diff options
Diffstat (limited to 'src/dynamic_fusion/runtime/gpu')
3 files changed, 171 insertions, 0 deletions
diff --git a/src/dynamic_fusion/runtime/gpu/cl/ClKernelRuntime.cpp b/src/dynamic_fusion/runtime/gpu/cl/ClKernelRuntime.cpp index b3ec39362c..6a57482bb2 100644 --- a/src/dynamic_fusion/runtime/gpu/cl/ClKernelRuntime.cpp +++ b/src/dynamic_fusion/runtime/gpu/cl/ClKernelRuntime.cpp @@ -59,6 +59,8 @@ void ClKernelRuntime::configure(const ClCompileContext &compile_ctx, const GpuKe inline void ClKernelRuntime::add_tensor_argument(unsigned int &idx, const GpuKernelArgumentInfo &arg, const ICLTensor *tensor, const Window &arg_slice, std::vector<cl::Image2D> &cl_images) { + ARM_COMPUTE_ERROR_ON_NULLPTR(tensor); + switch(arg.type) { case GpuKernelArgumentInfo::Type::Scalar: @@ -142,6 +144,18 @@ inline void ClKernelRuntime::add_tensor_argument(unsigned int &idx, const GpuKer add_4d_tensor_nhwc_argument(idx, tensor); break; } + case GpuKernelArgumentInfo::Type::Tensor_Special_0: + { + const ITensorInfo *info = tensor->info(); + const Strides &strides = info->strides_in_bytes(); + + _kernel.setArg(idx++, tensor->cl_buffer()); + const size_t dim1xdim2 = info->tensor_shape()[1] * info->tensor_shape()[2]; + _kernel.setArg<cl_int>(idx++, static_cast<int32_t>(dim1xdim2)); + const size_t stride1 = strides[1]; + _kernel.setArg<cl_int>(idx++, static_cast<int32_t>(stride1)); + break; + } default: { ARM_COMPUTE_ERROR("Unsupported"); diff --git a/src/dynamic_fusion/runtime/gpu/cl/ckw_driver/GpuCkwKernelArgumentsHelpers.cpp b/src/dynamic_fusion/runtime/gpu/cl/ckw_driver/GpuCkwKernelArgumentsHelpers.cpp new file mode 100644 index 0000000000..40934cf6e9 --- /dev/null +++ b/src/dynamic_fusion/runtime/gpu/cl/ckw_driver/GpuCkwKernelArgumentsHelpers.cpp @@ -0,0 +1,96 @@ +/* + * Copyright (c) 2023 Arm Limited. + * + * SPDX-License-Identifier: MIT + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in all + * copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ + +#include "GpuCkwKernelArgumentsHelpers.h" + +namespace arm_compute +{ +void cl_add_tensor_component_argument(cl::Kernel &kernel, unsigned int &idx, ICLTensor *tensor, ckw::TensorComponent component) +{ + ARM_COMPUTE_ERROR_ON(tensor == nullptr); + + const auto *info = tensor->info(); + const auto &strides = info->strides_in_bytes(); + + switch(component) + { + case ckw::TensorComponent::OffsetFirstElement: + kernel.setArg<cl_uint>(idx++, info->offset_first_element_in_bytes()); + break; + case ckw::TensorComponent::Stride0: + kernel.setArg<cl_uint>(idx++, strides[0]); + break; + case ckw::TensorComponent::Stride1: + kernel.setArg<cl_uint>(idx++, strides[1]); + break; + case ckw::TensorComponent::Stride2: + kernel.setArg<cl_uint>(idx++, strides[2]); + break; + case ckw::TensorComponent::Stride3: + kernel.setArg<cl_uint>(idx++, strides[3]); + break; + case ckw::TensorComponent::Stride4: + kernel.setArg<cl_uint>(idx++, strides[4]); + break; + case ckw::TensorComponent::Dim0: + kernel.setArg<cl_uint>(idx++, info->dimension(0)); + break; + case ckw::TensorComponent::Dim1: + kernel.setArg<cl_uint>(idx++, info->dimension(1)); + break; + case ckw::TensorComponent::Dim2: + kernel.setArg<cl_uint>(idx++, info->dimension(2)); + break; + case ckw::TensorComponent::Dim3: + kernel.setArg<cl_uint>(idx++, info->dimension(3)); + break; + case ckw::TensorComponent::Dim4: + kernel.setArg<cl_uint>(idx++, info->dimension(4)); + break; + case ckw::TensorComponent::Dim1xDim2: + kernel.setArg<cl_uint>(idx++, info->dimension(1) * info->dimension(2)); + break; + case ckw::TensorComponent::Dim2xDim3: + kernel.setArg<cl_uint>(idx++, info->dimension(2) * info->dimension(3)); + break; + case ckw::TensorComponent::Dim1xDim2xDim3: + kernel.setArg<cl_uint>(idx++, info->dimension(1) * info->dimension(2) * info->dimension(3)); + break; + case ckw::TensorComponent::Unknown: + default: + ARM_COMPUTE_ERROR("Unknown tensor component"); + } +} + +void cl_add_buffer_argument(cl::Kernel &kernel, unsigned int &idx, const cl::Buffer &buffer) +{ + kernel.setArg(idx++, buffer); +} + +void cl_add_texture_argument(cl::Kernel &kernel, unsigned int &idx, const cl::Image &image) +{ + kernel.setArg(idx++, image); +} + +} // namespace arm_compute diff --git a/src/dynamic_fusion/runtime/gpu/cl/ckw_driver/GpuCkwKernelArgumentsHelpers.h b/src/dynamic_fusion/runtime/gpu/cl/ckw_driver/GpuCkwKernelArgumentsHelpers.h new file mode 100644 index 0000000000..206ad2e401 --- /dev/null +++ b/src/dynamic_fusion/runtime/gpu/cl/ckw_driver/GpuCkwKernelArgumentsHelpers.h @@ -0,0 +1,61 @@ +/* + * Copyright (c) 2023 Arm Limited. + * + * SPDX-License-Identifier: MIT + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in all + * copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ + +#ifndef ACL_SRC_DYNAMIC_FUSION_RUNTIME_GPU_CL_CKW_DRIVER_GPUCKWKERNELARGUMENTSHELPERS +#define ACL_SRC_DYNAMIC_FUSION_RUNTIME_GPU_CL_CKW_DRIVER_GPUCKWKERNELARGUMENTSHELPERS + +#include "arm_compute/core/CL/ICLTensor.h" + +#include "ckw/TensorInfo.h" + +namespace arm_compute +{ +/** Select a Compute Kernel Writer tensor component from a tensor and add to the kernel's arguments at the specified index idx. + * + * @param[in,out] kernel OpenCL kernel to configure with the provided argument. + * @param[in,out] idx Index at which to add the argument. + * @param[in] tensor Tensor from which to access the tensor component. + * @param[in] component Tensor component to select such as tensor dimensions, strides, etc. + */ +void cl_add_tensor_component_argument(cl::Kernel &kernel, unsigned int &idx, ICLTensor *tensor, ckw::TensorComponent component); + +/** Add an OpenCL buffer object to the kernel's arguments at the specified index @p idx. + * + * @param[in,out] kernel OpenCL kernel to configure with the provided argument. + * @param[in,out] idx Index at which to add the argument. + * @param[in] buffer OpenCL buffer containing the tensor's data. + */ +void cl_add_buffer_argument(cl::Kernel &kernel, unsigned int &idx, const cl::Buffer &buffer); + +/** Add an OpenCL image object to the kernel's arguments at the specified index @p idx. + * + * @param[in,out] kernel OpenCL kernel to configure with the provided argument. + * @param[in,out] idx Index at which to add the argument. + * @param[in] image OpenCL image containing the image's data. + */ +void cl_add_texture_argument(cl::Kernel &kernel, unsigned int &idx, const cl::Image &image); + +} // namespace arm_compute + +#endif /* ACL_SRC_DYNAMIC_FUSION_RUNTIME_GPU_CL_CKW_DRIVER_GPUCKWKERNELARGUMENTSHELPERS */ |