From b4e313286b91ffb16a8d9d71696bee93a629143a Mon Sep 17 00:00:00 2001 From: "assistant-librarian[bot]" Date: Fri, 14 Nov 2025 20:13:54 +0000 Subject: [PATCH] Merge commit '0aadb4b2c4114a26147c30abc894f2693795b888' into develop --- profiler/include/profiler/common.hpp | 2 +- profiler/include/profiler/data_type_enum.hpp | 2 +- .../profiler/profile_avg_pool2d_bwd_impl.hpp | 2 +- .../profiler/profile_avg_pool3d_bwd_impl.hpp | 2 +- ...le_batched_gemm_add_relu_gemm_add_impl.hpp | 2 +- .../profile_batched_gemm_b_scale_impl.hpp | 2 +- ...ed_gemm_bias_softmax_gemm_permute_impl.hpp | 2 +- .../profile_batched_gemm_gemm_impl.hpp | 2 +- .../profiler/profile_batched_gemm_impl.hpp | 2 +- .../profile_batched_gemm_reduce_impl.hpp | 2 +- ...profile_batched_gemm_softmax_gemm_impl.hpp | 2 +- ...batched_gemm_softmax_gemm_permute_impl.hpp | 2 +- .../profile_batchnorm_backward_impl.hpp | 2 +- .../profile_batchnorm_forward_impl.hpp | 2 +- .../profiler/profile_batchnorm_infer_impl.hpp | 2 +- .../profiler/profile_contraction_impl.hpp | 2 +- .../profiler/profile_contraction_utils.hpp | 2 +- .../profiler/profile_conv_bwd_data_impl.hpp | 2 +- .../profile_conv_fwd_bias_relu_add_impl.hpp | 2 +- .../profile_conv_fwd_bias_relu_impl.hpp | 2 +- .../profiler/profile_conv_fwd_impl.hpp | 2 +- .../profile_conv_tensor_rearrange_impl.hpp | 2 +- .../profile_elementwise_layernorm_impl.hpp | 2 +- .../profiler/profile_gemm_ab_scale_impl.hpp | 2 +- .../profile_gemm_add_add_fastgelu_impl.hpp | 2 +- .../profile_gemm_add_fastgelu_impl.hpp | 2 +- .../profiler/profile_gemm_add_impl.hpp | 2 +- .../profile_gemm_add_multiply_impl.hpp | 2 +- ...ofile_gemm_add_relu_add_layernorm_impl.hpp | 2 +- .../profiler/profile_gemm_add_relu_impl.hpp | 2 +- .../profiler/profile_gemm_add_silu_impl.hpp | 2 +- .../profiler/profile_gemm_b_scale_impl.hpp | 2 +- .../profile_gemm_bias_add_reduce_impl.hpp | 2 +- .../profiler/profile_gemm_bilinear_impl.hpp | 2 +- .../profile_gemm_blockscale_wp_impl.hpp | 2 +- .../profiler/profile_gemm_fastgelu_impl.hpp | 2 +- .../include/profiler/profile_gemm_impl.hpp | 2 +- .../profiler/profile_gemm_multi_abd_impl.hpp | 2 +- .../profile_gemm_multiply_add_impl.hpp | 2 +- .../profile_gemm_multiply_multiply_impl.hpp | 2 +- ...profile_gemm_multiply_multiply_wp_impl.hpp | 2 +- .../include/profiler/profile_gemm_mx_impl.hpp | 2 +- .../profile_gemm_quantization_impl.hpp | 2 +- .../profiler/profile_gemm_reduce_impl.hpp | 2 +- .../profiler/profile_gemm_splitk_impl.hpp | 2 +- .../profiler/profile_gemm_streamk_impl.hpp | 2 +- .../profile_gemm_universal_batched_impl.hpp | 2 +- .../profiler/profile_gemm_universal_impl.hpp | 2 +- ...profile_gemm_universal_preshuffle_impl.hpp | 2 +- .../profile_gemm_universal_reduce_impl.hpp | 2 +- .../profile_gemm_universal_streamk_impl.hpp | 2 +- .../profile_grouped_conv_bwd_data_impl.hpp | 2 +- .../profile_grouped_conv_bwd_weight_impl.hpp | 2 +- ...grouped_conv_fwd_bias_bnorm_clamp_impl.hpp | 2 +- ...ofile_grouped_conv_fwd_bias_clamp_impl.hpp | 2 +- .../profile_grouped_conv_fwd_impl.hpp | 2 +- ...ile_grouped_conv_fwd_outelementop_impl.hpp | 3 + .../profile_grouped_gemm_fastgelu_impl.hpp | 2 +- .../profile_grouped_gemm_fixed_nk_impl.hpp | 2 +- .../profiler/profile_grouped_gemm_impl.hpp | 2 +- ...e_grouped_gemm_multiply_tile_loop_impl.hpp | 2 +- .../profile_grouped_gemm_tile_loop_impl.hpp | 2 +- .../profile_groupnorm_bwd_data_impl.hpp | 2 +- .../profile_groupnorm_bwd_gamma_beta_impl.hpp | 2 +- .../profiler/profile_groupnorm_fwd_impl.hpp | 2 +- .../profile_layernorm_bwd_data_impl.hpp | 2 +- .../profile_layernorm_bwd_gamma_beta_impl.hpp | 2 +- .../profiler/profile_layernorm_fwd_impl.hpp | 2 +- .../profiler/profile_max_pool2d_bwd_impl.hpp | 2 +- .../profiler/profile_max_pool3d_bwd_impl.hpp | 2 +- .../profiler/profile_permute_scale_impl.hpp | 2 +- .../profiler/profile_pool2d_fwd_impl.hpp | 2 +- .../profiler/profile_pool3d_fwd_impl.hpp | 2 +- .../include/profiler/profile_reduce_impl.hpp | 2 +- .../include/profiler/profile_softmax_impl.hpp | 2 +- .../profiler/profile_transpose_impl.hpp | 356 +++++++++--------- profiler/src/profile_avg_pool2d_bwd.cpp | 2 +- profiler/src/profile_avg_pool3d_bwd.cpp | 2 +- profiler/src/profile_batched_gemm.cpp | 2 +- ...profile_batched_gemm_add_relu_gemm_add.cpp | 2 +- profiler/src/profile_batched_gemm_b_scale.cpp | 2 +- profiler/src/profile_batched_gemm_gemm.cpp | 2 +- profiler/src/profile_batched_gemm_multi_d.cpp | 2 +- profiler/src/profile_batched_gemm_reduce.cpp | 2 +- profiler/src/profile_batchnorm_bwd.cpp | 2 +- profiler/src/profile_batchnorm_fwd.cpp | 2 +- profiler/src/profile_batchnorm_infer.cpp | 2 +- profiler/src/profile_contraction_bilinear.cpp | 2 +- profiler/src/profile_contraction_scale.cpp | 2 +- profiler/src/profile_conv_bwd_data.cpp | 2 +- profiler/src/profile_conv_fwd.cpp | 2 +- profiler/src/profile_conv_fwd_bias_relu.cpp | 2 +- .../src/profile_conv_fwd_bias_relu_add.cpp | 2 +- .../src/profile_conv_tensor_rearrange.cpp | 2 +- profiler/src/profile_gemm.cpp | 2 +- profiler/src/profile_gemm_ab_scale.cpp | 2 +- profiler/src/profile_gemm_add.cpp | 2 +- .../src/profile_gemm_add_add_fastgelu.cpp | 2 +- profiler/src/profile_gemm_add_fastgelu.cpp | 2 +- profiler/src/profile_gemm_add_multiply.cpp | 2 +- profiler/src/profile_gemm_add_relu.cpp | 2 +- .../profile_gemm_add_relu_add_layernorm.cpp | 2 +- profiler/src/profile_gemm_add_silu.cpp | 2 +- profiler/src/profile_gemm_b_scale.cpp | 2 +- profiler/src/profile_gemm_bias_add_reduce.cpp | 2 +- profiler/src/profile_gemm_bilinear.cpp | 2 +- profiler/src/profile_gemm_blockscale_wp.cpp | 2 +- profiler/src/profile_gemm_fastgelu.cpp | 2 +- profiler/src/profile_gemm_multi_abd.cpp | 2 +- profiler/src/profile_gemm_multiply_add.cpp | 2 +- .../src/profile_gemm_multiply_multiply.cpp | 2 +- .../src/profile_gemm_multiply_multiply_wp.cpp | 2 +- profiler/src/profile_gemm_mx.cpp | 2 +- profiler/src/profile_gemm_quantization.cpp | 2 +- profiler/src/profile_gemm_reduce.cpp | 2 +- profiler/src/profile_gemm_splitk.cpp | 2 +- profiler/src/profile_gemm_streamk.cpp | 2 +- profiler/src/profile_gemm_universal.cpp | 2 +- .../src/profile_gemm_universal_batched.cpp | 2 +- .../src/profile_gemm_universal_preshuffle.cpp | 2 +- .../src/profile_gemm_universal_reduce.cpp | 2 +- .../src/profile_gemm_universal_streamk.cpp | 2 +- .../src/profile_grouped_conv_bwd_data.cpp | 2 +- .../src/profile_grouped_conv_bwd_weight.cpp | 2 +- profiler/src/profile_grouped_conv_fwd.cpp | 2 +- .../profile_grouped_conv_fwd_bias_clamp.cpp | 2 +- .../src/profile_grouped_conv_fwd_clamp.cpp | 2 +- .../profile_grouped_conv_fwd_outelementop.cpp | 3 + profiler/src/profile_grouped_gemm.cpp | 2 +- .../src/profile_grouped_gemm_fastgelu.cpp | 2 +- .../src/profile_grouped_gemm_fixed_nk.cpp | 2 +- ...rofile_grouped_gemm_multiply_tile_loop.cpp | 2 +- .../src/profile_grouped_gemm_tile_loop.cpp | 2 +- profiler/src/profile_groupnorm_bwd_data.cpp | 2 +- .../src/profile_groupnorm_bwd_gamma_beta.cpp | 2 +- profiler/src/profile_groupnorm_fwd.cpp | 2 +- profiler/src/profile_layernorm_bwd_data.cpp | 2 +- .../src/profile_layernorm_bwd_gamma_beta.cpp | 2 +- profiler/src/profile_layernorm_fwd.cpp | 2 +- profiler/src/profile_max_pool2d_bwd.cpp | 2 +- profiler/src/profile_max_pool2d_fwd.cpp | 2 +- profiler/src/profile_max_pool3d_bwd.cpp | 2 +- profiler/src/profile_permute_scale.cpp | 2 +- profiler/src/profile_pool3d_fwd.cpp | 2 +- profiler/src/profile_reduce.cpp | 2 +- profiler/src/profile_softmax.cpp | 2 +- profiler/src/profile_transpose.cpp | 162 ++++---- profiler/src/profiler.cpp | 2 +- profiler/src/profiler_operation_registry.hpp | 2 +- 149 files changed, 410 insertions(+), 404 deletions(-) diff --git a/profiler/include/profiler/common.hpp b/profiler/include/profiler/common.hpp index 2f72e67c6b..21c0525a02 100644 --- a/profiler/include/profiler/common.hpp +++ b/profiler/include/profiler/common.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/data_type_enum.hpp b/profiler/include/profiler/data_type_enum.hpp index bbfb04df7d..cefe240bea 100644 --- a/profiler/include/profiler/data_type_enum.hpp +++ b/profiler/include/profiler/data_type_enum.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_avg_pool2d_bwd_impl.hpp b/profiler/include/profiler/profile_avg_pool2d_bwd_impl.hpp index 537a4703d3..36eaebc4ae 100644 --- a/profiler/include/profiler/profile_avg_pool2d_bwd_impl.hpp +++ b/profiler/include/profiler/profile_avg_pool2d_bwd_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_avg_pool3d_bwd_impl.hpp b/profiler/include/profiler/profile_avg_pool3d_bwd_impl.hpp index c97e42228d..6528c89ea8 100644 --- a/profiler/include/profiler/profile_avg_pool3d_bwd_impl.hpp +++ b/profiler/include/profiler/profile_avg_pool3d_bwd_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_batched_gemm_add_relu_gemm_add_impl.hpp b/profiler/include/profiler/profile_batched_gemm_add_relu_gemm_add_impl.hpp index 4b0b8e5bcb..c15b16e224 100644 --- a/profiler/include/profiler/profile_batched_gemm_add_relu_gemm_add_impl.hpp +++ b/profiler/include/profiler/profile_batched_gemm_add_relu_gemm_add_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_batched_gemm_b_scale_impl.hpp b/profiler/include/profiler/profile_batched_gemm_b_scale_impl.hpp index 357ab8d70f..2399f793b3 100644 --- a/profiler/include/profiler/profile_batched_gemm_b_scale_impl.hpp +++ b/profiler/include/profiler/profile_batched_gemm_b_scale_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2023-2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_batched_gemm_bias_softmax_gemm_permute_impl.hpp b/profiler/include/profiler/profile_batched_gemm_bias_softmax_gemm_permute_impl.hpp index ca0d031dba..c9eddd1553 100644 --- a/profiler/include/profiler/profile_batched_gemm_bias_softmax_gemm_permute_impl.hpp +++ b/profiler/include/profiler/profile_batched_gemm_bias_softmax_gemm_permute_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_batched_gemm_gemm_impl.hpp b/profiler/include/profiler/profile_batched_gemm_gemm_impl.hpp index a8571d0779..f8113a6714 100644 --- a/profiler/include/profiler/profile_batched_gemm_gemm_impl.hpp +++ b/profiler/include/profiler/profile_batched_gemm_gemm_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_batched_gemm_impl.hpp b/profiler/include/profiler/profile_batched_gemm_impl.hpp index 0fdda68c4d..11a9da365d 100644 --- a/profiler/include/profiler/profile_batched_gemm_impl.hpp +++ b/profiler/include/profiler/profile_batched_gemm_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_batched_gemm_reduce_impl.hpp b/profiler/include/profiler/profile_batched_gemm_reduce_impl.hpp index cb91d8090d..249e80309b 100644 --- a/profiler/include/profiler/profile_batched_gemm_reduce_impl.hpp +++ b/profiler/include/profiler/profile_batched_gemm_reduce_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_batched_gemm_softmax_gemm_impl.hpp b/profiler/include/profiler/profile_batched_gemm_softmax_gemm_impl.hpp index 183b0e183a..029793bb18 100644 --- a/profiler/include/profiler/profile_batched_gemm_softmax_gemm_impl.hpp +++ b/profiler/include/profiler/profile_batched_gemm_softmax_gemm_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_batched_gemm_softmax_gemm_permute_impl.hpp b/profiler/include/profiler/profile_batched_gemm_softmax_gemm_permute_impl.hpp index e953cc4b66..b09b27bd69 100644 --- a/profiler/include/profiler/profile_batched_gemm_softmax_gemm_permute_impl.hpp +++ b/profiler/include/profiler/profile_batched_gemm_softmax_gemm_permute_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_batchnorm_backward_impl.hpp b/profiler/include/profiler/profile_batchnorm_backward_impl.hpp index bf5a661407..2b0e8c3806 100644 --- a/profiler/include/profiler/profile_batchnorm_backward_impl.hpp +++ b/profiler/include/profiler/profile_batchnorm_backward_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_batchnorm_forward_impl.hpp b/profiler/include/profiler/profile_batchnorm_forward_impl.hpp index 078f6bff87..9eead80eb2 100644 --- a/profiler/include/profiler/profile_batchnorm_forward_impl.hpp +++ b/profiler/include/profiler/profile_batchnorm_forward_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_batchnorm_infer_impl.hpp b/profiler/include/profiler/profile_batchnorm_infer_impl.hpp index c866b88e8a..5ae150f269 100644 --- a/profiler/include/profiler/profile_batchnorm_infer_impl.hpp +++ b/profiler/include/profiler/profile_batchnorm_infer_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_contraction_impl.hpp b/profiler/include/profiler/profile_contraction_impl.hpp index 361861a6d1..f720fa6a7f 100644 --- a/profiler/include/profiler/profile_contraction_impl.hpp +++ b/profiler/include/profiler/profile_contraction_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2023-2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_contraction_utils.hpp b/profiler/include/profiler/profile_contraction_utils.hpp index adfd98a37e..cea16fded0 100644 --- a/profiler/include/profiler/profile_contraction_utils.hpp +++ b/profiler/include/profiler/profile_contraction_utils.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2023-2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_conv_bwd_data_impl.hpp b/profiler/include/profiler/profile_conv_bwd_data_impl.hpp index 8f7adebdd4..a0f9b9ac25 100644 --- a/profiler/include/profiler/profile_conv_bwd_data_impl.hpp +++ b/profiler/include/profiler/profile_conv_bwd_data_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_conv_fwd_bias_relu_add_impl.hpp b/profiler/include/profiler/profile_conv_fwd_bias_relu_add_impl.hpp index 436fbdbd75..3cda620831 100644 --- a/profiler/include/profiler/profile_conv_fwd_bias_relu_add_impl.hpp +++ b/profiler/include/profiler/profile_conv_fwd_bias_relu_add_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_conv_fwd_bias_relu_impl.hpp b/profiler/include/profiler/profile_conv_fwd_bias_relu_impl.hpp index 808c1a1c90..2a7ee6fd66 100644 --- a/profiler/include/profiler/profile_conv_fwd_bias_relu_impl.hpp +++ b/profiler/include/profiler/profile_conv_fwd_bias_relu_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_conv_fwd_impl.hpp b/profiler/include/profiler/profile_conv_fwd_impl.hpp index 200409fe61..ae92dc792c 100644 --- a/profiler/include/profiler/profile_conv_fwd_impl.hpp +++ b/profiler/include/profiler/profile_conv_fwd_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_conv_tensor_rearrange_impl.hpp b/profiler/include/profiler/profile_conv_tensor_rearrange_impl.hpp index 171ae1662b..dc534a17e1 100644 --- a/profiler/include/profiler/profile_conv_tensor_rearrange_impl.hpp +++ b/profiler/include/profiler/profile_conv_tensor_rearrange_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_elementwise_layernorm_impl.hpp b/profiler/include/profiler/profile_elementwise_layernorm_impl.hpp index ca08f48bcf..d587c3a6ff 100644 --- a/profiler/include/profiler/profile_elementwise_layernorm_impl.hpp +++ b/profiler/include/profiler/profile_elementwise_layernorm_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_ab_scale_impl.hpp b/profiler/include/profiler/profile_gemm_ab_scale_impl.hpp index 3c511469f2..5396a52e21 100644 --- a/profiler/include/profiler/profile_gemm_ab_scale_impl.hpp +++ b/profiler/include/profiler/profile_gemm_ab_scale_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024-2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_add_add_fastgelu_impl.hpp b/profiler/include/profiler/profile_gemm_add_add_fastgelu_impl.hpp index 81b8d8ddbf..2b0d2401bd 100644 --- a/profiler/include/profiler/profile_gemm_add_add_fastgelu_impl.hpp +++ b/profiler/include/profiler/profile_gemm_add_add_fastgelu_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_add_fastgelu_impl.hpp b/profiler/include/profiler/profile_gemm_add_fastgelu_impl.hpp index 9e4d30142b..c85c04197b 100644 --- a/profiler/include/profiler/profile_gemm_add_fastgelu_impl.hpp +++ b/profiler/include/profiler/profile_gemm_add_fastgelu_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_add_impl.hpp b/profiler/include/profiler/profile_gemm_add_impl.hpp index 502d2b2951..559cf7a5b2 100644 --- a/profiler/include/profiler/profile_gemm_add_impl.hpp +++ b/profiler/include/profiler/profile_gemm_add_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_add_multiply_impl.hpp b/profiler/include/profiler/profile_gemm_add_multiply_impl.hpp index fcb546fe96..72ff9ca5f7 100644 --- a/profiler/include/profiler/profile_gemm_add_multiply_impl.hpp +++ b/profiler/include/profiler/profile_gemm_add_multiply_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_add_relu_add_layernorm_impl.hpp b/profiler/include/profiler/profile_gemm_add_relu_add_layernorm_impl.hpp index 99076a20ec..c728aa9ddf 100644 --- a/profiler/include/profiler/profile_gemm_add_relu_add_layernorm_impl.hpp +++ b/profiler/include/profiler/profile_gemm_add_relu_add_layernorm_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_add_relu_impl.hpp b/profiler/include/profiler/profile_gemm_add_relu_impl.hpp index e7f4338ef0..27334d66f1 100644 --- a/profiler/include/profiler/profile_gemm_add_relu_impl.hpp +++ b/profiler/include/profiler/profile_gemm_add_relu_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_add_silu_impl.hpp b/profiler/include/profiler/profile_gemm_add_silu_impl.hpp index e8a96208f6..b9de475a87 100644 --- a/profiler/include/profiler/profile_gemm_add_silu_impl.hpp +++ b/profiler/include/profiler/profile_gemm_add_silu_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_b_scale_impl.hpp b/profiler/include/profiler/profile_gemm_b_scale_impl.hpp index 8ca1350523..d3db464559 100644 --- a/profiler/include/profiler/profile_gemm_b_scale_impl.hpp +++ b/profiler/include/profiler/profile_gemm_b_scale_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2023-2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_bias_add_reduce_impl.hpp b/profiler/include/profiler/profile_gemm_bias_add_reduce_impl.hpp index b265101f3f..1930cf9eb6 100644 --- a/profiler/include/profiler/profile_gemm_bias_add_reduce_impl.hpp +++ b/profiler/include/profiler/profile_gemm_bias_add_reduce_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_bilinear_impl.hpp b/profiler/include/profiler/profile_gemm_bilinear_impl.hpp index b540e938b5..7e593e9e42 100644 --- a/profiler/include/profiler/profile_gemm_bilinear_impl.hpp +++ b/profiler/include/profiler/profile_gemm_bilinear_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_blockscale_wp_impl.hpp b/profiler/include/profiler/profile_gemm_blockscale_wp_impl.hpp index da0dc60760..49fef5a0fc 100644 --- a/profiler/include/profiler/profile_gemm_blockscale_wp_impl.hpp +++ b/profiler/include/profiler/profile_gemm_blockscale_wp_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_fastgelu_impl.hpp b/profiler/include/profiler/profile_gemm_fastgelu_impl.hpp index 0fe8abe242..5add7b6073 100644 --- a/profiler/include/profiler/profile_gemm_fastgelu_impl.hpp +++ b/profiler/include/profiler/profile_gemm_fastgelu_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_impl.hpp b/profiler/include/profiler/profile_gemm_impl.hpp index 93eac048cd..458cba381e 100644 --- a/profiler/include/profiler/profile_gemm_impl.hpp +++ b/profiler/include/profiler/profile_gemm_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_multi_abd_impl.hpp b/profiler/include/profiler/profile_gemm_multi_abd_impl.hpp index 51922fde33..e94d5bb910 100644 --- a/profiler/include/profiler/profile_gemm_multi_abd_impl.hpp +++ b/profiler/include/profiler/profile_gemm_multi_abd_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_multiply_add_impl.hpp b/profiler/include/profiler/profile_gemm_multiply_add_impl.hpp index 2711d595d6..ab107cf997 100644 --- a/profiler/include/profiler/profile_gemm_multiply_add_impl.hpp +++ b/profiler/include/profiler/profile_gemm_multiply_add_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_multiply_multiply_impl.hpp b/profiler/include/profiler/profile_gemm_multiply_multiply_impl.hpp index 0fa6c62148..76368d9252 100644 --- a/profiler/include/profiler/profile_gemm_multiply_multiply_impl.hpp +++ b/profiler/include/profiler/profile_gemm_multiply_multiply_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_multiply_multiply_wp_impl.hpp b/profiler/include/profiler/profile_gemm_multiply_multiply_wp_impl.hpp index 21613e49c6..570aee7f85 100644 --- a/profiler/include/profiler/profile_gemm_multiply_multiply_wp_impl.hpp +++ b/profiler/include/profiler/profile_gemm_multiply_multiply_wp_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_mx_impl.hpp b/profiler/include/profiler/profile_gemm_mx_impl.hpp index 1fbe60c6cf..0f84744fd6 100644 --- a/profiler/include/profiler/profile_gemm_mx_impl.hpp +++ b/profiler/include/profiler/profile_gemm_mx_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_quantization_impl.hpp b/profiler/include/profiler/profile_gemm_quantization_impl.hpp index 02f374164e..24516284a0 100644 --- a/profiler/include/profiler/profile_gemm_quantization_impl.hpp +++ b/profiler/include/profiler/profile_gemm_quantization_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_reduce_impl.hpp b/profiler/include/profiler/profile_gemm_reduce_impl.hpp index c870a95cbe..163cbe204d 100644 --- a/profiler/include/profiler/profile_gemm_reduce_impl.hpp +++ b/profiler/include/profiler/profile_gemm_reduce_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_splitk_impl.hpp b/profiler/include/profiler/profile_gemm_splitk_impl.hpp index 744db27675..789a460ca7 100644 --- a/profiler/include/profiler/profile_gemm_splitk_impl.hpp +++ b/profiler/include/profiler/profile_gemm_splitk_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_streamk_impl.hpp b/profiler/include/profiler/profile_gemm_streamk_impl.hpp index f86e7ad447..01944dbbdd 100644 --- a/profiler/include/profiler/profile_gemm_streamk_impl.hpp +++ b/profiler/include/profiler/profile_gemm_streamk_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_universal_batched_impl.hpp b/profiler/include/profiler/profile_gemm_universal_batched_impl.hpp index 99e24cd205..9125da53a9 100644 --- a/profiler/include/profiler/profile_gemm_universal_batched_impl.hpp +++ b/profiler/include/profiler/profile_gemm_universal_batched_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_universal_impl.hpp b/profiler/include/profiler/profile_gemm_universal_impl.hpp index bee907dd76..7ece78ea7b 100644 --- a/profiler/include/profiler/profile_gemm_universal_impl.hpp +++ b/profiler/include/profiler/profile_gemm_universal_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2023-2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_universal_preshuffle_impl.hpp b/profiler/include/profiler/profile_gemm_universal_preshuffle_impl.hpp index 5ec056efd1..8bf8957343 100644 --- a/profiler/include/profiler/profile_gemm_universal_preshuffle_impl.hpp +++ b/profiler/include/profiler/profile_gemm_universal_preshuffle_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2023-2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_universal_reduce_impl.hpp b/profiler/include/profiler/profile_gemm_universal_reduce_impl.hpp index 554956ee88..f1057529c9 100644 --- a/profiler/include/profiler/profile_gemm_universal_reduce_impl.hpp +++ b/profiler/include/profiler/profile_gemm_universal_reduce_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_gemm_universal_streamk_impl.hpp b/profiler/include/profiler/profile_gemm_universal_streamk_impl.hpp index 035a1b77df..aa879eba9c 100644 --- a/profiler/include/profiler/profile_gemm_universal_streamk_impl.hpp +++ b/profiler/include/profiler/profile_gemm_universal_streamk_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024-2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_grouped_conv_bwd_data_impl.hpp b/profiler/include/profiler/profile_grouped_conv_bwd_data_impl.hpp index 2369b2eac8..b9e463dc1e 100644 --- a/profiler/include/profiler/profile_grouped_conv_bwd_data_impl.hpp +++ b/profiler/include/profiler/profile_grouped_conv_bwd_data_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/include/profiler/profile_grouped_conv_bwd_weight_impl.hpp b/profiler/include/profiler/profile_grouped_conv_bwd_weight_impl.hpp index 6654275fd0..bc7ecaa6ca 100644 --- a/profiler/include/profiler/profile_grouped_conv_bwd_weight_impl.hpp +++ b/profiler/include/profiler/profile_grouped_conv_bwd_weight_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_grouped_conv_fwd_bias_bnorm_clamp_impl.hpp b/profiler/include/profiler/profile_grouped_conv_fwd_bias_bnorm_clamp_impl.hpp index 2f7f3ae4d8..22ff02676a 100644 --- a/profiler/include/profiler/profile_grouped_conv_fwd_bias_bnorm_clamp_impl.hpp +++ b/profiler/include/profiler/profile_grouped_conv_fwd_bias_bnorm_clamp_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_grouped_conv_fwd_bias_clamp_impl.hpp b/profiler/include/profiler/profile_grouped_conv_fwd_bias_clamp_impl.hpp index 2dbadd8eb1..2a2f516727 100644 --- a/profiler/include/profiler/profile_grouped_conv_fwd_bias_clamp_impl.hpp +++ b/profiler/include/profiler/profile_grouped_conv_fwd_bias_clamp_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_grouped_conv_fwd_impl.hpp b/profiler/include/profiler/profile_grouped_conv_fwd_impl.hpp index d490cf4167..427d2b14df 100644 --- a/profiler/include/profiler/profile_grouped_conv_fwd_impl.hpp +++ b/profiler/include/profiler/profile_grouped_conv_fwd_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_grouped_conv_fwd_outelementop_impl.hpp b/profiler/include/profiler/profile_grouped_conv_fwd_outelementop_impl.hpp index ae12070014..50b97c3bae 100644 --- a/profiler/include/profiler/profile_grouped_conv_fwd_outelementop_impl.hpp +++ b/profiler/include/profiler/profile_grouped_conv_fwd_outelementop_impl.hpp @@ -1,3 +1,6 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. +// SPDX-License-Identifier: MIT + #pragma once #include "ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_convscale.hpp" diff --git a/profiler/include/profiler/profile_grouped_gemm_fastgelu_impl.hpp b/profiler/include/profiler/profile_grouped_gemm_fastgelu_impl.hpp index f05b13b749..227b494266 100644 --- a/profiler/include/profiler/profile_grouped_gemm_fastgelu_impl.hpp +++ b/profiler/include/profiler/profile_grouped_gemm_fastgelu_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_grouped_gemm_fixed_nk_impl.hpp b/profiler/include/profiler/profile_grouped_gemm_fixed_nk_impl.hpp index 8fb20f0135..f551a16a1b 100644 --- a/profiler/include/profiler/profile_grouped_gemm_fixed_nk_impl.hpp +++ b/profiler/include/profiler/profile_grouped_gemm_fixed_nk_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_grouped_gemm_impl.hpp b/profiler/include/profiler/profile_grouped_gemm_impl.hpp index 8314b9053f..03a2ed3186 100644 --- a/profiler/include/profiler/profile_grouped_gemm_impl.hpp +++ b/profiler/include/profiler/profile_grouped_gemm_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_grouped_gemm_multiply_tile_loop_impl.hpp b/profiler/include/profiler/profile_grouped_gemm_multiply_tile_loop_impl.hpp index 1b17f05760..f76460ffd8 100644 --- a/profiler/include/profiler/profile_grouped_gemm_multiply_tile_loop_impl.hpp +++ b/profiler/include/profiler/profile_grouped_gemm_multiply_tile_loop_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_grouped_gemm_tile_loop_impl.hpp b/profiler/include/profiler/profile_grouped_gemm_tile_loop_impl.hpp index cf3c3a6bae..282d5f22fd 100644 --- a/profiler/include/profiler/profile_grouped_gemm_tile_loop_impl.hpp +++ b/profiler/include/profiler/profile_grouped_gemm_tile_loop_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_groupnorm_bwd_data_impl.hpp b/profiler/include/profiler/profile_groupnorm_bwd_data_impl.hpp index c1647815ad..b9a9d139ac 100644 --- a/profiler/include/profiler/profile_groupnorm_bwd_data_impl.hpp +++ b/profiler/include/profiler/profile_groupnorm_bwd_data_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_groupnorm_bwd_gamma_beta_impl.hpp b/profiler/include/profiler/profile_groupnorm_bwd_gamma_beta_impl.hpp index 5e9d3df1b1..216e253737 100644 --- a/profiler/include/profiler/profile_groupnorm_bwd_gamma_beta_impl.hpp +++ b/profiler/include/profiler/profile_groupnorm_bwd_gamma_beta_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_groupnorm_fwd_impl.hpp b/profiler/include/profiler/profile_groupnorm_fwd_impl.hpp index 60982d18d5..e984dfcb12 100644 --- a/profiler/include/profiler/profile_groupnorm_fwd_impl.hpp +++ b/profiler/include/profiler/profile_groupnorm_fwd_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_layernorm_bwd_data_impl.hpp b/profiler/include/profiler/profile_layernorm_bwd_data_impl.hpp index 7704085048..8efe709f96 100644 --- a/profiler/include/profiler/profile_layernorm_bwd_data_impl.hpp +++ b/profiler/include/profiler/profile_layernorm_bwd_data_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_layernorm_bwd_gamma_beta_impl.hpp b/profiler/include/profiler/profile_layernorm_bwd_gamma_beta_impl.hpp index e36b20e1b5..b89702a51b 100644 --- a/profiler/include/profiler/profile_layernorm_bwd_gamma_beta_impl.hpp +++ b/profiler/include/profiler/profile_layernorm_bwd_gamma_beta_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_layernorm_fwd_impl.hpp b/profiler/include/profiler/profile_layernorm_fwd_impl.hpp index 51dcbb1275..1de8178d3d 100644 --- a/profiler/include/profiler/profile_layernorm_fwd_impl.hpp +++ b/profiler/include/profiler/profile_layernorm_fwd_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_max_pool2d_bwd_impl.hpp b/profiler/include/profiler/profile_max_pool2d_bwd_impl.hpp index a8efee3ef0..3fad040004 100644 --- a/profiler/include/profiler/profile_max_pool2d_bwd_impl.hpp +++ b/profiler/include/profiler/profile_max_pool2d_bwd_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_max_pool3d_bwd_impl.hpp b/profiler/include/profiler/profile_max_pool3d_bwd_impl.hpp index cf6050969f..35d0020418 100644 --- a/profiler/include/profiler/profile_max_pool3d_bwd_impl.hpp +++ b/profiler/include/profiler/profile_max_pool3d_bwd_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_permute_scale_impl.hpp b/profiler/include/profiler/profile_permute_scale_impl.hpp index 9ccbd67783..a2ffc48380 100644 --- a/profiler/include/profiler/profile_permute_scale_impl.hpp +++ b/profiler/include/profiler/profile_permute_scale_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_pool2d_fwd_impl.hpp b/profiler/include/profiler/profile_pool2d_fwd_impl.hpp index 962be4448c..389edfbf48 100644 --- a/profiler/include/profiler/profile_pool2d_fwd_impl.hpp +++ b/profiler/include/profiler/profile_pool2d_fwd_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_pool3d_fwd_impl.hpp b/profiler/include/profiler/profile_pool3d_fwd_impl.hpp index e1d0c1573d..851c3fbad5 100644 --- a/profiler/include/profiler/profile_pool3d_fwd_impl.hpp +++ b/profiler/include/profiler/profile_pool3d_fwd_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_reduce_impl.hpp b/profiler/include/profiler/profile_reduce_impl.hpp index 14a93af69d..191c57780c 100644 --- a/profiler/include/profiler/profile_reduce_impl.hpp +++ b/profiler/include/profiler/profile_reduce_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_softmax_impl.hpp b/profiler/include/profiler/profile_softmax_impl.hpp index d7a790803a..bdaeb52acf 100644 --- a/profiler/include/profiler/profile_softmax_impl.hpp +++ b/profiler/include/profiler/profile_softmax_impl.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #pragma once diff --git a/profiler/include/profiler/profile_transpose_impl.hpp b/profiler/include/profiler/profile_transpose_impl.hpp index 0baf2eac99..81ebbcdf17 100644 --- a/profiler/include/profiler/profile_transpose_impl.hpp +++ b/profiler/include/profiler/profile_transpose_impl.hpp @@ -1,178 +1,178 @@ -// SPDX-License-Identifier: MIT -// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved. - -#pragma once - -#include -#include -#include - -#include "ck/ck.hpp" -#include "ck/tensor_operation/gpu/device/tensor_layout.hpp" -#include "ck/tensor_operation/gpu/device/device_elementwise.hpp" -#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp" -#include "ck/tensor_operation/gpu/device/impl/device_elementwise_dynamic_vector_dims_impl.hpp" - -#include "ck/library/tensor_operation_instance/gpu/transpose_3d.hpp" - -#include "ck/library/utility/check_err.hpp" -#include "ck/library/utility/device_memory.hpp" -#include "ck/library/utility/host_tensor.hpp" -#include "ck/library/utility/host_tensor_generator.hpp" -#include "ck/library/utility/literals.hpp" - -namespace ck { -namespace profiler { - -template -void host_elementwise4D(HostTensorB& B_ndhwc, const HostTensorA& A_ncdhw, Functor functor) -{ - for(std::size_t n = 0; n < A_ncdhw.mDesc.GetLengths()[0]; ++n) - for(std::size_t c = 0; c < A_ncdhw.mDesc.GetLengths()[1]; ++c) - for(std::size_t d = 0; d < A_ncdhw.mDesc.GetLengths()[2]; ++d) - for(std::size_t h = 0; h < A_ncdhw.mDesc.GetLengths()[3]; ++h) - for(std::size_t w = 0; w < A_ncdhw.mDesc.GetLengths()[4]; ++w) - { - auto a_val = A_ncdhw(n, c, d, h, w); - functor(B_ndhwc(n, d, h, w, c), a_val); - } -} - -template -bool profile_transpose_impl(int do_verification, - int init_method, - bool do_log, - bool time_kernel, - std::vector lengths) -{ - bool pass = true; - - index_t N = lengths[0]; - index_t C = lengths[1]; - index_t D = lengths[2]; - index_t H = lengths[3]; - index_t W = lengths[4]; - - std::vector ncdhw = {N, C, D, H, W}; - std::vector ndhwc = {N, D, H, W, C}; - Tensor a(ncdhw); - Tensor b(ndhwc); - Tensor host_b(ndhwc); - - // a.GenerateTensorValue(GeneratorTensor_3{0.0, 1.0}); - - std::array ab_lengths{N, C, H, W, D}; - std::array a_strides = {C * D * H * W, H * W, W, 1, D * H * W}; // N, C, D, H, W - std::array b_strides = {C * H * W * D, H * W * D, W * D, D, 1}; // N, D, H, W, C - - std::cout << "A: " << a.mDesc << std::endl; - std::cout << "B: " << b.mDesc << std::endl; - - switch(init_method) - { - case 0: break; - case 1: a.GenerateTensorValue(GeneratorTensor_2{-1, 2}); break; - default: a.GenerateTensorValue(GeneratorTensor_3{0.0, 1.0}); - } - - using ElementOp = ck::tensor_operation::element_wise::PassThrough; - - DeviceMem a_device_buf(sizeof(ADataType) * a.mDesc.GetElementSpaceSize()); - DeviceMem b_device_buf(sizeof(BDataType) * b.mDesc.GetElementSpaceSize()); - - a_device_buf.ToDevice(a.mData.data()); - - std::array input = {a_device_buf.GetDeviceBuffer()}; - std::array output = {b_device_buf.GetDeviceBuffer()}; - using DeviceOp = ck::tensor_operation::device:: - DeviceElementwise, ck::Tuple, ElementOp, NumDim>; - - // get device op instances - const auto op_ptrs = ck::tensor_operation::device::instance::DeviceOperationInstanceFactory< - DeviceOp>::GetInstances(); - - std::cout << "found " << op_ptrs.size() << " instances" << std::endl; - - if(do_verification) - { - host_elementwise4D(host_b, a, ElementOp{}); - } - - std::string best_op_name; - float best_ave_time = 0; - float best_tflops = 0; - float best_gb_per_sec = 0; - - for(auto& op_ptr : op_ptrs) - { - auto argument_ptr = op_ptr->MakeArgumentPointer( - ab_lengths, {a_strides}, {b_strides}, input, output, ElementOp{}); - - auto invoker_ptr = op_ptr->MakeInvokerPointer(); - - if(op_ptr->IsSupportedArgument(argument_ptr.get())) - { - - // re-init C to zero before profiling next kernel - b_device_buf.SetZero(); - - // run for verification - invoker_ptr->Run(argument_ptr.get(), StreamConfig{nullptr, false}); - - if(do_verification) - { - b_device_buf.FromDevice(b.mData.data()); - - pass &= ck::utils::check_err( - b.mData, host_b.mData, "Error: Incorrect results b", 1e-3, 1e-3); - - if(do_log) - { - LogRangeAsType(std::cout << "a : ", a.mData, ",") << std::endl; - LogRangeAsType(std::cout << "b: ", b.mData, ",") << std::endl; - } - } - - std::string op_name = op_ptr->GetTypeString(); - - // run for timing purposes - float ave_time = - invoker_ptr->Run(argument_ptr.get(), StreamConfig{nullptr, time_kernel}); - - std::size_t flop = - std::size_t(2) * ncdhw[0] * ncdhw[1] * ncdhw[2] * ncdhw[3] * ncdhw[4]; - - std::size_t num_btype = - sizeof(ADataType) * (ncdhw[0] * ncdhw[1] * ncdhw[2] * ncdhw[3] * ncdhw[4]) + - sizeof(BDataType) * (ncdhw[0] * ncdhw[1] * ncdhw[2] * ncdhw[3] * ncdhw[4]); - - float tflops = static_cast(flop) / 1.E9 / ave_time; - - float gb_per_sec = num_btype / 1.E6 / ave_time; - - std::cout << "Perf: " << std::setw(10) << ave_time << " ms, " << tflops << " TFlops, " - << gb_per_sec << " GB/s, " << op_name << std::endl; - - if(tflops > best_tflops) - { - best_op_name = op_name; - best_tflops = tflops; - best_ave_time = ave_time; - best_gb_per_sec = gb_per_sec; - } - } - else - { - std::cout << op_ptr->GetTypeString() << " does not support this problem" << std::endl; - } - } - - std::cout << " N = " << N << " C = " << C << " D = " << D << " H = " << H << " W = " << W - << " : " << best_ave_time << " ms, " << best_tflops << " TFlops, " << best_gb_per_sec - << " GB/s, " << best_op_name << std::endl; - - return pass; -} - -} // namespace profiler -} // namespace ck +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. +// SPDX-License-Identifier: MIT + +#pragma once + +#include +#include +#include + +#include "ck/ck.hpp" +#include "ck/tensor_operation/gpu/device/tensor_layout.hpp" +#include "ck/tensor_operation/gpu/device/device_elementwise.hpp" +#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp" +#include "ck/tensor_operation/gpu/device/impl/device_elementwise_dynamic_vector_dims_impl.hpp" + +#include "ck/library/tensor_operation_instance/gpu/transpose_3d.hpp" + +#include "ck/library/utility/check_err.hpp" +#include "ck/library/utility/device_memory.hpp" +#include "ck/library/utility/host_tensor.hpp" +#include "ck/library/utility/host_tensor_generator.hpp" +#include "ck/library/utility/literals.hpp" + +namespace ck { +namespace profiler { + +template +void host_elementwise4D(HostTensorB& B_ndhwc, const HostTensorA& A_ncdhw, Functor functor) +{ + for(std::size_t n = 0; n < A_ncdhw.mDesc.GetLengths()[0]; ++n) + for(std::size_t c = 0; c < A_ncdhw.mDesc.GetLengths()[1]; ++c) + for(std::size_t d = 0; d < A_ncdhw.mDesc.GetLengths()[2]; ++d) + for(std::size_t h = 0; h < A_ncdhw.mDesc.GetLengths()[3]; ++h) + for(std::size_t w = 0; w < A_ncdhw.mDesc.GetLengths()[4]; ++w) + { + auto a_val = A_ncdhw(n, c, d, h, w); + functor(B_ndhwc(n, d, h, w, c), a_val); + } +} + +template +bool profile_transpose_impl(int do_verification, + int init_method, + bool do_log, + bool time_kernel, + std::vector lengths) +{ + bool pass = true; + + index_t N = lengths[0]; + index_t C = lengths[1]; + index_t D = lengths[2]; + index_t H = lengths[3]; + index_t W = lengths[4]; + + std::vector ncdhw = {N, C, D, H, W}; + std::vector ndhwc = {N, D, H, W, C}; + Tensor a(ncdhw); + Tensor b(ndhwc); + Tensor host_b(ndhwc); + + // a.GenerateTensorValue(GeneratorTensor_3{0.0, 1.0}); + + std::array ab_lengths{N, C, H, W, D}; + std::array a_strides = {C * D * H * W, H * W, W, 1, D * H * W}; // N, C, D, H, W + std::array b_strides = {C * H * W * D, H * W * D, W * D, D, 1}; // N, D, H, W, C + + std::cout << "A: " << a.mDesc << std::endl; + std::cout << "B: " << b.mDesc << std::endl; + + switch(init_method) + { + case 0: break; + case 1: a.GenerateTensorValue(GeneratorTensor_2{-1, 2}); break; + default: a.GenerateTensorValue(GeneratorTensor_3{0.0, 1.0}); + } + + using ElementOp = ck::tensor_operation::element_wise::PassThrough; + + DeviceMem a_device_buf(sizeof(ADataType) * a.mDesc.GetElementSpaceSize()); + DeviceMem b_device_buf(sizeof(BDataType) * b.mDesc.GetElementSpaceSize()); + + a_device_buf.ToDevice(a.mData.data()); + + std::array input = {a_device_buf.GetDeviceBuffer()}; + std::array output = {b_device_buf.GetDeviceBuffer()}; + using DeviceOp = ck::tensor_operation::device:: + DeviceElementwise, ck::Tuple, ElementOp, NumDim>; + + // get device op instances + const auto op_ptrs = ck::tensor_operation::device::instance::DeviceOperationInstanceFactory< + DeviceOp>::GetInstances(); + + std::cout << "found " << op_ptrs.size() << " instances" << std::endl; + + if(do_verification) + { + host_elementwise4D(host_b, a, ElementOp{}); + } + + std::string best_op_name; + float best_ave_time = 0; + float best_tflops = 0; + float best_gb_per_sec = 0; + + for(auto& op_ptr : op_ptrs) + { + auto argument_ptr = op_ptr->MakeArgumentPointer( + ab_lengths, {a_strides}, {b_strides}, input, output, ElementOp{}); + + auto invoker_ptr = op_ptr->MakeInvokerPointer(); + + if(op_ptr->IsSupportedArgument(argument_ptr.get())) + { + + // re-init C to zero before profiling next kernel + b_device_buf.SetZero(); + + // run for verification + invoker_ptr->Run(argument_ptr.get(), StreamConfig{nullptr, false}); + + if(do_verification) + { + b_device_buf.FromDevice(b.mData.data()); + + pass &= ck::utils::check_err( + b.mData, host_b.mData, "Error: Incorrect results b", 1e-3, 1e-3); + + if(do_log) + { + LogRangeAsType(std::cout << "a : ", a.mData, ",") << std::endl; + LogRangeAsType(std::cout << "b: ", b.mData, ",") << std::endl; + } + } + + std::string op_name = op_ptr->GetTypeString(); + + // run for timing purposes + float ave_time = + invoker_ptr->Run(argument_ptr.get(), StreamConfig{nullptr, time_kernel}); + + std::size_t flop = + std::size_t(2) * ncdhw[0] * ncdhw[1] * ncdhw[2] * ncdhw[3] * ncdhw[4]; + + std::size_t num_btype = + sizeof(ADataType) * (ncdhw[0] * ncdhw[1] * ncdhw[2] * ncdhw[3] * ncdhw[4]) + + sizeof(BDataType) * (ncdhw[0] * ncdhw[1] * ncdhw[2] * ncdhw[3] * ncdhw[4]); + + float tflops = static_cast(flop) / 1.E9 / ave_time; + + float gb_per_sec = num_btype / 1.E6 / ave_time; + + std::cout << "Perf: " << std::setw(10) << ave_time << " ms, " << tflops << " TFlops, " + << gb_per_sec << " GB/s, " << op_name << std::endl; + + if(tflops > best_tflops) + { + best_op_name = op_name; + best_tflops = tflops; + best_ave_time = ave_time; + best_gb_per_sec = gb_per_sec; + } + } + else + { + std::cout << op_ptr->GetTypeString() << " does not support this problem" << std::endl; + } + } + + std::cout << " N = " << N << " C = " << C << " D = " << D << " H = " << H << " W = " << W + << " : " << best_ave_time << " ms, " << best_tflops << " TFlops, " << best_gb_per_sec + << " GB/s, " << best_op_name << std::endl; + + return pass; +} + +} // namespace profiler +} // namespace ck diff --git a/profiler/src/profile_avg_pool2d_bwd.cpp b/profiler/src/profile_avg_pool2d_bwd.cpp index 61d0413d43..0b7936d57a 100644 --- a/profiler/src/profile_avg_pool2d_bwd.cpp +++ b/profiler/src/profile_avg_pool2d_bwd.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_avg_pool3d_bwd.cpp b/profiler/src/profile_avg_pool3d_bwd.cpp index 0ff50a5292..b279a45483 100644 --- a/profiler/src/profile_avg_pool3d_bwd.cpp +++ b/profiler/src/profile_avg_pool3d_bwd.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_batched_gemm.cpp b/profiler/src/profile_batched_gemm.cpp index d9da68b050..c2db49caf1 100644 --- a/profiler/src/profile_batched_gemm.cpp +++ b/profiler/src/profile_batched_gemm.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_batched_gemm_add_relu_gemm_add.cpp b/profiler/src/profile_batched_gemm_add_relu_gemm_add.cpp index 3d29c4b84a..55b8d1cd57 100644 --- a/profiler/src/profile_batched_gemm_add_relu_gemm_add.cpp +++ b/profiler/src/profile_batched_gemm_add_relu_gemm_add.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_batched_gemm_b_scale.cpp b/profiler/src/profile_batched_gemm_b_scale.cpp index 5ed673e127..ee17322e14 100644 --- a/profiler/src/profile_batched_gemm_b_scale.cpp +++ b/profiler/src/profile_batched_gemm_b_scale.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2025, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_batched_gemm_gemm.cpp b/profiler/src/profile_batched_gemm_gemm.cpp index 9a99874d1c..f59812eaad 100644 --- a/profiler/src/profile_batched_gemm_gemm.cpp +++ b/profiler/src/profile_batched_gemm_gemm.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_batched_gemm_multi_d.cpp b/profiler/src/profile_batched_gemm_multi_d.cpp index eb94c07900..f64e7ace1a 100644 --- a/profiler/src/profile_batched_gemm_multi_d.cpp +++ b/profiler/src/profile_batched_gemm_multi_d.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_batched_gemm_reduce.cpp b/profiler/src/profile_batched_gemm_reduce.cpp index 9620d63caf..2354c7938c 100644 --- a/profiler/src/profile_batched_gemm_reduce.cpp +++ b/profiler/src/profile_batched_gemm_reduce.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_batchnorm_bwd.cpp b/profiler/src/profile_batchnorm_bwd.cpp index 1738d53dbe..dabe97bfae 100644 --- a/profiler/src/profile_batchnorm_bwd.cpp +++ b/profiler/src/profile_batchnorm_bwd.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_batchnorm_fwd.cpp b/profiler/src/profile_batchnorm_fwd.cpp index 507fb4b450..6a87346c24 100644 --- a/profiler/src/profile_batchnorm_fwd.cpp +++ b/profiler/src/profile_batchnorm_fwd.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_batchnorm_infer.cpp b/profiler/src/profile_batchnorm_infer.cpp index f1c19bc36e..548cba9a42 100644 --- a/profiler/src/profile_batchnorm_infer.cpp +++ b/profiler/src/profile_batchnorm_infer.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_contraction_bilinear.cpp b/profiler/src/profile_contraction_bilinear.cpp index a64555fc66..f38e647c71 100644 --- a/profiler/src/profile_contraction_bilinear.cpp +++ b/profiler/src/profile_contraction_bilinear.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2023-2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_contraction_scale.cpp b/profiler/src/profile_contraction_scale.cpp index a168c09bcf..786db970e8 100644 --- a/profiler/src/profile_contraction_scale.cpp +++ b/profiler/src/profile_contraction_scale.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2023-2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_conv_bwd_data.cpp b/profiler/src/profile_conv_bwd_data.cpp index e08a39aeb0..f4685c1ed5 100644 --- a/profiler/src/profile_conv_bwd_data.cpp +++ b/profiler/src/profile_conv_bwd_data.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_conv_fwd.cpp b/profiler/src/profile_conv_fwd.cpp index 701999d8a9..59b9f11308 100644 --- a/profiler/src/profile_conv_fwd.cpp +++ b/profiler/src/profile_conv_fwd.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_conv_fwd_bias_relu.cpp b/profiler/src/profile_conv_fwd_bias_relu.cpp index 31055ec1d1..717b0f5aa4 100644 --- a/profiler/src/profile_conv_fwd_bias_relu.cpp +++ b/profiler/src/profile_conv_fwd_bias_relu.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_conv_fwd_bias_relu_add.cpp b/profiler/src/profile_conv_fwd_bias_relu_add.cpp index 8c2439a0c7..70a5f9e101 100644 --- a/profiler/src/profile_conv_fwd_bias_relu_add.cpp +++ b/profiler/src/profile_conv_fwd_bias_relu_add.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_conv_tensor_rearrange.cpp b/profiler/src/profile_conv_tensor_rearrange.cpp index 6420698a28..70239e8a80 100644 --- a/profiler/src/profile_conv_tensor_rearrange.cpp +++ b/profiler/src/profile_conv_tensor_rearrange.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm.cpp b/profiler/src/profile_gemm.cpp index c322c7054b..5acc558f4b 100644 --- a/profiler/src/profile_gemm.cpp +++ b/profiler/src/profile_gemm.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_ab_scale.cpp b/profiler/src/profile_gemm_ab_scale.cpp index c2889d5490..09dc3134c0 100644 --- a/profiler/src/profile_gemm_ab_scale.cpp +++ b/profiler/src/profile_gemm_ab_scale.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_add.cpp b/profiler/src/profile_gemm_add.cpp index 749966af1b..85b1770c69 100644 --- a/profiler/src/profile_gemm_add.cpp +++ b/profiler/src/profile_gemm_add.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_add_add_fastgelu.cpp b/profiler/src/profile_gemm_add_add_fastgelu.cpp index 8af3768a48..15fb999b9b 100644 --- a/profiler/src/profile_gemm_add_add_fastgelu.cpp +++ b/profiler/src/profile_gemm_add_add_fastgelu.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_add_fastgelu.cpp b/profiler/src/profile_gemm_add_fastgelu.cpp index f8335d8c05..ab15e2956e 100644 --- a/profiler/src/profile_gemm_add_fastgelu.cpp +++ b/profiler/src/profile_gemm_add_fastgelu.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_add_multiply.cpp b/profiler/src/profile_gemm_add_multiply.cpp index f8ec7abb66..00deaa0953 100644 --- a/profiler/src/profile_gemm_add_multiply.cpp +++ b/profiler/src/profile_gemm_add_multiply.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_add_relu.cpp b/profiler/src/profile_gemm_add_relu.cpp index 025fddc82b..efdd041053 100644 --- a/profiler/src/profile_gemm_add_relu.cpp +++ b/profiler/src/profile_gemm_add_relu.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_add_relu_add_layernorm.cpp b/profiler/src/profile_gemm_add_relu_add_layernorm.cpp index 558d255ce1..44e48acf2a 100644 --- a/profiler/src/profile_gemm_add_relu_add_layernorm.cpp +++ b/profiler/src/profile_gemm_add_relu_add_layernorm.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_add_silu.cpp b/profiler/src/profile_gemm_add_silu.cpp index a35c0a4092..aff9089275 100644 --- a/profiler/src/profile_gemm_add_silu.cpp +++ b/profiler/src/profile_gemm_add_silu.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_b_scale.cpp b/profiler/src/profile_gemm_b_scale.cpp index 7bcc96a434..61064bf21b 100644 --- a/profiler/src/profile_gemm_b_scale.cpp +++ b/profiler/src/profile_gemm_b_scale.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2023-2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_bias_add_reduce.cpp b/profiler/src/profile_gemm_bias_add_reduce.cpp index 76daffbc67..1562a28f1b 100644 --- a/profiler/src/profile_gemm_bias_add_reduce.cpp +++ b/profiler/src/profile_gemm_bias_add_reduce.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_bilinear.cpp b/profiler/src/profile_gemm_bilinear.cpp index 4527a2fa00..cc05be7d4b 100644 --- a/profiler/src/profile_gemm_bilinear.cpp +++ b/profiler/src/profile_gemm_bilinear.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_blockscale_wp.cpp b/profiler/src/profile_gemm_blockscale_wp.cpp index d5f66c0b65..b0f9d20344 100644 --- a/profiler/src/profile_gemm_blockscale_wp.cpp +++ b/profiler/src/profile_gemm_blockscale_wp.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2025, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_fastgelu.cpp b/profiler/src/profile_gemm_fastgelu.cpp index 93573002ef..94905ffb8b 100644 --- a/profiler/src/profile_gemm_fastgelu.cpp +++ b/profiler/src/profile_gemm_fastgelu.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_multi_abd.cpp b/profiler/src/profile_gemm_multi_abd.cpp index 157bcbc977..a1d1555422 100644 --- a/profiler/src/profile_gemm_multi_abd.cpp +++ b/profiler/src/profile_gemm_multi_abd.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2025, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_multiply_add.cpp b/profiler/src/profile_gemm_multiply_add.cpp index 88d3b5256a..46a2d8690a 100644 --- a/profiler/src/profile_gemm_multiply_add.cpp +++ b/profiler/src/profile_gemm_multiply_add.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_multiply_multiply.cpp b/profiler/src/profile_gemm_multiply_multiply.cpp index 87424c21a2..fa6b3832b0 100644 --- a/profiler/src/profile_gemm_multiply_multiply.cpp +++ b/profiler/src/profile_gemm_multiply_multiply.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024-2025, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_multiply_multiply_wp.cpp b/profiler/src/profile_gemm_multiply_multiply_wp.cpp index ff6cffb5f2..b65e0298ef 100644 --- a/profiler/src/profile_gemm_multiply_multiply_wp.cpp +++ b/profiler/src/profile_gemm_multiply_multiply_wp.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_mx.cpp b/profiler/src/profile_gemm_mx.cpp index 9fd6f29464..4734789f8c 100644 --- a/profiler/src/profile_gemm_mx.cpp +++ b/profiler/src/profile_gemm_mx.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2025, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_quantization.cpp b/profiler/src/profile_gemm_quantization.cpp index d28dd60dce..5de19dd1bb 100644 --- a/profiler/src/profile_gemm_quantization.cpp +++ b/profiler/src/profile_gemm_quantization.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2025, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_reduce.cpp b/profiler/src/profile_gemm_reduce.cpp index 48f6f5eb49..ce79102424 100644 --- a/profiler/src/profile_gemm_reduce.cpp +++ b/profiler/src/profile_gemm_reduce.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_splitk.cpp b/profiler/src/profile_gemm_splitk.cpp index 2a0467bc81..0130e8a9db 100644 --- a/profiler/src/profile_gemm_splitk.cpp +++ b/profiler/src/profile_gemm_splitk.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_streamk.cpp b/profiler/src/profile_gemm_streamk.cpp index a0a49eb36d..1f8bfba5d7 100644 --- a/profiler/src/profile_gemm_streamk.cpp +++ b/profiler/src/profile_gemm_streamk.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_universal.cpp b/profiler/src/profile_gemm_universal.cpp index d35cd27651..ded4d68891 100644 --- a/profiler/src/profile_gemm_universal.cpp +++ b/profiler/src/profile_gemm_universal.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2023-2025, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_universal_batched.cpp b/profiler/src/profile_gemm_universal_batched.cpp index d57511fbfc..d1396fa074 100644 --- a/profiler/src/profile_gemm_universal_batched.cpp +++ b/profiler/src/profile_gemm_universal_batched.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_universal_preshuffle.cpp b/profiler/src/profile_gemm_universal_preshuffle.cpp index d8d8f29ac6..3e5f661063 100644 --- a/profiler/src/profile_gemm_universal_preshuffle.cpp +++ b/profiler/src/profile_gemm_universal_preshuffle.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2023-2025, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_universal_reduce.cpp b/profiler/src/profile_gemm_universal_reduce.cpp index cd7423d5de..91187b7523 100644 --- a/profiler/src/profile_gemm_universal_reduce.cpp +++ b/profiler/src/profile_gemm_universal_reduce.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_gemm_universal_streamk.cpp b/profiler/src/profile_gemm_universal_streamk.cpp index 3fd61db138..c9691964ec 100644 --- a/profiler/src/profile_gemm_universal_streamk.cpp +++ b/profiler/src/profile_gemm_universal_streamk.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024-2025, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_grouped_conv_bwd_data.cpp b/profiler/src/profile_grouped_conv_bwd_data.cpp index 95098e2301..62d6e860f9 100644 --- a/profiler/src/profile_grouped_conv_bwd_data.cpp +++ b/profiler/src/profile_grouped_conv_bwd_data.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_grouped_conv_bwd_weight.cpp b/profiler/src/profile_grouped_conv_bwd_weight.cpp index 7d3f1ad6c0..a18aab41a5 100644 --- a/profiler/src/profile_grouped_conv_bwd_weight.cpp +++ b/profiler/src/profile_grouped_conv_bwd_weight.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_grouped_conv_fwd.cpp b/profiler/src/profile_grouped_conv_fwd.cpp index 8400b020f7..c94b77dd4f 100644 --- a/profiler/src/profile_grouped_conv_fwd.cpp +++ b/profiler/src/profile_grouped_conv_fwd.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_grouped_conv_fwd_bias_clamp.cpp b/profiler/src/profile_grouped_conv_fwd_bias_clamp.cpp index fb1eedf2a7..4eb12e6e19 100644 --- a/profiler/src/profile_grouped_conv_fwd_bias_clamp.cpp +++ b/profiler/src/profile_grouped_conv_fwd_bias_clamp.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2025, Advanced Micro Devices, Inc. All rights reserved. #include "ck/tensor_operation/gpu/device/tensor_layout.hpp" #include "profiler/profile_grouped_conv_fwd_bias_clamp_impl.hpp" diff --git a/profiler/src/profile_grouped_conv_fwd_clamp.cpp b/profiler/src/profile_grouped_conv_fwd_clamp.cpp index 1b100ff867..7df9fd6167 100644 --- a/profiler/src/profile_grouped_conv_fwd_clamp.cpp +++ b/profiler/src/profile_grouped_conv_fwd_clamp.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2025, Advanced Micro Devices, Inc. All rights reserved. #include "ck/tensor_operation/gpu/device/tensor_layout.hpp" #include "profiler/profile_grouped_conv_fwd_impl.hpp" diff --git a/profiler/src/profile_grouped_conv_fwd_outelementop.cpp b/profiler/src/profile_grouped_conv_fwd_outelementop.cpp index 196a2cf3f2..00b4bd8f13 100644 --- a/profiler/src/profile_grouped_conv_fwd_outelementop.cpp +++ b/profiler/src/profile_grouped_conv_fwd_outelementop.cpp @@ -1,3 +1,6 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. +// SPDX-License-Identifier: MIT + #include "ck/tensor_operation/gpu/device/tensor_layout.hpp" #include "profiler/profile_grouped_conv_fwd_outelementop_impl.hpp" diff --git a/profiler/src/profile_grouped_gemm.cpp b/profiler/src/profile_grouped_gemm.cpp index 2adcd6483a..32a390c475 100644 --- a/profiler/src/profile_grouped_gemm.cpp +++ b/profiler/src/profile_grouped_gemm.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_grouped_gemm_fastgelu.cpp b/profiler/src/profile_grouped_gemm_fastgelu.cpp index 50ecf25cae..7bcf2d40d2 100644 --- a/profiler/src/profile_grouped_gemm_fastgelu.cpp +++ b/profiler/src/profile_grouped_gemm_fastgelu.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_grouped_gemm_fixed_nk.cpp b/profiler/src/profile_grouped_gemm_fixed_nk.cpp index 827866ce69..c93725c3e7 100644 --- a/profiler/src/profile_grouped_gemm_fixed_nk.cpp +++ b/profiler/src/profile_grouped_gemm_fixed_nk.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_grouped_gemm_multiply_tile_loop.cpp b/profiler/src/profile_grouped_gemm_multiply_tile_loop.cpp index 5cf0af5ecf..e747cfd5fe 100644 --- a/profiler/src/profile_grouped_gemm_multiply_tile_loop.cpp +++ b/profiler/src/profile_grouped_gemm_multiply_tile_loop.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_grouped_gemm_tile_loop.cpp b/profiler/src/profile_grouped_gemm_tile_loop.cpp index 76ff9e162e..ba4b9ba708 100644 --- a/profiler/src/profile_grouped_gemm_tile_loop.cpp +++ b/profiler/src/profile_grouped_gemm_tile_loop.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_groupnorm_bwd_data.cpp b/profiler/src/profile_groupnorm_bwd_data.cpp index f9fea1db55..2e05fc5b2d 100644 --- a/profiler/src/profile_groupnorm_bwd_data.cpp +++ b/profiler/src/profile_groupnorm_bwd_data.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_groupnorm_bwd_gamma_beta.cpp b/profiler/src/profile_groupnorm_bwd_gamma_beta.cpp index 7fcef3a4e2..31878eb3c2 100644 --- a/profiler/src/profile_groupnorm_bwd_gamma_beta.cpp +++ b/profiler/src/profile_groupnorm_bwd_gamma_beta.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_groupnorm_fwd.cpp b/profiler/src/profile_groupnorm_fwd.cpp index 9a595bf7a7..c81a9f94a9 100644 --- a/profiler/src/profile_groupnorm_fwd.cpp +++ b/profiler/src/profile_groupnorm_fwd.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_layernorm_bwd_data.cpp b/profiler/src/profile_layernorm_bwd_data.cpp index 1f364d79ba..fb2d8d6419 100644 --- a/profiler/src/profile_layernorm_bwd_data.cpp +++ b/profiler/src/profile_layernorm_bwd_data.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_layernorm_bwd_gamma_beta.cpp b/profiler/src/profile_layernorm_bwd_gamma_beta.cpp index 0f3436c663..aea4d435fe 100644 --- a/profiler/src/profile_layernorm_bwd_gamma_beta.cpp +++ b/profiler/src/profile_layernorm_bwd_gamma_beta.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_layernorm_fwd.cpp b/profiler/src/profile_layernorm_fwd.cpp index 7031b36531..014a9d0164 100644 --- a/profiler/src/profile_layernorm_fwd.cpp +++ b/profiler/src/profile_layernorm_fwd.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_max_pool2d_bwd.cpp b/profiler/src/profile_max_pool2d_bwd.cpp index 26e84c880e..0058256acf 100644 --- a/profiler/src/profile_max_pool2d_bwd.cpp +++ b/profiler/src/profile_max_pool2d_bwd.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_max_pool2d_fwd.cpp b/profiler/src/profile_max_pool2d_fwd.cpp index 9ed96b6c40..c74f6ab8b9 100644 --- a/profiler/src/profile_max_pool2d_fwd.cpp +++ b/profiler/src/profile_max_pool2d_fwd.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_max_pool3d_bwd.cpp b/profiler/src/profile_max_pool3d_bwd.cpp index 45a64df423..dd890ef280 100644 --- a/profiler/src/profile_max_pool3d_bwd.cpp +++ b/profiler/src/profile_max_pool3d_bwd.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_permute_scale.cpp b/profiler/src/profile_permute_scale.cpp index 8ebb2289ed..e6908989b6 100644 --- a/profiler/src/profile_permute_scale.cpp +++ b/profiler/src/profile_permute_scale.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_pool3d_fwd.cpp b/profiler/src/profile_pool3d_fwd.cpp index 4ea1fbcf49..21896c7e48 100644 --- a/profiler/src/profile_pool3d_fwd.cpp +++ b/profiler/src/profile_pool3d_fwd.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_reduce.cpp b/profiler/src/profile_reduce.cpp index e4af5680a5..dbdc30cf2e 100644 --- a/profiler/src/profile_reduce.cpp +++ b/profiler/src/profile_reduce.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_softmax.cpp b/profiler/src/profile_softmax.cpp index dfe8d95c90..096a2d4eb4 100644 --- a/profiler/src/profile_softmax.cpp +++ b/profiler/src/profile_softmax.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profile_transpose.cpp b/profiler/src/profile_transpose.cpp index 5aa9b34a1c..c618135608 100644 --- a/profiler/src/profile_transpose.cpp +++ b/profiler/src/profile_transpose.cpp @@ -1,81 +1,81 @@ -// SPDX-License-Identifier: MIT -// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. - -#include -#include -#include -#include - -#include "profiler/profile_transpose_impl.hpp" -#include "profiler_operation_registry.hpp" - -enum struct DataType -{ - F32_F32_F32_F32_F32, // 0 - F16_F16_F16_F16_F16, // 1 -}; - -#define OP_NAME "transpose" -#define OP_DESC "Transpose" - -static void print_helper_msg() -{ - printf("arg1: tensor operation (" OP_NAME ": " OP_DESC ")\n"); - printf("arg2: data type (0: fp32; 1: fp16)\n"); - printf("arg3: verification (0: no; 1: yes)\n"); - printf("arg4: initialization (0: no init; 1: integer value; 2: decimal value)\n"); - printf("arg5: print tensor value (0: no; 1: yes)\n"); - printf("arg6: time kernel (0=no, 1=yes)\n"); - printf("arg7 to arg11: N, C, D, H, W\n"); -} - -int profile_transpose(int argc, char* argv[]) -{ - if(argc != 12) - { - print_helper_msg(); - exit(1); - } - - const auto data_type = static_cast(std::stoi(argv[2])); - const bool do_verification = std::stoi(argv[3]); - const int init_method = std::stoi(argv[4]); - const bool do_log = std::stoi(argv[5]); - const bool time_kernel = std::stoi(argv[6]); - const std::vector lengths = {std::stoi(argv[7]), - std::stoi(argv[8]), - std::stoi(argv[9]), - std::stoi(argv[10]), - std::stoi(argv[11])}; - - using F32 = float; - using F16 = ck::half_t; - - auto profile = [&](auto a_type, auto b_type) { - using ADataType = decltype(a_type); - using BDataType = decltype(b_type); - constexpr ck::index_t NumDim = 5; - - bool pass = ck::profiler::profile_transpose_impl( - do_verification, init_method, do_log, time_kernel, lengths); - - return pass ? 0 : 1; - }; - - if(data_type == DataType::F32_F32_F32_F32_F32) - { - return profile(F32{}, F32{}); - } - else if(data_type == DataType::F16_F16_F16_F16_F16) - { - return profile(F16{}, F16{}); - } - else - { - std::cout << "this data_type & layout is not implemented" << std::endl; - - return 1; - } -} - -REGISTER_PROFILER_OPERATION(OP_NAME, OP_DESC, profile_transpose); +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. +// SPDX-License-Identifier: MIT + +#include +#include +#include +#include + +#include "profiler/profile_transpose_impl.hpp" +#include "profiler_operation_registry.hpp" + +enum struct DataType +{ + F32_F32_F32_F32_F32, // 0 + F16_F16_F16_F16_F16, // 1 +}; + +#define OP_NAME "transpose" +#define OP_DESC "Transpose" + +static void print_helper_msg() +{ + printf("arg1: tensor operation (" OP_NAME ": " OP_DESC ")\n"); + printf("arg2: data type (0: fp32; 1: fp16)\n"); + printf("arg3: verification (0: no; 1: yes)\n"); + printf("arg4: initialization (0: no init; 1: integer value; 2: decimal value)\n"); + printf("arg5: print tensor value (0: no; 1: yes)\n"); + printf("arg6: time kernel (0=no, 1=yes)\n"); + printf("arg7 to arg11: N, C, D, H, W\n"); +} + +int profile_transpose(int argc, char* argv[]) +{ + if(argc != 12) + { + print_helper_msg(); + exit(1); + } + + const auto data_type = static_cast(std::stoi(argv[2])); + const bool do_verification = std::stoi(argv[3]); + const int init_method = std::stoi(argv[4]); + const bool do_log = std::stoi(argv[5]); + const bool time_kernel = std::stoi(argv[6]); + const std::vector lengths = {std::stoi(argv[7]), + std::stoi(argv[8]), + std::stoi(argv[9]), + std::stoi(argv[10]), + std::stoi(argv[11])}; + + using F32 = float; + using F16 = ck::half_t; + + auto profile = [&](auto a_type, auto b_type) { + using ADataType = decltype(a_type); + using BDataType = decltype(b_type); + constexpr ck::index_t NumDim = 5; + + bool pass = ck::profiler::profile_transpose_impl( + do_verification, init_method, do_log, time_kernel, lengths); + + return pass ? 0 : 1; + }; + + if(data_type == DataType::F32_F32_F32_F32_F32) + { + return profile(F32{}, F32{}); + } + else if(data_type == DataType::F16_F16_F16_F16_F16) + { + return profile(F16{}, F16{}); + } + else + { + std::cout << "this data_type & layout is not implemented" << std::endl; + + return 1; + } +} + +REGISTER_PROFILER_OPERATION(OP_NAME, OP_DESC, profile_transpose); diff --git a/profiler/src/profiler.cpp b/profiler/src/profiler.cpp index 0f528c008f..1631b89716 100644 --- a/profiler/src/profiler.cpp +++ b/profiler/src/profiler.cpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include diff --git a/profiler/src/profiler_operation_registry.hpp b/profiler/src/profiler_operation_registry.hpp index 7e6d22d4ce..de78a7f65d 100644 --- a/profiler/src/profiler_operation_registry.hpp +++ b/profiler/src/profiler_operation_registry.hpp @@ -1,5 +1,5 @@ +// Copyright (c) Advanced Micro Devices, Inc., or its affiliates. // SPDX-License-Identifier: MIT -// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved. #include #include