Merge fb9f77575a08918a682d7eb0781c6e5f1d4f0bf3 into 3e8f91d1a116060d3fedfe856f3721db970de030

* Enhancement for GLFW include and lib search

* Fixing issue #321: A potential bug in memMapIPCDrv/memMapIpc.cpp

* Update CMakelist.txt for the sample 0_Introduction/template

* Copy .dll to correct dir for 5_Domain_Specific/Mandelbrot

* Fix typo

* Update changelog for cudaNvSciBufMultiplanar

CHANGELOG.md

@@ -36,6 +36,7 @@
     * `cuDLALayerwiseStatsHybrid`
     * `cuDLALayerwiseStatsStandalone`
     * `cuDLAStandaloneMode`
+    * `cudaNvSciBufMultiplanar`
     * `cudaNvSciNvMedia`
     * `fluidsGLES`
     * `nbody_opengles`

Samples/0_Introduction/CMakeLists.txt

@@ -55,6 +55,7 @@ add_subdirectory(simpleTexture3D)
 add_subdirectory(simpleTextureDrv)
 add_subdirectory(simpleVoteIntrinsics)
 add_subdirectory(simpleZeroCopy)
+add_subdirectory(template)
 add_subdirectory(systemWideAtomics)
 add_subdirectory(vectorAdd)
 add_subdirectory(vectorAddDrv)

Samples/0_Introduction/template/CMakeLists.txt

@@ -20,7 +20,7 @@ include_directories(../../../Common)
 
 # Source file
 # Add target for template
-add_executable(template template.cu)
+add_executable(template template.cu template_cpu.cpp)
 
 target_compile_options(template PRIVATE $<$<COMPILE_LANGUAGE:CUDA>:--extended-lambda>)
 

Samples/3_CUDA_Features/bf16TensorCoreGemm/bf16TensorCoreGemm.cu

@@ -592,8 +592,8 @@ __global__ void simple_wmma_bf16gemm(__nv_bfloat16 *a, __nv_bfloat16 *b, float *
       int aCol = i; 
       int aRow = warpM * M;
 
-      int bCol = i;
-      int bRow = warpN * N;
+      int bCol = warpN * N;
+      int bRow = i;
 
       // Bounds checking
       if (aRow < m_ld && aCol < k_ld && bRow < k_ld && bCol < n_ld) {

Samples/5_Domain_Specific/Mandelbrot/CMakeLists.txt

@@ -65,14 +65,14 @@ target_compile_features(Mandelbrot PRIVATE cxx_std_17 cuda_std_17)
                 POST_BUILD
                 COMMAND ${CMAKE_COMMAND} -E copy
                 ${CMAKE_CURRENT_SOURCE_DIR}/../../../bin/win64/$<CONFIGURATION>/freeglut.dll
-                ${CMAKE_CURRENT_BINARY_DIR}
+                ${CMAKE_CURRENT_BINARY_DIR}/$<CONFIGURATION>
             )
 
             add_custom_command(TARGET Mandelbrot
                 POST_BUILD
                 COMMAND ${CMAKE_COMMAND} -E copy
                 ${CMAKE_CURRENT_SOURCE_DIR}/../../../bin/win64/$<CONFIGURATION>/glew64.dll
-                ${CMAKE_CURRENT_BINARY_DIR}
+                ${CMAKE_CURRENT_BINARY_DIR}/$<CONFIGURATION>
             )
         endif()