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

example/09_convnd_fwd/CMakeLists.txt

-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
-set(target 0)
-foreach(gpu IN LISTS GPU_TARGETS)
- if(gpu IN_LIST gpu_list AND target EQUAL 0)
-      add_example_executable(example_convnd_fwd_xdl_fp32 convnd_fwd_xdl_fp32.cpp)
-      add_example_executable(example_convnd_fwd_xdl_fp16 convnd_fwd_xdl_fp16.cpp)
-      add_example_executable(example_convnd_fwd_xdl_bf16 convnd_fwd_xdl_bf16.cpp)
-      add_example_executable(example_convnd_fwd_xdl_int8 convnd_fwd_xdl_int8.cpp)
-      add_example_executable(example_convnd_fwd_xdl_fp8 convnd_fwd_xdl_fp8.cpp)
-      add_example_executable(example_convnd_fwd_xdl_bf8 convnd_fwd_xdl_bf8.cpp)
-      # FIXME: re-enable this exampe as test when SWDEV-335738 is fixed
-      add_example_executable_no_testing(example_convnd_fwd_xdl_fp64 convnd_fwd_xdl_fp64.cpp)
-   set(target 1)
- endif()
-endforeach()
-
+add_example_executable(example_convnd_fwd_xdl_fp32 convnd_fwd_xdl_fp32.cpp)
+add_example_executable(example_convnd_fwd_xdl_fp16 convnd_fwd_xdl_fp16.cpp)
+add_example_executable(example_convnd_fwd_xdl_bf16 convnd_fwd_xdl_bf16.cpp)
+add_example_executable(example_convnd_fwd_xdl_int8 convnd_fwd_xdl_int8.cpp)
+add_example_executable(example_convnd_fwd_xdl_fp8 convnd_fwd_xdl_fp8.cpp)
+add_example_executable(example_convnd_fwd_xdl_fp64 convnd_fwd_xdl_fp64.cpp)
+add_example_executable(example_convnd_fwd_xdl_bf8 convnd_fwd_xdl_bf8.cpp)
 add_example_executable(example_convnd_fwd_dl_fp16 convnd_fwd_dl_fp16.cpp)
 add_example_executable(example_convnd_fwd_dl_fp32 convnd_fwd_dl_fp32.cpp)
 add_example_executable(example_convnd_fwd_dl_int8 convnd_fwd_dl_int8.cpp)

example/10_convnd_fwd_multiple_d_multiple_reduce/CMakeLists.txt

-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
-set(target 0)
-foreach(gpu IN LISTS GPU_TARGETS)
-    if(gpu IN_LIST gpu_list AND target EQUAL 0)
-        add_custom_target(example_convnd_fwd_reduce_xdl)
+add_custom_target(example_convnd_fwd_reduce_xdl)
+add_example_executable(example_convnd_fwd_max_xdl_int8 convnd_fwd_max_xdl_int8.cpp)
+add_example_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_int8)
 
-        add_example_executable(example_convnd_fwd_max_xdl_int8 convnd_fwd_max_xdl_int8.cpp)
-        add_example_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_int8)
+add_example_executable_no_testing(example_convnd_fwd_max_xdl_bf16 convnd_fwd_max_xdl_bf16.cpp)
+add_example_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_bf16)
 
-        add_example_executable_no_testing(example_convnd_fwd_max_xdl_bf16 convnd_fwd_max_xdl_bf16.cpp)
-        add_example_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_bf16)
+add_example_executable_no_testing(example_convnd_fwd_max_xdl_fp16 convnd_fwd_max_xdl_fp16.cpp)
+add_example_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_fp16)
 
-        add_example_executable_no_testing(example_convnd_fwd_max_xdl_fp16 convnd_fwd_max_xdl_fp16.cpp)
-        add_example_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_fp16)
+add_example_executable(example_convnd_fwd_max_xdl_fp32 convnd_fwd_max_xdl_fp32.cpp)
+add_example_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_fp32)
 
-        add_example_executable(example_convnd_fwd_max_xdl_fp32 convnd_fwd_max_xdl_fp32.cpp)
-        add_example_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_fp32)
-
-        if(USE_BITINT_EXTENSION_INT4)
-            add_example_executable(example_convnd_fwd_max_xdl_int4 convnd_fwd_max_xdl_int4.cpp)
-            add_example_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_int4)
-        endif(USE_BITINT_EXTENSION_INT4)
-        set(target 1)
-    endif()
-endforeach()
+if(USE_BITINT_EXTENSION_INT4)
+    add_example_executable(example_convnd_fwd_max_xdl_int4 convnd_fwd_max_xdl_int4.cpp)
+    add_example_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_int4)
+endif(USE_BITINT_EXTENSION_INT4)

example/14_gemm_quantization/CMakeLists.txt

-# dlops
 add_example_executable(example_gemm_dl_quantization_int8 gemm_dl_quantization_int8.cpp)
-# xdlops
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
-set(target 0)
-foreach(gpu IN LISTS GPU_TARGETS)
- if(gpu IN_LIST gpu_list AND target EQUAL 0)
-   add_example_executable(example_gemm_xdl_bias_relu_quantization_int8 gemm_xdl_bias_relu_quantization_int8.cpp)
-   add_example_executable(example_gemm_xdl_quantization_int8 gemm_xdl_quantization_int8.cpp)
-   set(target 1)
- endif()
-endforeach()
+add_example_executable(example_gemm_xdl_bias_relu_quantization_int8 gemm_xdl_bias_relu_quantization_int8.cpp)
+add_example_executable(example_gemm_xdl_quantization_int8 gemm_xdl_quantization_int8.cpp)

example/16_gemm_multi_d_multi_reduces/CMakeLists.txt

-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
-set(target 0)
-foreach(gpu IN LISTS GPU_TARGETS)
-    if(gpu IN_LIST gpu_list AND target EQUAL 0)
-        add_custom_target(example_gemm_reduce_xdl)
-        add_custom_target(example_gemm_reduce_xdl_max)
-        add_custom_target(example_gemm_reduce_xdl_mean_meansquare)
-        add_custom_target(example_gemm_add_add_mean_meansquare_xdl)
-
-        add_example_executable(example_gemm_max_xdl_fp16 gemm_max_xdl_fp16.cpp)
-        add_example_dependencies(example_gemm_reduce_xdl_max example_gemm_max_xdl_fp16)
-
-        add_example_executable(example_gemm_add_add_mean_meansquare_xdl_fp16 gemm_add_add_mean_meansquare_xdl_fp16.cpp)
-        add_example_dependencies(example_gemm_add_add_mean_meansquare_xdl example_gemm_add_add_mean_meansquare_xdl_fp16)
-
-        add_example_executable(example_gemm_mean_meansquare_xdl_fp16 gemm_mean_meansquare_xdl_fp16.cpp)
-        add_example_dependencies(example_gemm_reduce_xdl_mean_meansquare example_gemm_mean_meansquare_xdl_fp16)
-
-        add_example_executable(example_gemm_max_xdl_int8 gemm_max_xdl_int8.cpp)
-        add_example_dependencies(example_gemm_reduce_xdl_max example_gemm_max_xdl_int8)
-
-        add_example_executable(example_gemm_add_addsquare_xdl_int8 gemm_add_addsquare_xdl_int8.cpp)
-        add_example_dependencies(example_gemm_reduce_xdl_mean_meansquare example_gemm_add_addsquare_xdl_int8)
-
-        add_example_executable(example_gemm_max_xdl_fp32 gemm_max_xdl_fp32.cpp)
-        add_example_dependencies(example_gemm_reduce_xdl_max example_gemm_max_xdl_fp32)
-
-        add_example_executable(example_gemm_mean_meansquare_xdl_fp32 gemm_mean_meansquare_xdl_fp32.cpp)
-        add_example_dependencies(example_gemm_reduce_xdl_mean_meansquare example_gemm_mean_meansquare_xdl_fp32)
-
-        add_example_executable(example_gemm_max_xdl_bf16 gemm_max_xdl_bf16.cpp)
-        add_example_dependencies(example_gemm_reduce_xdl_max example_gemm_max_xdl_bf16)
-
-        add_example_executable(example_gemm_mean_meansquare_xdl_bf16 gemm_mean_meansquare_xdl_bf16.cpp)
-        add_example_dependencies(example_gemm_reduce_xdl_mean_meansquare example_gemm_mean_meansquare_xdl_bf16)
-
-        add_example_dependencies(example_gemm_reduce_xdl
-            example_gemm_reduce_xdl_mean_meansquare
-            example_gemm_reduce_xdl_max
-            example_gemm_add_add_mean_meansquare_xdl)
-
-        if(USE_BITINT_EXTENSION_INT4)
-            add_example_executable(example_gemm_max_xdl_int4 gemm_max_xdl_int4.cpp)
-            add_example_dependencies(example_gemm_reduce_xdl_max example_gemm_max_xdl_int4)
-        endif()
-        set(target 1)
-    endif()
-endforeach()
+add_custom_target(example_gemm_reduce_xdl)
+add_custom_target(example_gemm_reduce_xdl_max)
+add_custom_target(example_gemm_reduce_xdl_mean_meansquare)
+add_custom_target(example_gemm_add_add_mean_meansquare_xdl)
+
+add_example_executable(example_gemm_max_xdl_fp16 gemm_max_xdl_fp16.cpp)
+add_example_dependencies(example_gemm_reduce_xdl_max example_gemm_max_xdl_fp16)
+
+add_example_executable(example_gemm_add_add_mean_meansquare_xdl_fp16 gemm_add_add_mean_meansquare_xdl_fp16.cpp)
+add_example_dependencies(example_gemm_add_add_mean_meansquare_xdl example_gemm_add_add_mean_meansquare_xdl_fp16)
+
+add_example_executable(example_gemm_mean_meansquare_xdl_fp16 gemm_mean_meansquare_xdl_fp16.cpp)
+add_example_dependencies(example_gemm_reduce_xdl_mean_meansquare example_gemm_mean_meansquare_xdl_fp16)
+
+add_example_executable(example_gemm_max_xdl_int8 gemm_max_xdl_int8.cpp)
+add_example_dependencies(example_gemm_reduce_xdl_max example_gemm_max_xdl_int8)
+
+add_example_executable(example_gemm_add_addsquare_xdl_int8 gemm_add_addsquare_xdl_int8.cpp)
+add_example_dependencies(example_gemm_reduce_xdl_mean_meansquare example_gemm_add_addsquare_xdl_int8)
+
+add_example_executable(example_gemm_max_xdl_fp32 gemm_max_xdl_fp32.cpp)
+add_example_dependencies(example_gemm_reduce_xdl_max example_gemm_max_xdl_fp32)
+
+add_example_executable(example_gemm_mean_meansquare_xdl_fp32 gemm_mean_meansquare_xdl_fp32.cpp)
+add_example_dependencies(example_gemm_reduce_xdl_mean_meansquare example_gemm_mean_meansquare_xdl_fp32)
+
+add_example_executable(example_gemm_max_xdl_bf16 gemm_max_xdl_bf16.cpp)
+add_example_dependencies(example_gemm_reduce_xdl_max example_gemm_max_xdl_bf16)
+
+add_example_executable(example_gemm_mean_meansquare_xdl_bf16 gemm_mean_meansquare_xdl_bf16.cpp)
+add_example_dependencies(example_gemm_reduce_xdl_mean_meansquare example_gemm_mean_meansquare_xdl_bf16)
+
+add_example_dependencies(example_gemm_reduce_xdl
+    example_gemm_reduce_xdl_mean_meansquare
+    example_gemm_reduce_xdl_max
+    example_gemm_add_add_mean_meansquare_xdl)
+
+if(USE_BITINT_EXTENSION_INT4)
+    add_example_executable(example_gemm_max_xdl_int4 gemm_max_xdl_int4.cpp)
+    add_example_dependencies(example_gemm_reduce_xdl_max example_gemm_max_xdl_int4)
+endif()

example/17_convnd_bwd_data/CMakeLists.txt

-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
-set(target 0)
-foreach(gpu IN LISTS GPU_TARGETS)
- if(gpu IN_LIST gpu_list AND target EQUAL 0)
-   add_example_executable(example_convnd_bwd_data_xdl_fp16 convnd_bwd_data_xdl_fp16.cpp)
-   if(result EQUAL 0)
-      target_link_libraries(example_convnd_bwd_data_xdl_fp16 PRIVATE utility)
-   endif()
-   set(target 1)
- endif()
-endforeach()
+add_example_executable(example_convnd_bwd_data_xdl_fp16 convnd_bwd_data_xdl_fp16.cpp)
+if(result EQUAL 0)
+    target_link_libraries(example_convnd_bwd_data_xdl_fp16 PRIVATE utility)
+endif()
 
 add_example_executable(example_convnd_bwd_data_dl_fp16 convnd_bwd_data_dl_fp16.cpp)
 if(result EQUAL 0)

example/20_grouped_conv_bwd_weight/CMakeLists.txt

-list(APPEND gpu_list_xdl gfx908 gfx90a gfx940 gfx941 gfx942)
-list(APPEND gpu_list_wmma gfx1100 gfx1101 gfx1102)
-set(target 0)
-foreach(gpu IN LISTS GPU_TARGETS)
-    if(gpu IN_LIST gpu_list_xdl AND target EQUAL 0)
-        add_custom_target(example_grouped_conv_bwd_weight)
-        add_example_executable(example_grouped_conv_bwd_weight_xdl_fp16 grouped_conv_bwd_weight_xdl_fp16.cpp)
-        add_example_dependencies(example_grouped_conv_bwd_weight example_grouped_conv_bwd_weight_xdl_fp16)
+add_custom_target(example_grouped_conv_bwd_weight)
+add_example_executable(example_grouped_conv_bwd_weight_xdl_fp16 grouped_conv_bwd_weight_xdl_fp16.cpp)
+add_example_dependencies(example_grouped_conv_bwd_weight example_grouped_conv_bwd_weight_xdl_fp16)
 
-        add_example_executable(example_grouped_conv_bwd_weight_xdl_bf16 grouped_conv_bwd_weight_xdl_bf16.cpp)
-        add_example_dependencies(example_grouped_conv_bwd_weight example_grouped_conv_bwd_weight_xdl_bf16)
+add_example_executable(example_grouped_conv_bwd_weight_xdl_bf16 grouped_conv_bwd_weight_xdl_bf16.cpp)
+add_example_dependencies(example_grouped_conv_bwd_weight example_grouped_conv_bwd_weight_xdl_bf16)
 
-        add_example_executable(example_grouped_conv_bwd_weight_xdl_fp16_comp_bf8_fp8 grouped_conv_bwd_weight_xdl_fp16_comp_bf8_fp8.cpp)
-        add_example_dependencies(example_grouped_conv_bwd_weight example_grouped_conv_bwd_weight_xdl_fp16_comp_bf8_fp8)
-        set(target 1)
-    endif()
+add_example_executable(example_grouped_conv_bwd_weight_xdl_fp16_comp_bf8_fp8 grouped_conv_bwd_weight_xdl_fp16_comp_bf8_fp8.cpp)
+add_example_dependencies(example_grouped_conv_bwd_weight example_grouped_conv_bwd_weight_xdl_fp16_comp_bf8_fp8)
 
-    if(gpu IN_LIST gpu_list_wmma AND target EQUAL 0)
-        add_custom_target(example_grouped_conv_bwd_weight)
-        add_example_executable(example_grouped_conv_bwd_weight_wmma_fp16 grouped_conv_bwd_weight_wmma_fp16.cpp)
-        add_example_dependencies(example_grouped_conv_bwd_weight example_grouped_conv_bwd_weight_wmma_fp16)
-        set(target 1)
-    endif()
-endforeach()
-
-add_custom_target(example_grouped_conv_bwd_weight_dl)
+add_example_executable(example_grouped_conv_bwd_weight_wmma_fp16 grouped_conv_bwd_weight_wmma_fp16.cpp)
+add_example_dependencies(example_grouped_conv_bwd_weight example_grouped_conv_bwd_weight_wmma_fp16)
 
 add_example_executable(example_grouped_conv_bwd_weight_dl_fp16 grouped_conv_bwd_weight_dl_fp16.cpp)
-add_example_dependencies(example_grouped_conv_bwd_weight_dl example_grouped_conv_bwd_weight_dl_fp16)
+add_example_dependencies(example_grouped_conv_bwd_weight example_grouped_conv_bwd_weight_dl_fp16)

example/21_gemm_layernorm/CMakeLists.txt

-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
-set(target 0)
-foreach(gpu IN LISTS GPU_TARGETS)
- if(gpu IN_LIST gpu_list AND target EQUAL 0)
-   add_example_executable(example_gemm_bias_relu_add_layernorm_xdl_welford_fp16 gemm_bias_relu_add_layernorm_xdl_welford_fp16.cpp)
-   add_example_executable(example_gemm_bias_relu_add_layernorm_xdl_naive_fp16 gemm_bias_relu_add_layernorm_xdl_naive_fp16.cpp)
-   add_example_executable(example_gemm_layernorm_xdl_naive_fp16 gemm_layernorm_xdl_naive_fp16.cpp)
-   add_example_executable(example_gemm_xdl_layernorm_naive_single_kernel_fp16 gemm_xdl_layernorm_naive_single_kernel_fp16.cpp)
-   set(target 1)
- endif()
-endforeach()
-
+add_example_executable(example_gemm_bias_relu_add_layernorm_xdl_welford_fp16 gemm_bias_relu_add_layernorm_xdl_welford_fp16.cpp)
+add_example_executable(example_gemm_bias_relu_add_layernorm_xdl_naive_fp16 gemm_bias_relu_add_layernorm_xdl_naive_fp16.cpp)
+add_example_executable(example_gemm_layernorm_xdl_naive_fp16 gemm_layernorm_xdl_naive_fp16.cpp)
+add_example_executable(example_gemm_xdl_layernorm_naive_single_kernel_fp16 gemm_xdl_layernorm_naive_single_kernel_fp16.cpp)

example/26_contraction/CMakeLists.txt

@@ -4,49 +4,49 @@ add_custom_target(example_contraction_bilinear)
 
 # FP32
 add_example_executable(example_contraction_bilinear_xdl_fp32 contraction_bilinear_xdl_fp32.cpp)
-add_dependencies(example_contraction_bilinear example_contraction_bilinear_xdl_fp32)
+add_example_dependencies(example_contraction_bilinear example_contraction_bilinear_xdl_fp32)
 
 add_example_executable(example_contraction_scale_xdl_fp32 contraction_scale_xdl_fp32.cpp)
-add_dependencies(example_contraction_scale example_contraction_scale_xdl_fp32)
+add_example_dependencies(example_contraction_scale example_contraction_scale_xdl_fp32)
 
 add_example_executable(example_contraction_bilinear_xdl_fp32_compute_bf16 contraction_bilinear_xdl_fp32_compute_bf16.cpp)
-add_dependencies(example_contraction_bilinear example_contraction_bilinear_xdl_fp32_compute_bf16)
+add_example_dependencies(example_contraction_bilinear example_contraction_bilinear_xdl_fp32_compute_bf16)
 
 add_example_executable(example_contraction_scale_xdl_fp32_compute_bf16 contraction_scale_xdl_fp32_compute_bf16.cpp)
-add_dependencies(example_contraction_scale example_contraction_scale_xdl_fp32_compute_bf16)
+add_example_dependencies(example_contraction_scale example_contraction_scale_xdl_fp32_compute_bf16)
 
 add_example_executable(example_contraction_bilinear_xdl_fp32_compute_fp16 contraction_bilinear_xdl_fp32_compute_fp16.cpp)
-add_dependencies(example_contraction_bilinear example_contraction_bilinear_xdl_fp32_compute_fp16)
+add_example_dependencies(example_contraction_bilinear example_contraction_bilinear_xdl_fp32_compute_fp16)
 
 add_example_executable(example_contraction_scale_xdl_fp32_compute_fp16 contraction_scale_xdl_fp32_compute_fp16.cpp)
-add_dependencies(example_contraction_scale example_contraction_scale_xdl_fp32_compute_fp16)
+add_example_dependencies(example_contraction_scale example_contraction_scale_xdl_fp32_compute_fp16)
 
 # FP64
 add_example_executable(example_contraction_bilinear_xdl_fp64 contraction_bilinear_xdl_fp64.cpp)
-add_dependencies(example_contraction_bilinear example_contraction_bilinear_xdl_fp64)
+add_example_dependencies(example_contraction_bilinear example_contraction_bilinear_xdl_fp64)
 
 add_example_executable(example_contraction_scale_xdl_fp64 contraction_scale_xdl_fp64.cpp)
-add_dependencies(example_contraction_scale example_contraction_scale_xdl_fp64)
+add_example_dependencies(example_contraction_scale example_contraction_scale_xdl_fp64)
 
 add_example_executable(example_contraction_bilinear_xdl_fp64_compute_fp32 contraction_bilinear_xdl_fp64_compute_fp32.cpp)
-add_dependencies(example_contraction_bilinear example_contraction_bilinear_xdl_fp64_compute_fp32)
+add_example_dependencies(example_contraction_bilinear example_contraction_bilinear_xdl_fp64_compute_fp32)
 
 add_example_executable(example_contraction_scale_xdl_fp64_compute_fp32 contraction_scale_xdl_fp64_compute_fp32.cpp)
-add_dependencies(example_contraction_scale example_contraction_scale_xdl_fp64_compute_fp32)
+add_example_dependencies(example_contraction_scale example_contraction_scale_xdl_fp64_compute_fp32)
 
 # FP16
 add_example_executable(example_contraction_bilinear_xdl_fp16_compute_fp32 contraction_bilinear_xdl_fp16_compute_fp32.cpp)
-add_dependencies(example_contraction_bilinear example_contraction_bilinear_xdl_fp16_compute_fp32)
+add_example_dependencies(example_contraction_bilinear example_contraction_bilinear_xdl_fp16_compute_fp32)
 
 add_example_executable(example_contraction_scale_xdl_fp16_compute_fp32 contraction_scale_xdl_fp16_compute_fp32.cpp)
-add_dependencies(example_contraction_scale example_contraction_scale_xdl_fp16_compute_fp32)
+add_example_dependencies(example_contraction_scale example_contraction_scale_xdl_fp16_compute_fp32)
 
 # BF16
 add_example_executable(example_contraction_bilinear_xdl_bf16_compute_fp32 contraction_bilinear_xdl_bf16_compute_fp32.cpp)
-add_dependencies(example_contraction_bilinear example_contraction_bilinear_xdl_bf16_compute_fp32)
+add_example_dependencies(example_contraction_bilinear example_contraction_bilinear_xdl_bf16_compute_fp32)
 
 add_example_executable(example_contraction_scale_xdl_bf16_compute_fp32 contraction_scale_xdl_bf16_compute_fp32.cpp)
-add_dependencies(example_contraction_scale example_contraction_scale_xdl_bf16_compute_fp32)
+add_example_dependencies(example_contraction_scale example_contraction_scale_xdl_bf16_compute_fp32)
 
-add_dependencies(example_contraction example_contraction_scale)
-add_dependencies(example_contraction example_contraction_bilinear)
+add_example_dependencies(example_contraction example_contraction_scale)
+add_example_dependencies(example_contraction example_contraction_bilinear)

example/29_batched_gemm_bias_e_permute/CMakeLists.txt

 add_example_executable(example_batched_gemm_bias_e_permute_xdl_fp16 batched_gemm_bias_e_permute_xdl_fp16.cpp)
-
-if(GPU_TARGETS MATCHES "gfx11")
-    add_example_executable(example_batched_gemm_bias_e_permute_wmma_fp16 batched_gemm_bias_e_permute_wmma_fp16.cpp)
-endif()
+add_example_executable(example_batched_gemm_bias_e_permute_wmma_fp16 batched_gemm_bias_e_permute_wmma_fp16.cpp)

example/30_grouped_conv_fwd_multiple_d/CMakeLists.txt

-list(APPEND gpu_list1 gfx908 gfx90a gfx940 gfx941 gfx942)
-list(APPEND gpu_list2 gfx1100 gfx1101 gfx1102)
+add_custom_target(example_grouped_conv_fwd_multiple_d)
+add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_fp16 grouped_conv_fwd_bias_relu_add_xdl_fp16.cpp)
+add_example_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_fp16)
 
-set(target 0)
-foreach(gpu IN LISTS GPU_TARGETS)
-    if(gpu IN_LIST gpu_list1 AND target EQUAL 0)
-        add_custom_target(example_grouped_conv_fwd_multiple_d)
+add_example_executable(example_grouped_conv_fwd_xdl_fp16 grouped_conv_fwd_xdl_fp16.cpp)
+add_example_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_xdl_fp16)
 
-        add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_fp16 grouped_conv_fwd_bias_relu_add_xdl_fp16.cpp)
-        add_example_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_fp16)
+add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_fp32 grouped_conv_fwd_bias_relu_add_xdl_fp32.cpp)
+add_example_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_fp32)
 
-        add_example_executable(example_grouped_conv_fwd_xdl_fp16 grouped_conv_fwd_xdl_fp16.cpp)
-        add_example_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_xdl_fp16)
+add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_bf16 grouped_conv_fwd_bias_relu_add_xdl_bf16.cpp)
+add_example_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_bf16)
 
-        add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_fp32 grouped_conv_fwd_bias_relu_add_xdl_fp32.cpp)
-        add_example_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_fp32)
+add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_int8 grouped_conv_fwd_bias_relu_add_xdl_int8.cpp)
+add_example_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_int8)
 
-        add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_bf16 grouped_conv_fwd_bias_relu_add_xdl_bf16.cpp)
-        add_example_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_bf16)
+if(USE_BITINT_EXTENSION_INT4)
+    add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_int4 grouped_conv_fwd_bias_relu_add_xdl_int4.cpp)
+    add_example_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_int4)
+endif() # USE_BITINT_EXTENSION_INT4
 
-        add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_int8 grouped_conv_fwd_bias_relu_add_xdl_int8.cpp)
-        add_example_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_int8)
-
-        if(USE_BITINT_EXTENSION_INT4)
-            add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_int4 grouped_conv_fwd_bias_relu_add_xdl_int4.cpp)
-            add_example_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_int4)
-        endif() # USE_BITINT_EXTENSION_INT4
-
-        set(target 1)
-    endif()
-endforeach()
-
-set(target 0)
-foreach(gpu IN LISTS GPU_TARGETS)
-    if(gpu IN_LIST gpu_list2 AND target EQUAL 0)
-        add_example_executable(example_grouped_conv_fwd_bias_relu_add_wmma_fp16 grouped_conv_fwd_bias_relu_add_wmma_fp16.cpp)
-        add_example_executable(example_grouped_conv_fwd_bias_relu_add_wmma_int8 grouped_conv_fwd_bias_relu_add_wmma_int8.cpp)
-        set(target 1)
-    endif()
-endforeach()
+add_example_executable(example_grouped_conv_fwd_bias_relu_add_wmma_fp16 grouped_conv_fwd_bias_relu_add_wmma_fp16.cpp)
+add_example_executable(example_grouped_conv_fwd_bias_relu_add_wmma_int8 grouped_conv_fwd_bias_relu_add_wmma_int8.cpp)

example/31_batched_gemm_gemm/CMakeLists.txt

-list(APPEND gpu_list1 gfx908 gfx90a gfx940 gfx941 gfx942)
-
-set(target 0)
-foreach(gpu IN LISTS GPU_TARGETS)
- if(gpu IN_LIST gpu_list1 AND target EQUAL 0)
-   add_example_executable(example_batched_gemm_gemm_xdl_fp32 batched_gemm_gemm_xdl_fp32.cpp)
-   add_example_executable(example_batched_gemm_gemm_xdl_fp16 batched_gemm_gemm_xdl_fp16.cpp)
-   add_example_executable(example_batched_gemm_gemm_xdl_bf16 batched_gemm_gemm_xdl_bf16.cpp)
-   if(USE_BITINT_EXTENSION_INT4)
-      add_example_executable(example_batched_gemm_gemm_xdl_int4 batched_gemm_gemm_xdl_int4.cpp)
-   endif(USE_BITINT_EXTENSION_INT4)
-   set(target 1)
- endif()
-endforeach()
+add_example_executable(example_batched_gemm_gemm_xdl_fp32 batched_gemm_gemm_xdl_fp32.cpp)
+add_example_executable(example_batched_gemm_gemm_xdl_fp16 batched_gemm_gemm_xdl_fp16.cpp)
+add_example_executable(example_batched_gemm_gemm_xdl_bf16 batched_gemm_gemm_xdl_bf16.cpp)
+if(USE_BITINT_EXTENSION_INT4)
+   add_example_executable(example_batched_gemm_gemm_xdl_int4 batched_gemm_gemm_xdl_int4.cpp)
+endif(USE_BITINT_EXTENSION_INT4)
 
 if(NOT GPU_TARGETS MATCHES "gfx94" AND NOT GPU_TARGETS MATCHES "gfx1")
    add_example_executable(example_batched_gemm_gemm_xdl_int8 batched_gemm_gemm_xdl_int8.cpp)

example/32_batched_gemm_scale_softmax_gemm/CMakeLists.txt

-if(GPU_TARGETS MATCHES "gfx11")
-    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)
-    add_example_executable(example_self_attention_forward_wmma_fp16 self_attention_forward_wmma_fp16.cpp)
-    add_example_executable(example_cross_attention_forward_wmma_fp16 cross_attention_forward_wmma_fp16.cpp)
-    add_example_executable(example_multi_query_attention_forward_wmma_fp16 multi_query_attention_forward_wmma_fp16.cpp)
-    add_example_executable(example_grouped_query_attention_forward_wmma_fp16 grouped_query_attention_forward_wmma_fp16.cpp)
-endif()
+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)
+add_example_executable(example_self_attention_forward_wmma_fp16 self_attention_forward_wmma_fp16.cpp)
+add_example_executable(example_cross_attention_forward_wmma_fp16 cross_attention_forward_wmma_fp16.cpp)
+add_example_executable(example_multi_query_attention_forward_wmma_fp16 multi_query_attention_forward_wmma_fp16.cpp)
+add_example_executable(example_grouped_query_attention_forward_wmma_fp16 grouped_query_attention_forward_wmma_fp16.cpp)
 
 add_custom_target(example_gemm_scale_softmax_gemm)
 

example/35_splitK_gemm/CMakeLists.txt

-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
-set(target 0)
-foreach(gpu IN LISTS GPU_TARGETS)
- if(gpu IN_LIST gpu_list AND target EQUAL 0)
-   add_custom_target(example_splitK_gemm_xdl)
+add_custom_target(example_splitK_gemm_xdl)
+add_example_executable(example_splitK_gemm_xdl_fp32 splitK_gemm_xdl_fp32.cpp)
+add_example_dependencies(example_splitK_gemm_xdl example_splitK_gemm_xdl_fp32)
 
-   add_example_executable(example_splitK_gemm_xdl_fp32 splitK_gemm_xdl_fp32.cpp)
-   add_example_dependencies(example_splitK_gemm_xdl example_splitK_gemm_xdl_fp32)
+add_example_executable(example_splitK_gemm_xdl_fp16 splitK_gemm_xdl_fp16.cpp)
+add_example_dependencies(example_splitK_gemm_xdl example_splitK_gemm_xdl_fp16)
 
-   add_example_executable(example_splitK_gemm_xdl_fp16 splitK_gemm_xdl_fp16.cpp)
-   add_example_dependencies(example_splitK_gemm_xdl example_splitK_gemm_xdl_fp16)
+add_example_executable(example_splitK_gemm_xdl_fp16_fp8 splitK_gemm_xdl_fp16_fp8.cpp)
+add_example_dependencies(example_splitK_gemm_xdl example_splitK_gemm_xdl_fp16_fp8)
 
-   add_example_executable(example_splitK_gemm_xdl_fp16_fp8 splitK_gemm_xdl_fp16_fp8.cpp)
-   add_example_dependencies(example_splitK_gemm_xdl example_splitK_gemm_xdl_fp16_fp8)
+add_example_executable(example_splitK_gemm_xdl_lds_direct_load_fp16 splitK_gemm_xdl_lds_direct_load_fp16.cpp)
+add_example_dependencies(example_splitK_gemm_xdl example_splitK_gemm_xdl_lds_direct_load_fp16)
 
-   add_example_executable(example_splitK_gemm_xdl_lds_direct_load_fp16 splitK_gemm_xdl_lds_direct_load_fp16.cpp)
-   add_example_dependencies(example_splitK_gemm_xdl example_splitK_gemm_xdl_lds_direct_load_fp16)
+add_example_executable(example_splitK_gemm_xdl_bf16 splitK_gemm_xdl_bf16.cpp)
+add_example_dependencies(example_splitK_gemm_xdl example_splitK_gemm_xdl_bf16)
 
-   add_example_executable(example_splitK_gemm_xdl_bf16 splitK_gemm_xdl_bf16.cpp)
-   add_example_dependencies(example_splitK_gemm_xdl example_splitK_gemm_xdl_bf16)
+add_example_executable(example_splitK_gemm_xdl_int8 splitK_gemm_xdl_int8.cpp)
+add_example_dependencies(example_splitK_gemm_xdl example_splitK_gemm_xdl_int8)
 
-   add_example_executable(example_splitK_gemm_xdl_int8 splitK_gemm_xdl_int8.cpp)
-   add_example_dependencies(example_splitK_gemm_xdl example_splitK_gemm_xdl_int8)
-
-   if(USE_BITINT_EXTENSION_INT4)
-       add_example_executable(example_splitK_gemm_xdl_int4 splitK_gemm_xdl_int4.cpp)
-       add_example_dependencies(example_splitK_gemm_xdl example_splitK_gemm_xdl_int4)
-   endif()
-
-   set(target 1)
- endif()
-endforeach()
+if(USE_BITINT_EXTENSION_INT4)
+    add_example_executable(example_splitK_gemm_xdl_int4 splitK_gemm_xdl_int4.cpp)
+    add_example_dependencies(example_splitK_gemm_xdl example_splitK_gemm_xdl_int4)
+endif()

example/38_grouped_conv_bwd_data_multiple_d/CMakeLists.txt

-list(APPEND gpu_list_xdl gfx908 gfx90a gfx940 gfx941 gfx942)
-list(APPEND gpu_list_wmma gfx1100 gfx1101 gfx1102)
-set(target 0)
-foreach(gpu IN LISTS GPU_TARGETS)
-    if(gpu IN_LIST gpu_list_xdl AND target EQUAL 0)
-        add_custom_target(example_grouped_conv_bwd_data)
+add_custom_target(example_grouped_conv_bwd_data)
 
-        add_example_executable(example_grouped_conv_bwd_data_xdl_fp16 grouped_conv_bwd_data_xdl_fp16.cpp)
-        add_example_dependencies(example_grouped_conv_bwd_data example_grouped_conv_bwd_data_xdl_fp16)
+add_example_executable(example_grouped_conv_bwd_data_xdl_fp16 grouped_conv_bwd_data_xdl_fp16.cpp)
+add_example_dependencies(example_grouped_conv_bwd_data example_grouped_conv_bwd_data_xdl_fp16)
 
-        add_example_executable(example_grouped_conv_bwd_data_bias_relu_xdl_fp16 grouped_conv_bwd_data_bias_relu_xdl_fp16.cpp)
-        add_example_dependencies(example_grouped_conv_bwd_data example_grouped_conv_bwd_data_bias_relu_xdl_fp16)
+add_example_executable(example_grouped_conv_bwd_data_bias_relu_xdl_fp16 grouped_conv_bwd_data_bias_relu_xdl_fp16.cpp)
+add_example_dependencies(example_grouped_conv_bwd_data example_grouped_conv_bwd_data_bias_relu_xdl_fp16)
 
-        set(target 1)
-    endif()
-endforeach()
-
-foreach(gpu IN LISTS GPU_TARGETS)
-    if(gpu IN_LIST gpu_list_wmma AND target EQUAL 0)
-        add_custom_target(example_grouped_conv_bwd_data)
-
-        add_example_executable(example_grouped_conv_bwd_data_wmma_fp16 grouped_conv_bwd_data_wmma_fp16.cpp)
-        add_example_dependencies(example_grouped_conv_bwd_data example_grouped_conv_bwd_data_wmma_fp16)
-
-        set(target 1)
-    endif()
-endforeach()
+add_example_executable(example_grouped_conv_bwd_data_wmma_fp16 grouped_conv_bwd_data_wmma_fp16.cpp)
+add_example_dependencies(example_grouped_conv_bwd_data example_grouped_conv_bwd_data_wmma_fp16)

example/40_conv2d_fwd_quantization/CMakeLists.txt

-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
-set(target 0)
-foreach(gpu IN LISTS GPU_TARGETS)
- if(gpu IN_LIST gpu_list AND target EQUAL 0)
-   add_example_executable(example_conv2d_fwd_xdl_perlayer_quantization_int8 conv2d_fwd_xdl_perlayer_quantization_int8.cpp)
-   add_example_executable(example_conv2d_fwd_xdl_perchannel_quantization_int8 conv2d_fwd_xdl_perchannel_quantization_int8.cpp)
-   add_example_executable(example_conv2d_fwd_xdl_bias_relu_perlayer_quantization_int8 conv2d_fwd_xdl_bias_relu_perlayer_quantization_int8.cpp)
-   add_example_executable(example_conv2d_fwd_xdl_bias_relu_perchannel_quantization_int8 conv2d_fwd_xdl_bias_relu_perchannel_quantization_int8.cpp)
-   set(target 1)
- endif()
-endforeach()
+add_example_executable(example_conv2d_fwd_xdl_perlayer_quantization_int8 conv2d_fwd_xdl_perlayer_quantization_int8.cpp)
+add_example_executable(example_conv2d_fwd_xdl_perchannel_quantization_int8 conv2d_fwd_xdl_perchannel_quantization_int8.cpp)
+add_example_executable(example_conv2d_fwd_xdl_bias_relu_perlayer_quantization_int8 conv2d_fwd_xdl_bias_relu_perlayer_quantization_int8.cpp)
+add_example_executable(example_conv2d_fwd_xdl_bias_relu_perchannel_quantization_int8 conv2d_fwd_xdl_bias_relu_perchannel_quantization_int8.cpp)
 
-  # Conv perlayer quantization
-  add_example_executable(example_conv2d_fwd_dl_perlayer_quantization_int8 conv2d_fwd_dl_perlayer_quantization_int8.cpp)
-  # Conv perchannel quantization
-  add_example_executable(example_conv2d_fwd_dl_perchannel_quantization_int8 conv2d_fwd_dl_perchannel_quantization_int8.cpp)
-  # Conv + bias + relu perlayer quantization
-  add_example_executable(example_conv2d_fwd_dl_bias_relu_perlayer_quantization_int8 conv2d_fwd_dl_bias_relu_perlayer_quantization_int8.cpp)
-  # Conv + bias + relu perchannel quantization
-  add_example_executable(example_conv2d_fwd_dl_bias_relu_perchannel_quantization_int8 conv2d_fwd_dl_bias_relu_perchannel_quantization_int8.cpp)
-  # Conv + bias + tanh perlayer quantization
-  add_example_executable(example_conv2d_fwd_dl_bias_tanh_perlayer_quantization_int8 conv2d_fwd_dl_bias_tanh_perlayer_quantization_int8.cpp)
-  # Conv + bias + tanh perchannel quantization
-  add_example_executable(example_conv2d_fwd_dl_bias_tanh_perchannel_quantization_int8 conv2d_fwd_dl_bias_tanh_perchannel_quantization_int8.cpp)
+# Conv perlayer quantization
+add_example_executable(example_conv2d_fwd_dl_perlayer_quantization_int8 conv2d_fwd_dl_perlayer_quantization_int8.cpp)
+# Conv perchannel quantization
+add_example_executable(example_conv2d_fwd_dl_perchannel_quantization_int8 conv2d_fwd_dl_perchannel_quantization_int8.cpp)
+# Conv + bias + relu perlayer quantization
+add_example_executable(example_conv2d_fwd_dl_bias_relu_perlayer_quantization_int8 conv2d_fwd_dl_bias_relu_perlayer_quantization_int8.cpp)
+# Conv + bias + relu perchannel quantization
+add_example_executable(example_conv2d_fwd_dl_bias_relu_perchannel_quantization_int8 conv2d_fwd_dl_bias_relu_perchannel_quantization_int8.cpp)
+# Conv + bias + tanh perlayer quantization
+add_example_executable(example_conv2d_fwd_dl_bias_tanh_perlayer_quantization_int8 conv2d_fwd_dl_bias_tanh_perlayer_quantization_int8.cpp)
+# Conv + bias + tanh perchannel quantization
+add_example_executable(example_conv2d_fwd_dl_bias_tanh_perchannel_quantization_int8 conv2d_fwd_dl_bias_tanh_perchannel_quantization_int8.cpp)

example/41_grouped_conv_conv_fwd/CMakeLists.txt

-list(APPEND gpu_list1 gfx908 gfx90a gfx940 gfx941 gfx942)
-list(APPEND gpu_list2 gfx908 gfx90a)
-set(target 0)
-foreach(gpu IN LISTS GPU_TARGETS)
- if(gpu IN_LIST gpu_list1 AND target EQUAL 0)
-   add_example_executable(example_grouped_conv_conv_fwd_xdl_fp32 grouped_conv_conv_fwd_xdl_fp32.cpp)
-   add_example_executable(example_grouped_conv_conv_fwd_xdl_fp16 grouped_conv_conv_fwd_xdl_fp16.cpp)
-   add_example_executable(example_grouped_conv_conv_fwd_xdl_bf16 grouped_conv_conv_fwd_xdl_bf16.cpp)
-   if(USE_BITINT_EXTENSION_INT4)
-      add_example_executable(example_grouped_conv_conv_fwd_xdl_int4 grouped_conv_conv_fwd_xdl_int4.cpp)
-   endif(USE_BITINT_EXTENSION_INT4)
-   set(target 1)
- endif()
-endforeach()
+add_example_executable(example_grouped_conv_conv_fwd_xdl_fp32 grouped_conv_conv_fwd_xdl_fp32.cpp)
+add_example_executable(example_grouped_conv_conv_fwd_xdl_fp16 grouped_conv_conv_fwd_xdl_fp16.cpp)
+add_example_executable(example_grouped_conv_conv_fwd_xdl_bf16 grouped_conv_conv_fwd_xdl_bf16.cpp)
+if(USE_BITINT_EXTENSION_INT4)
+   add_example_executable(example_grouped_conv_conv_fwd_xdl_int4 grouped_conv_conv_fwd_xdl_int4.cpp)
+endif(USE_BITINT_EXTENSION_INT4)
 
 if(NOT GPU_TARGETS MATCHES "gfx94" AND NOT GPU_TARGETS MATCHES "gfx1")
    add_example_executable(example_grouped_conv_conv_fwd_xdl_int8 grouped_conv_conv_fwd_xdl_int8.cpp)

example/47_gemm_bias_softmax_gemm_permute/CMakeLists.txt

-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
-set(target 0)
-foreach(gpu IN LISTS GPU_TARGETS)
- if(gpu IN_LIST gpu_list AND target EQUAL 0)
-   add_example_executable(example_gemm_bias_softmax_gemm_permute gemm_bias_softmax_gemm_permute.cpp)
-   set(target 1)
- endif()
-endforeach()
+add_example_executable(example_gemm_bias_softmax_gemm_permute gemm_bias_softmax_gemm_permute_xdl.cpp)

example/52_im2col_col2im/CMakeLists.txt

-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
-set(target 0)
-foreach(gpu IN LISTS GPU_TARGETS)
-    if(gpu IN_LIST gpu_list AND target EQUAL 0)
-        add_custom_target(example_im2col_col2im)
+add_custom_target(example_im2col_col2im)
 
-        add_example_executable(example_image_to_column_f32 image_to_column_f32.cpp)
-        add_example_dependencies(example_im2col_col2im example_image_to_column_f32)
+add_example_executable(example_image_to_column_f32 image_to_column_f32.cpp)
+add_example_dependencies(example_im2col_col2im example_image_to_column_f32)
 
-        add_example_executable(example_column_to_image_f32 column_to_image_f32.cpp)
-        add_example_dependencies(example_im2col_col2im example_column_to_image_f32)
-
-        set(target 1)
-    endif()
-endforeach()
+add_example_executable(example_column_to_image_f32 column_to_image_f32.cpp)
+add_example_dependencies(example_im2col_col2im example_column_to_image_f32)

example/60_gemm_multi_ABD/CMakeLists.txt

-list(APPEND gpu_list2 gfx908 gfx90a gfx940 gfx941 gfx942)
-set(target 0)
-foreach(gpu IN LISTS GPU_TARGETS)
-    if(gpu IN_LIST gpu_list2 AND target EQUAL 0)
-        add_example_executable(example_gemm_multi_ABD_xdl_fp16 gemm_multi_ABD_xdl_fp16.cpp)
-        set(target 1)
-    endif()
-endforeach()
+add_example_executable(example_gemm_multi_ABD_xdl_fp16 gemm_multi_ABD_xdl_fp16.cpp)