CUDA 13.4 samples update - v13.4-public

Release 13.4 of the CUDA Samples supported by CUDA Toolkit 13.4.
See Changelog for more information.

cpp/4_CUDA_Libraries/cudaNvSci/.vscode/c_cpp_properties.json

@@ -1,18 +0,0 @@
-{
-    "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
-}

cpp/4_CUDA_Libraries/cudaNvSci/.vscode/extensions.json

@@ -1,7 +0,0 @@
-{
-    "recommendations": [
-        "nvidia.nsight-vscode-edition",
-        "ms-vscode.cpptools",
-        "ms-vscode.makefile-tools"
-    ]
-}

cpp/4_CUDA_Libraries/cudaNvSci/CMakeLists.txt

@@ -1,46 +1,32 @@
 cmake_minimum_required(VERSION 3.20)
 
-list(APPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/../../../cmake/Modules")
+include("${CMAKE_CURRENT_SOURCE_DIR}/../../../cmake/DetectCudaArch.cmake")
+
+if(SAMPLE_SKIP_BUILD)
+    return()
+endif()
 
 project(cudaNvSci LANGUAGES C CXX CUDA)
 
+include("${CMAKE_CURRENT_SOURCE_DIR}/../../../cmake/CudaSampleCommon.cmake")
+
 find_package(CUDAToolkit REQUIRED)
 
-set(CMAKE_POSITION_INDEPENDENT_CODE ON)
-
-set(CMAKE_CUDA_ARCHITECTURES 75 80 86 87 90 110)
-set(CMAKE_CUDA_FLAGS "${CMAKE_CUDA_FLAGS} -Wno-deprecated-gpu-targets")
-
-if(ENABLE_CUDA_DEBUG)
-    set(CMAKE_CUDA_FLAGS "${CMAKE_CUDA_FLAGS} -G")        # enable cuda-gdb (may significantly affect performance on some targets)
-else()
-    set(CMAKE_CUDA_FLAGS "${CMAKE_CUDA_FLAGS} -lineinfo") # add line information to all builds for debug tools (exclusive to -G option)
-endif()
-
-# Include directories and libraries
 include_directories(../../../Common)
 
 find_package(NVSCI)
 
-if(CMAKE_SYSTEM_NAME STREQUAL "Linux")
+if(CMAKE_SYSTEM_NAME STREQUAL "Linux" OR CMAKE_SYSTEM_NAME STREQUAL "QNX")
     if(NVSCI_FOUND)
         message(STATUS "FOUND NVSCI libs: ${NVSCIBUF_LIB} ${NVSCISYNC_LIB}")
         message(STATUS "Using NVSCI headers path: ${NVSCIBUF_INCLUDE_DIR} ${NVSCIBUF_INCLUDE_DIR}")
-        # Source file
-        # Add target for cudaNvSci
         add_executable(cudaNvSci imageKernels.cu cudaNvSci.cpp main.cpp)
-
         target_compile_options(cudaNvSci PRIVATE $<$<COMPILE_LANGUAGE:CUDA>:--extended-lambda>)
-
         target_compile_features(cudaNvSci PRIVATE cxx_std_17 cuda_std_17)
-
-        set_target_properties(cudaNvSci PROPERTIES CUDA_SEPARABLE_COMPILATION ON)
-
         target_include_directories(cudaNvSci PUBLIC
             ${CUDAToolkit_INCLUDE_DIRS}
             ${NVSCI_INCLUDE_DIRS}
         )
-
         target_link_libraries(cudaNvSci
             CUDA::cuda_driver
             ${NVSCI_LIBRARIES}
@@ -50,18 +36,15 @@ if(CMAKE_SYSTEM_NAME STREQUAL "Linux")
             COMMAND ${CMAKE_COMMAND} -E copy_if_different
             ${CMAKE_CURRENT_SOURCE_DIR}/teapot1024.ppm ${CMAKE_CURRENT_BINARY_DIR}/teapot1024.ppm
         )
-
         # Specify additional clean files
         set_target_properties(cudaNvSci PROPERTIES
             ADDITIONAL_CLEAN_FILES "teapot1024_out.ppm"
         )
+        include(InstallSamples)
+        setup_samples_install()
     else()
         message(STATUS "NvSCI not found - will not build sample 'cudaNvSci'")
     endif()
 else()
-    message(STATUS "Will not build sample cudaNvSci - requires Linux OS")
+    message(STATUS "Will not build sample cudaNvSci - requires Linux or QNX")
 endif()
-
-# Include installation configuration
-include(${CMAKE_CURRENT_SOURCE_DIR}/../../../cmake/InstallSamples.cmake)
-setup_samples_install()

cpp/4_CUDA_Libraries/cudaNvSci/README.md

@@ -2,7 +2,7 @@
 
 ## Description
 
-This sample demonstrates CUDA-NvSciBuf/NvSciSync Interop. Two CPU threads import the NvSciBuf and NvSciSync into CUDA to perform two image processing algorithms on a ppm image - image rotation in 1st thread & rgba to grayscale conversion of rotated image in 2nd thread. Currently only supported on Ubuntu 18.04
+This sample demonstrates CUDA-NvSciBuf/NvSciSync Interop. Two CPU threads import the NvSciBuf and NvSciSync into CUDA to perform two image processing algorithms on a ppm image - image rotation in 1st thread & rgba to grayscale conversion of rotated image in 2nd thread.
 
 ## Key Concepts
 
@@ -14,7 +14,7 @@ CUDA NvSci Interop, Data Parallel Algorithms, Image Processing
 
 ## Supported OSes
 
-Linux
+Linux, QNX (standard and safety cross-builds)
 
 ## Supported CPU Architecture
 
@@ -36,4 +36,6 @@ cudaExternalMemoryGetMappedBuffer, cudaImportExternalSemaphore, cudaDeviceGetAtt
 Download and install the [CUDA Toolkit](https://developer.nvidia.com/cuda-downloads) for your corresponding platform.
 Make sure the dependencies mentioned in [Dependencies]() section above are installed.
 
+On a QNX cross-build the NvSci headers and libraries come from the target filesystem, so pass `-DTARGET_FS=/path/to/qnx/targetfs` to cmake. Without it the build reports `NvSCI not found` and skips this sample. See the [QNX build instructions](../../../README.md#qnx).
+
 ## References (for more details)

cpp/4_CUDA_Libraries/cudaNvSci/imageKernels.cu

@@ -91,7 +91,7 @@ void launchGrayScaleKernel(unsigned int *d_rgbaImage,
     rgbToGrayscaleKernel<<<numOfBlocks, numThreadsPerBlock, 0, stream>>>(d_rgbaImage, imageWidth, imageHeight);
 
     unsigned int *outputData;
-    checkCudaErrors(cudaMallocHost((void **)&outputData, sizeof(unsigned int) * imageWidth * imageHeight));
+    checkCudaErrors(cudaMallocHost((void **)&outputData, sizeof(unsigned int) * imageWidth * imageHeight, 0));
     checkCudaErrors(cudaMemcpyAsync(
         outputData, d_rgbaImage, sizeof(unsigned int) * imageWidth * imageHeight, cudaMemcpyDeviceToHost, stream));
     checkCudaErrors(cudaStreamSynchronize(stream));