diff options
| author | Michalis Spyrou <michalis.spyrou@arm.com> | 2021-03-19 15:06:07 +0000 |
|---|---|---|
| committer | Georgios Pinitas <georgios.pinitas@arm.com> | 2021-03-29 19:48:22 +0000 |
| commit | 702dc0c71f2b2830b63e3b4079ede0ef76377f0a (patch) | |
| tree | afb0dacd74a8650350802aea8043822503c2f8b2 /src/core/CL/kernels/CLDepthConvertLayerKernel.cpp | |
| parent | 26c5c0fc10cbee1b523af0b27f884719dc50820a (diff) | |
| download | ComputeLibrary-702dc0c71f2b2830b63e3b4079ede0ef76377f0a.tar.gz | |
Remove usage of valid window region CL - NHWC
Resolves: COMPMID-4153
Change-Id: Ib0d60c9acaac8aaf3946c62fc2d740b5ec6cee5c
Signed-off-by: Michalis Spyrou <michalis.spyrou@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/5301
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/core/CL/kernels/CLDepthConvertLayerKernel.cpp')
| -rw-r--r-- | src/core/CL/kernels/CLDepthConvertLayerKernel.cpp | 10 |
1 files changed, 1 insertions, 9 deletions
diff --git a/src/core/CL/kernels/CLDepthConvertLayerKernel.cpp b/src/core/CL/kernels/CLDepthConvertLayerKernel.cpp index c98d66f390..0d5c7a4881 100644 --- a/src/core/CL/kernels/CLDepthConvertLayerKernel.cpp +++ b/src/core/CL/kernels/CLDepthConvertLayerKernel.cpp @@ -1,5 +1,5 @@ /* - * Copyright (c) 2016-2020 Arm Limited. + * Copyright (c) 2016-2021 Arm Limited. * * SPDX-License-Identifier: MIT * @@ -118,11 +118,6 @@ void CLDepthConvertLayerKernel::configure(const CLCompileContext &compile_contex unsigned int idx = 2 * num_arguments_per_3D_tensor(); // Skip the input and output parameters _kernel.setArg(idx++, shift); - // Since we have a leftover vector size calculated using the input tensor shape, it is required to - // have the input region equal to the tensor shape - ValidRegion input_valid_region = input->info()->valid_region(); - input->info()->set_valid_region(ValidRegion(Coordinates(0, 0), input->info()->tensor_shape())); - // Configure kernel Window win = calculate_max_window(*input->info(), Steps(num_elems_processed_per_iteration)); ICLKernel::configure_internal(win); @@ -132,9 +127,6 @@ void CLDepthConvertLayerKernel::configure(const CLCompileContext &compile_contex Window collapsed_window = full_window.collapse_if_possible(full_window, Window::DimZ); ICLKernel::configure_internal(collapsed_window); - // Restore the valid region - input->info()->set_valid_region(input_valid_region); - ARM_COMPUTE_ERROR_ON(has_padding_changed(padding_info)); // Set config_id for enabling LWS tuning |