commit d02d5edfa15ba6c04a9986a8a362a945cb38ac31
author Michele Di Giorgio <michele.digiorgio@arm.com> Fri Jan 22 09:47:04 2021 +0000
committer Michele Di Giorgio <michele.digiorgio@arm.com> Fri Jun 18 10:33:48 2021 +0000
tree ced4f49691d6c7038e347a8709b315bff59c64cf
parent b014c27ba6db9840e4a72519760d51a87a2af7e7

Integrate improved CPU depthwise convolution kernels

* Replace assembly kernels for depthwise convolution with more optimized
  ones.
* Add int8 assembly kernels.
* Fix implicit padding on optimized kernels

Resolves: COMPMID-3867, COMPMID-4361

Change-Id: I0b0867e05f61be4f368f62190d55e14d0ab3ebf2
Signed-off-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/5622
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>

src/core/utils/quantization/AsymmHelpers.cpp

diff --git a/src/core/utils/quantization/AsymmHelpers.cpp b/src/core/utils/quantization/AsymmHelpers.cpp
index 49e39f6..81e813c 100644
--- a/src/core/utils/quantization/AsymmHelpers.cpp
+++ b/src/core/utils/quantization/AsymmHelpers.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2017-2020 Arm Limited.
+ * Copyright (c) 2017-2021 Arm Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -185,16 +185,15 @@
 void compute_quantized_multipliers_and_shifts(const ITensorInfo *input,
                                               const ITensorInfo *weights,
                                               const ITensorInfo *output,
-                                              unsigned int       idx_ofms,
                                               int32_t           *output_multipliers_ptr,
                                               int32_t           *output_shifts_ptr)
 {
-    const unsigned int num_filters = is_data_type_quantized_per_channel(weights->data_type()) ? weights->dimension(idx_ofms) : 1;
-
     const UniformQuantizationInfo iq_info = input->quantization_info().uniform();
     const QuantizationInfo        wq_info = weights->quantization_info();
     const UniformQuantizationInfo oq_info = output->quantization_info().uniform();
 
+    const unsigned int num_filters = wq_info.scale().size();
+
     for(unsigned int i = 0; i < num_filters; ++i)
     {
         int32_t     output_multiplier = 0;