Clang format, Add gfx1101, gfx1102 support of FMHA example

example/32_batched_gemm_scale_softmax_gemm/CMakeLists.txt

@@ -5,7 +5,7 @@ add_example_executable(example_batched_gemm_scale_softmax_gemm_permute_xdl_bf16
 add_example_executable(example_grouped_gemm_scale_softmax_gemm_permute_xdl_fp16 grouped_gemm_scale_softmax_gemm_permute_xdl_fp16.cpp)
 add_example_executable(example_batched_gemm_lower_triangle_scale_softmax_gemm_permute_xdl_fp16 batched_gemm_lower_triangle_scale_softmax_gemm_permute_xdl_fp16.cpp)
 add_example_executable(example_grouped_gemm_lower_triangle_scale_softmax_gemm_permute_xdl_fp16 grouped_gemm_lower_triangle_scale_softmax_gemm_permute_xdl_fp16.cpp)
-if(GPU_TARGETS MATCHES "gfx1100")
+if(GPU_TARGETS MATCHES "gfx1100" OR GPU_TARGETS MATCHES "gfx1101" OR GPU_TARGETS MATCHES "gfx1102")
     add_example_executable(example_batched_gemm_lower_triangle_scale_softmax_gemm_permute_wmma_fp16 batched_gemm_lower_triangle_scale_softmax_gemm_permute_wmma_fp16.cpp)
     add_example_executable(example_batched_gemm_scale_softmax_gemm_permute_wmma_fp16 batched_gemm_scale_softmax_gemm_permute_wmma_fp16.cpp)
 endif()
@@ -19,7 +19,7 @@ add_dependencies(example_gemm_scale_softmax_gemm example_grouped_gemm_scale_soft
 add_dependencies(example_gemm_scale_softmax_gemm example_batched_gemm_lower_triangle_scale_softmax_gemm_permute_xdl_fp16)
 add_dependencies(example_gemm_scale_softmax_gemm example_grouped_gemm_lower_triangle_scale_softmax_gemm_permute_xdl_fp16)
 
-if(GPU_TARGETS MATCHES "gfx1100")
+if(GPU_TARGETS MATCHES "gfx1100" OR GPU_TARGETS MATCHES "gfx1101" OR GPU_TARGETS MATCHES "gfx1102")
     add_custom_target(example_gemm_scale_softmax_gemm_wmma)
     add_dependencies(example_gemm_scale_softmax_gemm_wmma example_batched_gemm_lower_triangle_scale_softmax_gemm_permute_wmma_fp16)
     add_dependencies(example_gemm_scale_softmax_gemm_wmma example_batched_gemm_scale_softmax_gemm_permute_wmma_fp16)

include/ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_d_wmma_cshuffle.hpp

@@ -180,7 +180,6 @@ struct DeviceGroupedConvFwdMultipleD_Wmma_CShuffle
     static constexpr auto AEnableLds = AEnableLds_auto || AEnableLds_manu;
     static constexpr auto BEnableLds = BEnableLds_auto || BEnableLds_manu;
 
-
     static constexpr auto conv_to_gemm_transformer =
         TransformConvFwdToGemm<NDimSpatial, ConvForwardSpecialization>{};
 

script/clang-format-overwrite.sh

-# find . -name deps -prune -o -name build -prune -o -iname '*.h' -o -iname '*.hpp' -o -iname '*.cpp' -o -iname '*.h.in' -o -iname '*.hpp.in' -o -iname '*.cpp.in' -o -iname '*.cl' -o -iname '*.cuh' -o -iname '*.cu' -o -iname '*.inc' | xargs -n 1 -P 16 -I{} -t sh -c 'clang-format-10 -i -style=file {}'
-git status --porcelain | awk '$1 != "D" && (match($2, "\\.cpp|hpp|inc")) {print $2}' | xargs -n 1 -P 16 -I{} -t sh -c 'clang-format-10 -i -style=file {}'
+find . -name deps -prune -o -name build -prune -o -iname '*.h' -o -iname '*.hpp' -o -iname '*.cpp' -o -iname '*.h.in' -o -iname '*.hpp.in' -o -iname '*.cpp.in' -o -iname '*.cl' -o -iname '*.cuh' -o -iname '*.cu' -o -iname '*.inc' | xargs -n 1 -P 16 -I{} -t sh -c 'clang-format-10 -i -style=file {}'
+# git status --porcelain | awk '$1 != "D" && (match($2, "\\.cpp|hpp|inc")) {print $2}' | xargs -n 1 -P 16 -I{} -t sh -c 'clang-format-10 -i -style=file {}'