Merge commit 'c54ecd905b07849076069d56c284472230564568' into develop

profiler/README.md

@@ -1,5 +1,23 @@
 [Back to the main page](../README.md)
 # Composable Kernel profiler
+
+## Building Specific Profilers
+To reduce build time, filter which operations to compile using CMake options:
+
+```bash
+# Build all grouped_gemm variants (grouped_gemm, grouped_gemm_fastgelu, grouped_gemm_tile_loop, etc.)
+cmake -DCK_PROFILER_OP_FILTER="grouped_gemm" <other options> ..
+
+# Build ONLY base grouped_gemm (excludes variants - use exact regex match with ^ and $)
+cmake -DCK_PROFILER_OP_FILTER="^grouped_gemm$" <other options> ..
+```
+
+Both `CK_PROFILER_OP_FILTER` and `CK_PROFILER_INSTANCE_FILTER` accept regex patterns. Default builds all operations.
+
+To find the complete list of operations, run the following command:
+```bash
+find profiler/src -name "profile_*.cpp" | sed 's|profiler/src/profile_||' | sed 's|.cpp||' | sort
+```
 ## Profiler GEMM UNIVERSAL kernels
 ```bash
 # arg1: tensor operation (gemm_universal: Universal GEMM)

script/check_copyright_year.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 
 current_year=$(date +%Y)
 exit_code=0

script/clang-format-overwrite.sh

@@ -1,2 +1,5 @@
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 find . -name deps -prune -o -name build -prune -o -iname '*.h' -o -iname '*.hpp' -o -iname '*.cpp' -o -iname '*.h.in' -o -iname '*.hpp.in' -o -iname '*.cpp.in' -o -iname '*.cl' -o -iname '*.cuh' -o -iname '*.cu' -o -iname '*.inc' | grep -v 'build/' | grep -v 'include/rapidjson'| xargs -n 1 -P 16 -I{} -t sh -c 'clang-format-18 -i -style=file {}'
 git status --porcelain | awk '$1 != "D" && (match($2, "\\.cpp|.hpp|.inc|include/rapidjson/")) {print $2}' | xargs -n 1 -P 16 -I{} -t sh -c 'clang-format-18 -i -style=file {}'

script/cmake-ck-dev.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 # exit when a command exits with non-zero status; also when an unbound variable is referenced
 set -eu
 # pipefail is supported by many shells, not supported by sh and dash

script/convert_miopen_driver_to_profiler.py

@@ -1,5 +1,6 @@
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
 # SPDX-License-Identifier: MIT
-# Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved.
+
 # Convert miopen driver command to ck Profiler
 # Example: python3 ../script/convert_miopen_driver_to_profiler.py
 # /opt/rocm/bin/MIOpenDriver conv -n 32 -c 64 -H 28 -W 28 -k 64 -y 3 -x 3

script/count_vgpr.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 FILE=$1
 
 for num in {0..255}

script/dependency-parser/generate_list_of_files_not_referenced_in_tests.py

@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
 
-## Copyright © Advanced Micro Devices, Inc. or its affiliates.
-## SPDX-License-Identifier: MIT
 
 # This script generate list of files that are not referenced from any test (list in JSON format)
 # Script only looks at not referenced files from three directories: include, library and profiler

script/dependency-parser/main.py

@@ -1,5 +1,5 @@
 #!/usr/bin/env python3
-# Copyright © Advanced Micro Devices, Inc., or its affiliates.
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
 # SPDX-License-Identifier: MIT
 
 """

script/dependency-parser/src/enhanced_ninja_parser.py

@@ -1,5 +1,5 @@
 #!/usr/bin/env python3
-# Copyright © Advanced Micro Devices, Inc., or its affiliates.
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
 # SPDX-License-Identifier: MIT
 
 """

script/dependency-parser/src/selective_test_filter.py

@@ -1,5 +1,5 @@
 #!/usr/bin/env python3
-# Copyright © Advanced Micro Devices, Inc., or its affiliates.
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
 # SPDX-License-Identifier: MIT
 
 """

script/gemm_profile.sh

@@ -1,5 +1,5 @@
 #!/bin/bash
-# Copyright © Advanced Micro Devices, Inc., or its affiliates.
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
 # SPDX-License-Identifier: MIT
 
 BIN=./bin/tile_example_gemm_weight_preshuffle

script/hipclang_opt.sh

@@ -1,3 +1,6 @@
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 rm *.ll *.s
 
 BC_FILE=$1

script/install_precommit.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 run_and_check() {
     "$@"
     status=$?

script/launch_tests.sh

@@ -1,5 +1,5 @@
 #!/bin/bash
-# Copyright © Advanced Micro Devices, Inc., or its affiliates.
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
 # SPDX-License-Identifier: MIT
 
 # Get the directory where the script is located

script/ninja_json_converter.py

@@ -1,5 +1,5 @@
 #!/usr/bin/env python3
-# Copyright © Advanced Micro Devices, Inc., or its affiliates.
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
 # SPDX-License-Identifier: MIT
 
 """

script/process_perf_data.py

@@ -1,4 +1,7 @@
 #!/usr/bin/env python3
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 import os
 import io
 import argparse

script/process_perf_data.sh

@@ -1,4 +1,7 @@
 #!/bin/bash 
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 #
 # in order to run this script you'd need the following python packages:
 

script/process_qa_data.sh

@@ -1,4 +1,7 @@
 #!/bin/bash 
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 #
 # in order to run this script you'd need the following python packages:
 

script/profile_batched_gemm.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
  
 ## GPU visibility
 export HIP_VISIBLE_DEVICES=0

script/profile_gemm.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 
 ## GPU visibility
 export HIP_VISIBLE_DEVICES=0

script/profile_gemm_bilinear.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 ## GPU visibility
 export HIP_VISIBLE_DEVICES=0
 DRIVER="../build/bin/ckProfiler"

script/profile_grouped_conv_bwd_data.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
  
 ## GPU visibility
 export HIP_VISIBLE_DEVICES=0

script/profile_grouped_conv_bwd_weight.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
  
 ## GPU visibility
 export HIP_VISIBLE_DEVICES=0

script/profile_grouped_conv_fwd.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
  
 ## GPU visibility
 export HIP_VISIBLE_DEVICES=0

script/profile_grouped_conv_fwd_outelementop.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 
 ## GPU visibility
 export HIP_VISIBLE_DEVICES=0

script/profile_grouped_gemm.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
  
 ## GPU visibility
 export HIP_VISIBLE_DEVICES=0

script/profile_mixed_gemm.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 
 ## GPU visibility
 export HIP_VISIBLE_DEVICES=0

script/profile_onnx_gemm.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 ## GPU visibility
 export HIP_VISIBLE_DEVICES=0
 DRIVER="../build/bin/ckProfiler"

script/profile_permute_scale.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 
 ## GPU visibility
 export HIP_VISIBLE_DEVICES=0

script/profile_reduce_no_index.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 DRIVER="../build/bin/ckProfiler"
 VERIFY="-v $1"
 INIT=$2

script/profile_reduce_with_index.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 DRIVER="../build/bin/ckProfiler"
 VERIFY="-v $1"
 INIT=$2

script/profile_resnet50.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 
 ## GPU visibility
 export HIP_VISIBLE_DEVICES=0

script/profile_splitK_gemm.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 
 ## GPU visibility
 export HIP_VISIBLE_DEVICES=0

script/remod_for_ck_tile.py

@@ -1,3 +1,6 @@
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 import os
 
 root_dir = os.getcwd()

script/remove_exec_bit.sh

@@ -1,5 +1,5 @@
 #!/usr/bin/env bash
-# Copyright © Advanced Micro Devices, Inc., or its affiliates.
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
 # SPDX-License-Identifier: MIT
 
 for file in $(git diff --cached --name-only --diff-filter=ACM | grep -E '\.(cpp|hpp|txt|inc)$'); do

script/run_ck_profiler_gemm_with_csv_shapes.py

@@ -1,6 +1,6 @@
 #!/usr/bin/env python3
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
 # SPDX-License-Identifier: MIT
-# Copyright (c) 2025, Advanced Micro Devices, Inc. All rights reserved.
 
 # -*- coding: utf-8 -*-
 

script/run_full_performance_tests.sh

@@ -1,4 +1,7 @@
 #!/bin/bash 
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 #
 # in order to run this script you'd first need to build the ckProfiler executable in ../build/bin/
 # you would also need to set up some environment variables in order to 

script/run_gemm_performance_tests.sh

@@ -1,4 +1,7 @@
 #!/bin/bash 
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 #
 # in order to run this script you'd first need to build the ckProfiler executable in ../build/bin/
 # run the script as "./run_gemm_performance_tests.sh <verification> <tag for your test environment> <branch name> <node name> <arch>

script/run_performance_tests.sh

@@ -1,4 +1,7 @@
 #!/bin/bash 
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 #
 # in order to run this script you'd first need to build the ckProfiler executable in ../build/bin/
 # run the script as "./run_performance_tests.sh <verification> <tag for your test environment> <branch name> <node name>

script/sccache_wrapper.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 set -e
 COMPILERS_HASH_DIR=${COMPILERS_HASH_DIR:-"/tmp/.sccache"}
 SCCACHE_EXTRAFILES=${SCCACHE_EXTRAFILES:-"${COMPILERS_HASH_DIR}/rocm_compilers_hash_file"}

script/test_convnd_fwd.sh

@@ -1,4 +1,7 @@
 #!/usr/bin/env bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 
 # set -e
 

script/test_reduce_no_index.sh

@@ -1,4 +1,7 @@
 #!/bin/bash
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 
 ## The following will be used for CI
 

script/uninstall_precommit.sh

@@ -1 +1,4 @@
+# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
+# SPDX-License-Identifier: MIT
+
 pre-commit uninstall