Merge branch 'transpose_opt' of https://github.com/ROCm/composable_kernel into rowwise_opt

4885c38a · aska-0096 · cbf14ee1 · 7c8e92fa · 4885c38a · 4885c38a
Commit 4885c38a authored Sep 03, 2024 by aska-0096
20 changed files
--- a/include/ck_tile/ops/fmha/pipeline/block_fmha_pipeline_qx_ks_vs_custom_policy.hpp
+++ b/include/ck_tile/ops/fmha/pipeline/block_fmha_pipeline_qx_ks_vs_custom_policy.hpp
--- a/include/ck_tile/ops/fmha/pipeline/tile_fmha_traits.hpp
+++ b/include/ck_tile/ops/fmha/pipeline/tile_fmha_traits.hpp
--- a/library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_outelementop_instance.hpp
+++ b/library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_outelementop_instance.hpp
--- a/library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_convscale.hpp
+++ b/library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_convscale.hpp
--- a/library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_convscale_relu.hpp
+++ b/library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_convscale_relu.hpp
--- a/library/include/ck/library/tensor_operation_instance/gpu/permute_scale.hpp
+++ b/library/include/ck/library/tensor_operation_instance/gpu/permute_scale.hpp
--- a/library/include/ck/library/tensor_operation_instance/gpu/permute_scale/device_permute_scale_instances.hpp
+++ b/library/include/ck/library/tensor_operation_instance/gpu/permute_scale/device_permute_scale_instances.hpp
--- a/library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_blockwise_f32_f32_f32_amax.hpp
+++ b/library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_blockwise_f32_f32_f32_amax.hpp
--- a/library/include/ck/library/utility/check_err.hpp
+++ b/library/include/ck/library/utility/check_err.hpp
--- a/library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd_convscale/CMakeLists.txt
+++ b/library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd_convscale/CMakeLists.txt
@@ -3,6 +3,7 @@ set(GROUPED_CONV3D_FWD_CONVSCALE
   xdl/device_grouped_conv3d_fwd_xdl_convscale_ndhwgc_gkzyxc_ndhwgk_f8_instance.cpp
   xdl/device_grouped_conv3d_fwd_xdl_convscale_ndhwgc_gkzyxc_ndhwgk_bf8_instance.cpp
   xdl/device_grouped_conv3d_fwd_xdl_convscale_ndhwgc_gkzyxc_ndhwgk_f8_bf8_instance.cpp
-   xdl/device_grouped_conv3d_fwd_xdl_convscale_ndhwgc_gkzyxc_ndhwgk_bf8_f8_instance.cpp)
+   xdl/device_grouped_conv3d_fwd_xdl_convscale_ndhwgc_gkzyxc_ndhwgk_bf8_f8_instance.cpp
+   xdl/device_grouped_conv3d_fwd_xdl_combconvscale_ndhwgc_gkzyxc_ndhwgk_f8_f8_f32_instance.cpp)
 add_instance_library(device_grouped_conv3d_fwd_convscale_instance ${GROUPED_CONV3D_FWD_CONVSCALE})
--- a/library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd_convscale/xdl/device_grouped_conv3d_fwd_xdl_combconvscale_ndhwgc_gkzyxc_ndhwgk_f8_f8_f32_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd_convscale/xdl/device_grouped_conv3d_fwd_xdl_combconvscale_ndhwgc_gkzyxc_ndhwgk_f8_f8_f32_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd_convscale_relu/CMakeLists.txt
+++ b/library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd_convscale_relu/CMakeLists.txt
--- a/library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd_convscale_relu/xdl/device_grouped_conv3d_fwd_xdl_combconvscale_relu_ndhwgc_gkzyxc_ndhwgk_f8_f8_f32_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd_convscale_relu/xdl/device_grouped_conv3d_fwd_xdl_combconvscale_relu_ndhwgc_gkzyxc_ndhwgk_f8_f8_f32_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd_convscale_relu/xdl/device_grouped_conv3d_fwd_xdl_convscale_relu_ndhwgc_gkzyxc_ndhwgk_f8_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd_convscale_relu/xdl/device_grouped_conv3d_fwd_xdl_convscale_relu_ndhwgc_gkzyxc_ndhwgk_f8_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/permute_scale/CMakeLists.txt
+++ b/library/src/tensor_operation_instance/gpu/permute_scale/CMakeLists.txt
--- a/library/src/tensor_operation_instance/gpu/permute_scale/device_permute_scale_6d_fp32_fp8_instances.cpp
+++ b/library/src/tensor_operation_instance/gpu/permute_scale/device_permute_scale_6d_fp32_fp8_instances.cpp
--- a/library/src/tensor_operation_instance/gpu/reduce/device_reduce_instance_blockwise_f32_f32_f32_amax.cpp
+++ b/library/src/tensor_operation_instance/gpu/reduce/device_reduce_instance_blockwise_f32_f32_f32_amax.cpp
--- a/profiler/include/profiler/profile_grouped_conv_bwd_weight_impl.hpp
+++ b/profiler/include/profiler/profile_grouped_conv_bwd_weight_impl.hpp
--- a/profiler/src/profile_grouped_conv_bwd_weight.cpp
+++ b/profiler/src/profile_grouped_conv_bwd_weight.cpp
--- a/script/convert_miopen_driver_to_profiler.py
+++ b/script/convert_miopen_driver_to_profiler.py