mirror of
https://github.com/ROCm/composable_kernel.git
synced 2026-05-04 13:41:24 +00:00
f2cfc6b94e
* Add missing copyright statements * Use ck_tile::host_tensor_descriptor instead of a custom lambda * Refactor use of check_data_type in test classes * Use TEST_SUITE_NAME with TYPED_TEST_SUITE * Remove an unused namespace * Make dim3 const * Add BF8 x BF8 tests for CompV3 in test_gemm_pipeline_kernel_types.hpp * Add F8 x BF8 tests for CompV3 in test_gemm_pipeline_kernel_types.hpp * Add BF16 x I4 tests for CompV3 in test_gemm_pipeline_kernel_types.hpp * Add BF16 x BF16 tests for CompV3 in test_gemm_pipeline_kernel_types.hpp * Add BF8 x I4 tests for CompV3 in test_gemm_pipeline_kernel_types.hpp * Add F8 x I4 tests for CompV3 in test_gemm_pipeline_kernel_types.hpp * Add F16 x I4 tests for CompV3 in test_gemm_pipeline_kernel_types.hpp * Skip failing tests of F16 x I4 for CompV3 with K == 2 * K_Tile * Add missing precision type combinations to CompV4 from CompV3 * Move the INT8 tests around for consistency with KernelTypesCompV3Wmma * Add missing precision type combinations to CompV3Wmma from CompV3 * Remove the basic and universal tests and their dependencies * On __gfx950__, avoid using transposed loading of A with datatype pk_int4_t of B * Use ADataType and BDataType instead of ComputeDataType for WarpGemm * Explicitly set some return types to void * Use more general typenames in InterleavedPKTypeLoader * Add load_interleaved_pk_type.hpp to common.hpp * Use std::is_same_v in load_int4_tile * Add handling of LoadTranspose to load_int4_tile * Factor out common code in several places using load_int4_tile * Add support for pk_int4_t using load_int4_tile * Fix formatting
59 lines
3.4 KiB
CMake
59 lines
3.4 KiB
CMake
# Currently ck_tile_gemm is only built on gfx94/gfx95
|
|
set(EXAMPLE_GEMM_COMPILE_OPTIONS)
|
|
if(CK_USE_OCP_FP8)
|
|
list(APPEND EXAMPLE_GEMM_COMPILE_OPTIONS -DCK_TILE_USE_OCP_FP8)
|
|
endif()
|
|
set(EXAMPLE_GEMM_COMPILE_COMPUTE_V4_OPTIONS)
|
|
if(CK_USE_OCP_FP8)
|
|
list(APPEND EXAMPLE_GEMM_COMPILE_COMPUTE_V4_OPTIONS -DCK_TILE_USE_OCP_FP8)
|
|
endif()
|
|
list(APPEND EXAMPLE_GEMM_COMPILE_COMPUTE_V4_OPTIONS
|
|
-mllvm
|
|
-enable-noalias-to-md-conversion=0
|
|
)
|
|
set(EXAMPLE_GEMM_COMPILE_COMPUTE_ASYNC_OPTIONS ${EXAMPLE_GEMM_COMPILE_COMPUTE_V4_OPTIONS})
|
|
|
|
if(GPU_TARGETS MATCHES "gfx94|gfx95|gfx90a|gfx11|gfx12")
|
|
if(GPU_TARGETS MATCHES "gfx94|gfx95")
|
|
add_gtest_executable(test_ck_tile_gemm_pipeline_mem test_gemm_pipeline_mem.cpp)
|
|
add_gtest_executable(test_ck_tile_gemm_pipeline_compv3 test_gemm_pipeline_compv3.cpp)
|
|
add_gtest_executable(test_ck_tile_gemm_pipeline_compv4 test_gemm_pipeline_compv4.cpp)
|
|
add_gtest_executable(test_ck_tile_gemm_pipeline_persistent test_gemm_pipeline_persistent.cpp)
|
|
add_gtest_executable(test_ck_tile_gemm_pipeline_compv6 test_gemm_pipeline_compv6.cpp)
|
|
|
|
target_compile_options(test_ck_tile_gemm_pipeline_mem PRIVATE ${EXAMPLE_GEMM_COMPILE_OPTIONS})
|
|
target_compile_options(test_ck_tile_gemm_pipeline_compv3 PRIVATE ${EXAMPLE_GEMM_COMPILE_OPTIONS})
|
|
target_compile_options(test_ck_tile_gemm_pipeline_compv4 PRIVATE ${EXAMPLE_GEMM_COMPILE_COMPUTE_V4_OPTIONS})
|
|
target_compile_options(test_ck_tile_gemm_pipeline_persistent PRIVATE ${EXAMPLE_GEMM_COMPILE_OPTIONS})
|
|
target_compile_options(test_ck_tile_gemm_pipeline_compv6 PRIVATE ${EXAMPLE_GEMM_COMPILE_OPTIONS})
|
|
endif()
|
|
|
|
if(GPU_TARGETS MATCHES "gfx95")
|
|
add_gtest_executable(test_ck_tile_gemm_pipeline_comp_async test_gemm_pipeline_comp_async.cpp)
|
|
target_compile_options(test_ck_tile_gemm_pipeline_comp_async PRIVATE ${EXAMPLE_GEMM_COMPILE_COMPUTE_ASYNC_OPTIONS})
|
|
endif()
|
|
|
|
if(GPU_TARGETS MATCHES "gfx11|gfx12")
|
|
# On Radeon devices, build the WMMA version instead
|
|
# Define architecture macros for compile-time detection
|
|
if(GPU_TARGETS MATCHES "gfx12")
|
|
list(APPEND EXAMPLE_GEMM_COMPILE_OPTIONS -DARCH_GFX12)
|
|
list(APPEND EXAMPLE_GEMM_COMPILE_COMPUTE_V4_OPTIONS -DARCH_GFX12)
|
|
elseif(GPU_TARGETS MATCHES "gfx11")
|
|
list(APPEND EXAMPLE_GEMM_COMPILE_OPTIONS -DARCH_GFX11)
|
|
list(APPEND EXAMPLE_GEMM_COMPILE_COMPUTE_V4_OPTIONS -DARCH_GFX11)
|
|
endif()
|
|
|
|
add_gtest_executable(test_ck_tile_gemm_pipeline_mem_wmma test_gemm_pipeline_mem_wmma.cpp)
|
|
add_gtest_executable(test_ck_tile_gemm_pipeline_compv3_wmma test_gemm_pipeline_compv3_wmma.cpp)
|
|
add_gtest_executable(test_ck_tile_gemm_pipeline_compv4_wmma test_gemm_pipeline_compv4_wmma.cpp)
|
|
add_gtest_executable(test_ck_tile_gemm_pipeline_persistent_wmma test_gemm_pipeline_persistent_wmma.cpp)
|
|
target_compile_options(test_ck_tile_gemm_pipeline_mem_wmma PRIVATE ${EXAMPLE_GEMM_COMPILE_OPTIONS})
|
|
target_compile_options(test_ck_tile_gemm_pipeline_compv3_wmma PRIVATE ${EXAMPLE_GEMM_COMPILE_OPTIONS})
|
|
target_compile_options(test_ck_tile_gemm_pipeline_compv4_wmma PRIVATE ${EXAMPLE_GEMM_COMPILE_COMPUTE_V4_OPTIONS})
|
|
target_compile_options(test_ck_tile_gemm_pipeline_persistent_wmma PRIVATE ${EXAMPLE_GEMM_COMPILE_OPTIONS})
|
|
endif()
|
|
else()
|
|
message(DEBUG "Skipping ck_tile_gemm tests for current target test_ck_tile_gemm_pipeline")
|
|
endif()
|