From ada6cbc057ff725e57d301a99a1816ce602485b9 Mon Sep 17 00:00:00 2001 From: Giorgio Arena Date: Fri, 16 Apr 2021 17:03:39 +0100 Subject: Remove OpenCL padding: CLPixelWiseMultiplicationKernel - Change kernel's vec_size to 16 / sizeof(output) - Change ICLKernel.cpp to handle broadcast without padding Resolve COMPMID-3913 Signed-off-by: Giorgio Arena Change-Id: I03e884b250ef5784dc109bff8cf2c96b345d119f Signed-off-by: Giorgio Arena Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/5450 Tested-by: Arm Jenkins Comments-Addressed: Arm Jenkins Reviewed-by: Michele Di Giorgio Reviewed-by: Gian Marco Iodice --- src/core/CL/ICLKernel.cpp | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'src/core/CL/ICLKernel.cpp') diff --git a/src/core/CL/ICLKernel.cpp b/src/core/CL/ICLKernel.cpp index 1c6963f3f1..9ba17d0e03 100644 --- a/src/core/CL/ICLKernel.cpp +++ b/src/core/CL/ICLKernel.cpp @@ -105,8 +105,8 @@ void ICLKernel::add_tensor_argument(unsigned &idx, const ICLTensor *tensor, cons for(unsigned int d = 0; d < dimension_size; ++d) { - _kernel.setArg(idx++, strides[d]); - _kernel.setArg(idx++, strides[d] * window[d].step()); + _kernel.setArg(idx++, window.is_broadcasted(d) ? 0 : strides[d]); + _kernel.setArg(idx++, window.is_broadcasted(d) ? 0 : (strides[d] * window[d].step())); } _kernel.setArg(idx++, offset_first_element); -- cgit v1.2.1