CMakeLists.txt

include_directories(BEFORE
    ${PROJECT_SOURCE_DIR}/
    ${PROJECT_SOURCE_DIR}/profiler/include
)

include(gtest)

add_custom_target(tests)

function(add_test_executable TEST_NAME)
    message("adding test ${TEST_NAME}")
    set(result 1)
    if(DEFINED DTYPES)
        foreach(source IN LISTS ARGN)
            set(test 0)
            if((source MATCHES "_fp16" OR source MATCHES "_f16") AND NOT "fp16" IN_LIST DTYPES)
                set(test 1)
            endif()
            if((source MATCHES "_fp32" OR source MATCHES "_f32") AND NOT "fp32" IN_LIST DTYPES)
                set(test 1)
            endif()
            if((source MATCHES "_fp64" OR source MATCHES "_f64") AND NOT "fp64" IN_LIST DTYPES)
                set(test 1)
            endif()
            if((source MATCHES "_fp8" OR source MATCHES "_f8") AND NOT "fp8" IN_LIST DTYPES)
                set(test 1)
            endif()
            if((source MATCHES "_bf8" OR source MATCHES "_bf8") AND NOT "bf8" IN_LIST DTYPES)
                set(test 1)
            endif()
            if((source MATCHES "_bf16" OR source MATCHES "_b16") AND NOT "bf16" IN_LIST DTYPES)
                set(test 1)
            endif()
            if((source MATCHES "_int8" OR source MATCHES "_i8") AND NOT "int8" IN_LIST DTYPES)
                set(test 1)
            endif()
            if(test EQUAL 1)
                message("removing test ${source} ")
                list(REMOVE_ITEM ARGN "${source}")
            endif()
        endforeach()
    endif()

    set(TEST_TARGETS ${SUPPORTED_GPU_TARGETS})

    foreach(source IN LISTS ARGN)
        if(NOT DEFINED DL_KERNELS AND source MATCHES "_dl")
            message("removing dl test ${source} ")
            list(REMOVE_ITEM ARGN "${source}")
        endif()
    endforeach()
    foreach(source IN LISTS ARGN)
        if(NOT TEST_TARGETS MATCHES "gfx9" AND source MATCHES "xdl")
            message("removing xdl test ${source} ")
            list(REMOVE_ITEM ARGN "${source}")
        endif()
    endforeach()
    foreach(source IN LISTS ARGN)
	if(NOT TEST_TARGETS MATCHES "gfx11" AND NOT TEST_TARGETS MATCHES "gfx12" AND source MATCHES "wmma")
            message("removing wmma test ${source} ")
            list(REMOVE_ITEM ARGN "${source}")
        endif()
    endforeach()
    #only continue if there are some source files left on the list
    if(ARGN)
        if(ARGN MATCHES "_xdl")
             list(REMOVE_ITEM TEST_TARGETS gfx900 gfx906 gfx906:xnack- gfx1030 gfx1100 gfx1101 gfx1102 gfx1103 gfx1200 gfx1201 gfx10.3-generic gfx11-generic gfx12-generic)
        elseif(ARGN MATCHES "_wmma")
             list(REMOVE_ITEM TEST_TARGETS gfx900 gfx906 gfx906:xnack- gfx908:xnack+ gfx908:xnack- gfx90a:xnack+ gfx90a:xnack- gfx908 gfx90a gfx940 gfx941 gfx942 gfx1030 gfx950)
        elseif(ARGN MATCHES "_smfmac")
             list(REMOVE_ITEM TEST_TARGETS gfx900 gfx906 gfx906:xnack- gfx1030 gfx1100 gfx1101 gfx1102 gfx1103 gfx908 gfx90a gfx1200 gfx1201 gfx10.3-generic gfx11-generic gfx12-generic)
        endif()
        set_source_files_properties(${ARGN} PROPERTIES LANGUAGE HIP)
        add_executable(${TEST_NAME} ${ARGN})
        set_property(TARGET ${TEST_NAME} PROPERTY HIP_ARCHITECTURES ${TEST_TARGETS} )
        target_link_libraries(${TEST_NAME} PRIVATE getopt::getopt)
        add_test(NAME ${TEST_NAME} COMMAND $<TARGET_FILE:${TEST_NAME}>)
        add_dependencies(tests ${TEST_NAME})
        add_dependencies(check ${TEST_NAME})
        rocm_install(TARGETS ${TEST_NAME} COMPONENT tests)
        set(result 0)
    endif()
    #message("add_test returns ${result}")
    set(result ${result} PARENT_SCOPE)
endfunction()

function(add_gtest_executable TEST_NAME)
    message("adding gtest ${TEST_NAME}")
    set(result 1)
    if(DEFINED DTYPES)
        foreach(source IN LISTS ARGN)
            set(test 0)
            if((source MATCHES "_fp16" OR source MATCHES "_f16") AND NOT "fp16" IN_LIST DTYPES)
                set(test 1)
            endif()
            if((source MATCHES "_fp32" OR source MATCHES "_f32") AND NOT "fp32" IN_LIST DTYPES)
                set(test 1)
            endif()
            if((source MATCHES "_fp64" OR source MATCHES "_f64") AND NOT "fp64" IN_LIST DTYPES)
                set(test 1)
            endif()
            if((source MATCHES "_fp8" OR source MATCHES "_f8") AND NOT "fp8" IN_LIST DTYPES)
                set(test 1)
            endif()
            if((source MATCHES "_bf8" OR source MATCHES "_bf8") AND NOT "bf8" IN_LIST DTYPES)
                set(test 1)
            endif()
            if((source MATCHES "_bf16" OR source MATCHES "_b16") AND NOT "bf16" IN_LIST DTYPES)
                set(test 1)
            endif()
            if((source MATCHES "_int8" OR source MATCHES "_i8") AND NOT "int8" IN_LIST DTYPES)
                set(test 1)
            endif()
            if(test EQUAL 1)
                message("removing gtest ${source} ")
                list(REMOVE_ITEM ARGN "${source}")
            endif()
        endforeach()
    endif()

    set(TEST_TARGETS ${SUPPORTED_GPU_TARGETS})

    foreach(source IN LISTS ARGN)
        if(NOT DEFINED DL_KERNELS AND source MATCHES "_dl")
            message("removing dl test ${source} ")
            list(REMOVE_ITEM ARGN "${source}")
        endif()
    endforeach()

    foreach(source IN LISTS ARGN)
        if(NOT TEST_TARGETS MATCHES "gfx9" AND source MATCHES "xdl")
            message("removing xdl test ${source} ")
            list(REMOVE_ITEM ARGN "${source}")
        endif()
    endforeach()

    foreach(source IN LISTS ARGN)
    if(NOT TEST_TARGETS MATCHES "gfx95" AND source MATCHES "mx_")
        message("removing microscaling test ${source} ")
        list(REMOVE_ITEM ARGN "${source}")
    endif()
    endforeach()

    foreach(source IN LISTS ARGN)
	if(NOT TEST_TARGETS MATCHES "gfx11" AND NOT TEST_TARGETS MATCHES "gfx12" AND source MATCHES "wmma")
            message("removing wmma test ${source} ")
            list(REMOVE_ITEM ARGN "${source}")
        endif()
    endforeach()

    #only continue if there are some source files left on the list
    if(ARGN)
        if(ARGN MATCHES "_xdl")
             list(REMOVE_ITEM TEST_TARGETS gfx900 gfx906 gfx906:xnack- gfx1030 gfx1100 gfx1101 gfx1102 gfx1103 gfx1200 gfx1201 gfx10.3-generic gfx11-generic gfx12-generic)
        elseif(ARGN MATCHES "_wmma")
             list(REMOVE_ITEM TEST_TARGETS gfx900 gfx906 gfx906:xnack- gfx908:xnack+ gfx908:xnack- gfx90a:xnack+ gfx90a:xnack- gfx908 gfx90a gfx940 gfx941 gfx942 gfx1030 gfx950)
        elseif(ARGN MATCHES "_smfmac")
             list(REMOVE_ITEM TEST_TARGETS gfx900 gfx906 gfx906:xnack- gfx1030 gfx1100 gfx1101 gfx1102 gfx1103 gfx908 gfx90a gfx1200 gfx1201 gfx10.3-generic gfx11-generic gfx12-generic)
        endif()
        set_source_files_properties(${ARGN} PROPERTIES LANGUAGE HIP)
        add_executable(${TEST_NAME} ${ARGN})
        set_property(TARGET ${TEST_NAME} PROPERTY HIP_ARCHITECTURES ${TEST_TARGETS} )
        add_dependencies(tests ${TEST_NAME})
        add_dependencies(check ${TEST_NAME})

        # suppress gtest warnings
        target_compile_options(${TEST_NAME} PRIVATE -Wno-global-constructors -Wno-undef)
        target_link_libraries(${TEST_NAME} PRIVATE gtest_main getopt::getopt)
        add_test(NAME ${TEST_NAME} COMMAND $<TARGET_FILE:${TEST_NAME}>)
        rocm_install(TARGETS ${TEST_NAME} COMPONENT tests)
        set(result 0)
    endif()
    #message("add_gtest returns ${result}")
    set(result ${result} PARENT_SCOPE)
endfunction()

add_compile_options(-Wno-c++20-extensions)
add_subdirectory(ck_tile)
add_subdirectory(magic_number_division)
add_subdirectory(space_filling_curve)
add_subdirectory(conv_util)
add_subdirectory(reference_conv_fwd)
add_subdirectory(gemm)
add_subdirectory(gemm_add)
add_subdirectory(gemm_layernorm)
add_subdirectory(gemm_split_k)
add_subdirectory(gemm_universal)
add_subdirectory(gemm_reduce)
add_subdirectory(batched_gemm)
add_subdirectory(batched_gemm_reduce)
add_subdirectory(batched_gemm_gemm)
add_subdirectory(batched_gemm_softmax_gemm)
add_subdirectory(batched_gemm_softmax_gemm_permute)
add_subdirectory(grouped_gemm)
add_subdirectory(reduce)
add_subdirectory(convnd_fwd)
add_subdirectory(convnd_bwd_data)
add_subdirectory(grouped_convnd_fwd)
add_subdirectory(grouped_convnd_bwd_weight)
add_subdirectory(block_to_ctile_map)
add_subdirectory(softmax)
add_subdirectory(normalization_fwd)
add_subdirectory(normalization_bwd_data)
add_subdirectory(normalization_bwd_gamma_beta)
add_subdirectory(data_type)
add_subdirectory(elementwise_normalization)
add_subdirectory(batchnorm)
add_subdirectory(contraction)
add_subdirectory(pool)
add_subdirectory(batched_gemm_multi_d)
add_subdirectory(grouped_convnd_bwd_data)
add_subdirectory(conv_tensor_rearrange)
add_subdirectory(transpose)
add_subdirectory(permute_scale)
add_subdirectory(wrapper)
if(SUPPORTED_GPU_TARGETS MATCHES "gfx11")
    add_subdirectory(wmma_op)
endif()
if(SUPPORTED_GPU_TARGETS MATCHES "gfx942" OR SUPPORTED_GPU_TARGETS MATCHES "gfx950") # smfmac needs ROCm6.2
    add_subdirectory(smfmac_op)
endif()
if(SUPPORTED_GPU_TARGETS MATCHES "gfx950") 
    add_subdirectory(mx_mfma_op)
endif()
add_subdirectory(position_embedding)
add_subdirectory(scatter_gather)