Skip to content

GPU: Fix linker warning with GCC 14, no need to make stack executable for including the binary code #13974

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
davidrohr merged 2 commits into from
Feb 17, 2025
Merged

GPU: Fix linker warning with GCC 14, no need to make stack executable for including the binary code #13974

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
11a7766
GPU: Preparation to make memset on host multi-threaded
davidrohr Feb 17, 2025
239cbf2
GPU: Fix linker warning with GCC 14, no need to make stack executable…
davidrohr Feb 17, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
33 changes: 22 additions & 11 deletions GPU/GPUTracking/Base/GPUReconstructionCPU.cxx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -60,6 +60,21 @@ GPUReconstructionCPU::~GPUReconstructionCPU()
Exit(); // Needs to be identical to GPU backend bahavior in order to avoid calling abstract methods later in the destructor
}

int32_t GPUReconstructionCPUBackend::getNOMPThreads()
{
int32_t ompThreads = 0;
if (mProcessingSettings.ompKernels == 2) {
ompThreads = mProcessingSettings.ompThreads / mNestedLoopOmpFactor;
if ((uint32_t)getOMPThreadNum() < mProcessingSettings.ompThreads % mNestedLoopOmpFactor) {
ompThreads++;
}
ompThreads = std::max(1, ompThreads);
} else {
ompThreads = mProcessingSettings.ompKernels ? mProcessingSettings.ompThreads : 1;
}
return ompThreads;
}

template <class T, int32_t I, typename... Args>
inline int32_t GPUReconstructionCPUBackend::runKernelBackendInternal(const krnlSetupTime& _xyz, const Args&... args)
{
Expand All @@ -73,16 +88,7 @@ inline int32_t GPUReconstructionCPUBackend::runKernelBackendInternal(const krnlS
}
uint32_t num = y.num == 0 || y.num == -1 ? 1 : y.num;
for (uint32_t k = 0; k < num; k++) {
int32_t ompThreads = 0;
if (mProcessingSettings.ompKernels == 2) {
ompThreads = mProcessingSettings.ompThreads / mNestedLoopOmpFactor;
if ((uint32_t)getOMPThreadNum() < mProcessingSettings.ompThreads % mNestedLoopOmpFactor) {
ompThreads++;
}
ompThreads = std::max(1, ompThreads);
} else {
ompThreads = mProcessingSettings.ompKernels ? mProcessingSettings.ompThreads : 1;
}
int32_t ompThreads = getNOMPThreads();
if (ompThreads > 1) {
if (mProcessingSettings.debugLevel >= 5) {
printf("Running %d ompThreads\n", ompThreads);
Expand All @@ -105,7 +111,12 @@ inline int32_t GPUReconstructionCPUBackend::runKernelBackendInternal(const krnlS
template <>
inline int32_t GPUReconstructionCPUBackend::runKernelBackendInternal<GPUMemClean16, 0>(const krnlSetupTime& _xyz, void* const& ptr, uint64_t const& size)
{
memset(ptr, 0, size);
int32_t ompThreads = std::max<int32_t>(1, std::min<int32_t>(size / (16 * 1024 * 1024), getNOMPThreads()));
if (ompThreads > 1) {
memset(ptr, 0, size);
} else {
memset(ptr, 0, size);
}
return 0;
}

Expand Down
1 change: 1 addition & 0 deletions GPU/GPUTracking/Base/GPUReconstructionCPU.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -46,6 +46,7 @@ class GPUReconstructionCPUBackend : public GPUReconstruction
uint32_t mNestedLoopOmpFactor = 1;
static int32_t getOMPThreadNum();
static int32_t getOMPMaxThreads();
int32_t getNOMPThreads();
};

class GPUReconstructionCPU : public GPUReconstructionKernels<GPUReconstructionCPUBackend>
Expand Down
2 changes: 1 addition & 1 deletion GPU/GPUTracking/Base/cuda/CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -160,7 +160,7 @@ elseif(GPUCA_CUDA_COMPILE_MODE STREQUAL "perkernel")
add_custom_command(
OUTPUT ${CMAKE_CURRENT_BINARY_DIR}/GPUTrackingCUDAKernelModules.o
COMMAND cp -u $<TARGET_OBJECTS:GPUTrackingCUDAKernels> ${CMAKE_CURRENT_BINARY_DIR}/cuda_kernel_module_fatbin/
COMMAND ${CMAKE_LINKER} --relocatable --format binary --output ${CMAKE_CURRENT_BINARY_DIR}/GPUTrackingCUDAKernelModules.o $<PATH:RELATIVE_PATH,$<LIST:TRANSFORM,$<PATH:GET_FILENAME,$<TARGET_OBJECTS:GPUTrackingCUDAKernels>>,PREPEND,${CMAKE_CURRENT_BINARY_DIR}/cuda_kernel_module_fatbin/>,${CMAKE_CURRENT_BINARY_DIR}>
COMMAND ${CMAKE_LINKER} -z noexecstack --relocatable --format binary --output ${CMAKE_CURRENT_BINARY_DIR}/GPUTrackingCUDAKernelModules.o $<PATH:RELATIVE_PATH,$<LIST:TRANSFORM,$<PATH:GET_FILENAME,$<TARGET_OBJECTS:GPUTrackingCUDAKernels>>,PREPEND,${CMAKE_CURRENT_BINARY_DIR}/cuda_kernel_module_fatbin/>,${CMAKE_CURRENT_BINARY_DIR}>
DEPENDS GPUTrackingCUDAKernels $<TARGET_OBJECTS:GPUTrackingCUDAKernels>
COMMENT "Compiling fatbin kernels ${CMAKE_CURRENT_BINARY_DIR}/GPUTrackingCUDAKernelModules.o"
VERBATIM
Expand Down
2 changes: 1 addition & 1 deletion GPU/GPUTracking/Base/hip/CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -217,7 +217,7 @@ elseif(GPUCA_HIP_COMPILE_MODE STREQUAL "perkernel")
add_custom_command(
OUTPUT ${CMAKE_CURRENT_BINARY_DIR}/GPUTrackingHIPKernelModules.o
COMMAND cp -u $<TARGET_OBJECTS:GPUTrackingHIPKernels> ${CMAKE_CURRENT_BINARY_DIR}/hip_kernel_module_fatbin/
COMMAND ${CMAKE_LINKER} --relocatable --format binary --output ${CMAKE_CURRENT_BINARY_DIR}/GPUTrackingHIPKernelModules.o $<PATH:RELATIVE_PATH,$<LIST:TRANSFORM,$<PATH:GET_FILENAME,$<TARGET_OBJECTS:GPUTrackingHIPKernels>>,PREPEND,${CMAKE_CURRENT_BINARY_DIR}/hip_kernel_module_fatbin/>,${CMAKE_CURRENT_BINARY_DIR}>
COMMAND ${CMAKE_LINKER} -z noexecstack --relocatable --format binary --output ${CMAKE_CURRENT_BINARY_DIR}/GPUTrackingHIPKernelModules.o $<PATH:RELATIVE_PATH,$<LIST:TRANSFORM,$<PATH:GET_FILENAME,$<TARGET_OBJECTS:GPUTrackingHIPKernels>>,PREPEND,${CMAKE_CURRENT_BINARY_DIR}/hip_kernel_module_fatbin/>,${CMAKE_CURRENT_BINARY_DIR}>
DEPENDS GPUTrackingHIPKernels $<TARGET_OBJECTS:GPUTrackingHIPKernels>
COMMENT "Compiling fatbin kernels ${CMAKE_CURRENT_BINARY_DIR}/GPUTrackingHIPKernelModules.o"
VERBATIM
Expand Down
2 changes: 1 addition & 1 deletion GPU/GPUTracking/cmake/helpers.cmake
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,7 +17,7 @@ function(create_binary_resource RESOURCE OUTPUTFILE)
FILE(RELATIVE_PATH input-file-rel ${CMAKE_CURRENT_BINARY_DIR} ${input-file-abs})
add_custom_command(
OUTPUT ${OUTPUTFILE}
COMMAND ${CMAKE_LINKER} --relocatable --format binary --output ${OUTPUTFILE} ${input-file-rel}
COMMAND ${CMAKE_LINKER} -z noexecstack --relocatable --format binary --output ${OUTPUTFILE} ${input-file-rel}
DEPENDS ${input-file-rel}
COMMENT "Adding binary resource ${input-file-rel}"
VERBATIM
Expand Down