Shuenhoy/vtk-m
1
0
Fork 0
mirror of https://gitlab.kitware.com/vtk/vtk-m synced 2024-10-05 01:49:02 +00:00
Code Issues 2 Actions Packages Projects Releases Wiki Activity

Merge topic 'correct_clang_cuda_warnings'

Browse Source 
b48c19f25 Correct warnings found by using clang as the host compiler for cuda

Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Kenneth Moreland <kmorel@sandia.gov>
Merge-request: !2232
This commit is contained in:
Robert Maynard 2020-08-24 14:01:39 +00:00 committed by Kitware Robot
commit db3b281da3
3 changed files with 16 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
vtkm/cont/cuda/internal/DeviceAdapterAlgorithmCuda.cu
View File

@ -207,8 +207,8 @@ void DeviceAdapterAlgorithm<vtkm::cont::DeviceAdapterTagCuda>::GetBlocksAndThrea
int deviceId; int deviceId;
VTKM_CUDA_CALL(cudaGetDevice(&deviceId)); //get deviceid from cuda VTKM_CUDA_CALL(cudaGetDevice(&deviceId)); //get deviceid from cuda
const auto& params = cuda::internal::scheduling_1d_parameters[static_cast<size_t>(deviceId)]; const auto& params = cuda::internal::scheduling_1d_parameters[static_cast<size_t>(deviceId)];
blocks = params.first; blocks = static_cast<vtkm::UInt32>(params.first);
threadsPerBlock = params.second; threadsPerBlock = static_cast<vtkm::UInt32>(params.second);
} }
void DeviceAdapterAlgorithm<vtkm::cont::DeviceAdapterTagCuda>::GetBlocksAndThreads( void DeviceAdapterAlgorithm<vtkm::cont::DeviceAdapterTagCuda>::GetBlocksAndThreads(
@ -223,13 +223,13 @@ void DeviceAdapterAlgorithm<vtkm::cont::DeviceAdapterTagCuda>::GetBlocksAndThrea
if (size.z <= 1) if (size.z <= 1)
{ //2d images { //2d images
const auto& params = cuda::internal::scheduling_2d_parameters[static_cast<size_t>(deviceId)]; const auto& params = cuda::internal::scheduling_2d_parameters[static_cast<size_t>(deviceId)];
blocks = params.first; blocks = static_cast<vtkm::UInt32>(params.first);
threadsPerBlock = params.second; threadsPerBlock = params.second;
} }
else else
{ //3d images { //3d images
const auto& params = cuda::internal::scheduling_3d_parameters[static_cast<size_t>(deviceId)]; const auto& params = cuda::internal::scheduling_3d_parameters[static_cast<size_t>(deviceId)];
blocks = params.first; blocks = static_cast<vtkm::UInt32>(params.first);
threadsPerBlock = params.second; threadsPerBlock = params.second;
} }
} }

14
vtkm/cont/cuda/internal/DeviceAdapterMemoryManagerCuda.cu
View File

@ -145,8 +145,11 @@ void DeviceAdapterMemoryManager<vtkm::cont::DeviceAdapterTagCuda>::CopyHostToDev
vtkm::cont::GetHumanReadableSize(static_cast<std::size_t>(size)).c_str(), vtkm::cont::GetHumanReadableSize(static_cast<std::size_t>(size)).c_str(),
size); size);
VTKM_CUDA_CALL(cudaMemcpyAsync( VTKM_CUDA_CALL(cudaMemcpyAsync(dest.GetPointer(),
dest.GetPointer(), src.GetPointer(), size, cudaMemcpyHostToDevice, cudaStreamPerThread)); src.GetPointer(),
static_cast<std::size_t>(size),
cudaMemcpyHostToDevice,
cudaStreamPerThread));
} }
} }
@ -203,8 +206,11 @@ void DeviceAdapterMemoryManager<vtkm::cont::DeviceAdapterTagCuda>::CopyDeviceToH
vtkm::cont::GetHumanReadableSize(static_cast<std::size_t>(size)).c_str(), vtkm::cont::GetHumanReadableSize(static_cast<std::size_t>(size)).c_str(),
size); size);
VTKM_CUDA_CALL(cudaMemcpyAsync( VTKM_CUDA_CALL(cudaMemcpyAsync(dest.GetPointer(),
dest.GetPointer(), src.GetPointer(), size, cudaMemcpyDeviceToHost, cudaStreamPerThread)); src.GetPointer(),
static_cast<std::size_t>(size),
cudaMemcpyDeviceToHost,
cudaStreamPerThread));
} }
//In all cases we have possibly multiple async calls queued up in //In all cases we have possibly multiple async calls queued up in

2
vtkm/cont/cuda/testing/UnitTestCudaMathEdgeCases.cu
View File

@ -22,7 +22,9 @@
#include <random> #include <random>
#include <vector> #include <vector>
VTKM_THIRDPARTY_PRE_INCLUDE
#include "curand_kernel.h" #include "curand_kernel.h"
VTKM_THIRDPARTY_POST_INCLUDE
namespace namespace
{ {