IVGCVSW-1784: Rename float32 workloads for ACL

Change-Id: I98bdfe9cb12c663d1d5cfa456e2cc967d70ab22b

1 files changed, 59 insertions, 0 deletions

diff --git a/src/armnn/backends/ClWorkloads/ClMultiplicationFloatWorkload.cpp b/src/armnn/backends/ClWorkloads/ClMultiplicationFloatWorkload.cpp
new file mode 100644
index 0000000000..e161a0a8fe
--- /dev/null
+++ b/src/armnn/backends/ClWorkloads/ClMultiplicationFloatWorkload.cpp
@@ -0,0 +1,59 @@
+//
+// Copyright © 2017 Arm Ltd. All rights reserved.
+// See LICENSE file in the project root for full license information.
+//
+
+#include "ClMultiplicationFloatWorkload.hpp"
+#include "backends/ClTensorHandle.hpp"
+#include "backends/CpuTensorHandle.hpp"
+
+namespace armnn
+{
+
+arm_compute::Status ClMultiplicationWorkloadValidate(const TensorInfo& input0,
+                                                     const TensorInfo& input1,
+                                                     const TensorInfo& output)
+{
+    const arm_compute::TensorInfo aclInput1 = armcomputetensorutils::BuildArmComputeTensorInfo(input0);
+    const arm_compute::TensorInfo aclInput2 = armcomputetensorutils::BuildArmComputeTensorInfo(input1);
+    const arm_compute::TensorInfo aclOutput = armcomputetensorutils::BuildArmComputeTensorInfo(output);
+
+    // At the time of writing, configure() will fail if a rounding policy other than TO_ZERO is supplied to it,
+    // when providing a scale of 1.0 for F32 tensors, even though the provided rounding policy appears to be
+    // ignored for F32 tensors.
+    return arm_compute::CLPixelWiseMultiplication::validate(&aclInput1,
+                                                            &aclInput2,
+                                                            &aclOutput,
+                                                            1.0f,
+                                                            arm_compute::ConvertPolicy::SATURATE,
+                                                            arm_compute::RoundingPolicy::TO_ZERO);
+}
+
+
+ClMultiplicationFloatWorkload::ClMultiplicationFloatWorkload(const MultiplicationQueueDescriptor& descriptor,
+                                                                 const WorkloadInfo& info)
+    : FloatWorkload<MultiplicationQueueDescriptor>(descriptor, info)
+{
+    m_Data.ValidateInputsOutputs("ClMultiplicationFloatWorkload", 2, 1);
+
+    arm_compute::ICLTensor& input0 = static_cast<IClTensorHandle*>(m_Data.m_Inputs[0])->GetTensor();
+    arm_compute::ICLTensor& input1 = static_cast<IClTensorHandle*>(m_Data.m_Inputs[1])->GetTensor();
+    arm_compute::ICLTensor& output = static_cast<IClTensorHandle*>(m_Data.m_Outputs[0])->GetTensor();
+    // Construct
+    m_PixelWiseMultiplication.configure(&input0,
+                                        &input1,
+                                        &output,
+                                        1.0f,
+                                        arm_compute::ConvertPolicy::SATURATE,
+                                        arm_compute::RoundingPolicy::TO_NEAREST_EVEN);
+}
+
+void ClMultiplicationFloatWorkload::Execute() const
+{
+    ARMNN_SCOPED_PROFILING_EVENT_CL("ClMultiplicationFloatWorkload_Execute");
+
+    // Executes the layer.
+    m_PixelWiseMultiplication.run();
+}
+
+} //namespace armnn