mirror of
https://github.com/ROCm/composable_kernel.git
synced 2026-04-19 22:39:03 +00:00
e6bcd192d4
* add fp6 data-type and support sync/async dwordx3 load/store * clang-format * pre-commit * 1st commit * default mnk pass ut * fix a distrubution * fix * fix bdram distr * update * pass ut * improve perf * update * clean code * resolve copilot comment * reslove comment * clang-format --------- Co-authored-by: ZheWang <zhewan@amd.com>
50 lines
2.2 KiB
CMake
50 lines
2.2 KiB
CMake
# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
|
|
# SPDX-License-Identifier: MIT
|
|
|
|
set(SUPPORTED_GPUS gfx908 gfx90a gfx942 gfx950)
|
|
|
|
set(has_supported_gpu FALSE)
|
|
foreach(gpu IN LISTS GPU_TARGETS)
|
|
if(gpu IN_LIST SUPPORTED_GPUS)
|
|
set(has_supported_gpu TRUE)
|
|
break()
|
|
endif()
|
|
endforeach()
|
|
|
|
if(has_supported_gpu)
|
|
# NOTE: we turn off undefined-func-template to let source compile without explicit declare function specializations
|
|
# ... because they are auto-generated
|
|
set(EXAMPLE_FLATMM_COMPILE_OPTIONS -Wno-undefined-func-template)
|
|
set(EXAMPLE_MOE_FLATMM_COMPILE_OPTIONS)
|
|
|
|
if(CK_USE_OCP_FP8)
|
|
list(APPEND EXAMPLE_FLATMM_COMPILE_OPTIONS -DCK_TILE_USE_OCP_FP8)
|
|
endif()
|
|
list(APPEND EXAMPLE_FLATMM_COMPILE_OPTIONS "SHELL: -mllvm -greedy-reverse-local-assignment=1")
|
|
add_executable(tile_example_flatmm_basic flatmm_basic.cpp)
|
|
target_compile_options(tile_example_flatmm_basic PRIVATE ${EXAMPLE_FLATMM_COMPILE_OPTIONS})
|
|
|
|
add_executable(tile_example_moe_flatmm moe_flatmm.cpp)
|
|
target_compile_options(tile_example_moe_flatmm PRIVATE ${EXAMPLE_FLATMM_COMPILE_OPTIONS})
|
|
|
|
add_executable(tile_example_grouped_flatmm grouped_flatmm.cpp)
|
|
target_compile_options(tile_example_grouped_flatmm PRIVATE ${EXAMPLE_FLATMM_COMPILE_OPTIONS})
|
|
|
|
if(GPU_TARGETS MATCHES "gfx95" OR GPU_TARGETS MATCHES "gfx94")
|
|
add_executable(tile_example_mixed_prec_flatmm mixed_prec/mixed_prec_flatmm.cpp)
|
|
target_compile_options(tile_example_mixed_prec_flatmm PRIVATE ${EXAMPLE_FLATMM_COMPILE_OPTIONS})
|
|
|
|
add_executable(tile_example_a16w4_moe_flatmm mixed_prec/a16w4_moe_flatmm.cpp)
|
|
target_compile_options(tile_example_a16w4_moe_flatmm PRIVATE ${EXAMPLE_FLATMM_COMPILE_OPTIONS})
|
|
endif()
|
|
if (GPU_TARGETS MATCHES "gfx95")
|
|
include(mxgemm/mx_flatmm_instance.cmake)
|
|
mx_flatmm_instance_generate(EXAMPLE_MX_FLATMM_FILES)
|
|
message(STATUS "Generated MX FlatMM kernel files: ${EXAMPLE_MX_FLATMM_FILES}")
|
|
|
|
add_executable(tile_example_mx_flatmm mxgemm/mx_flatmm.cpp ${EXAMPLE_MX_FLATMM_FILES})
|
|
target_include_directories(tile_example_mx_flatmm PRIVATE mxgemm)
|
|
target_compile_options(tile_example_mx_flatmm PRIVATE ${EXAMPLE_FLATMM_COMPILE_OPTIONS})
|
|
endif()
|
|
endif()
|