NASP/cuda-samples
47
0
Fork 0
mirror of https://github.com/NVIDIA/cuda-samples.git synced 2025-07-01 20:20:29 +08:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #368 from XSShawnZeng/master

Browse Source 
Update the vulkan headers include sequence and the transpose code format check
This commit is contained in:
Rob Armstrong 2025-05-21 09:27:13 -07:00 committed by GitHub
commit adacf1cffd
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
4 changed files with 14 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
Samples/5_Domain_Specific/simpleVulkan/VulkanBaseApp.h
View File

@ -34,8 +34,10 @@
#include <vulkan/vulkan.h>
#ifdef _WIN64
#define NOMINMAX
#include <vulkan/vulkan_win32.h>
// Add windows.h to the include path
#include <windows.h>
// Add vulkan_win32.h to the include path
#include <vulkan/vulkan_win32.h>
#endif /* _WIN64 */
/* remove _VK_TIMELINE_SEMAPHORE to use binary semaphores */

4
Samples/5_Domain_Specific/simpleVulkanMMAP/VulkanBaseApp.h
View File

@ -34,8 +34,10 @@
#include <vulkan/vulkan.h>
#ifdef _WIN64
#define NOMINMAX
#include <vulkan/vulkan_win32.h>
// Add windows.h to the include path firstly as dependency for other Windows headers
#include <windows.h>
// Add other Windows headers
#include <vulkan/vulkan_win32.h>
#endif /* _WIN64 */
struct GLFWwindow;

4
Samples/5_Domain_Specific/vulkanImageCUDA/vulkanImageCUDA.cu
View File

@ -27,10 +27,12 @@
#define GLFW_INCLUDE_VULKAN
#ifdef _WIN64
// Add windows.h to the include path firstly as dependency for other Windows headers
#include <windows.h>
// Add other Windows headers
#include <VersionHelpers.h>
#include <aclapi.h>
#include <dxgi1_2.h>
#include <windows.h>
#define _USE_MATH_DEFINES
#endif

11
Samples/6_Performance/transpose/transpose.cu
View File

@ -597,17 +597,16 @@ int main(int argc, char **argv)
1,
TILE_DIM * BLOCK_ROWS);
// Reset d_odata to zero before starting the next loop iteration to avoid
// carrying over results from previous kernels. Without this reset, residual
// data from a prior kernel (e.g., 'copy') could make a subsequent
// kernel (e.g., 'copySharedMem') appear correct even if it performs no work,
// Reset d_odata to zero before starting the next loop iteration to avoid
// carrying over results from previous kernels. Without this reset, residual
// data from a prior kernel (e.g., 'copy') could make a subsequent
// kernel (e.g., 'copySharedMem') appear correct even if it performs no work,
// leading to false positives in compareData.
for (int i = 0; i < (size_x * size_y); ++i) {
h_odata[i] = 0;
}
// copy host data to device
checkCudaErrors(
cudaMemcpy(d_odata, h_odata, mem_size, cudaMemcpyHostToDevice));
checkCudaErrors(cudaMemcpy(d_odata, h_odata, mem_size, cudaMemcpyHostToDevice));
}
// cleanup