merge from the public repo

example/09_convnd_fwd/convnd_fwd_xdl_fp32.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #include "convnd_fwd_common.hpp"
 

example/09_convnd_fwd/convnd_fwd_xdl_fp64.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #include "convnd_fwd_common.hpp"
 

example/09_convnd_fwd/convnd_fwd_xdl_int8.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #include "convnd_fwd_common.hpp"
 

example/09_convnd_fwd/run_convnd_fwd_dl_example.inc

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #pragma once
 

example/09_convnd_fwd/run_convnd_fwd_example.inc

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #pragma once
 

example/10_convnd_fwd_multiple_d_multiple_reduce/CMakeLists.txt

-add_custom_target(example_convnd_fwd_reduce_xdl)
+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_example_executable(example_convnd_fwd_max_xdl_int8 convnd_fwd_max_xdl_int8.cpp)
-add_example_executable_no_testing(example_convnd_fwd_max_xdl_bf16 convnd_fwd_max_xdl_bf16.cpp)
-add_example_executable_no_testing(example_convnd_fwd_max_xdl_fp16 convnd_fwd_max_xdl_fp16.cpp)
-add_example_executable(example_convnd_fwd_max_xdl_fp32 convnd_fwd_max_xdl_fp32.cpp)
+        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_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_int8)
-add_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_bf16)
-add_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_fp16)
-add_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_fp32)
+        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)
 
-if(USE_BITINT_EXTENSION_INT4)
-  add_example_executable(example_convnd_fwd_max_xdl_int4 convnd_fwd_max_xdl_int4.cpp)
-  add_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_int4)
-endif(USE_BITINT_EXTENSION_INT4)
+        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)
+
+        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()

example/10_convnd_fwd_multiple_d_multiple_reduce/common.hpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <algorithm>
 #include <cassert>

example/10_convnd_fwd_multiple_d_multiple_reduce/convnd_fwd_max_xdl_bf16.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #include "common.hpp"
 

example/10_convnd_fwd_multiple_d_multiple_reduce/convnd_fwd_max_xdl_fp16.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #include "common.hpp"
 

example/10_convnd_fwd_multiple_d_multiple_reduce/convnd_fwd_max_xdl_fp32.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #include "common.hpp"
 

example/10_convnd_fwd_multiple_d_multiple_reduce/convnd_fwd_max_xdl_int4.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #ifndef CK_EXPERIMENTAL_BIT_INT_EXTENSION_INT4
 #error Should compile this file with ck::int4_t support

example/10_convnd_fwd_multiple_d_multiple_reduce/convnd_fwd_max_xdl_int8.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #include "common.hpp"
 

example/10_convnd_fwd_multiple_d_multiple_reduce/run_convnd_fwd_max_example.inc

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #pragma once
 

example/12_reduce/README.md

@@ -2,7 +2,7 @@
 
 ## Run ```example_reduce_blockwise```
 ```bash
-# -D <xxx> : input 3d/4d/5d tensor lengths
+# -D <xxx> : input 3D/4D/5D tensor lengths
 # -R <xxx> : reduce dimension ids
 # -v <x> :   verification (0=no, 1=yes)
 #arg1: data type (0: fp16, 1: fp32, 3: int8, 5: bp16, 6: fp64, 7: int4)
@@ -22,7 +22,7 @@ Perf: 0.238063 ms, 264.285 GB/s, DeviceReduceBlockWise<256,M_C4_S1,K_C64_S1,InSr
 
 ## Run ```example_reduce_multiblock_atomic_add```
 ```bash
-# -D <xxx> : input 3d/4d/5d tensor lengths
+# -D <xxx> : input 3D/4D/5D tensor lengths
 # -R <xxx> : reduce dimension ids
 # -v <x> :   verification (0=no, 1=yes)
 #arg1: data type (0: fp32, 1: fp64)

example/12_reduce/reduce_blockwise.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iostream>
 #include <initializer_list>

example/12_reduce/reduce_blockwise_impl.hpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #pragma once
 

example/12_reduce/reduce_blockwise_two_call.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iostream>
 #include <numeric>

example/12_reduce/reduce_example_common.hpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #pragma once
 

example/12_reduce/reduce_multiblock_atomic_add.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iostream>
 #include <initializer_list>

example/12_reduce/reduce_multiblock_atomic_add_impl.hpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #pragma once