mirror of
https://github.com/ROCm/composable_kernel.git
synced 2026-04-19 22:39:03 +00:00
Split the instances by architecture. (#1223)
* parse examples inside the add_example_executable function * fix the example 64 cmake file * add xdl flag to the gemm_bias_softmax_gemm_permute example * add filtering of tests based on architecture type * enable test_grouped_gemm for gfx9 only * enable test_transpose only for gfx9 * only linnk test_transpose if it gets built * split the gemm instances by architectures * split gemm_bilinear,grouped_conv_bwd_weight instances by targets * split instances by architecture * split grouped_conv instances by architecture * fix clang format * fix the if-else logic in group_conv headers * small fix for grouped convolution instances * fix the grouped conv bwd weight dl instances * fix client examples * only enable client examples 3 and 4 on gfx9 * set the gfx9 macro * make sure the architecture macros are set by cmake * use separate set of xdl/wmma flags for host code * sinmplify the main cmake file * add conv_fwd_bf8 instance declaration
This commit is contained in:
@@ -1,27 +1,29 @@
|
||||
add_custom_target(client_gemm_fastgelu_examples)
|
||||
if(GPU_TARGETS MATCHES "gfx9")
|
||||
add_custom_target(client_gemm_fastgelu_examples)
|
||||
|
||||
add_executable(client_gemm_add_add_fastgelu gemm_add_add_fastgelu.cpp)
|
||||
target_link_libraries(client_gemm_add_add_fastgelu PRIVATE composable_kernel::device_gemm_operations)
|
||||
add_executable(client_gemm_add_add_fastgelu gemm_add_add_fastgelu.cpp)
|
||||
target_link_libraries(client_gemm_add_add_fastgelu PRIVATE composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(client_gemm_add_fastgelu gemm_add_fastgelu.cpp)
|
||||
target_link_libraries(client_gemm_add_fastgelu PRIVATE composable_kernel::device_gemm_operations)
|
||||
add_executable(client_gemm_add_fastgelu gemm_add_fastgelu.cpp)
|
||||
target_link_libraries(client_gemm_add_fastgelu PRIVATE composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(client_gemm_fastgelu gemm_fastgelu.cpp)
|
||||
target_link_libraries(client_gemm_fastgelu PRIVATE composable_kernel::device_gemm_operations)
|
||||
add_executable(client_gemm_fastgelu gemm_fastgelu.cpp)
|
||||
target_link_libraries(client_gemm_fastgelu PRIVATE composable_kernel::device_gemm_operations)
|
||||
|
||||
add_dependencies(client_gemm_fastgelu_examples client_gemm_add_add_fastgelu client_gemm_add_fastgelu
|
||||
add_dependencies(client_gemm_fastgelu_examples client_gemm_add_add_fastgelu client_gemm_add_fastgelu
|
||||
client_gemm_fastgelu)
|
||||
|
||||
add_custom_target(client_gemm_fastgelu_generic_examples)
|
||||
add_custom_target(client_gemm_fastgelu_generic_examples)
|
||||
|
||||
add_executable(client_gemm_add_add_fastgelu_generic gemm_add_add_fastgelu_generic.cpp)
|
||||
target_link_libraries(client_gemm_add_add_fastgelu_generic composable_kernel::device_gemm_operations)
|
||||
add_executable(client_gemm_add_add_fastgelu_generic gemm_add_add_fastgelu_generic.cpp)
|
||||
target_link_libraries(client_gemm_add_add_fastgelu_generic composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(client_gemm_add_fastgelu_generic gemm_add_fastgelu_generic.cpp)
|
||||
target_link_libraries(client_gemm_add_fastgelu_generic PRIVATE composable_kernel::device_gemm_operations)
|
||||
add_executable(client_gemm_add_fastgelu_generic gemm_add_fastgelu_generic.cpp)
|
||||
target_link_libraries(client_gemm_add_fastgelu_generic PRIVATE composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(client_gemm_fastgelu_generic gemm_fastgelu_generic.cpp)
|
||||
target_link_libraries(client_gemm_fastgelu_generic PRIVATE composable_kernel::device_gemm_operations)
|
||||
add_executable(client_gemm_fastgelu_generic gemm_fastgelu_generic.cpp)
|
||||
target_link_libraries(client_gemm_fastgelu_generic PRIVATE composable_kernel::device_gemm_operations)
|
||||
|
||||
add_dependencies(client_gemm_fastgelu_generic_examples client_gemm_add_add_fastgelu_generic
|
||||
add_dependencies(client_gemm_fastgelu_generic_examples client_gemm_add_add_fastgelu_generic
|
||||
client_gemm_add_fastgelu_generic client_gemm_fastgelu_generic)
|
||||
endif()
|
||||
|
||||
@@ -1,5 +1,7 @@
|
||||
add_executable(client_gemm_add_add_layernorm_naive gemm_add_add_layernorm_naive.cpp)
|
||||
target_link_libraries(client_gemm_add_add_layernorm_naive PRIVATE composable_kernel::device_gemm_operations composable_kernel::device_other_operations)
|
||||
if(GPU_TARGETS MATCHES "gfx9")
|
||||
add_executable(client_gemm_add_add_layernorm_naive gemm_add_add_layernorm_naive.cpp)
|
||||
target_link_libraries(client_gemm_add_add_layernorm_naive PRIVATE composable_kernel::device_gemm_operations composable_kernel::device_other_operations)
|
||||
|
||||
add_executable(client_gemm_add_relu_add_layernorm_welford gemm_add_relu_add_layernorm_welford.cpp)
|
||||
target_link_libraries(client_gemm_add_relu_add_layernorm_welford PRIVATE composable_kernel::device_gemm_operations composable_kernel::device_other_operations)
|
||||
add_executable(client_gemm_add_relu_add_layernorm_welford gemm_add_relu_add_layernorm_welford.cpp)
|
||||
target_link_libraries(client_gemm_add_relu_add_layernorm_welford PRIVATE composable_kernel::device_gemm_operations composable_kernel::device_other_operations)
|
||||
endif()
|
||||
|
||||
@@ -1,15 +1,16 @@
|
||||
add_executable(client_contraction_scale_fp32 contraction_scale_fp32.cpp)
|
||||
target_link_libraries(client_contraction_scale_fp32 PRIVATE composable_kernel::device_other_operations composable_kernel::device_contraction_operations composable_kernel::device_gemm_operations)
|
||||
if(GPU_TARGETS MATCHES "gfx9")
|
||||
add_executable(client_contraction_scale_fp32 contraction_scale_fp32.cpp)
|
||||
target_link_libraries(client_contraction_scale_fp32 PRIVATE composable_kernel::device_other_operations composable_kernel::device_contraction_operations composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(client_contraction_bilinear_fp32 contraction_bilinear_fp32.cpp)
|
||||
target_link_libraries(client_contraction_bilinear_fp32 PRIVATE composable_kernel::device_other_operations composable_kernel::device_contraction_operations composable_kernel::device_gemm_operations)
|
||||
add_executable(client_contraction_bilinear_fp32 contraction_bilinear_fp32.cpp)
|
||||
target_link_libraries(client_contraction_bilinear_fp32 PRIVATE composable_kernel::device_other_operations composable_kernel::device_contraction_operations composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(client_contraction_scale_fp64 contraction_scale_fp64.cpp)
|
||||
target_link_libraries(client_contraction_scale_fp64 PRIVATE composable_kernel::device_other_operations composable_kernel::device_contraction_operations composable_kernel::device_gemm_operations)
|
||||
add_executable(client_contraction_scale_fp64 contraction_scale_fp64.cpp)
|
||||
target_link_libraries(client_contraction_scale_fp64 PRIVATE composable_kernel::device_other_operations composable_kernel::device_contraction_operations composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(client_contraction_bilinear_fp64 contraction_bilinear_fp64.cpp)
|
||||
target_link_libraries(client_contraction_bilinear_fp64 PRIVATE composable_kernel::device_other_operations composable_kernel::device_contraction_operations composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(contraction_g1m2n3k1_add_xdl_fp16 contraction_g1m2n3k1_add_xdl_fp16.cpp)
|
||||
target_link_libraries(contraction_g1m2n3k1_add_xdl_fp16 PRIVATE composable_kernel::device_other_operations composable_kernel::device_contraction_operations composable_kernel::device_gemm_operations)
|
||||
add_executable(client_contraction_bilinear_fp64 contraction_bilinear_fp64.cpp)
|
||||
target_link_libraries(client_contraction_bilinear_fp64 PRIVATE composable_kernel::device_other_operations composable_kernel::device_contraction_operations composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(contraction_g1m2n3k1_add_xdl_fp16 contraction_g1m2n3k1_add_xdl_fp16.cpp)
|
||||
target_link_libraries(contraction_g1m2n3k1_add_xdl_fp16 PRIVATE composable_kernel::device_other_operations composable_kernel::device_contraction_operations composable_kernel::device_gemm_operations)
|
||||
endif()
|
||||
|
||||
@@ -1,5 +1,7 @@
|
||||
add_executable(client_grouped_conv2d_fwd grouped_conv2d_fwd.cpp)
|
||||
target_link_libraries(client_grouped_conv2d_fwd PRIVATE composable_kernel::device_conv_operations)
|
||||
if(GPU_TARGETS MATCHES "gfx9")
|
||||
add_executable(client_grouped_conv2d_fwd grouped_conv2d_fwd.cpp)
|
||||
target_link_libraries(client_grouped_conv2d_fwd PRIVATE composable_kernel::device_conv_operations)
|
||||
|
||||
add_executable(client_grouped_conv1d_fwd grouped_conv1d_fwd.cpp)
|
||||
target_link_libraries(client_grouped_conv1d_fwd PRIVATE composable_kernel::device_conv_operations)
|
||||
add_executable(client_grouped_conv1d_fwd grouped_conv1d_fwd.cpp)
|
||||
target_link_libraries(client_grouped_conv1d_fwd PRIVATE composable_kernel::device_conv_operations)
|
||||
endif()
|
||||
@@ -1,5 +1,7 @@
|
||||
add_executable(client_fused_attention fused_attention.cpp)
|
||||
target_link_libraries(client_fused_attention PRIVATE composable_kernel::device_other_operations composable_kernel::device_gemm_operations)
|
||||
if(GPU_TARGETS MATCHES "gfx9")
|
||||
add_executable(client_fused_attention fused_attention.cpp)
|
||||
target_link_libraries(client_fused_attention PRIVATE composable_kernel::device_other_operations composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(client_fused_attention_bias fused_attention_bias.cpp)
|
||||
target_link_libraries(client_fused_attention_bias PRIVATE composable_kernel::device_other_operations composable_kernel::device_gemm_operations)
|
||||
add_executable(client_fused_attention_bias fused_attention_bias.cpp)
|
||||
target_link_libraries(client_fused_attention_bias PRIVATE composable_kernel::device_other_operations composable_kernel::device_gemm_operations)
|
||||
endif()
|
||||
|
||||
@@ -1,22 +1,22 @@
|
||||
if(DTYPES MATCHES "int8" OR NOT DEFINED DTYPES)
|
||||
add_executable(client_conv2d_fwd_bias_tanh_perchannel_quantization conv2d_fwd_bias_tanh_perchannel_quantization.cpp)
|
||||
target_link_libraries(client_conv2d_fwd_bias_tanh_perchannel_quantization PRIVATE composable_kernel::device_conv_operations composable_kernel::device_other_operations composable_kernel::device_gemm_operations)
|
||||
if(GPU_TARGETS MATCHES "gfx9" AND (DTYPES MATCHES "int8" OR NOT DEFINED DTYPES))
|
||||
add_executable(client_conv2d_fwd_bias_tanh_perchannel_quantization conv2d_fwd_bias_tanh_perchannel_quantization.cpp)
|
||||
target_link_libraries(client_conv2d_fwd_bias_tanh_perchannel_quantization PRIVATE composable_kernel::device_conv_operations composable_kernel::device_other_operations composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(client_conv2d_fwd_bias_relu_perchannel_quantization conv2d_fwd_bias_relu_perchannel_quantization.cpp)
|
||||
target_link_libraries(client_conv2d_fwd_bias_relu_perchannel_quantization PRIVATE composable_kernel::device_conv_operations composable_kernel::device_other_operations composable_kernel::device_gemm_operations)
|
||||
add_executable(client_conv2d_fwd_bias_relu_perchannel_quantization conv2d_fwd_bias_relu_perchannel_quantization.cpp)
|
||||
target_link_libraries(client_conv2d_fwd_bias_relu_perchannel_quantization PRIVATE composable_kernel::device_conv_operations composable_kernel::device_other_operations composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(client_conv2d_fwd_bias_tanh_perlayer_quantization conv2d_fwd_bias_tanh_perlayer_quantization.cpp)
|
||||
target_link_libraries(client_conv2d_fwd_bias_tanh_perlayer_quantization PRIVATE composable_kernel::device_conv_operations composable_kernel::device_other_operations composable_kernel::device_gemm_operations)
|
||||
add_executable(client_conv2d_fwd_bias_tanh_perlayer_quantization conv2d_fwd_bias_tanh_perlayer_quantization.cpp)
|
||||
target_link_libraries(client_conv2d_fwd_bias_tanh_perlayer_quantization PRIVATE composable_kernel::device_conv_operations composable_kernel::device_other_operations composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(client_conv2d_fwd_bias_relu_perlayer_quantization conv2d_fwd_bias_relu_perlayer_quantization.cpp)
|
||||
target_link_libraries(client_conv2d_fwd_bias_relu_perlayer_quantization PRIVATE composable_kernel::device_conv_operations composable_kernel::device_other_operations composable_kernel::device_gemm_operations)
|
||||
add_executable(client_conv2d_fwd_bias_relu_perlayer_quantization conv2d_fwd_bias_relu_perlayer_quantization.cpp)
|
||||
target_link_libraries(client_conv2d_fwd_bias_relu_perlayer_quantization PRIVATE composable_kernel::device_conv_operations composable_kernel::device_other_operations composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(client_conv2d_fwd_perchannel_quantization conv2d_fwd_perchannel_quantization.cpp)
|
||||
target_link_libraries(client_conv2d_fwd_perchannel_quantization PRIVATE composable_kernel::device_conv_operations composable_kernel::device_other_operations composable_kernel::device_gemm_operations)
|
||||
add_executable(client_conv2d_fwd_perchannel_quantization conv2d_fwd_perchannel_quantization.cpp)
|
||||
target_link_libraries(client_conv2d_fwd_perchannel_quantization PRIVATE composable_kernel::device_conv_operations composable_kernel::device_other_operations composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(client_conv2d_fwd_perlayer_quantization conv2d_fwd_perlayer_quantization.cpp)
|
||||
target_link_libraries(client_conv2d_fwd_perlayer_quantization PRIVATE composable_kernel::device_conv_operations composable_kernel::device_other_operations composable_kernel::device_gemm_operations)
|
||||
add_executable(client_conv2d_fwd_perlayer_quantization conv2d_fwd_perlayer_quantization.cpp)
|
||||
target_link_libraries(client_conv2d_fwd_perlayer_quantization PRIVATE composable_kernel::device_conv_operations composable_kernel::device_other_operations composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(client_gemm_quantization gemm_quantization.cpp)
|
||||
target_link_libraries(client_gemm_quantization PRIVATE composable_kernel::device_conv_operations composable_kernel::device_other_operations composable_kernel::device_gemm_operations)
|
||||
add_executable(client_gemm_quantization gemm_quantization.cpp)
|
||||
target_link_libraries(client_gemm_quantization PRIVATE composable_kernel::device_conv_operations composable_kernel::device_other_operations composable_kernel::device_gemm_operations)
|
||||
endif()
|
||||
|
||||
@@ -1,5 +1,7 @@
|
||||
add_executable(client_conv3d_bwd_data_fp16 conv3d_bwd_data_fp16.cpp)
|
||||
add_executable(client_conv3d_bwd_data_fp32 conv3d_bwd_data_fp32.cpp)
|
||||
if(GPU_TARGETS MATCHES "gfx9")
|
||||
add_executable(client_conv3d_bwd_data_fp16 conv3d_bwd_data_fp16.cpp)
|
||||
add_executable(client_conv3d_bwd_data_fp32 conv3d_bwd_data_fp32.cpp)
|
||||
|
||||
target_link_libraries(client_conv3d_bwd_data_fp16 PRIVATE composable_kernel::device_conv_operations)
|
||||
target_link_libraries(client_conv3d_bwd_data_fp32 PRIVATE composable_kernel::device_conv_operations)
|
||||
target_link_libraries(client_conv3d_bwd_data_fp16 PRIVATE composable_kernel::device_conv_operations)
|
||||
target_link_libraries(client_conv3d_bwd_data_fp32 PRIVATE composable_kernel::device_conv_operations)
|
||||
endif()
|
||||
|
||||
@@ -1,3 +1,4 @@
|
||||
|
||||
add_executable(client_gemm_add_multiply gemm_add_multiply.cpp)
|
||||
target_link_libraries(client_gemm_add_multiply PRIVATE composable_kernel::device_gemm_operations)
|
||||
if(GPU_TARGETS MATCHES "gfx9")
|
||||
add_executable(client_gemm_add_multiply gemm_add_multiply.cpp)
|
||||
target_link_libraries(client_gemm_add_multiply PRIVATE composable_kernel::device_gemm_operations)
|
||||
endif()
|
||||
|
||||
@@ -1,2 +1,4 @@
|
||||
add_executable(client_grouped_gemm_fastgelu grouped_gemm_fastgelu.cpp)
|
||||
target_link_libraries(client_grouped_gemm_fastgelu PRIVATE composable_kernel::device_gemm_operations)
|
||||
if(GPU_TARGETS MATCHES "gfx9")
|
||||
add_executable(client_grouped_gemm_fastgelu grouped_gemm_fastgelu.cpp)
|
||||
target_link_libraries(client_grouped_gemm_fastgelu PRIVATE composable_kernel::device_gemm_operations)
|
||||
endif()
|
||||
|
||||
@@ -1,4 +1,4 @@
|
||||
if((DTYPES MATCHES "fp8" AND DTYPES MATCHES "fp16") OR NOT DEFINED DTYPES)
|
||||
if(GPU_TARGETS MATCHES "gfx9" AND ((DTYPES MATCHES "fp8" AND DTYPES MATCHES "fp16") OR NOT DEFINED DTYPES))
|
||||
add_executable(client_splitK_gemm splitK_gemm_fp16_f8.cpp)
|
||||
target_link_libraries(client_splitK_gemm PRIVATE composable_kernel::device_gemm_operations)
|
||||
endif()
|
||||
|
||||
@@ -1,2 +1,4 @@
|
||||
add_executable(client_grouped_gemm_fixed_nk_bias_fp16 grouped_gemm_fixed_nk_bias_fp16.cpp)
|
||||
target_link_libraries(client_grouped_gemm_fixed_nk_bias_fp16 PRIVATE composable_kernel::device_gemm_operations)
|
||||
if(GPU_TARGETS MATCHES "gfx9")
|
||||
add_executable(client_grouped_gemm_fixed_nk_bias_fp16 grouped_gemm_fixed_nk_bias_fp16.cpp)
|
||||
target_link_libraries(client_grouped_gemm_fixed_nk_bias_fp16 PRIVATE composable_kernel::device_gemm_operations)
|
||||
endif()
|
||||
|
||||
@@ -1,11 +1,13 @@
|
||||
add_executable(client_grouped_gemm_fixed_nk_fp16 grouped_gemm_fixed_nk_fp16.cpp)
|
||||
target_link_libraries(client_grouped_gemm_fixed_nk_fp16 PRIVATE composable_kernel::device_gemm_operations)
|
||||
if(GPU_TARGETS MATCHES "gfx9")
|
||||
add_executable(client_grouped_gemm_fixed_nk_fp16 grouped_gemm_fixed_nk_fp16.cpp)
|
||||
target_link_libraries(client_grouped_gemm_fixed_nk_fp16 PRIVATE composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(client_grouped_gemm_fixed_nk_fp8 grouped_gemm_fixed_nk_fp8.cpp)
|
||||
target_link_libraries(client_grouped_gemm_fixed_nk_fp8 PRIVATE composable_kernel::device_gemm_operations)
|
||||
add_executable(client_grouped_gemm_fixed_nk_fp8 grouped_gemm_fixed_nk_fp8.cpp)
|
||||
target_link_libraries(client_grouped_gemm_fixed_nk_fp8 PRIVATE composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(client_grouped_gemm_fixed_nk_i8 grouped_gemm_fixed_nk_i8.cpp)
|
||||
target_link_libraries(client_grouped_gemm_fixed_nk_i8 PRIVATE composable_kernel::device_gemm_operations)
|
||||
add_executable(client_grouped_gemm_fixed_nk_i8 grouped_gemm_fixed_nk_i8.cpp)
|
||||
target_link_libraries(client_grouped_gemm_fixed_nk_i8 PRIVATE composable_kernel::device_gemm_operations)
|
||||
|
||||
add_executable(client_grouped_gemm_fixed_nk_bf16 grouped_gemm_fixed_nk_bf16.cpp)
|
||||
target_link_libraries(client_grouped_gemm_fixed_nk_bf16 PRIVATE composable_kernel::device_gemm_operations)
|
||||
add_executable(client_grouped_gemm_fixed_nk_bf16 grouped_gemm_fixed_nk_bf16.cpp)
|
||||
target_link_libraries(client_grouped_gemm_fixed_nk_bf16 PRIVATE composable_kernel::device_gemm_operations)
|
||||
endif()
|
||||
|
||||
@@ -1,3 +1,4 @@
|
||||
if(GPU_TARGETS MATCHES "gfx9")
|
||||
# Fwd scaleadd scaleadd relu
|
||||
add_executable(client_grouped_convnd_fwd_scaleadd_scaleadd_relu_fp32
|
||||
grouped_convnd_fwd_scaleadd_scaleadd_relu/grouped_conv_fwd_scaleadd_scaleadd_relu_fp32.cpp)
|
||||
@@ -46,3 +47,4 @@ target_link_libraries(client_grouped_convnd_fwd_scale_fp16 PRIVATE composable_ke
|
||||
add_executable(client_grouped_convnd_bwd_data_scale_fp16
|
||||
grouped_convnd_bwd_data_scale/grouped_conv_bwd_data_scale_fp16.cpp)
|
||||
target_link_libraries(client_grouped_convnd_bwd_data_scale_fp16 PRIVATE composable_kernel::device_conv_operations)
|
||||
endif()
|
||||
|
||||
@@ -2,9 +2,7 @@ add_executable(client_tensor_transform_using_wrapper tensor_transform_using_wrap
|
||||
target_link_libraries(client_tensor_transform_using_wrapper PRIVATE composable_kernel::device_other_operations)
|
||||
add_executable(client_wrapper_img2col wrapper_img2col.cpp)
|
||||
target_link_libraries(client_wrapper_img2col PRIVATE composable_kernel::device_other_operations)
|
||||
if(GPU_TARGETS MATCHES "gfx908" OR GPU_TARGETS MATCHES "gfx90a" OR
|
||||
GPU_TARGETS MATCHES "gfx940" OR GPU_TARGETS MATCHES "gfx941" OR
|
||||
GPU_TARGETS MATCHES "gfx942")
|
||||
if(GPU_TARGETS MATCHES "gfx9")
|
||||
add_executable(client_wrapper_basic_gemm wrapper_basic_gemm.cpp)
|
||||
target_link_libraries(client_wrapper_basic_gemm PRIVATE composable_kernel::device_other_operations)
|
||||
add_executable(client_wrapper_optimized_gemm wrapper_optimized_gemm.cpp)
|
||||
|
||||
@@ -48,7 +48,10 @@ else()
|
||||
endif()
|
||||
endif()
|
||||
|
||||
find_package(composable_kernel COMPONENTS device_other_operations device_gemm_operations device_conv_operations device_contraction_operations device_reduction_operations)
|
||||
find_package(composable_kernel COMPONENTS device_other_operations device_gemm_operations device_conv_operations device_reduction_operations)
|
||||
if(GPU_TARGETS MATCHES "gfx9")
|
||||
find_package(composable_kernel COMPONENTS device_contraction_operations)
|
||||
endif()
|
||||
find_package(hip REQUIRED PATHS /opt/rocm)
|
||||
message(STATUS "Build with HIP ${hip_VERSION}")
|
||||
|
||||
|
||||
Reference in New Issue
Block a user