{lang}[GCC/11.3.0] CUDA-Samples v11.8, CUDA-Samples v12.2 w/ CUDA 11.7.0, CUDA 12.2.0

easybuild/easyconfigs/c/CUDA-Samples/CUDA-Samples-11.3-GCC-10.3.0-CUDA-11.3.1.eb

@@ -11,7 +11,11 @@ toolchain = {'name': 'GCC', 'version': '10.3.0'}
 
 source_urls = ['https://github.com/NVIDIA/cuda-samples/archive/']
 sources = ['v%(version)s.tar.gz']
-checksums = ['2bee5f7c89347259aaab75aa6df6e10375059bdbbaf04cc7936f5db7d54fa3ac']
+patches = ['cuda-samples-11.3_multiple-sms.patch']
+checksums = [
+    {'v11.3.tar.gz': '2bee5f7c89347259aaab75aa6df6e10375059bdbbaf04cc7936f5db7d54fa3ac'},
+    {'cuda-samples-11.3_multiple-sms.patch': 'b31613f4160456f0d0abf82999c7fb7eee781f0efadc8b9bbb5a02ef0f37e21d'},
+]
 
 dependencies = [
     ('CUDA', '11.3.1', '', SYSTEM),
@@ -32,7 +36,7 @@ local_filters += "Samples/simpleVulkan/Makefile "
 local_filters += "Samples/simpleVulkanMMAP/Makefile "
 local_filters += "Samples/streamOrderedAllocationIPC/Makefile "
 local_filters += "Samples/vulkanImageCUDA/Makefile"
-buildopts = "HOST_COMPILER=g++ FILTER_OUT='%s'" % local_filters
+buildopts = "HOST_COMPILER=g++ SMS='%%(cuda_cc_space_sep_no_period)s' FILTER_OUT='%s'" % local_filters
 
 files_to_copy = [
     (['bin/%s/linux/release/*' % ARCH], 'bin'),

easybuild/easyconfigs/c/CUDA-Samples/CUDA-Samples-11.6-GCC-11.3.0-CUDA-11.7.0.eb

@@ -11,7 +11,11 @@ toolchain = {'name': 'GCC', 'version': '11.3.0'}
 
 source_urls = ['https://github.com/NVIDIA/cuda-samples/archive/']
 sources = ['v%(version)s.tar.gz']
-checksums = ['75b858bcf9e534eaa0f129c418e661b83872d743de218df8a5278cc429f9ea98']
+patches = ['cuda-samples-11.6_multiple-sms.patch']
+checksums = [
+    {'v11.6.tar.gz': '75b858bcf9e534eaa0f129c418e661b83872d743de218df8a5278cc429f9ea98'},
+    {'cuda-samples-11.6_multiple-sms.patch': '8849e4882d797d155d6ebb71377fa1409205361776ade8da699452a4ecb94a0a'},
+]
 
 dependencies = [
     ('CUDA', '11.7.0', '', SYSTEM),
@@ -33,7 +37,7 @@ local_filters += "Samples/simpleVulkanMMAP/Makefile "
 local_filters += "Samples/streamOrderedAllocationIPC/Makefile "
 local_filters += "Samples/vulkanImageCUDA/Makefile"
 
-buildopts = "HOST_COMPILER=g++ FILTER_OUT='%s'" % local_filters
+buildopts = "HOST_COMPILER=g++ SMS='%%(cuda_cc_space_sep_no_period)s' FILTER_OUT='%s'" % local_filters
 
 files_to_copy = [
     (['bin/%s/linux/release/*' % ARCH], 'bin'),

easybuild/easyconfigs/c/CUDA-Samples/CUDA-Samples-11.8-GCC-11.3.0-CUDA-11.7.0.eb

@@ -0,0 +1,59 @@
+easyblock = 'MakeCp'
+
+name = 'CUDA-Samples'
+version = '11.8'
+versionsuffix = '-CUDA-%(cudaver)s'
+
+homepage = 'https://github.com/NVIDIA/cuda-samples'
+description = "Samples for CUDA Developers which demonstrates features in CUDA Toolkit"
+
+toolchain = {'name': 'GCC', 'version': '11.3.0'}
+
+source_urls = ['https://github.com/NVIDIA/cuda-samples/archive/']
+sources = ['v%(version)s.tar.gz']
+patches = ['cuda-samples-11.6_multiple-sms.patch']
+checksums = [
+    {'v11.8.tar.gz': '1bc02c0ca42a323f3c7a05b5682eae703681a91e95b135bfe81f848b2d6a2c51'},
+    {'cuda-samples-11.6_multiple-sms.patch': '8849e4882d797d155d6ebb71377fa1409205361776ade8da699452a4ecb94a0a'},
+]
+
+dependencies = [
+    ('CUDA', '11.7.0', '', SYSTEM),
+]
+
+# Get rid of pre-built Windows DLLs and only build deviceQuery for now.
+prebuildopts = "rm -r bin/win64 && "
+
+# Filter out samples that require extensive dependencies.
+local_filters = "Samples/2_Concepts_and_Techniques/EGLStream_CUDA_Interop/Makefile "
+local_filters += "Samples/4_CUDA_Libraries/boxFilterNPP/Makefile "
+local_filters += "Samples/4_CUDA_Libraries/cannyEdgeDetectorNPP/Makefile "
+local_filters += "Samples/4_CUDA_Libraries/cudaNvSci/Makefile "
+local_filters += "Samples/4_CUDA_Libraries/cudaNvSciNvMedia/Makefile "
+local_filters += "Samples/5_Domain_Specific/simpleGL/Makefile "
+local_filters += "Samples/3_CUDA_Features/warpAggregatedAtomicsCG/Makefile "
+local_filters += "Samples/5_Domain_Specific/simpleVulkan/Makefile "
+local_filters += "Samples/5_Domain_Specific/simpleVulkanMMAP/Makefile "
+local_filters += "Samples/2_Concepts_and_Techniques/streamOrderedAllocationIPC/Makefile "
+local_filters += "Samples/5_Domain_Specific/vulkanImageCUDA/Makefile "
+local_filters += "Samples/6_Performance/LargeKernelParameter/Makefile "
+
+buildopts = "HOST_COMPILER=g++ SMS='%%(cuda_cc_space_sep_no_period)s' FILTER_OUT='%s'" % local_filters
+
+files_to_copy = [
+    (['bin/%s/linux/release/*' % ARCH], 'bin'),
+    'LICENSE',
+]
+
+local_binaries = ['deviceQuery', 'matrixMul', 'bandwidthTest', 'cudaOpenMP']
+
+# Only paths are used for sanity checks.
+# Commands may fail due to missing compatibility libraries that might be needed
+# to be able to use this specific CUDA version in combination with the available
+# NVIDIA drivers.
+sanity_check_paths = {
+    'files': ['bin/%s' % x for x in local_binaries],
+    'dirs': [],
+}
+
+moduleclass = 'lang'

easybuild/easyconfigs/c/CUDA-Samples/CUDA-Samples-12.1-GCC-12.3.0-CUDA-12.1.1.eb

@@ -11,7 +11,11 @@ toolchain = {'name': 'GCC', 'version': '12.3.0'}
 
 source_urls = ['https://github.com/NVIDIA/cuda-samples/archive/']
 sources = ['v%(version)s.tar.gz']
-checksums = ['f758160645b366d79c2638d8dfd389f01029b8d179ab0c11726b9ef58aecebd9']
+patches = ['cuda-samples-11.6_multiple-sms.patch']
+checksums = [
+    {'v12.1.tar.gz': 'f758160645b366d79c2638d8dfd389f01029b8d179ab0c11726b9ef58aecebd9'},
+    {'cuda-samples-11.6_multiple-sms.patch': '8849e4882d797d155d6ebb71377fa1409205361776ade8da699452a4ecb94a0a'},
+]
 
 dependencies = [
     ('CUDA', '12.1.1', '', SYSTEM),
@@ -58,7 +62,7 @@ if ARCH == 'aarch64':
     local_filters += "Samples/3_CUDA_Features/cdpQuadtree/Makefile "
     local_filters += "Samples/3_CUDA_Features/cdpAdvancedQuicksort/Makefile "
 
-buildopts = "HOST_COMPILER=g++ FILTER_OUT='%s'" % local_filters
+buildopts = "HOST_COMPILER=g++ SMS='%%(cuda_cc_space_sep_no_period)s' FILTER_OUT='%s'" % local_filters
 
 # Remove libraries in the bin dir after a successful 'make'
 buildopts += " && rm bin/*/linux/release/lib*.so.*"

easybuild/easyconfigs/c/CUDA-Samples/CUDA-Samples-12.2-GCC-11.3.0-CUDA-12.2.0.eb

@@ -0,0 +1,63 @@
+easyblock = 'MakeCp'
+
+name = 'CUDA-Samples'
+version = '12.2'
+versionsuffix = '-CUDA-%(cudaver)s'
+
+homepage = 'https://github.com/NVIDIA/cuda-samples'
+description = "Samples for CUDA Developers which demonstrates features in CUDA Toolkit"
+
+toolchain = {'name': 'GCC', 'version': '11.3.0'}
+
+source_urls = ['https://github.com/NVIDIA/cuda-samples/archive/']
+sources = ['v%(version)s.tar.gz']
+patches = ['cuda-samples-11.6_multiple-sms.patch']
+checksums = [
+    {'v12.2.tar.gz': '1823cfe28e97a9230107aa72b231f78952c0f178b71a920f036d360518480bdc'},
+    {'cuda-samples-11.6_multiple-sms.patch': '8849e4882d797d155d6ebb71377fa1409205361776ade8da699452a4ecb94a0a'},
+]
+
+builddependencies = [
+    ('CMake', '3.24.3'),
+]
+
+dependencies = [
+    ('CUDA', '12.2.0', '', SYSTEM),
+]
+
+# Get rid of pre-built Windows DLLs and only build deviceQuery for now.
+prebuildopts = "rm -r bin/win64 && "
+
+# Filter out samples that require extensive dependencies.
+local_filters = "Samples/2_Concepts_and_Techniques/EGLStream_CUDA_Interop/Makefile "
+local_filters += "Samples/4_CUDA_Libraries/boxFilterNPP/Makefile "
+local_filters += "Samples/4_CUDA_Libraries/cannyEdgeDetectorNPP/Makefile "
+local_filters += "Samples/4_CUDA_Libraries/cudaNvSci/Makefile "
+local_filters += "Samples/4_CUDA_Libraries/cudaNvSciNvMedia/Makefile "
+local_filters += "Samples/5_Domain_Specific/simpleGL/Makefile "
+local_filters += "Samples/3_CUDA_Features/warpAggregatedAtomicsCG/Makefile "
+local_filters += "Samples/5_Domain_Specific/simpleVulkan/Makefile "
+local_filters += "Samples/5_Domain_Specific/simpleVulkanMMAP/Makefile "
+local_filters += "Samples/2_Concepts_and_Techniques/streamOrderedAllocationIPC/Makefile "
+local_filters += "Samples/5_Domain_Specific/vulkanImageCUDA/Makefile "
+local_filters += "Samples/6_Performance/LargeKernelParameter/Makefile "
+
+buildopts = "HOST_COMPILER=g++ SMS='%%(cuda_cc_space_sep_no_period)s' FILTER_OUT='%s'" % local_filters
+
+files_to_copy = [
+    (['bin/%s/linux/release/*' % ARCH], 'bin'),
+    'LICENSE',
+]
+
+local_binaries = ['deviceQuery', 'matrixMul', 'bandwidthTest', 'cudaOpenMP']
+
+# Only paths are used for sanity checks.
+# Commands may fail due to missing compatibility libraries that might be needed
+# to be able to use this specific CUDA version in combination with the available
+# NVIDIA drivers.
+sanity_check_paths = {
+    'files': ['bin/%s' % x for x in local_binaries],
+    'dirs': [],
+}
+
+moduleclass = 'lang'

easybuild/easyconfigs/c/CUDA-Samples/cuda-samples-11.3_multiple-sms.patch

@@ -0,0 +1,31 @@
+# Fixes "nvcc fatal: Option '--ptx (-ptx)' is not allowed when compiling for multiple GPU architectures"
+# fatal compilation issue when building for multiple SM architectures
+# More info, see https://github.com/NVIDIA/cuda-samples/issues/289
+
+# Author: Caspar van Leeuwen
+
+diff -Nru cuda-samples-11.3.orig/Samples/memMapIPCDrv/Makefile cuda-samples-11.3/Samples/memMapIPCDrv/Makefile
+--- cuda-samples-11.3.orig/Samples/memMapIPCDrv/Makefile	2024-07-29 13:17:10.330743000 +0200
++++ cuda-samples-11.3/Samples/memMapIPCDrv/Makefile	2024-07-29 13:19:13.158507504 +0200
+@@ -321,6 +321,12 @@
+ ifneq ($(HIGHEST_SM),)
+ GENCODE_FLAGS += -gencode arch=compute_$(HIGHEST_SM),code=compute_$(HIGHEST_SM)
+ endif
++
++# Generate the explicit PTX file for the lowest SM architecture in $(SMS), so it works on all SMS listed there
++LOWEST_SM := $(firstword $(sort $(SMS)))
++ifneq ($(LOWEST_SM),)
++GENCODE_FLAGS_LOWEST_SM += -gencode arch=compute_$(LOWEST_SM),code=compute_$(LOWEST_SM)
++endif
+ endif
+
+ ifeq ($(TARGET_OS),darwin)
+@@ -401,7 +407,7 @@
+ endif
+
+ $(PTX_FILE): memMapIpc_kernel.cu
+-	$(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -ptx $<
++	$(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS_LOWEST_SM) -o $@ -ptx $<
+ 	$(EXEC) mkdir -p data
+ 	$(EXEC) cp -f $@ ./data
+ 	$(EXEC) mkdir -p ../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)

easybuild/easyconfigs/c/CUDA-Samples/cuda-samples-11.6_multiple-sms.patch

@@ -0,0 +1,56 @@
+# Fixes "nvcc fatal: Option '--ptx (-ptx)' is not allowed when compiling for multiple GPU architectures"
+# fatal compilation issue when building for multiple SM architectures
+# More info, see https://github.com/NVIDIA/cuda-samples/issues/289
+
+# Author: Caspar van Leeuwen
+
+diff -Nru cuda-samples-12.2.orig/Samples/3_CUDA_Features/memMapIPCDrv/Makefile cuda-samples-12.2/Samples/3_CUDA_Features/memMapIPCDrv/Makefile
+--- cuda-samples-12.2.orig/Samples/3_CUDA_Features/memMapIPCDrv/Makefile	2024-07-29 12:14:28.538848000 +0200
++++ cuda-samples-12.2/Samples/3_CUDA_Features/memMapIPCDrv/Makefile	2024-07-29 13:02:45.134261829 +0200
+@@ -313,6 +313,12 @@
+ ifneq ($(HIGHEST_SM),)
+ GENCODE_FLAGS += -gencode arch=compute_$(HIGHEST_SM),code=compute_$(HIGHEST_SM)
+ endif
++
++# Generate the explicit PTX file for the lowest SM architecture in $(SMS), so it works on all SMS listed there
++LOWEST_SM := $(firstword $(sort $(SMS)))
++ifneq ($(LOWEST_SM),)
++GENCODE_FLAGS_LOWEST_SM += -gencode arch=compute_$(LOWEST_SM),code=compute_$(LOWEST_SM)
++endif
+ endif
+
+ ifeq ($(TARGET_OS),darwin)
+@@ -394,7 +400,7 @@
+ endif
+
+ $(PTX_FILE): memMapIpc_kernel.cu
+-	$(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -ptx $<
++	$(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS_LOWEST_SM) -o $@ -ptx $<
+ 	$(EXEC) mkdir -p data
+ 	$(EXEC) cp -f $@ ./data
+ 	$(EXEC) mkdir -p ../../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)
+diff -Nru cuda-samples-12.2.orig/Samples/3_CUDA_Features/ptxjit/Makefile cuda-samples-12.2/Samples/3_CUDA_Features/ptxjit/Makefile
+--- cuda-samples-12.2.orig/Samples/3_CUDA_Features/ptxjit/Makefile	2024-07-29 12:14:28.546771000 +0200
++++ cuda-samples-12.2/Samples/3_CUDA_Features/ptxjit/Makefile	2024-07-29 13:02:38.741961008 +0200
+@@ -307,6 +307,12 @@
+ ifneq ($(HIGHEST_SM),)
+ GENCODE_FLAGS += -gencode arch=compute_$(HIGHEST_SM),code=compute_$(HIGHEST_SM)
+ endif
++
++# Generate the explicit PTX file for the lowest SM architecture in $(SMS), so it works on all SMS listed there
++LOWEST_SM := $(firstword $(sort $(SMS)))
++ifneq ($(LOWEST_SM),)
++GENCODE_FLAGS_LOWEST_SM += -gencode arch=compute_$(LOWEST_SM),code=compute_$(LOWEST_SM)
++endif
+ endif
+
+ ifeq ($(TARGET_OS),darwin)
+@@ -390,7 +396,7 @@
+ endif
+
+ $(PTX_FILE): ptxjit_kernel.cu
+-	$(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -ptx $<
++	$(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS_LOWEST_SM) -o $@ -ptx $<
+ 	$(EXEC) mkdir -p data
+ 	$(EXEC) cp -f $@ ./data
+ 	$(EXEC) mkdir -p ../../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)