COMPMID-675 - Reworked NEGEMMLowp interface/function
The new interface makes NEGEMMLowp able to work with ASYMM8 data types.
Implemented 2 new functions:
- NEGEMMLowpMatrixMultiplyCore
- NEGEMMLowpOutputStage
These functions should make the integration in android NN doable
For more information about GEMMLowp:
https://github.com/google/gemmlowp/blob/master/doc/low-precision.md
Change-Id: Ie2c775f45234f68ca53dba644b3a912b997fd890
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95504
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Pablo Tello <pablo.tello@arm.com>
diff --git a/arm_compute/runtime/NEON/NEFunctions.h b/arm_compute/runtime/NEON/NEFunctions.h
index 563ade2..118603b 100644
--- a/arm_compute/runtime/NEON/NEFunctions.h
+++ b/arm_compute/runtime/NEON/NEFunctions.h
@@ -60,8 +60,8 @@
#include "arm_compute/runtime/NEON/functions/NEFullyConnectedLayer.h"
#include "arm_compute/runtime/NEON/functions/NEGEMM.h"
#include "arm_compute/runtime/NEON/functions/NEGEMMInterleave4x4.h"
-#include "arm_compute/runtime/NEON/functions/NEGEMMLowp.h"
#include "arm_compute/runtime/NEON/functions/NEGEMMLowpMatrixMultiplyCore.h"
+#include "arm_compute/runtime/NEON/functions/NEGEMMLowpOutputStage.h"
#include "arm_compute/runtime/NEON/functions/NEGEMMTranspose1xW.h"
#include "arm_compute/runtime/NEON/functions/NEGaussian3x3.h"
#include "arm_compute/runtime/NEON/functions/NEGaussian5x5.h"
diff --git a/arm_compute/runtime/NEON/functions/NEGEMMLowp.h b/arm_compute/runtime/NEON/functions/NEGEMMLowp.h
deleted file mode 100644
index 59c919e..0000000
--- a/arm_compute/runtime/NEON/functions/NEGEMMLowp.h
+++ /dev/null
@@ -1,94 +0,0 @@
-/*
- * Copyright (c) 2017 ARM Limited.
- *
- * SPDX-License-Identifier: MIT
- *
- * Permission is hereby granted, free of charge, to any person obtaining a copy
- * of this software and associated documentation files (the "Software"), to
- * deal in the Software without restriction, including without limitation the
- * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
- * sell copies of the Software, and to permit persons to whom the Software is
- * furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in all
- * copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
- * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
- * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
- * SOFTWARE.
- */
-#ifndef __ARM_COMPUTE_NEGEMMLOWP_H__
-#define __ARM_COMPUTE_NEGEMMLOWP_H__
-
-#include "arm_compute/core/NEON/INEKernel.h"
-#include "arm_compute/core/NEON/kernels/NEGEMMLowpFinalizeKernel.h"
-#include "arm_compute/core/NEON/kernels/NEGEMMLowpReductionKernel.h"
-#include "arm_compute/runtime/IFunction.h"
-#include "arm_compute/runtime/IMemoryManager.h"
-#include "arm_compute/runtime/MemoryGroup.h"
-#include "arm_compute/runtime/NEON/functions/NEGEMMLowpMatrixMultiplyCore.h"
-#include "arm_compute/runtime/Tensor.h"
-
-#include <memory>
-
-namespace arm_compute
-{
-class ITensor;
-
-/** Basic function to execute GEMMLowp on NEON. This function calls the following NEON kernels/function:
- *
- * -# @ref NEGEMMLowpMatrixAReductionKernel
- * -# @ref NEGEMMLowpMatrixBReductionKernel
- * -# @ref NEGEMMLowpMatrixMultiplyCore
- * -# @ref NEGEMMLowpFinalizeKernel
- *
-*/
-class NEGEMMLowp : public IFunction
-{
-public:
- /** Constructor */
- NEGEMMLowp(std::shared_ptr<IMemoryManager> memory_manager = nullptr);
- /** Initialise the kernel's inputs, output
- *
- * @note GEMM_LOWP: low precision GEMM kernel
- * This kernel performs the following computations:
- *
- * -# Convert a values from int8 to int32 and add a_offset to each of them.
- * -# Convert b values from int8 to int32 and add b_offset to each of them.
- * -# Compute the int32 matrix product of the resulting a * b.
- * -# Add output_offset to each entry of the result.
- * -# Multiply each entry of the result and round to the nearest integer
- * -# Clamp the resulting int32 values to the [0..255] range and cast to int8.
- *
- * @param[in] a First input tensor (Matrix A). Data type supported: S8.
- * @param[in] b Second input tensor (Matrix B). Data type supported: same as @p a
- * @param[out] output Output tensor. Data type supported: same as @p a.
- * @param[in] a_offset Offset to be added to each element of the matrix A.
- * @param[in] b_offset Offset to be added to each element of the matrix B.
- * @param[in] c_offset Offset to be added to each element of the output matrix
- * @param[in] output_mult_int Value to be multiplied to each element of the output matrix
- * @param[in] shift Number of bits to shift right the result.
- */
- void configure(const ITensor *a, const ITensor *b, ITensor *output, int32_t a_offset, int32_t b_offset, int32_t c_offset, int32_t output_mult_int, int32_t shift);
-
- // Inherited methods overridden:
- void run() override;
-
-private:
- MemoryGroup _memory_group;
- NEGEMMLowpMatrixMultiplyCore _mm_func;
- NEGEMMLowpMatrixAReductionKernel _mtx_a_reduction_kernel;
- NEGEMMLowpMatrixBReductionKernel _mtx_b_reduction_kernel;
- NEGEMMLowpFinalizeKernel _finalize_kernel;
- Tensor _vector_sum_col;
- Tensor _vector_sum_row;
- Tensor _mm_output;
- int32_t _a_offset;
- int32_t _b_offset;
-};
-}
-#endif /*__ARM_COMPUTE_NEGEMMLOWP_H__ */
diff --git a/arm_compute/runtime/NEON/functions/NEGEMMLowpMatrixMultiplyCore.h b/arm_compute/runtime/NEON/functions/NEGEMMLowpMatrixMultiplyCore.h
index c81a432..0c441df 100644
--- a/arm_compute/runtime/NEON/functions/NEGEMMLowpMatrixMultiplyCore.h
+++ b/arm_compute/runtime/NEON/functions/NEGEMMLowpMatrixMultiplyCore.h
@@ -25,6 +25,8 @@
#define __ARM_COMPUTE_NEGEMMLOWPMATRIXMULTIPLYCORE_H__
#include "arm_compute/core/NEON/INEKernel.h"
+#include "arm_compute/core/NEON/kernels/NEGEMMLowpOffsetContributionKernel.h"
+#include "arm_compute/core/NEON/kernels/NEGEMMLowpReductionKernel.h"
#include "arm_compute/runtime/IFunction.h"
#include "arm_compute/runtime/IMemoryManager.h"
#include "arm_compute/runtime/MemoryGroup.h"
@@ -41,11 +43,13 @@
* -# @ref NEGEMMInterleave4x4Kernel
* -# @ref NEGEMMTranspose1xWKernel
* -# @ref NEGEMMLowpMatrixMultiplyKernel
+ * -# @ref NEGEMMLowpOffsetContributionKernel
*
* otherwise if the DOT product instruction is available:
*
* -# @ref NEGEMMInterleaveBlockedKernel
* -# @ref NEGEMMLowpAArch64V8P4Kernel
+ * -# @ref NEGEMMLowpOffsetContributionKernel
*
*/
class NEGEMMLowpMatrixMultiplyCore : public IFunction
@@ -58,11 +62,11 @@
* @note GEMM_LOWP: low precision GEMM kernel
* This kernel performs the following computations:
*
- * -# Convert a values from uint8 to int32
- * -# Convert b values from uint8 to int32
- * -# Compute the int32 matrix product of the resulting a * b.
+ * -# Convert a values from QASYMM8 to int32 and add a_offset to each of them.
+ * -# Convert b values from QASYMM8 to int32 add b_offset to each of them.
+ * -# Compute the matrix product of the resulting a * b in int32.
*
- * @param[in] a First input tensor (Matrix A). Data type supported: U8.
+ * @param[in] a First input tensor (Matrix A). Data type supported: QASYMM8.
* @param[in] b Second input tensor (Matrix B). Data type supported: same as @p a
* @param[out] output Output tensor. Data type supported: Data type supported: S32
*/
@@ -72,13 +76,20 @@
void run() override;
private:
- MemoryGroup _memory_group;
- std::unique_ptr<INEKernel> _mm_kernel;
- std::unique_ptr<INEKernel> _mtx_a_reshape_kernel;
- std::unique_ptr<INEKernel> _mtx_b_reshape_kernel;
- Tensor _tmp_a;
- Tensor _tmp_b;
- Tensor _workspace;
+ MemoryGroup _memory_group;
+ std::unique_ptr<INEKernel> _mm_kernel;
+ std::unique_ptr<INEKernel> _mtx_a_reshape_kernel;
+ std::unique_ptr<INEKernel> _mtx_b_reshape_kernel;
+ NEGEMMLowpMatrixAReductionKernel _mtx_a_reduction_kernel;
+ NEGEMMLowpMatrixBReductionKernel _mtx_b_reduction_kernel;
+ NEGEMMLowpOffsetContributionKernel _offset_contribution_kernel;
+ Tensor _vector_sum_col;
+ Tensor _vector_sum_row;
+ Tensor _tmp_a;
+ Tensor _tmp_b;
+ Tensor _workspace;
+ int32_t _a_offset;
+ int32_t _b_offset;
};
}
#endif /*__ARM_COMPUTE_NEGEMMLOWPMATRIXMULTIPLYCORE_H__ */
diff --git a/arm_compute/runtime/NEON/functions/NEGEMMLowpOutputStage.h b/arm_compute/runtime/NEON/functions/NEGEMMLowpOutputStage.h
new file mode 100644
index 0000000..8557ef4
--- /dev/null
+++ b/arm_compute/runtime/NEON/functions/NEGEMMLowpOutputStage.h
@@ -0,0 +1,69 @@
+/*
+ * Copyright (c) 2017 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#ifndef __ARM_COMPUTE_NEGEMMLOWPOUTPUTSTAGE_H__
+#define __ARM_COMPUTE_NEGEMMLOWPOUTPUTSTAGE_H__
+
+#include "arm_compute/runtime/NEON/INESimpleFunction.h"
+
+/** This file contains all available output stages for GEMMLowp on NEON.
+ *
+ * In gemmlowp, the "output stage" is the process that takes a final int32 accumulator value (the output of @ref NEGEMMLowpMatrixMultiplyCore),
+ * and processes it to obtain the final ASYMM8 value.
+ *
+ * More information about the GEMMLowp output stage can be found at https://github.com/google/gemmlowp/blob/master/doc/output.md
+ */
+
+namespace arm_compute
+{
+class ITensor;
+
+/** Basic function to execute NEGEMMLowpQuantizeDownInt32ToUint8Scale on NEON.
+ *
+ * NEGEMMLowpQuantizeDownInt32ToUint8Scale depends on 3 parameters: result_offset, result_mult_int, result_shift
+ * The final result is:
+ *
+ * ((input[i][k] + result_offset) * result_mult_int + rounding) >> result_shift
+ *
+ * where rounding = (result_shift < 1) ? 0 : (1 << (result_shift - 1))
+ *
+ * This function calls the following NEON kernels:
+ *
+ * -# @ref NEGEMMLowpQuantizeDownInt32ToUint8ScaleKernel
+ *
+*/
+class NEGEMMLowpQuantizeDownInt32ToUint8Scale : public INESimpleFunction
+{
+public:
+ /** Initialise the kernel's inputs, output
+ *
+ * @param[in] input Input tensor. It is the output of @ref NEGEMMLowpMatrixMultiplyCore function. Data type supported: S32
+ * @param[out] output Output tensor. Data type supported: Data type supported: QASYMM8
+ * @param[in] result_offset Offset to be added to each element of the input matrix
+ * @param[in] result_mult_int Value to be multiplied to each element of the input matrix when once the result_offset has been add
+ * @param[in] result_shift Number of bits to shift right the result before converting back to QASYMM8
+ */
+ void configure(const ITensor *input, ITensor *output, int result_offset, int result_mult_int, int result_shift);
+};
+}
+#endif /*__ARM_COMPUTE_NEGEMMLOWPOUTPUTSTAGE_H__ */
\ No newline at end of file