Commit Graph

3 Commits

Author SHA1 Message Date
Max Podkorytov
c8ff442803 [CK-tile] add more tests for batched transpose testing the rectangular block tile sizes (#2634)
* add failing tests

* swap out and reference

* add constraint assert to transpose input distribution

* test both pipelines with rectangular block tile

* print mismatched indices

* add a smaller failing test for old pipeline

* print grid and block

* fill output before operating on it

* swap m/n tile sizes and make one test pass

* add device syncs

* add one more flipped test case

* flip block tile at host arg init

* fix tiles for lds pipeline

* clang-format

* rename tests

* roll back error check

* remove device syncs

* reduce large test case's size

[ROCm/composable_kernel commit: ab26026835]
2025-08-07 16:51:53 -07:00
Max Podkorytov
b13f01345d remove std::format (#2604)
[ROCm/composable_kernel commit: 0d9439760f]
2025-08-01 19:22:07 -07:00
Max Podkorytov
e18a5f6859 [CK-tile] add gtest for ck-tile batched transpose kernels (#2585)
* add a dummy test file

* add kernel launch logic to the test

* transfer all test cases into gtest params

* factor kernel out into test config

* add load transpose pipeline tests

* add padded tests and skip invalid kernels at runtime

* enum class for pipeline type

* add multiwarp test cases

* fix type

* try to solve the problem

---------

Co-authored-by: ThomasNing <thomas.ning@amd.com>

[ROCm/composable_kernel commit: de0cdb4c31]
2025-07-30 07:31:05 -07:00