From 275f99cb09606191c5589952d57175be655de74a Mon Sep 17 00:00:00 2001 From: Georgios Pinitas Date: Fri, 23 Aug 2019 12:44:11 +0100 Subject: COMPMID-2451: Use kernel lws_hint() on enqueue Avoid querying device's default lws on kernel enqueue as this is already cached in the kernel during configuration. Change-Id: Ia26ecb712caeb8f042356815e0cfd23522764d27 Signed-off-by: Georgios Pinitas Reviewed-on: https://review.mlplatform.org/c/1803 Reviewed-by: Michele Di Giorgio Comments-Addressed: Arm Jenkins Tested-by: Arm Jenkins --- src/core/CL/kernels/CLReductionOperationKernel.cpp | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'src/core/CL/kernels/CLReductionOperationKernel.cpp') diff --git a/src/core/CL/kernels/CLReductionOperationKernel.cpp b/src/core/CL/kernels/CLReductionOperationKernel.cpp index 9db8ae6cde..ee33a72792 100644 --- a/src/core/CL/kernels/CLReductionOperationKernel.cpp +++ b/src/core/CL/kernels/CLReductionOperationKernel.cpp @@ -285,7 +285,7 @@ void CLReductionOperationKernel::run(const Window &window, cl::CommandQueue &que unsigned int idx = 0; add_1D_tensor_argument(idx, _input, in_slice); add_1D_tensor_argument(idx, _output, out_slice); - enqueue(queue, *this, in_slice); + enqueue(queue, *this, in_slice, lws_hint()); } while(window_in.slide_window_slice_1D(in_slice) && window.slide_window_slice_1D(out_slice)); } @@ -331,7 +331,7 @@ void CLReductionOperationKernel::run(const Window &window, cl::CommandQueue &que unsigned int idx = 0; add_2D_tensor_argument(idx, _input, in_slice); add_2D_tensor_argument(idx, _output, out_slice); - enqueue(queue, *this, in_slice); + enqueue(queue, *this, in_slice, lws_hint()); } while(window_in.slide_window_slice_2D(in_slice) && window.slide_window_slice_2D(out_slice)); } @@ -349,7 +349,7 @@ void CLReductionOperationKernel::run(const Window &window, cl::CommandQueue &que unsigned int idx = 0; add_3D_tensor_argument(idx, _input, in_slice); add_3D_tensor_argument(idx, _output, out_slice); - enqueue(queue, *this, in_slice); + enqueue(queue, *this, in_slice, lws_hint()); } while(window_in.slide_window_slice_3D(in_slice) && window.slide_window_slice_3D(out_slice)); } @@ -367,7 +367,7 @@ void CLReductionOperationKernel::run(const Window &window, cl::CommandQueue &que unsigned int idx = 0; add_4D_tensor_argument(idx, _input, in_slice); add_4D_tensor_argument(idx, _output, out_slice); - enqueue(queue, *this, in_slice); + enqueue(queue, *this, in_slice, lws_hint()); } while(window_in.slide_window_slice_4D(in_slice) && window.slide_window_slice_4D(out_slice)); } -- cgit v1.2.1