fix enum

library/include/ck/library/tensor_operation_instance/add_device_operation_instance.hpp

@@ -34,14 +34,16 @@ enum struct ArchitectureEnum
     None,
     Gfx908,
     Gfx90a,
+    Gfx940,
     Gfx1030,
     All
 };
-enum struct GemmFeatureEnum
+enum struct ArchFeatureEnum
 {
+    None,
     Xdl,
     Dl,
-    Wmd
+    Wmma
 };
 template <ArchitectureEnum... Is>
 struct ArchitectureEnumSequence
@@ -55,7 +57,7 @@ struct ArchitectureEnumSequence
         return mData[I];
     }
 };
-template <GemmFeatureEnum Feature, typename DeviceOp>
+template <ArchFeatureEnum Feature, typename DeviceOp>
 struct DeviceOperationInstances;
 
 template <typename Arch, typename DeviceOp>

library/include/ck/library/tensor_operation_instance/gpu/batched_gemm_softmax_gemm_permute/batched_gemm_softmax_gemm_permute.hpp

@@ -74,11 +74,11 @@ struct DeviceOperationInstanceCreator<Arch,
                                                          MaskingSpec>;
     static void add_device_instances(std::vector<std::unique_ptr<DeviceOp>>& instances)
     {
-        if constexpr(DeviceOperationInstances<GemmFeatureEnum::Xdl,
+        if constexpr(DeviceOperationInstances<ArchFeatureEnum::Xdl,
                                               DeviceOp>::template is_surport<Arch>())
             add_device_operation_instances(
                 instances,
-                DeviceOperationInstances<GemmFeatureEnum::Xdl, DeviceOp>::get_device_instances());
+                DeviceOperationInstances<ArchFeatureEnum::Xdl, DeviceOp>::get_device_instances());
     }
 };
 

library/include/ck/library/tensor_operation_instance/gpu/batched_gemm_softmax_gemm_permute/device_batched_gemm_multiple_d_softmax_gemm_permute_xdl_cshuffle_bf16_gmk_gnk_gno_gmo_instance.hpp

@@ -27,7 +27,7 @@ template <index_t NumDimG,
           typename B1ElementwiseOperation,
           typename C1DEElementwiseOperation,
           MaskingSpecialization MaskingSpec>
-struct DeviceOperationInstances<GemmFeatureEnum::Xdl,
+struct DeviceOperationInstances<ArchFeatureEnum::Xdl,
                                 DeviceBatchedGemmSoftmaxGemmPermute<NumDimG,
                                                                     NumDimM,
                                                                     NumDimN,
@@ -95,7 +95,8 @@ struct DeviceOperationInstances<GemmFeatureEnum::Xdl,
         ck::static_for<0, Archs::mSize, 1>{}([&](auto I) {
             if constexpr(Archs::At(I) == ArchitectureEnum::All ||
                          Archs::At(I) == ArchitectureEnum::Gfx908 ||
-                         Archs::At(I) == ArchitectureEnum::Gfx90a)
+                         Archs::At(I) == ArchitectureEnum::Gfx90a ||
+                         Archs::At(I) == ArchitectureEnum::Gfx940)
                 is_surport = true;
         });
         return is_surport;

library/include/ck/library/tensor_operation_instance/gpu/batched_gemm_softmax_gemm_permute/device_batched_gemm_multiple_d_softmax_gemm_permute_xdl_cshuffle_fp16_gmk_gnk_gno_gmo_instance.hpp

@@ -27,7 +27,7 @@ template <index_t NumDimG,
           typename B1ElementwiseOperation,
           typename C1DEElementwiseOperation,
           MaskingSpecialization MaskingSpec>
-struct DeviceOperationInstances<GemmFeatureEnum::Xdl,
+struct DeviceOperationInstances<ArchFeatureEnum::Xdl,
                                 DeviceBatchedGemmSoftmaxGemmPermute<NumDimG,
                                                                     NumDimM,
                                                                     NumDimN,
@@ -96,7 +96,8 @@ struct DeviceOperationInstances<GemmFeatureEnum::Xdl,
         ck::static_for<0, Archs::mSize, 1>{}([&](auto I) {
             if constexpr(Archs::At(I) == ArchitectureEnum::All ||
                          Archs::At(I) == ArchitectureEnum::Gfx908 ||
-                         Archs::At(I) == ArchitectureEnum::Gfx90a)
+                         Archs::At(I) == ArchitectureEnum::Gfx90a ||
+                         Archs::At(I) == ArchitectureEnum::Gfx940)
                 is_surport = true;
         });
         return is_surport;