diff options
author | Mike Kelly <mike.kelly@arm.com> | 2023-03-08 13:47:17 +0000 |
---|---|---|
committer | Francis Murtagh <francis.murtagh@arm.com> | 2023-03-14 16:40:09 +0000 |
commit | 3ec3077b4eaedcc0c20ab5774bdbe365da541445 (patch) | |
tree | d601d2000897dec8691bf64cbddc9036f26b8034 /src/backends/neon/NeonBackend.cpp | |
parent | a088cd00b3cce672d26cdcb4965fc2a86b48f339 (diff) | |
download | armnn-3ec3077b4eaedcc0c20ab5774bdbe365da541445.tar.gz |
IVGCVSW-3808 Add ElementwiseBinaryLayer
!android-nn-driver:9329
* Added ElementwiseBinaryLayer that can represent all ElementwiseBinary
operations including Add, Div, Sub, Maximum, Mul and Minimum.
* Updated Delegate to use ElementwiseBinaryLayer instead of the Add,
Div, Sub, Maximum, Mul and Minimum layers.
* Updated Deserializer to use ElementwiseBinaryLayer instead of the Add,
Div, Sub, Maximum, Mul and Minimum layers.
* Updated OnnxParser to use ElementwiseBinaryLayer instead of the Add
layer.
* Updated TfLiteParser to use ElementwiseBinaryLayer instead of the Add,
Div, Sub, Maximum, Mul and Minimum layers.
* Updated CL and Neon tests to use ElementwiseBinaryLayer.
* Updated CL and Neon Backend Specific Optimizations to accept
ElementBinaryLayers as well as Add, Div, Mul, Sub, Maximum and Minimum
layers.
Signed-off-by: Teresa Charlin <teresa.charlinreyes@arm.com>
Signed-off-by: Mike Kelly <mike.kelly@arm.com>
Change-Id: I7cbb96b60eb01f0e2b57b0541016d48a08b86c75
Diffstat (limited to 'src/backends/neon/NeonBackend.cpp')
-rw-r--r-- | src/backends/neon/NeonBackend.cpp | 88 |
1 files changed, 87 insertions, 1 deletions
diff --git a/src/backends/neon/NeonBackend.cpp b/src/backends/neon/NeonBackend.cpp index 968bce48c8..cea2aa3eba 100644 --- a/src/backends/neon/NeonBackend.cpp +++ b/src/backends/neon/NeonBackend.cpp @@ -1,5 +1,5 @@ // -// Copyright © 2017 Arm Ltd. All rights reserved. +// Copyright © 2017-2023 Arm Ltd and Contributors. All rights reserved. // SPDX-License-Identifier: MIT // @@ -395,6 +395,92 @@ OptimizationViews NeonBackend::OptimizeSubgraphView(const SubgraphView& subgraph untouched.erase(activationLayer->GetGuid()); } } + else if (base.GetType() == LayerType::ElementwiseBinary) + { + ElementwiseBinaryLayer* baseLayer = PolymorphicDowncast<ElementwiseBinaryLayer*>(&base); + + if (baseLayer->GetParameters().m_Operation == BinaryOperation::Add) + { + arm_compute::Status status = NeonAdditionWorkloadValidate( + baseLayer->GetInputSlot(0).GetConnectedOutputSlot()->GetTensorInfo(), + baseLayer->GetInputSlot(1).GetConnectedOutputSlot()->GetTensorInfo(), + activationLayer->GetInputSlot(0).GetConnectedOutputSlot()->GetTensorInfo(), + &activationDesc); + + if (status) + { + FuseElementwiseBinaryLayer<ElementwiseBinaryLayer>(optimizationViews, + baseLayer, + activationLayer, + activationDesc, + BinaryOperation::Add, + name); + untouched.erase(baseLayer->GetGuid()); + untouched.erase(activationLayer->GetGuid()); + } + } + else if (baseLayer->GetParameters().m_Operation == BinaryOperation::Div) + { + arm_compute::Status status = NeonDivisionWorkloadValidate( + baseLayer->GetInputSlot(0).GetConnectedOutputSlot()->GetTensorInfo(), + baseLayer->GetInputSlot(1).GetConnectedOutputSlot()->GetTensorInfo(), + activationLayer->GetInputSlot(0).GetConnectedOutputSlot()->GetTensorInfo(), + &activationDesc); + + if (status) + { + FuseElementwiseBinaryLayer<ElementwiseBinaryLayer>(optimizationViews, + baseLayer, + activationLayer, + activationDesc, + BinaryOperation::Div, + name); + untouched.erase(baseLayer->GetGuid()); + untouched.erase(activationLayer->GetGuid()); + } + } + else if (baseLayer->GetParameters().m_Operation == BinaryOperation::Mul) + { + arm_compute::Status status = NeonMultiplicationWorkloadValidate( + baseLayer->GetInputSlot(0).GetConnectedOutputSlot()->GetTensorInfo(), + baseLayer->GetInputSlot(1).GetConnectedOutputSlot()->GetTensorInfo(), + activationLayer->GetInputSlot(0).GetConnectedOutputSlot()->GetTensorInfo(), + &activationDesc); + + if (status) + { + FuseElementwiseBinaryLayer<ElementwiseBinaryLayer>(optimizationViews, + baseLayer, + activationLayer, + activationDesc, + BinaryOperation::Mul, + name); + untouched.erase(baseLayer->GetGuid()); + untouched.erase(activationLayer->GetGuid()); + } + } + else if (baseLayer->GetParameters().m_Operation == BinaryOperation::Sub) + { + arm_compute::Status status = NeonSubtractionWorkloadValidate( + baseLayer->GetInputSlot(0).GetConnectedOutputSlot()->GetTensorInfo(), + baseLayer->GetInputSlot(1).GetConnectedOutputSlot()->GetTensorInfo(), + activationLayer->GetInputSlot(0).GetConnectedOutputSlot()->GetTensorInfo(), + &activationDesc); + + if (status) + { + FuseElementwiseBinaryLayer<ElementwiseBinaryLayer>(optimizationViews, + baseLayer, + activationLayer, + activationDesc, + BinaryOperation::Sub, + name); + untouched.erase(baseLayer->GetGuid()); + untouched.erase(activationLayer->GetGuid()); + } + } + // No fusion available for other BinaryOperations + } } } } |