diff options
author | Michalis Spyrou <michalis.spyrou@arm.com> | 2018-01-26 15:06:19 +0000 |
---|---|---|
committer | Anthony Barbier <anthony.barbier@arm.com> | 2018-11-02 16:44:23 +0000 |
commit | f6402dd37092c842d1de9998b23640caf12f227b (patch) | |
tree | 217284ffe083ea6af5cb5bc2e7f348090122ff16 /src/core/CL/kernels | |
parent | bd0e61238b2126e990d7811750ad4511ec2ccbd1 (diff) | |
download | ComputeLibrary-f6402dd37092c842d1de9998b23640caf12f227b.tar.gz |
COMPMID-834 Fix arm_compute_nightly_validation getting killed
Changed CLReductionOperationKernel: Now each kernel computes
a 2D slice instead of 1D. This reduces the memory footprint
from around 1.6Gb for a 4k input image to a few Mb, which was
caused by the __local memory and was probably the cause for this bug.
Change-Id: I71ac71ff09b041c945a134177600f0f3475e48cf
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/117835
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Anthony Barbier <anthony.barbier@arm.com>
Diffstat (limited to 'src/core/CL/kernels')
-rw-r--r-- | src/core/CL/kernels/CLReductionOperationKernel.cpp | 14 |
1 files changed, 7 insertions, 7 deletions
diff --git a/src/core/CL/kernels/CLReductionOperationKernel.cpp b/src/core/CL/kernels/CLReductionOperationKernel.cpp index 18a8e353d7..1dd5eb97ec 100644 --- a/src/core/CL/kernels/CLReductionOperationKernel.cpp +++ b/src/core/CL/kernels/CLReductionOperationKernel.cpp @@ -1,5 +1,5 @@ /* - * Copyright (c) 2017 ARM Limited. + * Copyright (c) 2017-2018 ARM Limited. * * SPDX-License-Identifier: MIT * @@ -118,8 +118,8 @@ void CLReductionOperationKernel::run(const Window &window, cl::CommandQueue &que out_window.set(Window::DimX, Window::Dimension(0, 0, 0)); // Get first input and output slices - Window in_slice = window.first_slice_window_1D(); - Window out_slice = out_window.first_slice_window_1D(); + Window in_slice = window.first_slice_window_2D(); + Window out_slice = out_window.first_slice_window_2D(); // Reshape window const unsigned int border_width = ((in_slice.x().end() % 128) != 0) ? 128 - in_slice.x().end() % 128 : 0; @@ -127,14 +127,14 @@ void CLReductionOperationKernel::run(const Window &window, cl::CommandQueue &que // Set local sums buffer unsigned int local_sum_size = _lws_hint[0] * _input->info()->element_size(); - _kernel.setArg(num_arguments_per_1D_tensor() * 2, local_sum_size, nullptr); + _kernel.setArg(num_arguments_per_2D_tensor() * 2, local_sum_size, nullptr); do { unsigned int idx = 0; - add_1D_tensor_argument(idx, _input, in_slice); - add_1D_tensor_argument(idx, _output, out_slice); + add_2D_tensor_argument(idx, _input, in_slice); + add_2D_tensor_argument(idx, _output, out_slice); enqueue(queue, *this, in_slice, _lws_hint); } - while(window.slide_window_slice_1D(in_slice) && window.slide_window_slice_1D(out_slice)); + while(window.slide_window_slice_2D(in_slice) && window.slide_window_slice_2D(out_slice)); } |