VideoEditorBeta/OKLCh.cpp at master · erfan-ops/VideoEditorBeta · GitHub

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
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
#include "OKLCh.h"
#include "OKLCh_CUDA_launcher.cuh"
#include "OKLCh_OpenCL.h"
#include "globals.h"
#include "utils.h"

bool OKLChProcessor::firstTime = true;

void (OKLChProcessor::* OKLChProcessor::processFunc)() const = nullptr;
void (OKLChProcessor::* OKLChProcessor::processFuncRGBA)() const = nullptr;

cl_kernel OKLChProcessor::m_openclKernel = nullptr;
cl_kernel OKLChProcessor::m_openclKernelRGBA = nullptr;

OKLChProcessor::OKLChProcessor(int size, int nPixels, const float Ladd, const float Cadd, const float hadd)
    : m_nPixels(nPixels), m_Ladd(Ladd), m_Cadd(Cadd), m_hadd(hadd) {
    imgSize = size;

    if (OKLChProcessor::firstTime) {
        OKLChProcessor::init();
    }

    // Allocate buffers for CUDA or OpenCL
    if (isCudaAvailable()) {
        cudaStreamCreate(&this->m_cudaStream);
        allocateCUDA();

        blockSize = 256;
        gridSize = (m_nPixels + blockSize - 1) / blockSize;
    }
    else {
        globalWorkSize = m_nPixels;
        allocateOpenCL();
    }
}

OKLChProcessor::~OKLChProcessor() {
    if (isCudaAvailable()) {
        cudaFree(d_img);
        cudaStreamDestroy(m_cudaStream);
    }
    else {
        clReleaseMemObject(m_imgBuf);
    }
}

void OKLChProcessor::process() const {
    (this->*processFunc)();
}

void OKLChProcessor::processRGBA() const {
    (this->*processFuncRGBA)();
}

void OKLChProcessor::processCUDA() const {
    OKLCh_CUDA(gridSize, blockSize, m_cudaStream, d_img, m_nPixels, m_Ladd, m_Cadd, m_hadd);
}

void OKLChProcessor::processRGBA_CUDA() const {
    OKLCh_CUDA(gridSize, blockSize, m_cudaStream, d_img, m_nPixels, m_Ladd, m_Cadd, m_hadd); // TODO: implement RGBA version
}

void OKLChProcessor::processOpenCL() const {
    clSetKernelArg(m_openclKernel, 0, sizeof(cl_mem), &m_imgBuf);
    clSetKernelArg(m_openclKernel, 1, sizeof(int), &m_nPixels);
    clSetKernelArg(m_openclKernel, 2, sizeof(float), &m_Ladd);
    clSetKernelArg(m_openclKernel, 3, sizeof(float), &m_Cadd);
    clSetKernelArg(m_openclKernel, 4, sizeof(float), &m_hadd);

    cl_event kernelEvent;
    clEnqueueNDRangeKernel(globalQueueOpenCL,
        m_openclKernel,
        1,
        nullptr,
        &globalWorkSize,
        nullptr,
        0, nullptr, &kernelEvent);

    // ⏳ Wait for kernel to complete
    clWaitForEvents(1, &kernelEvent);

    // Release the kernel event
    clReleaseEvent(kernelEvent);
}

void OKLChProcessor::processRGBA_OpenCL() const {
    clSetKernelArg(m_openclKernel, 0, sizeof(cl_mem), &m_imgBuf);
    clSetKernelArg(m_openclKernel, 1, sizeof(int), &m_nPixels);
    clSetKernelArg(m_openclKernel, 2, sizeof(float), &m_Ladd);
    clSetKernelArg(m_openclKernel, 3, sizeof(float), &m_Cadd);
    clSetKernelArg(m_openclKernel, 4, sizeof(float), &m_hadd);

    cl_event kernelEvent;
    clEnqueueNDRangeKernel(globalQueueOpenCL,
        m_openclKernelRGBA,
        1,
        nullptr,
        &globalWorkSize,
        nullptr,
        0, nullptr, &kernelEvent);

    // ⏳ Wait for kernel to complete
    clWaitForEvents(1, &kernelEvent);

    // Release the kernel event
    clReleaseEvent(kernelEvent);
}

void OKLChProcessor::init() {
    if (OKLChProcessor::firstTime) {
        if (isCudaAvailable()) {
            OKLChProcessor::processFunc = &OKLChProcessor::processCUDA;
            OKLChProcessor::processFuncRGBA = &OKLChProcessor::processRGBA_CUDA;
        }
        else {
            OKLChProcessor::processFunc = &OKLChProcessor::processOpenCL;
            OKLChProcessor::processFuncRGBA = &OKLChProcessor::processRGBA_OpenCL;

            OKLChProcessor::m_openclKernel = openclUtils::createKernelFromSource(globalContextOpenCL, globalDeviceOpenCL, OKLChOpenCLKernelSource, "OKLCh_kernel");
            //OKLChProcessor::m_openclKernelRGBA = openclUtils::createKernelFromSource(globalContextOpenCL, globalDeviceOpenCL, OKLChRGBAOpenCLKernelSource, "glitchFocusRGBA_kernel");
        }
        OKLChProcessor::firstTime = false;
        if (BaseProcessor::firstTime) {
            BaseProcessor::init();
        }
    }
}