17:11 -04:00

utility

[CK_TILE] Optimize static_ford and sequence compile-time infrastructure (#5938 )

2026-04-02 15:25:14 -06:00

config.hpp

[CK] Disable compilation of problematic bwd weight conv instances for gfx90a (#6343 )

2026-04-13 13:40:27 +02:00

README.md

introducing ck_tile! (#1216 )

2024-04-15 19:27:12 -05:00

README.md

ck_tile/core

ck_tile/core contains every basic functions and structures to create a GPU kernel using ck_tile. User should only include ck_tile/core.hpp this single header to use all the functionality. Everything is under ck_tile namespace. The coding style under this folder should be similar to std (snake_case for structure/function, Camel for template types...)

algorithm/
    coordinate transform and some other reusable algorithm
arch/
    contains some basic device building block like mma, buffer addressing, etc...
container/
    contains basic container data structure, array/sequence/tuple/...
numeric/
    data type, and data type related math
tensor/
    tensor descriptors and tile level API
utility/
    other utility function for both host/device