24.04
|
#include <GpuPool2d.h>
Public Types | |
using | Attributes = Pool2dAttributes |
Attributes are a set of backend-agnostic parameters that define what an operator does. More... | |
using | Settings = GpuPool2dSettings |
Settings are a set of backend-specific parameters that influence the implementation of a operator. More... | |
Static Public Member Functions | |
static ITensorInfo * | create_op (GpuWorkloadSketch &sketch, ITensorInfo *src, const Attributes &attributes, const Settings &settings) |
Create an operator and fuse it into the workload sketch. More... | |
static Status | is_supported_op (const GpuWorkloadContext &context, const ITensorInfo *src, const Attributes &attributes, const Settings &settings) |
Check if the operator configuration is supported, irrespective of fusion. More... | |
static Status | validate_op (const GpuWorkloadSketch &sketch, const ITensorInfo *src, const Attributes &attributes, const Settings &settings) |
Validate the operator and check if it can be fused into the workload sketch. More... | |
Operator interface.
Definition at line 56 of file GpuPool2d.h.
using Attributes = Pool2dAttributes |
Attributes are a set of backend-agnostic parameters that define what an operator does.
Definition at line 60 of file GpuPool2d.h.
using Settings = GpuPool2dSettings |
Settings are a set of backend-specific parameters that influence the implementation of a operator.
Definition at line 62 of file GpuPool2d.h.
|
static |
Create an operator and fuse it into the workload sketch.
sketch
remains unchanged and valid.Valid data type configurations:
src | dst |
---|---|
F16 | F16 |
F32 | F32 |
Valid data layouts:
[in,out] | sketch | Workload sketch into which the operator will be fused |
[in] | src | Source tensor |
[in] | attributes | Operator attributes |
[in] | settings | Operator settings |
Definition at line 150 of file GpuPool2d.cpp.
References arm_compute::ACL_DST_0, arm_compute::ACL_SRC_0, ArgumentPack< T >::add_const_tensor(), GpuOperatorGroup::add_operator(), ArgumentPack< T >::add_tensor(), ARM_COMPUTE_ERROR, ARM_COMPUTE_ERROR_ON, ARM_COMPUTE_ERROR_ON_NULLPTR, ARM_COMPUTE_ERROR_THROW_ON, ARM_COMPUTE_LOG_PARAMS, ARM_COMPUTE_UNUSED, GpuWorkloadContext::cl_compile_context(), GpuWorkloadSketch::Implementation::component_graph(), GpuWorkloadSketch::Implementation::context(), GpuWorkloadSketch::Implementation::create_virtual_tensor(), arm_compute::test::validation::dst, GpuWorkloadSketch::implementation(), GpuOperatorGroup::new_operator(), arm_compute::experimental::dynamic_fusion::OpenCL, GpuWorkloadSketch::Implementation::operator_group(), UnitWorkloadStage::Run, arm_compute::test::validation::sketch, arm_compute::test::validation::src, and GpuPool2d::validate_op().
|
static |
Check if the operator configuration is supported, irrespective of fusion.
[in] | context | Workload context within which the operator is running |
[in] | src | Left hand side tensor info. Data types supported: F16/F32. |
[in] | attributes | Operator attributes |
[in] | settings | Operator settings |
Definition at line 107 of file GpuPool2d.cpp.
References arm_compute::ACL_DST_0, arm_compute::ACL_SRC_0, ArgumentPack< T >::add_const_tensor(), ARM_COMPUTE_RETURN_ERROR_MSG, ARM_COMPUTE_RETURN_ERROR_ON, ARM_COMPUTE_RETURN_ERROR_ON_DATA_LAYOUT_NOT_IN, ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN, ARM_COMPUTE_RETURN_ERROR_ON_MSG, ARM_COMPUTE_RETURN_ERROR_ON_NULLPTR, ARM_COMPUTE_RETURN_ON_ERROR, arm_compute::test::validation::context, Pool2dAttributes::exclude_padding(), arm_compute::F16, arm_compute::F32, arm_compute::NHWC, arm_compute::experimental::dynamic_fusion::OpenCL, UnitWorkloadStage::Run, arm_compute::test::validation::src, KernelProperties::stage(), and ClComponentPool2d::validate().
Referenced by GpuPool2d::validate_op().
|
static |
Validate the operator and check if it can be fused into the workload sketch.
Similar to GpuPool2d::create_op()
Definition at line 80 of file GpuPool2d.cpp.
References arm_compute::ACL_DST_0, arm_compute::ACL_SRC_0, ArgumentPack< T >::add_const_tensor(), ARM_COMPUTE_RETURN_ERROR_ON, ARM_COMPUTE_RETURN_ERROR_ON_MSG, ARM_COMPUTE_RETURN_ERROR_ON_NULLPTR, GpuWorkloadSketch::gpu_context(), GpuWorkloadSketch::implementation(), GpuPool2d::is_supported_op(), GpuOperatorGroup::new_operator(), GpuWorkloadSketch::Implementation::operator_group(), arm_compute::test::validation::sketch, arm_compute::test::validation::src, and GpuOperatorGroup::try_add_operator().
Referenced by GpuPool2d::create_op().