opencv/modules/dnn/src/cuda/array.hpp
Yashas Samaga B L 613c12e590 Merge pull request #14827 from YashasSamaga:cuda4dnn-csl-low
CUDA backend for the DNN module

* stub cuda4dnn design

* minor fixes for tests and doxygen

* add csl public api directory to module headers

* add low-level CSL components

* add high-level CSL components

* integrate csl::Tensor into backbone code

* switch to CPU iff unsupported; otherwise, fail on error

* add fully connected layer

* add softmax layer

* add activation layers

* support arbitary rank TensorDescriptor

* pass input wrappers to `initCUDA()`

* add 1d/2d/3d-convolution

* add pooling layer

* reorganize and refactor code

* fixes for gcc, clang and doxygen; remove cxx14/17 code

* add blank_layer

* add LRN layer

* add rounding modes for pooling layer

* split tensor.hpp into tensor.hpp and tensor_ops.hpp

* add concat layer

* add scale layer

* add batch normalization layer

* split math.cu into activations.cu and math.hpp

* add eltwise layer

* add flatten layer

* add tensor transform api

* add asymmetric padding support for convolution layer

* add reshape layer

* fix rebase issues

* add permute layer

* add padding support for concat layer

* refactor and reorganize code

* add normalize layer

* optimize bias addition in scale layer

* add prior box layer

* fix and optimize normalize layer

* add asymmetric padding support for pooling layer

* add event API

* improve pooling performance for some padding scenarios

* avoid over-allocation of compute resources to kernels

* improve prior box performance

* enable layer fusion

* add const layer

* add resize layer

* add slice layer

* add padding layer

* add deconvolution layer

* fix channelwise  ReLU initialization

* add vector traits

* add vectorized versions of relu, clipped_relu, power

* add vectorized concat kernels

* improve concat_with_offsets performance

* vectorize scale and bias kernels

* add support for multi-billion element tensors

* vectorize prior box kernels

* fix address alignment check

* improve bias addition performance of conv/deconv/fc layers

* restructure code for supporting multiple targets

* add DNN_TARGET_CUDA_FP64

* add DNN_TARGET_FP16

* improve vectorization

* add region layer

* improve tensor API, add dynamic ranks

1. use ManagedPtr instead of a Tensor in backend wrapper
2. add new methods to tensor classes
  - size_range: computes the combined size of for a given axis range
  - tensor span/view can be constructed from a raw pointer and shape
3. the tensor classes can change their rank at runtime (previously rank was fixed at compile-time)
4. remove device code from tensor classes (as they are unused)
5. enforce strict conditions on tensor class APIs to improve debugging ability

* fix parametric relu activation

* add squeeze/unsqueeze tensor API

* add reorg layer

* optimize permute and enable 2d permute

* enable 1d and 2d slice

* add split layer

* add shuffle channel layer

* allow tensors of different ranks in reshape primitive

* patch SliceOp to allow Crop Layer

* allow extra shape inputs in reshape layer

* use `std::move_backward` instead of `std::move` for insert in resizable_static_array

* improve workspace management

* add spatial LRN

* add nms (cpu) to region layer

* add max pooling with argmax ( and a fix to limits.hpp)

* add max unpooling layer

* rename DNN_TARGET_CUDA_FP32 to DNN_TARGET_CUDA

* update supportBackend to be more rigorous

* remove stray include from preventing non-cuda build

* include op_cuda.hpp outside condition #if

* refactoring, fixes and many optimizations

* drop DNN_TARGET_CUDA_FP64

* fix gcc errors

* increase max. tensor rank limit to six

* add Interp layer

* drop custom layers; use BackendNode

* vectorize activation kernels

* fixes for gcc

* remove wrong assertion

* fix broken assertion in unpooling primitive

* fix build errors in non-CUDA build

* completely remove workspace from public API

* fix permute layer

* enable accuracy and perf. tests for DNN_TARGET_CUDA

* add asynchronous forward

* vectorize eltwise ops

* vectorize fill kernel

* fixes for gcc

* remove CSL headers from public API

* remove csl header source group from cmake

* update min. cudnn version in cmake

* add numerically stable FP32 log1pexp

* refactor code

* add FP16 specialization to cudnn based tensor addition

* vectorize scale1 and bias1 + minor refactoring

* fix doxygen build

* fix invalid alignment assertion

* clear backend wrappers before allocateLayers

* ignore memory lock failures

* do not allocate internal blobs

* integrate NVTX

* add numerically stable half precision log1pexp

* fix indentation, following coding style,  improve docs

* remove accidental modification of IE code

* Revert "add asynchronous forward"

This reverts commit 1154b9da9da07e9b52f8a81bdcea48cf31c56f70.

* [cmake] throw error for unsupported CC versions

* fix rebase issues

* add more docs, refactor code, fix bugs

* minor refactoring and fixes

* resolve warnings/errors from clang

* remove haveCUDA() checks from supportBackend()

* remove NVTX integration

* changes based on review comments

* avoid exception when no CUDA device is present

* add color code for CUDA in Net::dump
2019-10-21 14:28:00 +03:00

74 lines
3.3 KiB
C++

// This file is part of OpenCV project.
// It is subject to the license terms in the LICENSE file found in the top-level directory
// of this distribution and at http://opencv.org/license.html.
#ifndef OPENCV_DNN_SRC_CUDA_ARRAY_HPP
#define OPENCV_DNN_SRC_CUDA_ARRAY_HPP
#include <cuda_runtime.h>
#include "types.hpp"
#include <cstddef>
#include <type_traits>
#include <iterator>
namespace cv { namespace dnn { namespace cuda4dnn { namespace csl { namespace device {
template <class T, std::size_t N>
struct array {
using value_type = T;
using size_type = device::size_type;
using difference_type = std::ptrdiff_t;
using reference = typename std::add_lvalue_reference<value_type>::type;
using const_reference = typename std::add_lvalue_reference<typename std::add_const<value_type>::type>::type;
using pointer = typename std::add_pointer<value_type>::type;
using const_pointer = typename std::add_pointer<typename std::add_const<value_type>::type>::type;
using iterator = pointer;
using const_iterator = const_pointer;
using reverse_iterator = std::reverse_iterator<iterator>;
using const_reverse_iterator = std::reverse_iterator<const_iterator>;
__host__ __device__ bool empty() const noexcept { return N == 0; }
__host__ __device__ size_type size() const noexcept { return N; }
__host__ __device__ iterator begin() noexcept { return ptr; }
__host__ __device__ iterator end() noexcept { return ptr + N; }
__host__ __device__ const_iterator begin() const noexcept { return ptr; }
__host__ __device__ const_iterator end() const noexcept { return ptr + N; }
__host__ __device__ const_iterator cbegin() const noexcept { return ptr; }
__host__ __device__ const_iterator cend() const noexcept { return ptr + N; }
__host__ __device__ reverse_iterator rbegin() noexcept { return ptr + N; }
__host__ __device__ reverse_iterator rend() noexcept { return ptr; }
__host__ __device__ const_reverse_iterator rbegin() const noexcept { return ptr + N; }
__host__ __device__ const_reverse_iterator rend() const noexcept { return ptr; }
__host__ __device__ const_reverse_iterator crbegin() const noexcept { return ptr + N; }
__host__ __device__ const_reverse_iterator crend() const noexcept { return ptr; }
template <class InputItr>
__host__ void assign(InputItr first, InputItr last) {
std::copy(first, last, std::begin(ptr));
}
__host__ __device__ reference operator[](int idx) { return ptr[idx]; }
__host__ __device__ const_reference operator[](int idx) const { return ptr[idx]; }
__host__ __device__ reference front() { return ptr[0]; }
__host__ __device__ const_reference front() const { return ptr[0]; }
__host__ __device__ reference back() { return ptr[N - 1]; }
__host__ __device__ const_reference back() const { return ptr[N - 1]; }
__host__ __device__ pointer data() noexcept { return ptr; }
__host__ __device__ const_pointer data() const noexcept { return ptr; }
T ptr[N];
};
}}}}} /* namespace cv::dnn::cuda4dnn::csl::device */
#endif /* OPENCV_DNN_SRC_CUDA_ARRAY_HPP */