mirror of
https://github.com/NVIDIA/cuda-samples.git
synced 2026-10-23 13:18:16 +08:00
add and update samples for CUDA 11.5
This commit is contained in:
Rutwik Choughule
18
Samples/FunctionPointers/.vscode/c_cpp_properties.json
vendored
Normal file
18
Samples/FunctionPointers/.vscode/c_cpp_properties.json
vendored
Normal file
@@ -0,0 +1,18 @@
|
||||
{
|
||||
"configurations": [
|
||||
{
|
||||
"name": "Linux",
|
||||
"includePath": [
|
||||
"${workspaceFolder}/**",
|
||||
"${workspaceFolder}/../../Common"
|
||||
],
|
||||
"defines": [],
|
||||
"compilerPath": "/usr/local/cuda/bin/nvcc",
|
||||
"cStandard": "gnu17",
|
||||
"cppStandard": "gnu++14",
|
||||
"intelliSenseMode": "linux-gcc-x64",
|
||||
"configurationProvider": "ms-vscode.makefile-tools"
|
||||
}
|
||||
],
|
||||
"version": 4
|
||||
}
|
||||
7
Samples/FunctionPointers/.vscode/extensions.json
vendored
Normal file
7
Samples/FunctionPointers/.vscode/extensions.json
vendored
Normal file
@@ -0,0 +1,7 @@
|
||||
{
|
||||
"recommendations": [
|
||||
"nvidia.nsight-vscode-edition",
|
||||
"ms-vscode.cpptools",
|
||||
"ms-vscode.makefile-tools"
|
||||
]
|
||||
}
|
||||
10
Samples/FunctionPointers/.vscode/launch.json
vendored
Normal file
10
Samples/FunctionPointers/.vscode/launch.json
vendored
Normal file
@@ -0,0 +1,10 @@
|
||||
{
|
||||
"configurations": [
|
||||
{
|
||||
"name": "CUDA C++: Launch",
|
||||
"type": "cuda-gdb",
|
||||
"request": "launch",
|
||||
"program": "${workspaceFolder}/FunctionPointers"
|
||||
}
|
||||
]
|
||||
}
|
||||
15
Samples/FunctionPointers/.vscode/tasks.json
vendored
Normal file
15
Samples/FunctionPointers/.vscode/tasks.json
vendored
Normal file
@@ -0,0 +1,15 @@
|
||||
{
|
||||
"version": "2.0.0",
|
||||
"tasks": [
|
||||
{
|
||||
"label": "sample",
|
||||
"type": "shell",
|
||||
"command": "make dbg=1",
|
||||
"problemMatcher": ["$nvcc"],
|
||||
"group": {
|
||||
"kind": "build",
|
||||
"isDefault": true
|
||||
}
|
||||
}
|
||||
]
|
||||
}
|
||||
510
Samples/FunctionPointers/FunctionPointers.cpp
Normal file
510
Samples/FunctionPointers/FunctionPointers.cpp
Normal file
@@ -0,0 +1,510 @@
|
||||
/* Copyright (c) 2021, NVIDIA CORPORATION. All rights reserved.
|
||||
*
|
||||
* Redistribution and use in source and binary forms, with or without
|
||||
* modification, are permitted provided that the following conditions
|
||||
* are met:
|
||||
* * Redistributions of source code must retain the above copyright
|
||||
* notice, this list of conditions and the following disclaimer.
|
||||
* * Redistributions in binary form must reproduce the above copyright
|
||||
* notice, this list of conditions and the following disclaimer in the
|
||||
* documentation and/or other materials provided with the distribution.
|
||||
* * Neither the name of NVIDIA CORPORATION nor the names of its
|
||||
* contributors may be used to endorse or promote products derived
|
||||
* from this software without specific prior written permission.
|
||||
*
|
||||
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
|
||||
* EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
||||
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
||||
* PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
|
||||
* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
|
||||
* EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
|
||||
* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
|
||||
* PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
|
||||
* OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
||||
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
||||
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
||||
*/
|
||||
|
||||
// OpenGL Graphics includes
|
||||
#include <helper_gl.h>
|
||||
#if defined(__APPLE__) || defined(MACOSX)
|
||||
#pragma clang diagnostic ignored "-Wdeprecated-declarations"
|
||||
#include <GLUT/glut.h>
|
||||
#ifndef glutCloseFunc
|
||||
#define glutCloseFunc glutWMCloseFunc
|
||||
#endif
|
||||
#else
|
||||
#include <GL/freeglut.h>
|
||||
#endif
|
||||
|
||||
// Includes
|
||||
#include <stdlib.h>
|
||||
#include <stdio.h>
|
||||
#include <string.h>
|
||||
#include <helper_functions.h> // helper functions for timing, string parsing
|
||||
|
||||
#include <cuda_runtime.h> // CUDA Runtime
|
||||
#include <cuda_gl_interop.h> // CUDA OpenGL interop
|
||||
|
||||
#include <helper_cuda.h> // includes for CUDA initialization and error checking
|
||||
|
||||
#include "FunctionPointers_kernels.h"
|
||||
|
||||
#define EXIT_WAIVED 2
|
||||
|
||||
//
|
||||
// Cuda example code that implements the Sobel edge detection
|
||||
// filter. This code works for 8-bit monochrome images.
|
||||
//
|
||||
// Use the '-' and '=' keys to change the scale factor.
|
||||
//
|
||||
// Other keys:
|
||||
// I: display image
|
||||
// T: display Sobel Edge Detection (computed solely with texture)
|
||||
// S: display Sobel Edge Detection (computed with texture and shared memory)
|
||||
|
||||
void cleanup(void);
|
||||
void initializeData(char *file);
|
||||
|
||||
#define MAX_EPSILON_ERROR 5.0f
|
||||
|
||||
static const char *sSDKsample = "CUDA Function Pointers (SobelFilter)";
|
||||
|
||||
const char *filterMode[] = {"No Filtering", "Sobel Texture",
|
||||
"Sobel SMEM+Texture", NULL};
|
||||
|
||||
static int wWidth = 512; // Window width
|
||||
static int wHeight = 512; // Window height
|
||||
static int imWidth = 0; // Image width
|
||||
static int imHeight = 0; // Image height
|
||||
static int blockOp = 0;
|
||||
static int pointOp = 1;
|
||||
|
||||
// Code to handle Auto verification
|
||||
const int frameCheckNumber = 4;
|
||||
int fpsCount = 0; // FPS count for averaging
|
||||
int fpsLimit = 8; // FPS limit for sampling
|
||||
unsigned int frameCount = 0;
|
||||
StopWatchInterface *timer = NULL;
|
||||
unsigned int g_Bpp;
|
||||
|
||||
int g_TotalErrors = 0;
|
||||
|
||||
int *pArgc = NULL;
|
||||
char **pArgv = NULL;
|
||||
|
||||
bool g_bQAReadback = false;
|
||||
|
||||
// Display Data
|
||||
static GLuint pbo_buffer = 0; // Front and back CA buffers
|
||||
struct cudaGraphicsResource
|
||||
*cuda_pbo_resource; // CUDA Graphics Resource (to transfer PBO)
|
||||
|
||||
static GLuint texid = 0; // Texture for display
|
||||
unsigned char *pixels = NULL; // Image pixel data on the host
|
||||
float imageScale = 1.f; // Image exposure
|
||||
enum SobelDisplayMode g_SobelDisplayMode;
|
||||
|
||||
#define OFFSET(i) ((char *)NULL + (i))
|
||||
#define MAX(a, b) ((a > b) ? a : b)
|
||||
#define REFRESH_DELAY 10 // ms
|
||||
|
||||
void computeFPS() {
|
||||
frameCount++;
|
||||
fpsCount++;
|
||||
|
||||
if (fpsCount == fpsLimit) {
|
||||
char fps[256];
|
||||
float ifps = 1.f / (sdkGetAverageTimerValue(&timer) / 1000.f);
|
||||
sprintf(fps, "FunctionPointers [CUDA Edge Detection] (%s): %3.1f fps",
|
||||
filterMode[g_SobelDisplayMode], ifps);
|
||||
|
||||
glutSetWindowTitle(fps);
|
||||
fpsCount = 0;
|
||||
|
||||
fpsLimit = (int)MAX(ifps, 1.f);
|
||||
sdkResetTimer(&timer);
|
||||
}
|
||||
}
|
||||
|
||||
// This is the normal display path
|
||||
void display(void) {
|
||||
sdkStartTimer(&timer);
|
||||
|
||||
// Sobel operation
|
||||
Pixel *data = NULL;
|
||||
|
||||
// map PBO to get CUDA device pointer
|
||||
checkCudaErrors(cudaGraphicsMapResources(1, &cuda_pbo_resource, 0));
|
||||
size_t num_bytes;
|
||||
checkCudaErrors(cudaGraphicsResourceGetMappedPointer(
|
||||
(void **)&data, &num_bytes, cuda_pbo_resource));
|
||||
// printf("CUDA mapped PBO: May access %ld bytes\n", num_bytes);
|
||||
|
||||
sobelFilter(data, imWidth, imHeight, g_SobelDisplayMode, imageScale, blockOp,
|
||||
pointOp);
|
||||
checkCudaErrors(cudaGraphicsUnmapResources(1, &cuda_pbo_resource, 0));
|
||||
|
||||
glClear(GL_COLOR_BUFFER_BIT);
|
||||
|
||||
glBindTexture(GL_TEXTURE_2D, texid);
|
||||
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, pbo_buffer);
|
||||
glTexSubImage2D(GL_TEXTURE_2D, 0, 0, 0, imWidth, imHeight, GL_LUMINANCE,
|
||||
GL_UNSIGNED_BYTE, OFFSET(0));
|
||||
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, 0);
|
||||
|
||||
glDisable(GL_DEPTH_TEST);
|
||||
glEnable(GL_TEXTURE_2D);
|
||||
glTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_LINEAR);
|
||||
glTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_LINEAR);
|
||||
glTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, GL_REPEAT);
|
||||
glTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, GL_REPEAT);
|
||||
|
||||
glBegin(GL_QUADS);
|
||||
glVertex2f(0, 0);
|
||||
glTexCoord2f(0, 0);
|
||||
glVertex2f(0, 1);
|
||||
glTexCoord2f(1, 0);
|
||||
glVertex2f(1, 1);
|
||||
glTexCoord2f(1, 1);
|
||||
glVertex2f(1, 0);
|
||||
glTexCoord2f(0, 1);
|
||||
glEnd();
|
||||
glBindTexture(GL_TEXTURE_2D, 0);
|
||||
|
||||
glutSwapBuffers();
|
||||
|
||||
sdkStopTimer(&timer);
|
||||
computeFPS();
|
||||
}
|
||||
|
||||
void timerEvent(int value) {
|
||||
if (glutGetWindow()) {
|
||||
glutPostRedisplay();
|
||||
glutTimerFunc(REFRESH_DELAY, timerEvent, 0);
|
||||
}
|
||||
}
|
||||
|
||||
void keyboard(unsigned char key, int /*x*/, int /*y*/) {
|
||||
char temp[256];
|
||||
|
||||
switch (key) {
|
||||
case 27:
|
||||
#if defined(__APPLE__) || defined(MACOSX)
|
||||
exit(EXIT_SUCCESS);
|
||||
#else
|
||||
glutDestroyWindow(glutGetWindow());
|
||||
return;
|
||||
#endif
|
||||
break;
|
||||
|
||||
case '-':
|
||||
imageScale -= 0.1f;
|
||||
printf("brightness = %4.2f\n", imageScale);
|
||||
break;
|
||||
|
||||
case '=':
|
||||
imageScale += 0.1f;
|
||||
printf("brightness = %4.2f\n", imageScale);
|
||||
break;
|
||||
|
||||
case 'i':
|
||||
case 'I':
|
||||
g_SobelDisplayMode = SOBELDISPLAY_IMAGE;
|
||||
sprintf(temp, "Function Pointers [CUDA Edge Detection] (%s)",
|
||||
filterMode[g_SobelDisplayMode]);
|
||||
glutSetWindowTitle(temp);
|
||||
break;
|
||||
|
||||
case 's':
|
||||
case 'S':
|
||||
g_SobelDisplayMode = SOBELDISPLAY_SOBELSHARED;
|
||||
sprintf(temp, "Function Pointers [CUDA Edge Detection] (%s)",
|
||||
filterMode[g_SobelDisplayMode]);
|
||||
glutSetWindowTitle(temp);
|
||||
break;
|
||||
|
||||
case 't':
|
||||
case 'T':
|
||||
g_SobelDisplayMode = SOBELDISPLAY_SOBELTEX;
|
||||
sprintf(temp, "Function Pointers [CUDA Edge Detection] (%s)",
|
||||
filterMode[g_SobelDisplayMode]);
|
||||
glutSetWindowTitle(temp);
|
||||
break;
|
||||
|
||||
case 'b':
|
||||
case 'B':
|
||||
blockOp = (blockOp + 1) % LAST_BLOCK_FILTER;
|
||||
break;
|
||||
|
||||
case 'p':
|
||||
case 'P':
|
||||
pointOp = (pointOp + 1) % LAST_POINT_FILTER;
|
||||
break;
|
||||
|
||||
default:
|
||||
break;
|
||||
}
|
||||
}
|
||||
|
||||
void reshape(int x, int y) {
|
||||
glViewport(0, 0, x, y);
|
||||
glMatrixMode(GL_PROJECTION);
|
||||
glLoadIdentity();
|
||||
glOrtho(0, 1, 0, 1, 0, 1);
|
||||
glMatrixMode(GL_MODELVIEW);
|
||||
glLoadIdentity();
|
||||
}
|
||||
|
||||
void cleanup(void) {
|
||||
cudaGraphicsUnregisterResource(cuda_pbo_resource);
|
||||
|
||||
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, 0);
|
||||
glDeleteBuffers(1, &pbo_buffer);
|
||||
glDeleteTextures(1, &texid);
|
||||
deleteTexture();
|
||||
|
||||
sdkDeleteTimer(&timer);
|
||||
}
|
||||
|
||||
void initializeData(char *file) {
|
||||
GLint bsize;
|
||||
unsigned int w, h;
|
||||
size_t file_length = strlen(file);
|
||||
|
||||
if (!strcmp(&file[file_length - 3], "pgm")) {
|
||||
if (sdkLoadPGM<unsigned char>(file, &pixels, &w, &h) != true) {
|
||||
printf("Failed to load PGM image file: %s\n", file);
|
||||
exit(EXIT_FAILURE);
|
||||
}
|
||||
|
||||
g_Bpp = 1;
|
||||
} else if (!strcmp(&file[file_length - 3], "ppm")) {
|
||||
if (sdkLoadPPM4(file, &pixels, &w, &h) != true) {
|
||||
printf("Failed to load PPM image file: %s\n", file);
|
||||
exit(EXIT_FAILURE);
|
||||
}
|
||||
|
||||
g_Bpp = 4;
|
||||
} else {
|
||||
exit(EXIT_FAILURE);
|
||||
}
|
||||
|
||||
imWidth = (int)w;
|
||||
imHeight = (int)h;
|
||||
setupTexture(imWidth, imHeight, pixels, g_Bpp);
|
||||
|
||||
// copy function pointer tables to host side for later use
|
||||
setupFunctionTables();
|
||||
|
||||
memset(pixels, 0x0, g_Bpp * sizeof(Pixel) * imWidth * imHeight);
|
||||
|
||||
if (!g_bQAReadback) {
|
||||
// use OpenGL Path
|
||||
glGenBuffers(1, &pbo_buffer);
|
||||
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, pbo_buffer);
|
||||
glBufferData(GL_PIXEL_UNPACK_BUFFER,
|
||||
g_Bpp * sizeof(Pixel) * imWidth * imHeight, pixels,
|
||||
GL_STREAM_DRAW);
|
||||
|
||||
glGetBufferParameteriv(GL_PIXEL_UNPACK_BUFFER, GL_BUFFER_SIZE, &bsize);
|
||||
|
||||
if ((GLuint)bsize != (g_Bpp * sizeof(Pixel) * imWidth * imHeight)) {
|
||||
printf("Buffer object (%d) has incorrect size (%d).\n",
|
||||
(unsigned)pbo_buffer, (unsigned)bsize);
|
||||
exit(EXIT_FAILURE);
|
||||
}
|
||||
|
||||
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, 0);
|
||||
|
||||
// register this buffer object with CUDA
|
||||
checkCudaErrors(cudaGraphicsGLRegisterBuffer(
|
||||
&cuda_pbo_resource, pbo_buffer, cudaGraphicsMapFlagsWriteDiscard));
|
||||
|
||||
glGenTextures(1, &texid);
|
||||
glBindTexture(GL_TEXTURE_2D, texid);
|
||||
glTexImage2D(GL_TEXTURE_2D, 0, ((g_Bpp == 1) ? GL_LUMINANCE : GL_BGRA),
|
||||
imWidth, imHeight, 0, GL_LUMINANCE, GL_UNSIGNED_BYTE, NULL);
|
||||
glBindTexture(GL_TEXTURE_2D, 0);
|
||||
|
||||
glPixelStorei(GL_UNPACK_ALIGNMENT, 1);
|
||||
glPixelStorei(GL_PACK_ALIGNMENT, 1);
|
||||
}
|
||||
}
|
||||
|
||||
void loadDefaultImage(char *loc_exec) {
|
||||
printf("Reading image: lena.pgm\n");
|
||||
const char *image_filename = "lena.pgm";
|
||||
char *image_path = sdkFindFilePath(image_filename, loc_exec);
|
||||
|
||||
if (image_path == NULL) {
|
||||
printf("Failed to read image file: <%s>\n", image_filename);
|
||||
exit(EXIT_FAILURE);
|
||||
}
|
||||
|
||||
initializeData(image_path);
|
||||
free(image_path);
|
||||
}
|
||||
|
||||
void initGL(int *argc, char **argv) {
|
||||
glutInit(argc, argv);
|
||||
glutInitDisplayMode(GLUT_RGBA | GLUT_DOUBLE);
|
||||
glutInitWindowSize(wWidth, wHeight);
|
||||
glutCreateWindow("Function Pointers [CUDA Edge Detection]n");
|
||||
|
||||
if (!isGLVersionSupported(1, 5) ||
|
||||
!areGLExtensionsSupported(
|
||||
"GL_ARB_vertex_buffer_object GL_ARB_pixel_buffer_object")) {
|
||||
fprintf(stderr, "Error: failed to get minimal extensions for demo\n");
|
||||
fprintf(stderr, "This sample requires:\n");
|
||||
fprintf(stderr, " OpenGL version 1.5\n");
|
||||
fprintf(stderr, " GL_ARB_vertex_buffer_object\n");
|
||||
fprintf(stderr, " GL_ARB_pixel_buffer_object\n");
|
||||
|
||||
exit(EXIT_WAIVED);
|
||||
}
|
||||
}
|
||||
|
||||
void runAutoTest(int argc, char *argv[]) {
|
||||
printf("[%s] (automated testing w/ readback)\n", sSDKsample);
|
||||
int devID = findCudaDevice(argc, (const char **)argv);
|
||||
|
||||
loadDefaultImage(argv[0]);
|
||||
|
||||
Pixel *d_result;
|
||||
checkCudaErrors(
|
||||
cudaMalloc((void **)&d_result, imWidth * imHeight * sizeof(Pixel)));
|
||||
|
||||
char *ref_file = NULL;
|
||||
char dump_file[256];
|
||||
|
||||
int mode = 0;
|
||||
mode = getCmdLineArgumentInt(argc, (const char **)argv, "mode");
|
||||
getCmdLineArgumentString(argc, (const char **)argv, "file", &ref_file);
|
||||
|
||||
switch (mode) {
|
||||
case 0:
|
||||
g_SobelDisplayMode = SOBELDISPLAY_IMAGE;
|
||||
sprintf(dump_file, "lena_orig.pgm");
|
||||
break;
|
||||
|
||||
case 1:
|
||||
g_SobelDisplayMode = SOBELDISPLAY_SOBELTEX;
|
||||
sprintf(dump_file, "lena_tex.pgm");
|
||||
break;
|
||||
|
||||
case 2:
|
||||
g_SobelDisplayMode = SOBELDISPLAY_SOBELSHARED;
|
||||
sprintf(dump_file, "lena_shared.pgm");
|
||||
break;
|
||||
|
||||
default:
|
||||
printf("Invalid Filter Mode File\n");
|
||||
exit(EXIT_FAILURE);
|
||||
break;
|
||||
}
|
||||
|
||||
printf("AutoTest: %s <%s>\n", sSDKsample, filterMode[g_SobelDisplayMode]);
|
||||
sobelFilter(d_result, imWidth, imHeight, g_SobelDisplayMode, imageScale,
|
||||
blockOp, pointOp);
|
||||
checkCudaErrors(cudaDeviceSynchronize());
|
||||
|
||||
unsigned char *h_result =
|
||||
(unsigned char *)malloc(imWidth * imHeight * sizeof(Pixel));
|
||||
checkCudaErrors(cudaMemcpy(h_result, d_result,
|
||||
imWidth * imHeight * sizeof(Pixel),
|
||||
cudaMemcpyDeviceToHost));
|
||||
sdkSavePGM(dump_file, h_result, imWidth, imHeight);
|
||||
|
||||
if (!sdkComparePGM(dump_file, sdkFindFilePath(ref_file, argv[0]),
|
||||
MAX_EPSILON_ERROR, 0.15f, false)) {
|
||||
g_TotalErrors++;
|
||||
}
|
||||
|
||||
checkCudaErrors(cudaFree(d_result));
|
||||
free(h_result);
|
||||
|
||||
if (g_TotalErrors != 0) {
|
||||
printf("Test failed!\n");
|
||||
exit(EXIT_FAILURE);
|
||||
}
|
||||
|
||||
printf("Test passed!\n");
|
||||
exit(EXIT_SUCCESS);
|
||||
}
|
||||
|
||||
int main(int argc, char **argv) {
|
||||
pArgc = &argc;
|
||||
pArgv = argv;
|
||||
|
||||
#if defined(__linux__)
|
||||
setenv("DISPLAY", ":0", 0);
|
||||
#endif
|
||||
|
||||
printf("%s Starting...\n\n", argv[0]);
|
||||
|
||||
if (checkCmdLineFlag(argc, (const char **)argv, "help")) {
|
||||
printf("\nUsage: FunctionPointers (SobelFilter) <options>\n");
|
||||
printf("\t\t-mode=n (0=original, 1=texture, 2=smem + texture)\n");
|
||||
printf("\t\t-file=ref_orig.pgm (ref_tex.pgm, ref_shared.pgm)\n\n");
|
||||
|
||||
exit(EXIT_WAIVED);
|
||||
}
|
||||
|
||||
if (checkCmdLineFlag(argc, (const char **)argv, "file")) {
|
||||
g_bQAReadback = true;
|
||||
runAutoTest(argc, argv);
|
||||
}
|
||||
|
||||
// use command-line specified CUDA device, otherwise use device with highest
|
||||
// Gflops/s
|
||||
if (checkCmdLineFlag(argc, (const char **)argv, "device")) {
|
||||
printf(
|
||||
" This SDK does not explicitly support -device=n when running with "
|
||||
"OpenGL.\n");
|
||||
printf(
|
||||
" When specifying -device=n (n=0,1,2,....) the sample must not use "
|
||||
"OpenGL.\n");
|
||||
printf(" See details below to run without OpenGL:\n\n");
|
||||
printf(" > %s -device=n\n\n", argv[0]);
|
||||
printf("exiting...\n");
|
||||
|
||||
exit(EXIT_WAIVED);
|
||||
}
|
||||
|
||||
if (!g_bQAReadback) {
|
||||
// First initialize OpenGL context, so we can properly set the GL for CUDA.
|
||||
// This is necessary in order to achieve optimal performance with
|
||||
// OpenGL/CUDA interop.
|
||||
initGL(&argc, argv);
|
||||
|
||||
int dev = findCudaDevice(argc, (const char **)argv);
|
||||
|
||||
sdkCreateTimer(&timer);
|
||||
sdkResetTimer(&timer);
|
||||
|
||||
glutDisplayFunc(display);
|
||||
glutKeyboardFunc(keyboard);
|
||||
glutReshapeFunc(reshape);
|
||||
|
||||
loadDefaultImage(argv[0]);
|
||||
|
||||
// If code is not printing the USage, then we execute this path.
|
||||
printf("I: display Image (no filtering)\n");
|
||||
printf("T: display Sobel Edge Detection (Using Texture)\n");
|
||||
printf("S: display Sobel Edge Detection (Using SMEM+Texture)\n");
|
||||
printf("Use the '-' and '=' keys to change the brightness.\n");
|
||||
printf("b: switch block filter operation (Mean/Sobel)\n");
|
||||
printf("p: switch point filter operation (Threshold ON/OFF)\n");
|
||||
fflush(stdout);
|
||||
|
||||
#if defined(__APPLE__) || defined(MACOSX)
|
||||
atexit(cleanup);
|
||||
#else
|
||||
glutCloseFunc(cleanup);
|
||||
#endif
|
||||
|
||||
glutTimerFunc(REFRESH_DELAY, timerEvent, 0);
|
||||
glutMainLoop();
|
||||
}
|
||||
}
|
||||
402
Samples/FunctionPointers/FunctionPointers_kernels.cu
Normal file
402
Samples/FunctionPointers/FunctionPointers_kernels.cu
Normal file
@@ -0,0 +1,402 @@
|
||||
/* Copyright (c) 2021, NVIDIA CORPORATION. All rights reserved.
|
||||
*
|
||||
* Redistribution and use in source and binary forms, with or without
|
||||
* modification, are permitted provided that the following conditions
|
||||
* are met:
|
||||
* * Redistributions of source code must retain the above copyright
|
||||
* notice, this list of conditions and the following disclaimer.
|
||||
* * Redistributions in binary form must reproduce the above copyright
|
||||
* notice, this list of conditions and the following disclaimer in the
|
||||
* documentation and/or other materials provided with the distribution.
|
||||
* * Neither the name of NVIDIA CORPORATION nor the names of its
|
||||
* contributors may be used to endorse or promote products derived
|
||||
* from this software without specific prior written permission.
|
||||
*
|
||||
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
|
||||
* EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
||||
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
||||
* PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
|
||||
* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
|
||||
* EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
|
||||
* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
|
||||
* PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
|
||||
* OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
||||
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
||||
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
||||
*/
|
||||
|
||||
#include <stdio.h>
|
||||
#include <stdlib.h>
|
||||
#include <cooperative_groups.h>
|
||||
|
||||
namespace cg = cooperative_groups;
|
||||
|
||||
#include <helper_cuda.h>
|
||||
|
||||
#include "FunctionPointers_kernels.h"
|
||||
|
||||
// Texture object for reading image
|
||||
cudaTextureObject_t tex;
|
||||
extern __shared__ unsigned char LocalBlock[];
|
||||
static cudaArray *array = NULL;
|
||||
|
||||
#define RADIUS 1
|
||||
|
||||
// pixel value used for thresholding function,
|
||||
// works well with sample image 'lena'
|
||||
#define THRESHOLD 150.0f
|
||||
|
||||
#ifdef FIXED_BLOCKWIDTH
|
||||
#define BlockWidth 80
|
||||
#define SharedPitch 384
|
||||
#endif
|
||||
|
||||
// A function pointer can be declared explicitly like this line:
|
||||
//__device__ unsigned char (*pointFunction)(unsigned char, float ) = NULL;
|
||||
// or by using typedef's like below:
|
||||
|
||||
typedef unsigned char (*blockFunction_t)(unsigned char, unsigned char,
|
||||
unsigned char, unsigned char,
|
||||
unsigned char, unsigned char,
|
||||
unsigned char, unsigned char,
|
||||
unsigned char, float);
|
||||
|
||||
typedef unsigned char (*pointFunction_t)(unsigned char, float);
|
||||
|
||||
__device__ blockFunction_t blockFunction;
|
||||
|
||||
__device__ unsigned char ComputeSobel(unsigned char ul, // upper left
|
||||
unsigned char um, // upper middle
|
||||
unsigned char ur, // upper right
|
||||
unsigned char ml, // middle left
|
||||
unsigned char mm, // middle (unused)
|
||||
unsigned char mr, // middle right
|
||||
unsigned char ll, // lower left
|
||||
unsigned char lm, // lower middle
|
||||
unsigned char lr, // lower right
|
||||
float fScale) {
|
||||
short Horz = ur + 2 * mr + lr - ul - 2 * ml - ll;
|
||||
short Vert = ul + 2 * um + ur - ll - 2 * lm - lr;
|
||||
short Sum = (short)(fScale * (abs((int)Horz) + abs((int)Vert)));
|
||||
return (unsigned char)((Sum < 0) ? 0 : ((Sum > 255) ? 255 : Sum));
|
||||
}
|
||||
|
||||
// define a function pointer and initialize to NULL
|
||||
__device__ unsigned char (*varFunction)(unsigned char, unsigned char,
|
||||
unsigned char, unsigned char,
|
||||
unsigned char, unsigned char,
|
||||
unsigned char, unsigned char,
|
||||
unsigned char, float x) = NULL;
|
||||
|
||||
__device__ unsigned char ComputeBox(unsigned char ul, // upper left
|
||||
unsigned char um, // upper middle
|
||||
unsigned char ur, // upper right
|
||||
unsigned char ml, // middle left
|
||||
unsigned char mm, // middle...middle
|
||||
unsigned char mr, // middle right
|
||||
unsigned char ll, // lower left
|
||||
unsigned char lm, // lower middle
|
||||
unsigned char lr, // lower right
|
||||
float fscale) {
|
||||
short Sum = (short)(ul + um + ur + ml + mm + mr + ll + lm + lr) / 9;
|
||||
Sum *= fscale;
|
||||
return (unsigned char)((Sum < 0) ? 0 : ((Sum > 255) ? 255 : Sum));
|
||||
}
|
||||
__device__ unsigned char Threshold(unsigned char in, float thresh) {
|
||||
if (in > thresh) {
|
||||
return 0xFF;
|
||||
} else {
|
||||
return 0;
|
||||
}
|
||||
}
|
||||
|
||||
// Declare function tables, one for the point function chosen, one for the
|
||||
// block function chosen. The number of entries is determined by the
|
||||
// enum in FunctionPointers_kernels.h
|
||||
__device__ blockFunction_t blockFunction_table[LAST_BLOCK_FILTER];
|
||||
__device__ pointFunction_t pointFunction_table[LAST_POINT_FILTER];
|
||||
|
||||
// Declare device side function pointers. We retrieve them later with
|
||||
// cudaMemcpyFromSymbol to set our function tables above in some
|
||||
// particular order specified at runtime.
|
||||
__device__ blockFunction_t pComputeSobel = ComputeSobel;
|
||||
__device__ blockFunction_t pComputeBox = ComputeBox;
|
||||
__device__ pointFunction_t pComputeThreshold = Threshold;
|
||||
|
||||
// Allocate host side tables to mirror the device side, and later, we
|
||||
// fill these tables with the function pointers. This lets us send
|
||||
// the pointers to the kernel on invocation, as a method of choosing
|
||||
// which function to run.
|
||||
blockFunction_t h_blockFunction_table[2];
|
||||
pointFunction_t h_pointFunction_table[2];
|
||||
|
||||
// Perform a filter operation on the data, using shared memory
|
||||
// The actual operation performed is
|
||||
// determined by the function pointer "blockFunction" and selected
|
||||
// by the integer argument "blockOperation" and has access
|
||||
// to an apron around the current pixel being processed.
|
||||
// Following the block operation, a per-pixel operation,
|
||||
// pointed to by pPointFunction is performed before the final
|
||||
// pixel is produced.
|
||||
__global__ void SobelShared(uchar4 *pSobelOriginal, unsigned short SobelPitch,
|
||||
#ifndef FIXED_BLOCKWIDTH
|
||||
short BlockWidth, short SharedPitch,
|
||||
#endif
|
||||
short w, short h, float fScale, int blockOperation,
|
||||
pointFunction_t pPointFunction,
|
||||
cudaTextureObject_t tex) {
|
||||
// Handle to thread block group
|
||||
cg::thread_block cta = cg::this_thread_block();
|
||||
short u = 4 * blockIdx.x * BlockWidth;
|
||||
short v = blockIdx.y * blockDim.y + threadIdx.y;
|
||||
short ib;
|
||||
|
||||
int SharedIdx = threadIdx.y * SharedPitch;
|
||||
|
||||
for (ib = threadIdx.x; ib < BlockWidth + 2 * RADIUS; ib += blockDim.x) {
|
||||
LocalBlock[SharedIdx + 4 * ib + 0] = tex2D<unsigned char>(
|
||||
tex, (float)(u + 4 * ib - RADIUS + 0), (float)(v - RADIUS));
|
||||
LocalBlock[SharedIdx + 4 * ib + 1] = tex2D<unsigned char>(
|
||||
tex, (float)(u + 4 * ib - RADIUS + 1), (float)(v - RADIUS));
|
||||
LocalBlock[SharedIdx + 4 * ib + 2] = tex2D<unsigned char>(
|
||||
tex, (float)(u + 4 * ib - RADIUS + 2), (float)(v - RADIUS));
|
||||
LocalBlock[SharedIdx + 4 * ib + 3] = tex2D<unsigned char>(
|
||||
tex, (float)(u + 4 * ib - RADIUS + 3), (float)(v - RADIUS));
|
||||
}
|
||||
|
||||
if (threadIdx.y < RADIUS * 2) {
|
||||
//
|
||||
// copy trailing RADIUS*2 rows of pixels into shared
|
||||
//
|
||||
SharedIdx = (blockDim.y + threadIdx.y) * SharedPitch;
|
||||
|
||||
for (ib = threadIdx.x; ib < BlockWidth + 2 * RADIUS; ib += blockDim.x) {
|
||||
LocalBlock[SharedIdx + 4 * ib + 0] =
|
||||
tex2D<unsigned char>(tex, (float)(u + 4 * ib - RADIUS + 0),
|
||||
(float)(v + blockDim.y - RADIUS));
|
||||
LocalBlock[SharedIdx + 4 * ib + 1] =
|
||||
tex2D<unsigned char>(tex, (float)(u + 4 * ib - RADIUS + 1),
|
||||
(float)(v + blockDim.y - RADIUS));
|
||||
LocalBlock[SharedIdx + 4 * ib + 2] =
|
||||
tex2D<unsigned char>(tex, (float)(u + 4 * ib - RADIUS + 2),
|
||||
(float)(v + blockDim.y - RADIUS));
|
||||
LocalBlock[SharedIdx + 4 * ib + 3] =
|
||||
tex2D<unsigned char>(tex, (float)(u + 4 * ib - RADIUS + 3),
|
||||
(float)(v + blockDim.y - RADIUS));
|
||||
}
|
||||
}
|
||||
|
||||
cg::sync(cta);
|
||||
|
||||
u >>= 2; // index as uchar4 from here
|
||||
uchar4 *pSobel = (uchar4 *)(((char *)pSobelOriginal) + v * SobelPitch);
|
||||
SharedIdx = threadIdx.y * SharedPitch;
|
||||
|
||||
blockFunction = blockFunction_table[blockOperation];
|
||||
|
||||
for (ib = threadIdx.x; ib < BlockWidth; ib += blockDim.x) {
|
||||
uchar4 out;
|
||||
|
||||
unsigned char pix00 = LocalBlock[SharedIdx + 4 * ib + 0 * SharedPitch + 0];
|
||||
unsigned char pix01 = LocalBlock[SharedIdx + 4 * ib + 0 * SharedPitch + 1];
|
||||
unsigned char pix02 = LocalBlock[SharedIdx + 4 * ib + 0 * SharedPitch + 2];
|
||||
unsigned char pix10 = LocalBlock[SharedIdx + 4 * ib + 1 * SharedPitch + 0];
|
||||
unsigned char pix11 = LocalBlock[SharedIdx + 4 * ib + 1 * SharedPitch + 1];
|
||||
unsigned char pix12 = LocalBlock[SharedIdx + 4 * ib + 1 * SharedPitch + 2];
|
||||
unsigned char pix20 = LocalBlock[SharedIdx + 4 * ib + 2 * SharedPitch + 0];
|
||||
unsigned char pix21 = LocalBlock[SharedIdx + 4 * ib + 2 * SharedPitch + 1];
|
||||
unsigned char pix22 = LocalBlock[SharedIdx + 4 * ib + 2 * SharedPitch + 2];
|
||||
|
||||
out.x = (*blockFunction)(pix00, pix01, pix02, pix10, pix11, pix12, pix20,
|
||||
pix21, pix22, fScale);
|
||||
|
||||
pix00 = LocalBlock[SharedIdx + 4 * ib + 0 * SharedPitch + 3];
|
||||
pix10 = LocalBlock[SharedIdx + 4 * ib + 1 * SharedPitch + 3];
|
||||
pix20 = LocalBlock[SharedIdx + 4 * ib + 2 * SharedPitch + 3];
|
||||
out.y = (*blockFunction)(pix01, pix02, pix00, pix11, pix12, pix10, pix21,
|
||||
pix22, pix20, fScale);
|
||||
|
||||
pix01 = LocalBlock[SharedIdx + 4 * ib + 0 * SharedPitch + 4];
|
||||
pix11 = LocalBlock[SharedIdx + 4 * ib + 1 * SharedPitch + 4];
|
||||
pix21 = LocalBlock[SharedIdx + 4 * ib + 2 * SharedPitch + 4];
|
||||
out.z = (*blockFunction)(pix02, pix00, pix01, pix12, pix10, pix11, pix22,
|
||||
pix20, pix21, fScale);
|
||||
|
||||
pix02 = LocalBlock[SharedIdx + 4 * ib + 0 * SharedPitch + 5];
|
||||
pix12 = LocalBlock[SharedIdx + 4 * ib + 1 * SharedPitch + 5];
|
||||
pix22 = LocalBlock[SharedIdx + 4 * ib + 2 * SharedPitch + 5];
|
||||
out.w = (*blockFunction)(pix00, pix01, pix02, pix10, pix11, pix12, pix20,
|
||||
pix21, pix22, fScale);
|
||||
|
||||
if (pPointFunction != NULL) {
|
||||
out.x = (*pPointFunction)(out.x, THRESHOLD);
|
||||
out.y = (*pPointFunction)(out.y, THRESHOLD);
|
||||
out.z = (*pPointFunction)(out.z, THRESHOLD);
|
||||
out.w = (*pPointFunction)(out.w, THRESHOLD);
|
||||
}
|
||||
|
||||
if (u + ib < w / 4 && v < h) {
|
||||
pSobel[u + ib] = out;
|
||||
}
|
||||
}
|
||||
|
||||
cg::sync(cta);
|
||||
}
|
||||
|
||||
__global__ void SobelCopyImage(Pixel *pSobelOriginal, unsigned int Pitch, int w,
|
||||
int h, float fscale, cudaTextureObject_t tex) {
|
||||
unsigned char *pSobel =
|
||||
(unsigned char *)(((char *)pSobelOriginal) + blockIdx.x * Pitch);
|
||||
|
||||
for (int i = threadIdx.x; i < w; i += blockDim.x) {
|
||||
pSobel[i] = min(
|
||||
max((tex2D<unsigned char>(tex, (float)i, (float)blockIdx.x) * fscale),
|
||||
0.f),
|
||||
255.f);
|
||||
}
|
||||
}
|
||||
|
||||
// Perform block and pointer filtering using texture lookups.
|
||||
// The block and point operations are determined by the
|
||||
// input argument (see comment above for "SobelShared" function)
|
||||
__global__ void SobelTex(Pixel *pSobelOriginal, unsigned int Pitch, int w,
|
||||
int h, float fScale, int blockOperation,
|
||||
pointFunction_t pPointOperation,
|
||||
cudaTextureObject_t tex) {
|
||||
unsigned char *pSobel =
|
||||
(unsigned char *)(((char *)pSobelOriginal) + blockIdx.x * Pitch);
|
||||
unsigned char tmp = 0;
|
||||
|
||||
for (int i = threadIdx.x; i < w; i += blockDim.x) {
|
||||
unsigned char pix00 =
|
||||
tex2D<unsigned char>(tex, (float)i - 1, (float)blockIdx.x - 1);
|
||||
unsigned char pix01 =
|
||||
tex2D<unsigned char>(tex, (float)i + 0, (float)blockIdx.x - 1);
|
||||
unsigned char pix02 =
|
||||
tex2D<unsigned char>(tex, (float)i + 1, (float)blockIdx.x - 1);
|
||||
unsigned char pix10 =
|
||||
tex2D<unsigned char>(tex, (float)i - 1, (float)blockIdx.x + 0);
|
||||
unsigned char pix11 =
|
||||
tex2D<unsigned char>(tex, (float)i + 0, (float)blockIdx.x + 0);
|
||||
unsigned char pix12 =
|
||||
tex2D<unsigned char>(tex, (float)i + 1, (float)blockIdx.x + 0);
|
||||
unsigned char pix20 =
|
||||
tex2D<unsigned char>(tex, (float)i - 1, (float)blockIdx.x + 1);
|
||||
unsigned char pix21 =
|
||||
tex2D<unsigned char>(tex, (float)i + 0, (float)blockIdx.x + 1);
|
||||
unsigned char pix22 =
|
||||
tex2D<unsigned char>(tex, (float)i + 1, (float)blockIdx.x + 1);
|
||||
tmp = (*(blockFunction_table[blockOperation]))(
|
||||
pix00, pix01, pix02, pix10, pix11, pix12, pix20, pix21, pix22, fScale);
|
||||
|
||||
if (pPointOperation != NULL) {
|
||||
tmp = (*pPointOperation)(tmp, 150.0);
|
||||
}
|
||||
|
||||
pSobel[i] = tmp;
|
||||
}
|
||||
}
|
||||
|
||||
extern "C" void setupTexture(int iw, int ih, Pixel *data, int Bpp) {
|
||||
cudaChannelFormatDesc desc;
|
||||
|
||||
if (Bpp == 1) {
|
||||
desc = cudaCreateChannelDesc<unsigned char>();
|
||||
} else {
|
||||
desc = cudaCreateChannelDesc<uchar4>();
|
||||
}
|
||||
|
||||
checkCudaErrors(cudaMallocArray(&array, &desc, iw, ih));
|
||||
checkCudaErrors(cudaMemcpy2DToArray(
|
||||
array, 0, 0, data, iw * Bpp * sizeof(Pixel), iw * Bpp * sizeof(Pixel), ih,
|
||||
cudaMemcpyHostToDevice));
|
||||
|
||||
cudaResourceDesc texRes;
|
||||
memset(&texRes, 0, sizeof(cudaResourceDesc));
|
||||
|
||||
texRes.resType = cudaResourceTypeArray;
|
||||
texRes.res.array.array = array;
|
||||
|
||||
cudaTextureDesc texDescr;
|
||||
memset(&texDescr, 0, sizeof(cudaTextureDesc));
|
||||
|
||||
checkCudaErrors(cudaCreateTextureObject(&tex, &texRes, &texDescr, NULL));
|
||||
}
|
||||
|
||||
extern "C" void deleteTexture(void) {
|
||||
checkCudaErrors(cudaFreeArray(array));
|
||||
checkCudaErrors(cudaDestroyTextureObject(tex));
|
||||
}
|
||||
|
||||
// Copy the pointers from the function tables to the host side
|
||||
void setupFunctionTables() {
|
||||
// Dynamically assign the function table.
|
||||
// Copy the function pointers to their appropriate locations according to the
|
||||
// enum
|
||||
checkCudaErrors(cudaMemcpyFromSymbol(&h_blockFunction_table[SOBEL_FILTER],
|
||||
pComputeSobel, sizeof(blockFunction_t)));
|
||||
checkCudaErrors(cudaMemcpyFromSymbol(&h_blockFunction_table[BOX_FILTER],
|
||||
pComputeBox, sizeof(blockFunction_t)));
|
||||
|
||||
// do the same for the point function, where the 2nd function is NULL ("no-op"
|
||||
// filter, skipped in kernel code)
|
||||
checkCudaErrors(cudaMemcpyFromSymbol(&h_pointFunction_table[THRESHOLD_FILTER],
|
||||
pComputeThreshold,
|
||||
sizeof(pointFunction_t)));
|
||||
h_pointFunction_table[NULL_FILTER] = NULL;
|
||||
|
||||
// now copy the function tables back to the device, so if we wish we can use
|
||||
// an index into the table to choose them
|
||||
// We have now set the order in the function table according to our enum.
|
||||
checkCudaErrors(
|
||||
cudaMemcpyToSymbol(blockFunction_table, h_blockFunction_table,
|
||||
sizeof(blockFunction_t) * LAST_BLOCK_FILTER));
|
||||
checkCudaErrors(
|
||||
cudaMemcpyToSymbol(pointFunction_table, h_pointFunction_table,
|
||||
sizeof(pointFunction_t) * LAST_POINT_FILTER));
|
||||
}
|
||||
|
||||
// Wrapper for the __global__ call that sets up the texture and threads
|
||||
// Below two methods for selecting the image processing function to run are
|
||||
// shown.
|
||||
// BlockOperation is an integer kernel argument used as an index into the
|
||||
// blockFunction_table on the device side
|
||||
// pPointOp is itself a function pointer passed as a kernel argument, retrieved
|
||||
// from a host side copy of the function table
|
||||
extern "C" void sobelFilter(Pixel *odata, int iw, int ih,
|
||||
enum SobelDisplayMode mode, float fScale,
|
||||
int blockOperation, int pointOperation) {
|
||||
pointFunction_t pPointOp = h_pointFunction_table[pointOperation];
|
||||
|
||||
switch (mode) {
|
||||
case SOBELDISPLAY_IMAGE:
|
||||
SobelCopyImage<<<ih, 384>>>(odata, iw, iw, ih, fScale, tex);
|
||||
break;
|
||||
|
||||
case SOBELDISPLAY_SOBELTEX:
|
||||
SobelTex<<<ih, 384>>>(odata, iw, iw, ih, fScale, blockOperation, pPointOp,
|
||||
tex);
|
||||
break;
|
||||
|
||||
case SOBELDISPLAY_SOBELSHARED: {
|
||||
dim3 threads(16, 4);
|
||||
#ifndef FIXED_BLOCKWIDTH
|
||||
int BlockWidth = 80; // must be divisible by 16 for coalescing
|
||||
#endif
|
||||
dim3 blocks = dim3(iw / (4 * BlockWidth) + (0 != iw % (4 * BlockWidth)),
|
||||
ih / threads.y + (0 != ih % threads.y));
|
||||
int SharedPitch = ~0x3f & (4 * (BlockWidth + 2 * RADIUS) + 0x3f);
|
||||
int sharedMem = SharedPitch * (threads.y + 2 * RADIUS);
|
||||
|
||||
// for the shared kernel, width must be divisible by 4
|
||||
iw &= ~3;
|
||||
|
||||
SobelShared<<<blocks, threads, sharedMem>>>(
|
||||
(uchar4 *)odata, iw,
|
||||
#ifndef FIXED_BLOCKWIDTH
|
||||
BlockWidth, SharedPitch,
|
||||
#endif
|
||||
iw, ih, fScale, blockOperation, pPointOp, tex);
|
||||
} break;
|
||||
}
|
||||
}
|
||||
58
Samples/FunctionPointers/FunctionPointers_kernels.h
Normal file
58
Samples/FunctionPointers/FunctionPointers_kernels.h
Normal file
@@ -0,0 +1,58 @@
|
||||
/* Copyright (c) 2021, NVIDIA CORPORATION. All rights reserved.
|
||||
*
|
||||
* Redistribution and use in source and binary forms, with or without
|
||||
* modification, are permitted provided that the following conditions
|
||||
* are met:
|
||||
* * Redistributions of source code must retain the above copyright
|
||||
* notice, this list of conditions and the following disclaimer.
|
||||
* * Redistributions in binary form must reproduce the above copyright
|
||||
* notice, this list of conditions and the following disclaimer in the
|
||||
* documentation and/or other materials provided with the distribution.
|
||||
* * Neither the name of NVIDIA CORPORATION nor the names of its
|
||||
* contributors may be used to endorse or promote products derived
|
||||
* from this software without specific prior written permission.
|
||||
*
|
||||
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
|
||||
* EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
||||
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
||||
* PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
|
||||
* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
|
||||
* EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
|
||||
* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
|
||||
* PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
|
||||
* OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
||||
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
||||
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
||||
*/
|
||||
|
||||
#ifndef __SOBELFILTER_KERNELS_H_
|
||||
#define __SOBELFILTER_KERNELS_H_
|
||||
|
||||
typedef unsigned char Pixel;
|
||||
|
||||
// global determines which filter to invoke
|
||||
enum SobelDisplayMode {
|
||||
SOBELDISPLAY_IMAGE = 0,
|
||||
SOBELDISPLAY_SOBELTEX,
|
||||
SOBELDISPLAY_SOBELSHARED
|
||||
};
|
||||
|
||||
// Enums to set up the function table
|
||||
// note: if you change these be sure to recompile those files
|
||||
// that include this header or ensure the .h is in the
|
||||
// dependencies for the related object files
|
||||
enum POINT_ENUM { SOBEL_FILTER = 0, BOX_FILTER, LAST_POINT_FILTER };
|
||||
|
||||
enum BLOCK_ENUM { THRESHOLD_FILTER = 0, NULL_FILTER, LAST_BLOCK_FILTER };
|
||||
|
||||
extern enum SobelDisplayMode g_SobelDisplayMode;
|
||||
|
||||
extern "C" void sobelFilter(Pixel *odata, int iw, int ih,
|
||||
enum SobelDisplayMode mode, float fScale,
|
||||
int blockOperation, int pointOperation);
|
||||
extern "C" void setupTexture(int iw, int ih, Pixel *data, int Bpp);
|
||||
extern "C" void deleteTexture(void);
|
||||
extern "C" void initFilter(void);
|
||||
void setupFunctionTables();
|
||||
|
||||
#endif
|
||||
20
Samples/FunctionPointers/FunctionPointers_vs2017.sln
Normal file
20
Samples/FunctionPointers/FunctionPointers_vs2017.sln
Normal file
@@ -0,0 +1,20 @@
|
||||
|
||||
Microsoft Visual Studio Solution File, Format Version 12.00
|
||||
# Visual Studio 2017
|
||||
Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "FunctionPointers", "FunctionPointers_vs2017.vcxproj", "{997E0757-EA74-4A4E-A0FC-47D8C8831A15}"
|
||||
EndProject
|
||||
Global
|
||||
GlobalSection(SolutionConfigurationPlatforms) = preSolution
|
||||
Debug|x64 = Debug|x64
|
||||
Release|x64 = Release|x64
|
||||
EndGlobalSection
|
||||
GlobalSection(ProjectConfigurationPlatforms) = postSolution
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Debug|x64.ActiveCfg = Debug|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Debug|x64.Build.0 = Debug|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Release|x64.ActiveCfg = Release|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Release|x64.Build.0 = Release|x64
|
||||
EndGlobalSection
|
||||
GlobalSection(SolutionProperties) = preSolution
|
||||
HideSolutionNode = FALSE
|
||||
EndGlobalSection
|
||||
EndGlobal
|
||||
124
Samples/FunctionPointers/FunctionPointers_vs2017.vcxproj
Normal file
124
Samples/FunctionPointers/FunctionPointers_vs2017.vcxproj
Normal file
@@ -0,0 +1,124 @@
|
||||
<?xml version="1.0" encoding="utf-8"?>
|
||||
<Project DefaultTargets="Build" ToolsVersion="4.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
|
||||
<PropertyGroup>
|
||||
<CUDAPropsPath Condition="'$(CUDAPropsPath)'==''">$(VCTargetsPath)\BuildCustomizations</CUDAPropsPath>
|
||||
</PropertyGroup>
|
||||
<ItemGroup Label="ProjectConfigurations">
|
||||
<ProjectConfiguration Include="Debug|x64">
|
||||
<Configuration>Debug</Configuration>
|
||||
<Platform>x64</Platform>
|
||||
</ProjectConfiguration>
|
||||
<ProjectConfiguration Include="Release|x64">
|
||||
<Configuration>Release</Configuration>
|
||||
<Platform>x64</Platform>
|
||||
</ProjectConfiguration>
|
||||
</ItemGroup>
|
||||
<PropertyGroup Label="Globals">
|
||||
<ProjectGuid>{997E0757-EA74-4A4E-A0FC-47D8C8831A15}</ProjectGuid>
|
||||
<RootNamespace>FunctionPointers_vs2017</RootNamespace>
|
||||
<ProjectName>FunctionPointers</ProjectName>
|
||||
<CudaToolkitCustomDir />
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(WindowsTargetPlatformVersion)'==''">
|
||||
<LatestTargetPlatformVersion>$([Microsoft.Build.Utilities.ToolLocationHelper]::GetLatestSDKTargetPlatformVersion('Windows', '10.0'))</LatestTargetPlatformVersion>
|
||||
<WindowsTargetPlatformVersion Condition="'$(WindowsTargetPlatformVersion)' == ''">$(LatestTargetPlatformVersion)</WindowsTargetPlatformVersion>
|
||||
<TargetPlatformVersion>$(WindowsTargetPlatformVersion)</TargetPlatformVersion>
|
||||
</PropertyGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" />
|
||||
<PropertyGroup>
|
||||
<ConfigurationType>Application</ConfigurationType>
|
||||
<CharacterSet>MultiByte</CharacterSet>
|
||||
<PlatformToolset>v141</PlatformToolset>
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Configuration)'=='Debug'">
|
||||
<UseDebugLibraries>true</UseDebugLibraries>
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Configuration)'=='Release'">
|
||||
<WholeProgramOptimization>true</WholeProgramOptimization>
|
||||
</PropertyGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.props" />
|
||||
<ImportGroup Label="ExtensionSettings">
|
||||
<Import Project="$(CUDAPropsPath)\CUDA 11.5.props" />
|
||||
</ImportGroup>
|
||||
<ImportGroup Label="PropertySheets">
|
||||
<Import Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" />
|
||||
</ImportGroup>
|
||||
<PropertyGroup Label="UserMacros" />
|
||||
<PropertyGroup>
|
||||
<IntDir>$(Platform)/$(Configuration)/</IntDir>
|
||||
<IncludePath>$(IncludePath)</IncludePath>
|
||||
<CodeAnalysisRuleSet>AllRules.ruleset</CodeAnalysisRuleSet>
|
||||
<CodeAnalysisRules />
|
||||
<CodeAnalysisRuleAssemblies />
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Platform)'=='x64'">
|
||||
<OutDir>../../bin/win64/$(Configuration)/</OutDir>
|
||||
</PropertyGroup>
|
||||
<ItemDefinitionGroup>
|
||||
<ClCompile>
|
||||
<WarningLevel>Level3</WarningLevel>
|
||||
<PreprocessorDefinitions>WIN32;_MBCS;%(PreprocessorDefinitions)</PreprocessorDefinitions>
|
||||
<AdditionalIncludeDirectories>./;$(CudaToolkitDir)/include;../../Common;</AdditionalIncludeDirectories>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<SubSystem>Console</SubSystem>
|
||||
<AdditionalDependencies>cudart_static.lib;freeglut.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
|
||||
<AdditionalLibraryDirectories>$(CudaToolkitLibDir);../../Common/lib/$(PlatformName);</AdditionalLibraryDirectories>
|
||||
<OutputFile>$(OutDir)/FunctionPointers.exe</OutputFile>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<CodeGeneration>compute_35,sm_35;compute_37,sm_37;compute_50,sm_50;compute_52,sm_52;compute_60,sm_60;compute_61,sm_61;compute_70,sm_70;compute_75,sm_75;compute_80,sm_80;compute_86,sm_86;</CodeGeneration>
|
||||
<AdditionalOptions>-Xcompiler "/wd 4819" %(AdditionalOptions)</AdditionalOptions>
|
||||
<Include>./;../../Common</Include>
|
||||
<Defines>WIN32</Defines>
|
||||
<AdditionalOptions>--threads 0</AdditionalOptions>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemDefinitionGroup Condition="'$(Configuration)'=='Debug'">
|
||||
<ClCompile>
|
||||
<Optimization>Disabled</Optimization>
|
||||
<RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<GenerateDebugInformation>true</GenerateDebugInformation>
|
||||
<LinkTimeCodeGeneration>Default</LinkTimeCodeGeneration>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<Runtime>MTd</Runtime>
|
||||
<TargetMachinePlatform>64</TargetMachinePlatform>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemDefinitionGroup Condition="'$(Configuration)'=='Release'">
|
||||
<ClCompile>
|
||||
<Optimization>MaxSpeed</Optimization>
|
||||
<RuntimeLibrary>MultiThreaded</RuntimeLibrary>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<GenerateDebugInformation>false</GenerateDebugInformation>
|
||||
<LinkTimeCodeGeneration>UseLinkTimeCodeGeneration</LinkTimeCodeGeneration>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<Runtime>MT</Runtime>
|
||||
<TargetMachinePlatform>64</TargetMachinePlatform>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemDefinitionGroup Condition="'$(Platform)'=='Win32'">
|
||||
<Link>
|
||||
|
||||
</Link>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemDefinitionGroup Condition="'$(Platform)'=='x64'">
|
||||
<Link>
|
||||
<AdditionalDependencies>glew64.lib;%(AdditionalDependencies)</AdditionalDependencies>
|
||||
</Link>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemGroup>
|
||||
<ClCompile Include="FunctionPointers.cpp" />
|
||||
<CudaCompile Include="FunctionPointers_kernels.cu" />
|
||||
<ClInclude Include="FunctionPointers_kernels.h" />
|
||||
</ItemGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" />
|
||||
<ImportGroup Label="ExtensionTargets">
|
||||
<Import Project="$(CUDAPropsPath)\CUDA 11.5.targets" />
|
||||
</ImportGroup>
|
||||
</Project>
|
||||
20
Samples/FunctionPointers/FunctionPointers_vs2019.sln
Normal file
20
Samples/FunctionPointers/FunctionPointers_vs2019.sln
Normal file
@@ -0,0 +1,20 @@
|
||||
|
||||
Microsoft Visual Studio Solution File, Format Version 12.00
|
||||
# Visual Studio 2019
|
||||
Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "FunctionPointers", "FunctionPointers_vs2019.vcxproj", "{997E0757-EA74-4A4E-A0FC-47D8C8831A15}"
|
||||
EndProject
|
||||
Global
|
||||
GlobalSection(SolutionConfigurationPlatforms) = preSolution
|
||||
Debug|x64 = Debug|x64
|
||||
Release|x64 = Release|x64
|
||||
EndGlobalSection
|
||||
GlobalSection(ProjectConfigurationPlatforms) = postSolution
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Debug|x64.ActiveCfg = Debug|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Debug|x64.Build.0 = Debug|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Release|x64.ActiveCfg = Release|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Release|x64.Build.0 = Release|x64
|
||||
EndGlobalSection
|
||||
GlobalSection(SolutionProperties) = preSolution
|
||||
HideSolutionNode = FALSE
|
||||
EndGlobalSection
|
||||
EndGlobal
|
||||
120
Samples/FunctionPointers/FunctionPointers_vs2019.vcxproj
Normal file
120
Samples/FunctionPointers/FunctionPointers_vs2019.vcxproj
Normal file
@@ -0,0 +1,120 @@
|
||||
<?xml version="1.0" encoding="utf-8"?>
|
||||
<Project DefaultTargets="Build" ToolsVersion="4.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
|
||||
<PropertyGroup>
|
||||
<CUDAPropsPath Condition="'$(CUDAPropsPath)'==''">$(VCTargetsPath)\BuildCustomizations</CUDAPropsPath>
|
||||
</PropertyGroup>
|
||||
<ItemGroup Label="ProjectConfigurations">
|
||||
<ProjectConfiguration Include="Debug|x64">
|
||||
<Configuration>Debug</Configuration>
|
||||
<Platform>x64</Platform>
|
||||
</ProjectConfiguration>
|
||||
<ProjectConfiguration Include="Release|x64">
|
||||
<Configuration>Release</Configuration>
|
||||
<Platform>x64</Platform>
|
||||
</ProjectConfiguration>
|
||||
</ItemGroup>
|
||||
<PropertyGroup Label="Globals">
|
||||
<ProjectGuid>{997E0757-EA74-4A4E-A0FC-47D8C8831A15}</ProjectGuid>
|
||||
<RootNamespace>FunctionPointers_vs2019</RootNamespace>
|
||||
<ProjectName>FunctionPointers</ProjectName>
|
||||
<CudaToolkitCustomDir />
|
||||
</PropertyGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" />
|
||||
<PropertyGroup>
|
||||
<ConfigurationType>Application</ConfigurationType>
|
||||
<CharacterSet>MultiByte</CharacterSet>
|
||||
<PlatformToolset>v142</PlatformToolset>
|
||||
<WindowsTargetPlatformVersion>10.0</WindowsTargetPlatformVersion>
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Configuration)'=='Debug'">
|
||||
<UseDebugLibraries>true</UseDebugLibraries>
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Configuration)'=='Release'">
|
||||
<WholeProgramOptimization>true</WholeProgramOptimization>
|
||||
</PropertyGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.props" />
|
||||
<ImportGroup Label="ExtensionSettings">
|
||||
<Import Project="$(CUDAPropsPath)\CUDA 11.5.props" />
|
||||
</ImportGroup>
|
||||
<ImportGroup Label="PropertySheets">
|
||||
<Import Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" />
|
||||
</ImportGroup>
|
||||
<PropertyGroup Label="UserMacros" />
|
||||
<PropertyGroup>
|
||||
<IntDir>$(Platform)/$(Configuration)/</IntDir>
|
||||
<IncludePath>$(IncludePath)</IncludePath>
|
||||
<CodeAnalysisRuleSet>AllRules.ruleset</CodeAnalysisRuleSet>
|
||||
<CodeAnalysisRules />
|
||||
<CodeAnalysisRuleAssemblies />
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Platform)'=='x64'">
|
||||
<OutDir>../../bin/win64/$(Configuration)/</OutDir>
|
||||
</PropertyGroup>
|
||||
<ItemDefinitionGroup>
|
||||
<ClCompile>
|
||||
<WarningLevel>Level3</WarningLevel>
|
||||
<PreprocessorDefinitions>WIN32;_MBCS;%(PreprocessorDefinitions)</PreprocessorDefinitions>
|
||||
<AdditionalIncludeDirectories>./;$(CudaToolkitDir)/include;../../Common;</AdditionalIncludeDirectories>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<SubSystem>Console</SubSystem>
|
||||
<AdditionalDependencies>cudart_static.lib;freeglut.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
|
||||
<AdditionalLibraryDirectories>$(CudaToolkitLibDir);../../Common/lib/$(PlatformName);</AdditionalLibraryDirectories>
|
||||
<OutputFile>$(OutDir)/FunctionPointers.exe</OutputFile>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<CodeGeneration>compute_35,sm_35;compute_37,sm_37;compute_50,sm_50;compute_52,sm_52;compute_60,sm_60;compute_61,sm_61;compute_70,sm_70;compute_75,sm_75;compute_80,sm_80;compute_86,sm_86;</CodeGeneration>
|
||||
<AdditionalOptions>-Xcompiler "/wd 4819" %(AdditionalOptions)</AdditionalOptions>
|
||||
<Include>./;../../Common</Include>
|
||||
<Defines>WIN32</Defines>
|
||||
<AdditionalOptions>--threads 0</AdditionalOptions>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemDefinitionGroup Condition="'$(Configuration)'=='Debug'">
|
||||
<ClCompile>
|
||||
<Optimization>Disabled</Optimization>
|
||||
<RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<GenerateDebugInformation>true</GenerateDebugInformation>
|
||||
<LinkTimeCodeGeneration>Default</LinkTimeCodeGeneration>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<Runtime>MTd</Runtime>
|
||||
<TargetMachinePlatform>64</TargetMachinePlatform>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemDefinitionGroup Condition="'$(Configuration)'=='Release'">
|
||||
<ClCompile>
|
||||
<Optimization>MaxSpeed</Optimization>
|
||||
<RuntimeLibrary>MultiThreaded</RuntimeLibrary>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<GenerateDebugInformation>false</GenerateDebugInformation>
|
||||
<LinkTimeCodeGeneration>UseLinkTimeCodeGeneration</LinkTimeCodeGeneration>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<Runtime>MT</Runtime>
|
||||
<TargetMachinePlatform>64</TargetMachinePlatform>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemDefinitionGroup Condition="'$(Platform)'=='Win32'">
|
||||
<Link>
|
||||
|
||||
</Link>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemDefinitionGroup Condition="'$(Platform)'=='x64'">
|
||||
<Link>
|
||||
<AdditionalDependencies>glew64.lib;%(AdditionalDependencies)</AdditionalDependencies>
|
||||
</Link>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemGroup>
|
||||
<ClCompile Include="FunctionPointers.cpp" />
|
||||
<CudaCompile Include="FunctionPointers_kernels.cu" />
|
||||
<ClInclude Include="FunctionPointers_kernels.h" />
|
||||
</ItemGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" />
|
||||
<ImportGroup Label="ExtensionTargets">
|
||||
<Import Project="$(CUDAPropsPath)\CUDA 11.5.targets" />
|
||||
</ImportGroup>
|
||||
</Project>
|
||||
385
Samples/FunctionPointers/Makefile
Normal file
385
Samples/FunctionPointers/Makefile
Normal file
@@ -0,0 +1,385 @@
|
||||
################################################################################
|
||||
# Copyright (c) 2019, NVIDIA CORPORATION. All rights reserved.
|
||||
#
|
||||
# Redistribution and use in source and binary forms, with or without
|
||||
# modification, are permitted provided that the following conditions
|
||||
# are met:
|
||||
# * Redistributions of source code must retain the above copyright
|
||||
# notice, this list of conditions and the following disclaimer.
|
||||
# * Redistributions in binary form must reproduce the above copyright
|
||||
# notice, this list of conditions and the following disclaimer in the
|
||||
# documentation and/or other materials provided with the distribution.
|
||||
# * Neither the name of NVIDIA CORPORATION nor the names of its
|
||||
# contributors may be used to endorse or promote products derived
|
||||
# from this software without specific prior written permission.
|
||||
#
|
||||
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
|
||||
# EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
||||
# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
||||
# PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
|
||||
# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
|
||||
# EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
|
||||
# PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
|
||||
# PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
|
||||
# OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
||||
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
||||
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
||||
#
|
||||
################################################################################
|
||||
#
|
||||
# Makefile project only supported on Mac OS X and Linux Platforms)
|
||||
#
|
||||
################################################################################
|
||||
|
||||
# Location of the CUDA Toolkit
|
||||
CUDA_PATH ?= /usr/local/cuda
|
||||
|
||||
##############################
|
||||
# start deprecated interface #
|
||||
##############################
|
||||
ifeq ($(x86_64),1)
|
||||
$(info WARNING - x86_64 variable has been deprecated)
|
||||
$(info WARNING - please use TARGET_ARCH=x86_64 instead)
|
||||
TARGET_ARCH ?= x86_64
|
||||
endif
|
||||
ifeq ($(ARMv7),1)
|
||||
$(info WARNING - ARMv7 variable has been deprecated)
|
||||
$(info WARNING - please use TARGET_ARCH=armv7l instead)
|
||||
TARGET_ARCH ?= armv7l
|
||||
endif
|
||||
ifeq ($(aarch64),1)
|
||||
$(info WARNING - aarch64 variable has been deprecated)
|
||||
$(info WARNING - please use TARGET_ARCH=aarch64 instead)
|
||||
TARGET_ARCH ?= aarch64
|
||||
endif
|
||||
ifeq ($(ppc64le),1)
|
||||
$(info WARNING - ppc64le variable has been deprecated)
|
||||
$(info WARNING - please use TARGET_ARCH=ppc64le instead)
|
||||
TARGET_ARCH ?= ppc64le
|
||||
endif
|
||||
ifneq ($(GCC),)
|
||||
$(info WARNING - GCC variable has been deprecated)
|
||||
$(info WARNING - please use HOST_COMPILER=$(GCC) instead)
|
||||
HOST_COMPILER ?= $(GCC)
|
||||
endif
|
||||
ifneq ($(abi),)
|
||||
$(error ERROR - abi variable has been removed)
|
||||
endif
|
||||
############################
|
||||
# end deprecated interface #
|
||||
############################
|
||||
|
||||
# architecture
|
||||
HOST_ARCH := $(shell uname -m)
|
||||
TARGET_ARCH ?= $(HOST_ARCH)
|
||||
ifneq (,$(filter $(TARGET_ARCH),x86_64 aarch64 sbsa ppc64le armv7l))
|
||||
ifneq ($(TARGET_ARCH),$(HOST_ARCH))
|
||||
ifneq (,$(filter $(TARGET_ARCH),x86_64 aarch64 sbsa ppc64le))
|
||||
TARGET_SIZE := 64
|
||||
else ifneq (,$(filter $(TARGET_ARCH),armv7l))
|
||||
TARGET_SIZE := 32
|
||||
endif
|
||||
else
|
||||
TARGET_SIZE := $(shell getconf LONG_BIT)
|
||||
endif
|
||||
else
|
||||
$(error ERROR - unsupported value $(TARGET_ARCH) for TARGET_ARCH!)
|
||||
endif
|
||||
|
||||
# sbsa and aarch64 systems look similar. Need to differentiate them at host level for now.
|
||||
ifeq ($(HOST_ARCH),aarch64)
|
||||
ifeq ($(CUDA_PATH)/targets/sbsa-linux,$(shell ls -1d $(CUDA_PATH)/targets/sbsa-linux 2>/dev/null))
|
||||
HOST_ARCH := sbsa
|
||||
TARGET_ARCH := sbsa
|
||||
endif
|
||||
endif
|
||||
|
||||
ifneq ($(TARGET_ARCH),$(HOST_ARCH))
|
||||
ifeq (,$(filter $(HOST_ARCH)-$(TARGET_ARCH),aarch64-armv7l x86_64-armv7l x86_64-aarch64 x86_64-sbsa x86_64-ppc64le))
|
||||
$(error ERROR - cross compiling from $(HOST_ARCH) to $(TARGET_ARCH) is not supported!)
|
||||
endif
|
||||
endif
|
||||
|
||||
# When on native aarch64 system with userspace of 32-bit, change TARGET_ARCH to armv7l
|
||||
ifeq ($(HOST_ARCH)-$(TARGET_ARCH)-$(TARGET_SIZE),aarch64-aarch64-32)
|
||||
TARGET_ARCH = armv7l
|
||||
endif
|
||||
|
||||
# operating system
|
||||
HOST_OS := $(shell uname -s 2>/dev/null | tr "[:upper:]" "[:lower:]")
|
||||
TARGET_OS ?= $(HOST_OS)
|
||||
ifeq (,$(filter $(TARGET_OS),linux darwin qnx android))
|
||||
$(error ERROR - unsupported value $(TARGET_OS) for TARGET_OS!)
|
||||
endif
|
||||
|
||||
# host compiler
|
||||
ifeq ($(TARGET_OS),darwin)
|
||||
ifeq ($(shell expr `xcodebuild -version | grep -i xcode | awk '{print $$2}' | cut -d'.' -f1` \>= 5),1)
|
||||
HOST_COMPILER ?= clang++
|
||||
endif
|
||||
else ifneq ($(TARGET_ARCH),$(HOST_ARCH))
|
||||
ifeq ($(HOST_ARCH)-$(TARGET_ARCH),x86_64-armv7l)
|
||||
ifeq ($(TARGET_OS),linux)
|
||||
HOST_COMPILER ?= arm-linux-gnueabihf-g++
|
||||
else ifeq ($(TARGET_OS),qnx)
|
||||
ifeq ($(QNX_HOST),)
|
||||
$(error ERROR - QNX_HOST must be passed to the QNX host toolchain)
|
||||
endif
|
||||
ifeq ($(QNX_TARGET),)
|
||||
$(error ERROR - QNX_TARGET must be passed to the QNX target toolchain)
|
||||
endif
|
||||
export QNX_HOST
|
||||
export QNX_TARGET
|
||||
HOST_COMPILER ?= $(QNX_HOST)/usr/bin/arm-unknown-nto-qnx6.6.0eabi-g++
|
||||
else ifeq ($(TARGET_OS),android)
|
||||
HOST_COMPILER ?= arm-linux-androideabi-g++
|
||||
endif
|
||||
else ifeq ($(TARGET_ARCH),aarch64)
|
||||
ifeq ($(TARGET_OS), linux)
|
||||
HOST_COMPILER ?= aarch64-linux-gnu-g++
|
||||
else ifeq ($(TARGET_OS),qnx)
|
||||
ifeq ($(QNX_HOST),)
|
||||
$(error ERROR - QNX_HOST must be passed to the QNX host toolchain)
|
||||
endif
|
||||
ifeq ($(QNX_TARGET),)
|
||||
$(error ERROR - QNX_TARGET must be passed to the QNX target toolchain)
|
||||
endif
|
||||
export QNX_HOST
|
||||
export QNX_TARGET
|
||||
HOST_COMPILER ?= $(QNX_HOST)/usr/bin/q++
|
||||
else ifeq ($(TARGET_OS), android)
|
||||
HOST_COMPILER ?= aarch64-linux-android-clang++
|
||||
endif
|
||||
else ifeq ($(TARGET_ARCH),sbsa)
|
||||
HOST_COMPILER ?= aarch64-linux-gnu-g++
|
||||
else ifeq ($(TARGET_ARCH),ppc64le)
|
||||
HOST_COMPILER ?= powerpc64le-linux-gnu-g++
|
||||
endif
|
||||
endif
|
||||
HOST_COMPILER ?= g++
|
||||
NVCC := $(CUDA_PATH)/bin/nvcc -ccbin $(HOST_COMPILER)
|
||||
|
||||
# internal flags
|
||||
NVCCFLAGS := -m${TARGET_SIZE}
|
||||
CCFLAGS :=
|
||||
LDFLAGS :=
|
||||
|
||||
# build flags
|
||||
ifeq ($(TARGET_OS),darwin)
|
||||
LDFLAGS += -rpath $(CUDA_PATH)/lib
|
||||
CCFLAGS += -arch $(HOST_ARCH)
|
||||
else ifeq ($(HOST_ARCH)-$(TARGET_ARCH)-$(TARGET_OS),x86_64-armv7l-linux)
|
||||
LDFLAGS += --dynamic-linker=/lib/ld-linux-armhf.so.3
|
||||
CCFLAGS += -mfloat-abi=hard
|
||||
else ifeq ($(TARGET_OS),android)
|
||||
LDFLAGS += -pie
|
||||
CCFLAGS += -fpie -fpic -fexceptions
|
||||
endif
|
||||
|
||||
ifneq ($(TARGET_ARCH),$(HOST_ARCH))
|
||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-linux)
|
||||
ifneq ($(TARGET_FS),)
|
||||
GCCVERSIONLTEQ46 := $(shell expr `$(HOST_COMPILER) -dumpversion` \<= 4.6)
|
||||
ifeq ($(GCCVERSIONLTEQ46),1)
|
||||
CCFLAGS += --sysroot=$(TARGET_FS)
|
||||
endif
|
||||
LDFLAGS += --sysroot=$(TARGET_FS)
|
||||
LDFLAGS += -rpath-link=$(TARGET_FS)/lib
|
||||
LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib
|
||||
LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib/arm-linux-gnueabihf
|
||||
endif
|
||||
endif
|
||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-linux)
|
||||
ifneq ($(TARGET_FS),)
|
||||
GCCVERSIONLTEQ46 := $(shell expr `$(HOST_COMPILER) -dumpversion` \<= 4.6)
|
||||
ifeq ($(GCCVERSIONLTEQ46),1)
|
||||
CCFLAGS += --sysroot=$(TARGET_FS)
|
||||
endif
|
||||
LDFLAGS += --sysroot=$(TARGET_FS)
|
||||
LDFLAGS += -rpath-link=$(TARGET_FS)/lib -L$(TARGET_FS)/lib
|
||||
LDFLAGS += -rpath-link=$(TARGET_FS)/lib/aarch64-linux-gnu -L$(TARGET_FS)/lib/aarch64-linux-gnu
|
||||
LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib -L$(TARGET_FS)/usr/lib
|
||||
LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib/aarch64-linux-gnu -L$(TARGET_FS)/usr/lib/aarch64-linux-gnu
|
||||
LDFLAGS += --unresolved-symbols=ignore-in-shared-libs
|
||||
CCFLAGS += -isystem=$(TARGET_FS)/usr/include -I$(TARGET_FS)/usr/include -I$(TARGET_FS)/usr/include/libdrm
|
||||
CCFLAGS += -isystem=$(TARGET_FS)/usr/include/aarch64-linux-gnu -I$(TARGET_FS)/usr/include/aarch64-linux-gnu
|
||||
endif
|
||||
endif
|
||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-qnx)
|
||||
NVCCFLAGS += --qpp-config 5.4.0,gcc_ntoaarch64le
|
||||
CCFLAGS += -DWIN_INTERFACE_CUSTOM -I/usr/include/aarch64-qnx-gnu
|
||||
LDFLAGS += -lsocket
|
||||
LDFLAGS += -L/usr/lib/aarch64-qnx-gnu
|
||||
CCFLAGS += "-Wl\,-rpath-link\,/usr/lib/aarch64-qnx-gnu"
|
||||
ifdef TARGET_OVERRIDE
|
||||
LDFLAGS += -lslog2
|
||||
endif
|
||||
|
||||
ifneq ($(TARGET_FS),)
|
||||
LDFLAGS += -L$(TARGET_FS)/usr/lib
|
||||
CCFLAGS += "-Wl\,-rpath-link\,$(TARGET_FS)/usr/lib"
|
||||
LDFLAGS += -L$(TARGET_FS)/usr/libnvidia
|
||||
CCFLAGS += "-Wl\,-rpath-link\,$(TARGET_FS)/usr/libnvidia"
|
||||
CCFLAGS += -I$(TARGET_FS)/../include
|
||||
endif
|
||||
endif
|
||||
endif
|
||||
|
||||
ifdef TARGET_OVERRIDE # cuda toolkit targets override
|
||||
NVCCFLAGS += -target-dir $(TARGET_OVERRIDE)
|
||||
endif
|
||||
|
||||
# Install directory of different arch
|
||||
CUDA_INSTALL_TARGET_DIR :=
|
||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-linux)
|
||||
CUDA_INSTALL_TARGET_DIR = targets/armv7-linux-gnueabihf/
|
||||
else ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-linux)
|
||||
CUDA_INSTALL_TARGET_DIR = targets/aarch64-linux/
|
||||
else ifeq ($(TARGET_ARCH)-$(TARGET_OS),sbsa-linux)
|
||||
CUDA_INSTALL_TARGET_DIR = targets/sbsa-linux/
|
||||
else ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-android)
|
||||
CUDA_INSTALL_TARGET_DIR = targets/armv7-linux-androideabi/
|
||||
else ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-android)
|
||||
CUDA_INSTALL_TARGET_DIR = targets/aarch64-linux-androideabi/
|
||||
else ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-qnx)
|
||||
CUDA_INSTALL_TARGET_DIR = targets/ARMv7-linux-QNX/
|
||||
else ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-qnx)
|
||||
CUDA_INSTALL_TARGET_DIR = targets/aarch64-qnx/
|
||||
else ifeq ($(TARGET_ARCH),ppc64le)
|
||||
CUDA_INSTALL_TARGET_DIR = targets/ppc64le-linux/
|
||||
endif
|
||||
|
||||
# Debug build flags
|
||||
ifeq ($(dbg),1)
|
||||
NVCCFLAGS += -g -G
|
||||
BUILD_TYPE := debug
|
||||
else
|
||||
BUILD_TYPE := release
|
||||
endif
|
||||
|
||||
ALL_CCFLAGS :=
|
||||
ALL_CCFLAGS += $(NVCCFLAGS)
|
||||
ALL_CCFLAGS += $(EXTRA_NVCCFLAGS)
|
||||
ALL_CCFLAGS += $(addprefix -Xcompiler ,$(CCFLAGS))
|
||||
ALL_CCFLAGS += $(addprefix -Xcompiler ,$(EXTRA_CCFLAGS))
|
||||
|
||||
SAMPLE_ENABLED := 1
|
||||
|
||||
# This sample is not supported on QNX
|
||||
ifeq ($(TARGET_OS),qnx)
|
||||
$(info >>> WARNING - FunctionPointers is not supported on QNX - waiving sample <<<)
|
||||
SAMPLE_ENABLED := 0
|
||||
endif
|
||||
|
||||
ALL_LDFLAGS :=
|
||||
ALL_LDFLAGS += $(ALL_CCFLAGS)
|
||||
ALL_LDFLAGS += $(addprefix -Xlinker ,$(LDFLAGS))
|
||||
ALL_LDFLAGS += $(addprefix -Xlinker ,$(EXTRA_LDFLAGS))
|
||||
|
||||
# Common includes and paths for CUDA
|
||||
INCLUDES := -I../../Common
|
||||
LIBRARIES :=
|
||||
|
||||
################################################################################
|
||||
|
||||
# Makefile include to help find GL Libraries
|
||||
include ./findgllib.mk
|
||||
|
||||
# OpenGL specific libraries
|
||||
ifeq ($(TARGET_OS),darwin)
|
||||
# Mac OSX specific libraries and paths to include
|
||||
LIBRARIES += -L/System/Library/Frameworks/OpenGL.framework/Libraries
|
||||
LIBRARIES += -lGL -lGLU
|
||||
ALL_LDFLAGS += -Xlinker -framework -Xlinker GLUT
|
||||
else
|
||||
LIBRARIES += $(GLLINK)
|
||||
LIBRARIES += -lGL -lGLU -lglut
|
||||
endif
|
||||
|
||||
#Detect if installed version of GCC supports required C++11
|
||||
ifeq ($(TARGET_OS),linux)
|
||||
empty :=
|
||||
space := $(empty) $(empty)
|
||||
GCCVERSIONSTRING := $(shell expr `$(HOST_COMPILER) -dumpversion`)
|
||||
#Create version number without "."
|
||||
GCCVERSION := $(shell expr `echo $(GCCVERSIONSTRING)` | cut -f1 -d.)
|
||||
GCCVERSION += $(shell expr `echo $(GCCVERSIONSTRING)` | cut -f2 -d.)
|
||||
GCCVERSION += $(shell expr `echo $(GCCVERSIONSTRING)` | cut -f3 -d.)
|
||||
# Make sure the version number has at least 3 decimals
|
||||
GCCVERSION += 00
|
||||
# Remove spaces from the version number
|
||||
GCCVERSION := $(subst $(space),$(empty),$(GCCVERSION))
|
||||
#$(warning $(GCCVERSION))
|
||||
|
||||
IS_MIN_VERSION := $(shell expr `echo $(GCCVERSION)` \>= 51000)
|
||||
|
||||
ifeq ($(IS_MIN_VERSION), 1)
|
||||
$(info >>> GCC Version is greater or equal to 5.1.0 <<<)
|
||||
else
|
||||
$(info >>> Waiving build. Minimum GCC version required is 5.1.0<<<)
|
||||
SAMPLE_ENABLED := 0
|
||||
endif
|
||||
endif
|
||||
|
||||
# Gencode arguments
|
||||
ifeq ($(TARGET_ARCH),$(filter $(TARGET_ARCH),armv7l aarch64))
|
||||
SMS ?= 35 37 50 52 60 61 70 72 75 80 86
|
||||
else
|
||||
SMS ?= 35 37 50 52 60 61 70 75 80 86
|
||||
endif
|
||||
|
||||
ifeq ($(SMS),)
|
||||
$(info >>> WARNING - no SM architectures have been specified - waiving sample <<<)
|
||||
SAMPLE_ENABLED := 0
|
||||
endif
|
||||
|
||||
ifeq ($(GENCODE_FLAGS),)
|
||||
# Generate SASS code for each SM architecture listed in $(SMS)
|
||||
$(foreach sm,$(SMS),$(eval GENCODE_FLAGS += -gencode arch=compute_$(sm),code=sm_$(sm)))
|
||||
|
||||
# Generate PTX code from the highest SM architecture in $(SMS) to guarantee forward-compatibility
|
||||
HIGHEST_SM := $(lastword $(sort $(SMS)))
|
||||
ifneq ($(HIGHEST_SM),)
|
||||
GENCODE_FLAGS += -gencode arch=compute_$(HIGHEST_SM),code=compute_$(HIGHEST_SM)
|
||||
endif
|
||||
endif
|
||||
|
||||
ALL_CCFLAGS += --threads 0 --std=c++11
|
||||
|
||||
ifeq ($(SAMPLE_ENABLED),0)
|
||||
EXEC ?= @echo "[@]"
|
||||
endif
|
||||
|
||||
################################################################################
|
||||
|
||||
# Target rules
|
||||
all: build
|
||||
|
||||
build: FunctionPointers
|
||||
|
||||
check.deps:
|
||||
ifeq ($(SAMPLE_ENABLED),0)
|
||||
@echo "Sample will be waived due to the above missing dependencies"
|
||||
else
|
||||
@echo "Sample is ready - all dependencies have been met"
|
||||
endif
|
||||
|
||||
FunctionPointers.o:FunctionPointers.cpp
|
||||
$(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -c $<
|
||||
|
||||
FunctionPointers_kernels.o:FunctionPointers_kernels.cu
|
||||
$(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -c $<
|
||||
|
||||
FunctionPointers: FunctionPointers.o FunctionPointers_kernels.o
|
||||
$(EXEC) $(NVCC) $(ALL_LDFLAGS) $(GENCODE_FLAGS) -o $@ $+ $(LIBRARIES)
|
||||
$(EXEC) mkdir -p ../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)
|
||||
$(EXEC) cp $@ ../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)
|
||||
|
||||
run: build
|
||||
$(EXEC) ./FunctionPointers
|
||||
|
||||
clean:
|
||||
rm -f FunctionPointers FunctionPointers.o FunctionPointers_kernels.o
|
||||
rm -rf ../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)/FunctionPointers
|
||||
|
||||
clobber: clean
|
||||
93
Samples/FunctionPointers/NsightEclipse.xml
Normal file
93
Samples/FunctionPointers/NsightEclipse.xml
Normal file
@@ -0,0 +1,93 @@
|
||||
<?xml version="1.0" encoding="UTF-8"?>
|
||||
<!DOCTYPE entry SYSTEM "SamplesInfo.dtd">
|
||||
<entry>
|
||||
<name>FunctionPointers</name>
|
||||
<description><![CDATA[This sample illustrates how to use function pointers and implements the Sobel Edge Detection filter for 8-bit monochrome images.]]></description>
|
||||
<devicecompilation>whole</devicecompilation>
|
||||
<files>
|
||||
<file>data\lena.pgm</file>
|
||||
<file>data\ref_orig.pgm</file>
|
||||
<file>data\ref_tex.pgm</file>
|
||||
<file>data\ref_shared.pgm</file>
|
||||
</files>
|
||||
<includepaths>
|
||||
<path>./</path>
|
||||
<path>../</path>
|
||||
<path>../../Common</path>
|
||||
</includepaths>
|
||||
<keyconcepts>
|
||||
<concept level="advanced">Graphics Interop</concept>
|
||||
<concept level="advanced">Image Processing</concept>
|
||||
</keyconcepts>
|
||||
<keywords>
|
||||
<keyword>CUDA</keyword>
|
||||
<keyword>GPGPU</keyword>
|
||||
<keyword>convolution</keyword>
|
||||
<keyword>separable convolution</keyword>
|
||||
<keyword>Image Processing</keyword>
|
||||
<keyword>filter</keyword>
|
||||
<keyword>finite impulse-response filter</keyword>
|
||||
<keyword>openGL</keyword>
|
||||
</keywords>
|
||||
<libraries>
|
||||
<library>GLU</library>
|
||||
<library>GL</library>
|
||||
<library framework="true" os="macosx">GLUT</library>
|
||||
<library os="linux">GLEW</library>
|
||||
<library os="linux">glut</library>
|
||||
<library os="linux">X11</library>
|
||||
</libraries>
|
||||
<librarypaths>
|
||||
<path arch="x86_64" os="linux">../../common/lib/linux/x86_64</path>
|
||||
<path arch="armv7l" os="linux">../../common/lib/linux/armv7l</path>
|
||||
<path os="macosx">../../common/lib/darwin</path>
|
||||
</librarypaths>
|
||||
<nsight_eclipse>true</nsight_eclipse>
|
||||
<primary_file>FunctionPointers.cpp</primary_file>
|
||||
<qatests>
|
||||
<qatest>-mode=0 -file=ref_orig.pgm</qatest>
|
||||
<qatest>-mode=1 -file=ref_tex.pgm</qatest>
|
||||
<qatest>-mode=2 -file=ref_shared.pgm</qatest>
|
||||
</qatests>
|
||||
<required_dependencies>
|
||||
<dependency>X11</dependency>
|
||||
<dependency>GL</dependency>
|
||||
</required_dependencies>
|
||||
<scopes>
|
||||
<scope>1:CUDA Advanced Topics</scope>
|
||||
<scope>2:Graphics Interop</scope>
|
||||
<scope>2:Image Processing</scope>
|
||||
</scopes>
|
||||
<sm-arch>sm35</sm-arch>
|
||||
<sm-arch>sm37</sm-arch>
|
||||
<sm-arch>sm50</sm-arch>
|
||||
<sm-arch>sm52</sm-arch>
|
||||
<sm-arch>sm60</sm-arch>
|
||||
<sm-arch>sm61</sm-arch>
|
||||
<sm-arch>sm70</sm-arch>
|
||||
<sm-arch>sm72</sm-arch>
|
||||
<sm-arch>sm75</sm-arch>
|
||||
<sm-arch>sm80</sm-arch>
|
||||
<sm-arch>sm86</sm-arch>
|
||||
<supported_envs>
|
||||
<env>
|
||||
<arch>x86_64</arch>
|
||||
<platform>linux</platform>
|
||||
</env>
|
||||
<env>
|
||||
<platform>windows7</platform>
|
||||
</env>
|
||||
<env>
|
||||
<arch>x86_64</arch>
|
||||
<platform>macosx</platform>
|
||||
</env>
|
||||
<env>
|
||||
<arch>arm</arch>
|
||||
</env>
|
||||
</supported_envs>
|
||||
<supported_sm_architectures>
|
||||
<include>all</include>
|
||||
</supported_sm_architectures>
|
||||
<title>Function Pointers</title>
|
||||
<type>exe</type>
|
||||
</entry>
|
||||
71
Samples/FunctionPointers/README.md
Normal file
71
Samples/FunctionPointers/README.md
Normal file
@@ -0,0 +1,71 @@
|
||||
# FunctionPointers - Function Pointers
|
||||
|
||||
## Description
|
||||
|
||||
This sample illustrates how to use function pointers and implements the Sobel Edge Detection filter for 8-bit monochrome images.
|
||||
|
||||
## Key Concepts
|
||||
|
||||
Graphics Interop, Image Processing
|
||||
|
||||
## Supported SM Architectures
|
||||
|
||||
[SM 3.5 ](https://developer.nvidia.com/cuda-gpus) [SM 3.7 ](https://developer.nvidia.com/cuda-gpus) [SM 5.0 ](https://developer.nvidia.com/cuda-gpus) [SM 5.2 ](https://developer.nvidia.com/cuda-gpus) [SM 6.0 ](https://developer.nvidia.com/cuda-gpus) [SM 6.1 ](https://developer.nvidia.com/cuda-gpus) [SM 7.0 ](https://developer.nvidia.com/cuda-gpus) [SM 7.2 ](https://developer.nvidia.com/cuda-gpus) [SM 7.5 ](https://developer.nvidia.com/cuda-gpus) [SM 8.0 ](https://developer.nvidia.com/cuda-gpus) [SM 8.6 ](https://developer.nvidia.com/cuda-gpus)
|
||||
|
||||
## Supported OSes
|
||||
|
||||
Linux, Windows
|
||||
|
||||
## Supported CPU Architecture
|
||||
|
||||
x86_64, armv7l
|
||||
|
||||
## CUDA APIs involved
|
||||
|
||||
## Dependencies needed to build/run
|
||||
[X11](../../README.md#x11), [GL](../../README.md#gl)
|
||||
|
||||
## Prerequisites
|
||||
|
||||
Download and install the [CUDA Toolkit 11.5](https://developer.nvidia.com/cuda-downloads) for your corresponding platform.
|
||||
Make sure the dependencies mentioned in [Dependencies]() section above are installed.
|
||||
|
||||
## Build and Run
|
||||
|
||||
### Windows
|
||||
The Windows samples are built using the Visual Studio IDE. Solution files (.sln) are provided for each supported version of Visual Studio, using the format:
|
||||
```
|
||||
*_vs<version>.sln - for Visual Studio <version>
|
||||
```
|
||||
Each individual sample has its own set of solution files in its directory:
|
||||
|
||||
To build/examine all the samples at once, the complete solution files should be used. To build/examine a single sample, the individual sample solution files should be used.
|
||||
> **Note:** Some samples require that the Microsoft DirectX SDK (June 2010 or newer) be installed and that the VC++ directory paths are properly set up (**Tools > Options...**). Check DirectX Dependencies section for details."
|
||||
|
||||
### Linux
|
||||
The Linux samples are built using makefiles. To use the makefiles, change the current directory to the sample directory you wish to build, and run make:
|
||||
```
|
||||
$ cd <sample_dir>
|
||||
$ make
|
||||
```
|
||||
The samples makefiles can take advantage of certain options:
|
||||
* **TARGET_ARCH=<arch>** - cross-compile targeting a specific architecture. Allowed architectures are x86_64, armv7l.
|
||||
By default, TARGET_ARCH is set to HOST_ARCH. On a x86_64 machine, not setting TARGET_ARCH is the equivalent of setting TARGET_ARCH=x86_64.<br/>
|
||||
`$ make TARGET_ARCH=x86_64` <br/> `$ make TARGET_ARCH=armv7l` <br/>
|
||||
See [here](http://docs.nvidia.com/cuda/cuda-samples/index.html#cross-samples) for more details.
|
||||
* **dbg=1** - build with debug symbols
|
||||
```
|
||||
$ make dbg=1
|
||||
```
|
||||
* **SMS="A B ..."** - override the SM architectures for which the sample will be built, where `"A B ..."` is a space-delimited list of SM architectures. For example, to generate SASS for SM 50 and SM 60, use `SMS="50 60"`.
|
||||
```
|
||||
$ make SMS="50 60"
|
||||
```
|
||||
|
||||
* **HOST_COMPILER=<host_compiler>** - override the default g++ host compiler. See the [Linux Installation Guide](http://docs.nvidia.com/cuda/cuda-installation-guide-linux/index.html#system-requirements) for a list of supported host compilers.
|
||||
```
|
||||
$ make HOST_COMPILER=g++
|
||||
```
|
||||
|
||||
## References (for more details)
|
||||
|
||||
5
Samples/FunctionPointers/data/lena.pgm
Normal file
5
Samples/FunctionPointers/data/lena.pgm
Normal file
File diff suppressed because one or more lines are too long
5
Samples/FunctionPointers/data/ref_orig.pgm
Normal file
5
Samples/FunctionPointers/data/ref_orig.pgm
Normal file
File diff suppressed because one or more lines are too long
BIN
Samples/FunctionPointers/data/ref_shared.pgm
Normal file
BIN
Samples/FunctionPointers/data/ref_shared.pgm
Normal file
Binary file not shown.
BIN
Samples/FunctionPointers/data/ref_tex.pgm
Normal file
BIN
Samples/FunctionPointers/data/ref_tex.pgm
Normal file
Binary file not shown.
145
Samples/FunctionPointers/findgllib.mk
Normal file
145
Samples/FunctionPointers/findgllib.mk
Normal file
@@ -0,0 +1,145 @@
|
||||
################################################################################
|
||||
# Copyright (c) 2020, NVIDIA CORPORATION. All rights reserved.
|
||||
#
|
||||
# Redistribution and use in source and binary forms, with or without
|
||||
# modification, are permitted provided that the following conditions
|
||||
# are met:
|
||||
# * Redistributions of source code must retain the above copyright
|
||||
# notice, this list of conditions and the following disclaimer.
|
||||
# * Redistributions in binary form must reproduce the above copyright
|
||||
# notice, this list of conditions and the following disclaimer in the
|
||||
# documentation and/or other materials provided with the distribution.
|
||||
# * Neither the name of NVIDIA CORPORATION nor the names of its
|
||||
# contributors may be used to endorse or promote products derived
|
||||
# from this software without specific prior written permission.
|
||||
#
|
||||
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
|
||||
# EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
||||
# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
||||
# PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
|
||||
# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
|
||||
# EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
|
||||
# PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
|
||||
# PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
|
||||
# OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
||||
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
||||
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
||||
#
|
||||
################################################################################
|
||||
#
|
||||
# findgllib.mk is used to find the necessary GL Libraries for specific distributions
|
||||
# this is supported on Mac OSX and Linux Platforms
|
||||
#
|
||||
################################################################################
|
||||
|
||||
# Determine OS platform and unix distribution
|
||||
ifeq ("$(TARGET_OS)","linux")
|
||||
# first search lsb_release
|
||||
DISTRO := $(shell lsb_release -i -s 2>/dev/null | tr "[:upper:]" "[:lower:]")
|
||||
ifeq ("$(DISTRO)","")
|
||||
# second search and parse /etc/issue
|
||||
DISTRO := $(shell awk '{print $$1}' /etc/issue | tr -d "[:space:]" | sed -e "/^$$/d" | tr "[:upper:]" "[:lower:]")
|
||||
# ensure data from /etc/issue is valid
|
||||
ifeq (,$(filter $(DISTRO),ubuntu fedora red rhel centos suse))
|
||||
DISTRO :=
|
||||
endif
|
||||
ifeq ("$(DISTRO)","")
|
||||
# third, we can search in /etc/os-release or /etc/{distro}-release
|
||||
DISTRO := $(shell awk '/ID/' /etc/*-release | sed 's/ID=//' | grep -v "VERSION" | grep -v "ID" | grep -v "DISTRIB")
|
||||
endif
|
||||
endif
|
||||
endif
|
||||
|
||||
ifeq ("$(TARGET_OS)","linux")
|
||||
# $(info) >> findgllib.mk -> LINUX path <<<)
|
||||
# Each set of Linux Distros have different paths for where to find their OpenGL libraries reside
|
||||
UBUNTU = $(shell echo $(DISTRO) | grep -i ubuntu >/dev/null 2>&1; echo $$?)
|
||||
FEDORA = $(shell echo $(DISTRO) | grep -i fedora >/dev/null 2>&1; echo $$?)
|
||||
RHEL = $(shell echo $(DISTRO) | grep -i 'red\|rhel' >/dev/null 2>&1; echo $$?)
|
||||
CENTOS = $(shell echo $(DISTRO) | grep -i centos >/dev/null 2>&1; echo $$?)
|
||||
SUSE = $(shell echo $(DISTRO) | grep -i 'suse\|sles' >/dev/null 2>&1; echo $$?)
|
||||
ifeq ("$(UBUNTU)","0")
|
||||
ifeq ($(HOST_ARCH)-$(TARGET_ARCH),x86_64-armv7l)
|
||||
GLPATH := /usr/arm-linux-gnueabihf/lib
|
||||
GLLINK := -L/usr/arm-linux-gnueabihf/lib
|
||||
ifneq ($(TARGET_FS),)
|
||||
GLPATH += $(TARGET_FS)/usr/lib/arm-linux-gnueabihf
|
||||
GLLINK += -L$(TARGET_FS)/usr/lib/arm-linux-gnueabihf
|
||||
endif
|
||||
else ifeq ($(HOST_ARCH)-$(TARGET_ARCH),x86_64-aarch64)
|
||||
GLPATH := /usr/aarch64-linux-gnu/lib
|
||||
GLLINK := -L/usr/aarch64-linux-gnu/lib
|
||||
ifneq ($(TARGET_FS),)
|
||||
GLPATH += $(TARGET_FS)/usr/lib
|
||||
GLPATH += $(TARGET_FS)/usr/lib/aarch64-linux-gnu
|
||||
GLLINK += -L$(TARGET_FS)/usr/lib/aarch64-linux-gnu
|
||||
endif
|
||||
else ifeq ($(HOST_ARCH)-$(TARGET_ARCH),x86_64-ppc64le)
|
||||
GLPATH := /usr/powerpc64le-linux-gnu/lib
|
||||
GLLINK := -L/usr/powerpc64le-linux-gnu/lib
|
||||
else
|
||||
UBUNTU_PKG_NAME = $(shell which dpkg >/dev/null 2>&1 && dpkg -l 'nvidia-*' | grep '^ii' | awk '{print $$2}' | head -1)
|
||||
ifneq ("$(UBUNTU_PKG_NAME)","")
|
||||
GLPATH ?= /usr/lib/$(UBUNTU_PKG_NAME)
|
||||
GLLINK ?= -L/usr/lib/$(UBUNTU_PKG_NAME)
|
||||
endif
|
||||
|
||||
DFLT_PATH ?= /usr/lib
|
||||
endif
|
||||
endif
|
||||
ifeq ("$(SUSE)","0")
|
||||
GLPATH ?= /usr/X11R6/lib64
|
||||
GLLINK ?= -L/usr/X11R6/lib64
|
||||
DFLT_PATH ?= /usr/lib64
|
||||
endif
|
||||
ifeq ("$(FEDORA)","0")
|
||||
GLPATH ?= /usr/lib64/nvidia
|
||||
GLLINK ?= -L/usr/lib64/nvidia
|
||||
DFLT_PATH ?= /usr/lib64
|
||||
endif
|
||||
ifeq ("$(RHEL)","0")
|
||||
GLPATH ?= /usr/lib64/nvidia
|
||||
GLLINK ?= -L/usr/lib64/nvidia
|
||||
DFLT_PATH ?= /usr/lib64
|
||||
endif
|
||||
ifeq ("$(CENTOS)","0")
|
||||
GLPATH ?= /usr/lib64/nvidia
|
||||
GLLINK ?= -L/usr/lib64/nvidia
|
||||
DFLT_PATH ?= /usr/lib64
|
||||
endif
|
||||
|
||||
# find libGL, libGLU
|
||||
GLLIB := $(shell find -L $(GLPATH) $(DFLT_PATH) -name libGL.so -print 2>/dev/null)
|
||||
GLULIB := $(shell find -L $(GLPATH) $(DFLT_PATH) -name libGLU.so -print 2>/dev/null)
|
||||
|
||||
ifeq ("$(GLLIB)","")
|
||||
$(info >>> WARNING - libGL.so not found, refer to CUDA Getting Started Guide for how to find and install them. <<<)
|
||||
SAMPLE_ENABLED := 0
|
||||
endif
|
||||
ifeq ("$(GLULIB)","")
|
||||
$(info >>> WARNING - libGLU.so not found, refer to CUDA Getting Started Guide for how to find and install them. <<<)
|
||||
SAMPLE_ENABLED := 0
|
||||
endif
|
||||
|
||||
HEADER_SEARCH_PATH ?= $(TARGET_FS)/usr/include
|
||||
ifeq ($(HOST_ARCH)-$(TARGET_ARCH)-$(TARGET_OS),x86_64-armv7l-linux)
|
||||
HEADER_SEARCH_PATH += /usr/arm-linux-gnueabihf/include
|
||||
else ifeq ($(HOST_ARCH)-$(TARGET_ARCH)-$(TARGET_OS),x86_64-aarch64-linux)
|
||||
HEADER_SEARCH_PATH += /usr/aarch64-linux-gnu/include
|
||||
endif
|
||||
|
||||
GLHEADER := $(shell find -L $(HEADER_SEARCH_PATH) -name gl.h -print 2>/dev/null)
|
||||
GLUHEADER := $(shell find -L $(HEADER_SEARCH_PATH) -name glu.h -print 2>/dev/null)
|
||||
|
||||
ifeq ("$(GLHEADER)","")
|
||||
$(info >>> WARNING - gl.h not found, refer to CUDA Getting Started Guide for how to find and install them. <<<)
|
||||
SAMPLE_ENABLED := 0
|
||||
endif
|
||||
ifeq ("$(GLUHEADER)","")
|
||||
$(info >>> WARNING - glu.h not found, refer to CUDA Getting Started Guide for how to find and install them. <<<)
|
||||
SAMPLE_ENABLED := 0
|
||||
endif
|
||||
else
|
||||
# This would be the Mac OS X path if we had to do anything special
|
||||
endif
|
||||
|
||||
Reference in New Issue
Block a user