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

library/src/tensor_operation_instance/gpu/CMakeLists.txt

@@ -36,12 +36,27 @@ function(add_instance_library INSTANCE_NAME)
             endif()
         endforeach()
     endif()
+    # Do not build DL instances if DL_KERNELS macro is not set
     foreach(source IN LISTS ARGN)
         if(NOT DEFINED DL_KERNELS AND source MATCHES "_dl")
             message("removing dl instance ${source} ")
             list(REMOVE_ITEM ARGN "${source}")
         endif()
     endforeach()
+    # Do not build XDL instances if gfx9 targets are not on the target list
+    foreach(source IN LISTS ARGN)
+        if(NOT GPU_TARGETS MATCHES "gfx9" AND source MATCHES "_xdl")
+            message("removing xdl instance ${source} ")
+            list(REMOVE_ITEM ARGN "${source}")
+        endif()
+    endforeach()
+    # Do not build WMMA instances if gfx11 targets are not on the target list
+    foreach(source IN LISTS ARGN)
+        if(NOT GPU_TARGETS MATCHES "gfx11" AND source MATCHES "_wmma")
+            message("removing wmma instance ${source} ")
+            list(REMOVE_ITEM ARGN "${source}")
+        endif()
+    endforeach()
     #only continue if there are some source files left on the list
     if(ARGN)
         add_library(${INSTANCE_NAME} OBJECT ${ARGN})
@@ -124,6 +139,26 @@ FOREACH(subdir_path ${dir_list})
             message("Found only dl instances, but DL_KERNELS is not set. Skipping.")
             set(add_inst 0)
         endif()
+        if(("${cmake_instance}" MATCHES "ONLY XDL_KERNELS") AND (NOT GPU_TARGETS MATCHES "gfx9"))
+            message("Found only xdl instances, but gfx9 is not on the targets list. Skipping.")
+            set(add_inst 0)
+        endif()
+        if(("${cmake_instance}" MATCHES "ONLY WMMA_KERNELS") AND (NOT GPU_TARGETS MATCHES "gfx11"))
+            message("Found only wmma instances, but gfx11 is not on the targets list. Skipping.")
+            set(add_inst 0)
+        endif()
+        if(("${cmake_instance}" MATCHES "ONLY XDL_AND_DL_KERNELS") AND (NOT DEFINED DL_KERNELS) AND (NOT GPU_TARGETS MATCHES "gfx9"))
+            message("Found only xdl and dl instances, but gfx9 is not on the targets listand DL_KERNELS is not set. Skipping.")
+            set(add_inst 0)
+        endif()
+        if(("${cmake_instance}" MATCHES "ONLY XDL_AND_WMMA_KERNELS") AND (NOT GPU_TARGETS MATCHES "gfx11") AND (NOT GPU_TARGETS MATCHES "gfx9"))
+            message("Found only xdl and wmma instances, but gfx11 and gfx9 are not on the targets list. Skipping.")
+            set(add_inst 0)
+        endif()
+        if(("${cmake_instance}" MATCHES "XDL_DL_WMMA_KERNELS") AND (NOT GPU_TARGETS MATCHES "gfx11") AND (NOT GPU_TARGETS MATCHES "gfx9") AND (NOT DEFINED DL_KERNELS))
+            message("Found xdl, dl, and wmma instances, but none of those meet the target list. Skipping.")
+            set(add_inst 0)
+        endif()
         if((add_inst EQUAL 1))
             get_filename_component(target_dir ${subdir_path} NAME)
             add_subdirectory(${target_dir})