Merge branch 'develop' into amd-develop

include/ck/host_utility/device_prop.hpp

@@ -84,4 +84,9 @@ inline bool is_gfx11_supported()
            ck::get_device_name() == "gfx1102" || ck::get_device_name() == "gfx1103";
 }
 
+inline bool is_gfx12_supported()
+{
+    return ck::get_device_name() == "gfx1200" || ck::get_device_name() == "gfx1201";
+}
+
 } // namespace ck

include/ck/tensor_operation/gpu/block/blockwise_gemm_dpp.hpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #pragma once
 
@@ -300,7 +300,7 @@ struct BlockwiseGemmDpp_ak0mak1_bk0nbk1_m0n0m1n1m2n2
                     constexpr index_t c_offset =
                         c_thread_desc_.CalculateOffset(make_tuple(m0, n0, 0));
 
-                    dpp_gemm.template Run(a_thread_vec.template AsType<dpp_input_type>(),
+                    dpp_gemm.Run(a_thread_vec.template AsType<dpp_input_type>(),
                                  b_thread_vec.template AsType<dpp_input_type>(),
                                  c_thread_buf.GetVectorTypeReference(Number<c_offset>{}));
                 });

include/ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops.hpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #pragma once
 
@@ -613,7 +613,7 @@ struct BlockwiseGemmXdlops_pipeline_v4
                             constexpr index_t c_offset =
                                 c_thread_desc_.CalculateOffset(make_tuple(m0, n0, 0));
 
-                            xdlops_gemm.template Run(
+                            xdlops_gemm.Run(
                                 a_thread_vec.template AsType<mfma_input_type>(),
                                 b_thread_vec.template AsType<mfma_input_type>(),
                                 c_thread_buf.GetVectorTypeReference(Number<c_offset>{}));
@@ -681,7 +681,7 @@ struct BlockwiseGemmXdlops_pipeline_v4
                             constexpr index_t c_offset =
                                 c_thread_desc_.CalculateOffset(make_tuple(m0, n0, 0));
 
-                            xdlops_gemm.template Run(
+                            xdlops_gemm.Run(
                                 a_thread_vec.template AsType<mfma_input_type>(),
                                 b_thread_vec.template AsType<mfma_input_type>(),
                                 c_thread_buf.GetVectorTypeReference(Number<c_offset>{}));
@@ -749,8 +749,7 @@ struct BlockwiseGemmXdlops_pipeline_v4
                         constexpr index_t c_offset =
                             c_thread_desc_.CalculateOffset(make_tuple(m0, n0, 0));
 
-                        xdlops_gemm.template Run(
-                            a_thread_vec.template AsType<mfma_input_type>(),
+                        xdlops_gemm.Run(a_thread_vec.template AsType<mfma_input_type>(),
                                         b_thread_vec.template AsType<mfma_input_type>(),
                                         c_thread_buf.GetVectorTypeReference(Number<c_offset>{}));
                     });
@@ -808,8 +807,7 @@ struct BlockwiseGemmXdlops_pipeline_v4
                         constexpr index_t c_offset =
                             c_thread_desc_.CalculateOffset(make_tuple(m0, n0, 0));
 
-                        xdlops_gemm.template Run(
-                            a_thread_vec.template AsType<mfma_input_type>(),
+                        xdlops_gemm.Run(a_thread_vec.template AsType<mfma_input_type>(),
                                         b_thread_vec.template AsType<mfma_input_type>(),
                                         c_thread_buf.GetVectorTypeReference(Number<c_offset>{}));
                     });
@@ -840,8 +838,7 @@ struct BlockwiseGemmXdlops_pipeline_v4
                         constexpr index_t c_offset =
                             c_thread_desc_.CalculateOffset(make_tuple(m0, n0, 0));
 
-                        xdlops_gemm.template Run(
-                            a_thread_vec.template AsType<mfma_input_type>(),
+                        xdlops_gemm.Run(a_thread_vec.template AsType<mfma_input_type>(),
                                         b_thread_vec.template AsType<mfma_input_type>(),
                                         c_thread_buf.GetVectorTypeReference(Number<c_offset>{}));
                     });
@@ -901,8 +898,7 @@ struct BlockwiseGemmXdlops_pipeline_v4
                         constexpr index_t c_offset =
                             c_thread_desc_.CalculateOffset(make_tuple(m0, n0, 0));
 
-                        xdlops_gemm.template Run(
-                            a_thread_vec.template AsType<mfma_input_type>(),
+                        xdlops_gemm.Run(a_thread_vec.template AsType<mfma_input_type>(),
                                         b_thread_vec.template AsType<mfma_input_type>(),
                                         c_thread_buf.GetVectorTypeReference(Number<c_offset>{}));
                     });
@@ -939,8 +935,7 @@ struct BlockwiseGemmXdlops_pipeline_v4
                         constexpr index_t c_offset =
                             c_thread_desc_.CalculateOffset(make_tuple(m0, n0, 0));
 
-                        xdlops_gemm.template Run(
-                            a_thread_vec.template AsType<mfma_input_type>(),
+                        xdlops_gemm.Run(a_thread_vec.template AsType<mfma_input_type>(),
                                         b_thread_vec.template AsType<mfma_input_type>(),
                                         c_thread_buf.GetVectorTypeReference(Number<c_offset>{}));
                     });