diff options
| author | ramelg01 <ramy.elgammal@arm.com> | 2022-02-04 20:49:14 +0000 |
|---|---|---|
| committer | Ramy Elgammal <ramy.elgammal@arm.com> | 2022-02-09 10:14:47 +0000 |
| commit | 2a86a30a5d9c047b0ec73a699b09a128f3fcb55e (patch) | |
| tree | 4421ab41ba31bd916a0331b1011184110bd51ff4 /src/gpu/cl/kernels/ClWinogradInputTransformKernel.h | |
| parent | ba8690b4f0f82dfb66e3da819361e4032e9fa4db (diff) | |
| download | ComputeLibrary-2a86a30a5d9c047b0ec73a699b09a128f3fcb55e.tar.gz | |
Improve start-up time for winograd_input_transform_*_nhwc
- pass tensor's dimensions at runtime rather than compile time
- Add guard macro to compile only kernel(s) of internest
Resolves: COMPMID-5119
Signed-off-by: Ramy Elgammal <ramy.elgammal@arm.com>
Change-Id: Ib01098e397011a1201c2800c62a8954ec70e63e8
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/7083
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Giorgio Arena <giorgio.arena@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/gpu/cl/kernels/ClWinogradInputTransformKernel.h')
| -rw-r--r-- | src/gpu/cl/kernels/ClWinogradInputTransformKernel.h | 4 |
1 files changed, 3 insertions, 1 deletions
diff --git a/src/gpu/cl/kernels/ClWinogradInputTransformKernel.h b/src/gpu/cl/kernels/ClWinogradInputTransformKernel.h index 631f427b82..c10c528b9b 100644 --- a/src/gpu/cl/kernels/ClWinogradInputTransformKernel.h +++ b/src/gpu/cl/kernels/ClWinogradInputTransformKernel.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2018-2021 Arm Limited. + * Copyright (c) 2018-2022 Arm Limited. * * SPDX-License-Identifier: MIT * @@ -80,6 +80,8 @@ private: int _num_tiles_x{ 0 }; int _num_tiles_y{ 0 }; unsigned int _step_z{ 1 }; + int32_t _src_width{ 0 }; + int32_t _src_height{ 0 }; }; } // namespace kernels } // namespace opencl |