From 6eb73458c4869165c88d33c6a745a91cdc73a36a Mon Sep 17 00:00:00 2001 From: Michalis Spyrou Date: Thu, 2 Jul 2020 17:39:25 +0100 Subject: COMPMID-3373: Async support to NEArithmetic* kernels/functions (Pt. 2) Signed-off-by: Michalis Spyrou Change-Id: Iec06adb535aaf7efb1838d921e8d6bb978b7b215 Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/3498 Tested-by: Arm Jenkins Reviewed-by: Michele Di Giorgio Comments-Addressed: Arm Jenkins --- .../NEON/functions/NEPixelWiseMultiplication.h | 139 ++++++++++++++++++++- 1 file changed, 134 insertions(+), 5 deletions(-) (limited to 'arm_compute/runtime/NEON/functions/NEPixelWiseMultiplication.h') diff --git a/arm_compute/runtime/NEON/functions/NEPixelWiseMultiplication.h b/arm_compute/runtime/NEON/functions/NEPixelWiseMultiplication.h index d84dff2c13..3b1209356a 100644 --- a/arm_compute/runtime/NEON/functions/NEPixelWiseMultiplication.h +++ b/arm_compute/runtime/NEON/functions/NEPixelWiseMultiplication.h @@ -25,15 +25,17 @@ #define ARM_COMPUTE_NEPIXELWISEMULTIPLICATION_H #include "arm_compute/core/Types.h" -#include "arm_compute/runtime/NEON/INESimpleFunction.h" -#include "arm_compute/runtime/NEON/INESimpleFunctionNoBorder.h" +#include "arm_compute/runtime/IFunction.h" +#include "arm_compute/runtime/NEON/INEOperator.h" namespace arm_compute { class ITensor; +namespace experimental +{ /** Basic function to run @ref NEPixelWiseMultiplicationKernel */ -class NEPixelWiseMultiplication : public INESimpleFunctionNoBorder +class NEPixelWiseMultiplication : public INEOperator { public: /** Initialise the kernel's inputs, output and convertion policy. @@ -60,7 +62,7 @@ public: * @param[in] rounding_policy Rounding policy. * @param[in] act_info (Optional) Activation layer information in case of a fused activation. Currently not supported. */ - void configure(ITensor *input1, ITensor *input2, ITensor *output, float scale, ConvertPolicy overflow_policy, RoundingPolicy rounding_policy, + void configure(ITensorInfo *input1, ITensorInfo *input2, ITensorInfo *output, float scale, ConvertPolicy overflow_policy, RoundingPolicy rounding_policy, const ActivationLayerInfo &act_info = ActivationLayerInfo()); /** Static function to check if given info will lead to a valid configuration of @ref NEPixelWiseMultiplication * @@ -88,12 +90,132 @@ public: */ static Status validate(const ITensorInfo *input1, const ITensorInfo *input2, const ITensorInfo *output, float scale, ConvertPolicy overflow_policy, RoundingPolicy rounding_policy, const ActivationLayerInfo &act_info = ActivationLayerInfo()); + + // Inherited methods overridden: + MemoryRequirements workspace() const override; }; /** Basic function to run @ref NEComplexPixelWiseMultiplicationKernel. */ -class NEComplexPixelWiseMultiplication : public INESimpleFunction +class NEComplexPixelWiseMultiplication : public INEOperator { public: + /** Initialise the kernel's inputs, output. + * + * @param[in, out] input1 An input tensor. Data types supported: F32. Number of channels supported: 2 (complex tensor). + * 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. Number of channels 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. Number of channels: same as @p input1. + * @param[in] act_info (Optional) Activation layer information in case of a fused activation. Currently not supported. + */ + void configure(ITensorInfo *input1, ITensorInfo *input2, ITensorInfo *output, const ActivationLayerInfo &act_info = ActivationLayerInfo()); + /** Static function to check if given info will lead to a valid configuration of @ref NEComplexPixelWiseMultiplication + * + * @param[in] input1 An input tensor info. Data types supported: F32. Number of channels supported: 2 (complex tensor). + * @param[in] input2 An input tensor info. Data types supported: same as @p input1. Number of channels supported: same as @p input1. + * @param[in] output The output tensor info. Data types supported: same as @p input1. Number of channels supported: same as @p input1. + * @param[in] act_info (Optional) Activation layer information in case of a fused activation. Currently not supported. + */ + static Status validate(const ITensorInfo *input1, const ITensorInfo *input2, const ITensorInfo *output, const ActivationLayerInfo &act_info = ActivationLayerInfo()); + + // Inherited methods overridden: + MemoryRequirements workspace() const override; +}; +} // namespace experimental + +/** Basic function to run @ref NEPixelWiseMultiplicationKernel */ +class NEPixelWiseMultiplication : public IFunction +{ +public: + /** Default Constructor */ + NEPixelWiseMultiplication(); + /** Default Destructor */ + ~NEPixelWiseMultiplication(); + /** Prevent instances of this class from being copied (As this class contains pointers) */ + NEPixelWiseMultiplication(const NEPixelWiseMultiplication &) = delete; + /** Default move constructor */ + NEPixelWiseMultiplication(NEPixelWiseMultiplication &&); + /** Prevent instances of this class from being copied (As this class contains pointers) */ + NEPixelWiseMultiplication &operator=(const NEPixelWiseMultiplication &) = delete; + /** Default move assignment operator */ + NEPixelWiseMultiplication &operator=(NEPixelWiseMultiplication &&); + /** Initialise the kernel's inputs, output and convertion policy. + * + * @note For @p scale equal to 1/255 only round to nearest even (implemented as round half up) is supported. + * For all other scale values only round to zero (implemented as round towards minus infinity) is supported. + * + * @param[in, out] input1 An input tensor. Data types supported: U8/QASYMM8/QASYMM8_SIGNED/S16/QSYMM16/F16/F32 + * This 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: U8, QASYMM8 (only if @p input1 is QASYMM8), QASYMM8_SIGNED (only if @p input1 is QASYMM8_SIGNED), S16, QSYMM16 (only if @p input1 is QSYMM16), F16 (only if @p input1 is F16), F32 (only if @p input1 is F32). + * This input tensor is [in, out] because its TensorInfo might be modified inside the kernel in case of broadcasting of dimension 0. + * @param[out] output Output tensor. Data types supported: + * - U8, only if both inputs are U8. + * - QASYMM8, only if both inputs are QASYMM8. + * - QASYMM8_SIGNED, only if @p input1 is QASYMM8_SIGNED. + * - S16. + * - QSYMM16, only if both inputs are QSYMM16. + * - S32, only if both inputs are QSYMM16. + * - F16, only if @p input1 is F16. + * - F32, only if both inputs are F32. + * @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. + * @param[in] overflow_policy Overflow policy. ConvertPolicy cannot be WRAP if datatype is QASYMM8, QASYMM8_SIGNED or QSYMM16. + * @param[in] rounding_policy Rounding policy. + * @param[in] act_info (Optional) Activation layer information in case of a fused activation. Currently not supported. + */ + void configure(const ITensor *input1, const ITensor *input2, ITensor *output, float scale, ConvertPolicy overflow_policy, RoundingPolicy rounding_policy, + const ActivationLayerInfo &act_info = ActivationLayerInfo()); + /** Static function to check if given info will lead to a valid configuration of @ref NEPixelWiseMultiplication + * + * @note For @p scale equal to 1/255 only round to nearest even (implemented as round half up) is supported. + * For all other scale values only round to zero (implemented as round towards minus infinity) is supported. + * + * @param[in] input1 An input tensor info. Data types supported: U8/QASYMM8/QASYMM8_SIGNED/S16/QSYMM16/F16/F32 + * @param[in] input2 An input tensor info. Data types supported: U8, QASYMM8 (only if @p input1 is QASYMM8), QASYMM8_SIGNED (only if @p input1 is QASYMM8_SIGNED), S16, QSYMM16 (only if both inputs are QSYMM16), F16 (only if @p input1 is F16), F32 (only if @p input1 is F32). + * @param[in] output Output tensor info. Data types supported: + * - U8, only if both inputs are U8. + * - QASYMM8, only if both inputs are QASYMM8. + * - QASYMM8_SIGNED, only if @p input1 is QASYMM8_SIGNED. + * - S16. + * - QSYMM16, only if both inputs are QSYMM16. + * - S32, only if both inputs are QSYMM16. + * - F16, only if @p input1 is F16. + * - F32, only if both inputs are F32. + * @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. + * @param[in] overflow_policy Overflow policy. ConvertPolicy cannot be WRAP if datatype is QASYMM8, QASYMM8_SIGNED or QSYMM16. + * @param[in] rounding_policy Rounding policy. + * @param[in] act_info (Optional) Activation layer information in case of a fused activation. Currently not supported. + * + * @return a status + */ + static Status validate(const ITensorInfo *input1, const ITensorInfo *input2, const ITensorInfo *output, float scale, ConvertPolicy overflow_policy, RoundingPolicy rounding_policy, + const ActivationLayerInfo &act_info = ActivationLayerInfo()); + + // Inherited methods overridden: + void run() override; + +private: + struct Impl; + std::unique_ptr _impl; +}; + +/** Basic function to run @ref NEComplexPixelWiseMultiplicationKernel. */ +class NEComplexPixelWiseMultiplication : public IFunction +{ +public: + /** Default Constructor */ + NEComplexPixelWiseMultiplication(); + /** Default Destructor */ + ~NEComplexPixelWiseMultiplication(); + /** Prevent instances of this class from being copied (As this class contains pointers) */ + NEComplexPixelWiseMultiplication(const NEComplexPixelWiseMultiplication &) = delete; + /** Default move constructor */ + NEComplexPixelWiseMultiplication(NEComplexPixelWiseMultiplication &&); + /** Prevent instances of this class from being copied (As this class contains pointers) */ + NEComplexPixelWiseMultiplication &operator=(const NEComplexPixelWiseMultiplication &) = delete; + /** Default move assignment operator */ + NEComplexPixelWiseMultiplication &operator=(NEComplexPixelWiseMultiplication &&); /** Initialise the kernel's inputs, output. * * @param[in, out] input1 An input tensor. Data types supported: F32. Number of channels supported: 2 (complex tensor). @@ -112,6 +234,13 @@ public: * @param[in] act_info (Optional) Activation layer information in case of a fused activation. Currently not supported. */ static Status validate(const ITensorInfo *input1, const ITensorInfo *input2, const ITensorInfo *output, const ActivationLayerInfo &act_info = ActivationLayerInfo()); + + // Inherited methods overridden: + void run() override; + +private: + struct Impl; + std::unique_ptr _impl; }; } #endif /*ARM_COMPUTE_NEPIXELWISEMULTIPLICATION_H */ -- cgit v1.2.1