diff options
author | Michele Di Giorgio <michele.digiorgio@arm.com> | 2018-01-17 17:29:33 +0000 |
---|---|---|
committer | Anthony Barbier <anthony.barbier@arm.com> | 2018-11-02 16:45:42 +0000 |
commit | 6259e5f9204abf31b811b1d002f68ce6504197bd (patch) | |
tree | 2dac943b3c794b66ccd90c8dc8e15d47699c5ea8 /arm_compute/runtime/CL/functions | |
parent | 19d0547aa8c60b95766c195822769c7fea78aeaa (diff) | |
download | ComputeLibrary-6259e5f9204abf31b811b1d002f68ce6504197bd.tar.gz |
COMPMID-787: Add CL support for broadcast multiply
Change-Id: I71f67789648ef05ccdedce77c7427bc0127b3a69
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/116741
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Diffstat (limited to 'arm_compute/runtime/CL/functions')
-rw-r--r-- | arm_compute/runtime/CL/functions/CLPixelWiseMultiplication.h | 20 |
1 files changed, 11 insertions, 9 deletions
diff --git a/arm_compute/runtime/CL/functions/CLPixelWiseMultiplication.h b/arm_compute/runtime/CL/functions/CLPixelWiseMultiplication.h index d57bfda2c1..75b67cd17c 100644 --- a/arm_compute/runtime/CL/functions/CLPixelWiseMultiplication.h +++ b/arm_compute/runtime/CL/functions/CLPixelWiseMultiplication.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2016, 2017 ARM Limited. + * Copyright (c) 2016-2018 ARM Limited. * * SPDX-License-Identifier: MIT * @@ -37,15 +37,17 @@ class CLPixelWiseMultiplication : public ICLSimpleFunction public: /** Initialise the kernel's inputs, output and convertion policy. * - * @param[in] input1 An input tensor. Data types supported: U8/QS8/QS16/S16/F16/F32. - * @param[in] input2 An input tensor. Data types supported: same as @p input1. - * @param[out] output The output tensor, Data types supported: same as @p input1. Note: U8 (QS8, QS16) requires both inputs to be U8 (QS8, QS16). - * @param[in] scale Scale to apply after multiplication. - * Scale must be positive and its value must be either 1/255 or 1/2^n where n is between 0 and 15. For QS8 and QS16 scale must be 1. - * @param[in] overflow_policy Overflow policy. Supported overflow policies: Wrap, Saturate - * @param[in] rounding_policy Rounding policy. Supported rounding modes: to zero, to nearest even. + * @param[in, out] input1 An input tensor. Data types supported: U8/QS8/QS16/S16/F16/F32. + * The input tensor is [in, out] because its TensorInfo might be modified inside the kernel in case of broadcasting of dimension 0. + * @param[in, out] input2 An input tensor. Data types supported: same as @p input1. + * The input tensor is [in, out] because its TensorInfo might be modified inside the kernel in case of broadcasting of dimension 0. + * @param[out] output The output tensor, Data types supported: same as @p input1. Note: U8 (QS8, QS16) requires both inputs to be U8 (QS8, QS16). + * @param[in] scale Scale to apply after multiplication. + * Scale must be positive and its value must be either 1/255 or 1/2^n where n is between 0 and 15. For QS8 and QS16 scale must be 1. + * @param[in] overflow_policy Overflow policy. Supported overflow policies: Wrap, Saturate + * @param[in] rounding_policy Rounding policy. Supported rounding modes: to zero, to nearest even. */ - void configure(const ICLTensor *input1, const ICLTensor *input2, ICLTensor *output, float scale, + void configure(ICLTensor *input1, ICLTensor *input2, ICLTensor *output, float scale, ConvertPolicy overflow_policy, RoundingPolicy rounding_policy); /** Static function to check if given info will lead to a valid configuration of @ref CLPixelWiseMultiplication * |