blob: 9b699ef4f1e1a95518efb6803e97efd9c237fe54 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
|
//
// Copyright © 2023 Arm Ltd and Contributors. All rights reserved.
// SPDX-License-Identifier: MIT
//
#include "NeonTileWorkload.hpp"
#include "NeonWorkloadUtils.hpp"
#include <aclCommon/ArmComputeUtils.hpp>
#include <vector>
#include <algorithm>
using namespace armnn::armcomputetensorutils;
namespace armnn
{
arm_compute::Status NeonTileWorkloadValidate(const TensorInfo& input,
const TensorInfo& output,
const TileDescriptor& descriptor)
{
const arm_compute::TensorInfo aclInput = BuildArmComputeTensorInfo(input);
const arm_compute::TensorInfo aclOutput = BuildArmComputeTensorInfo(output);
std::vector<unsigned int> aclMultiples = descriptor.m_Multiples;
std::reverse(aclMultiples.begin(),aclMultiples.end());
return arm_compute::NETile::validate(&aclInput, &aclOutput, aclMultiples);
}
NeonTileWorkload::NeonTileWorkload(const armnn::TileQueueDescriptor& descriptor,
const armnn::WorkloadInfo& info)
: BaseWorkload<TileQueueDescriptor>(descriptor, info)
{
m_Data.ValidateInputsOutputs("NeonTileWorkload", 1, 1);
std::vector<unsigned int> aclMultiples = descriptor.m_Parameters.m_Multiples;
std::reverse(aclMultiples.begin(),aclMultiples.end());
arm_compute::ITensor& input = static_cast<IAclTensorHandle*>(m_Data.m_Inputs[0])->GetTensor();
arm_compute::ITensor& output = static_cast<IAclTensorHandle*>(m_Data.m_Outputs[0])->GetTensor();
m_Layer.configure(&input, &output, aclMultiples);
}
void NeonTileWorkload::Execute() const
{
ARMNN_SCOPED_PROFILING_EVENT_NEON_GUID("NeonTileWorkload_Execute", this->GetGuid());
m_Layer.run();
}
} //namespace armnn
|
