portelli/Grid
1
0
Fork 0
mirror of https://github.com/paboyle/Grid.git synced 2024-11-14 01:35:36 +00:00
Code Releases Activity

More verbosity. The 16bit limit on Grid.y, Grid.z is annoying

Browse Source
This commit is contained in:
Peter Boyle 2021-03-09 04:29:37 +01:00
parent c90beee774
commit 4d1ea15c79
2 changed files with 5 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
Grid/threads/Accelerator.cc
View File

@ -53,7 +53,6 @@ void acceleratorInit(void)
prop = gpu_props[i]; prop = gpu_props[i];
totalDeviceMem = prop.totalGlobalMem; totalDeviceMem = prop.totalGlobalMem;
if ( world_rank == 0) { if ( world_rank == 0) {
#ifndef GRID_DEFAULT_GPU
if ( i==rank ) { if ( i==rank ) {
printf("AcceleratorCudaInit[%d]: ========================\n",rank); printf("AcceleratorCudaInit[%d]: ========================\n",rank);
printf("AcceleratorCudaInit[%d]: Device Number : %d\n", rank,i); printf("AcceleratorCudaInit[%d]: Device Number : %d\n", rank,i);
@ -67,8 +66,8 @@ void acceleratorInit(void)
GPU_PROP(warpSize); GPU_PROP(warpSize);
GPU_PROP(pciBusID); GPU_PROP(pciBusID);
GPU_PROP(pciDeviceID); GPU_PROP(pciDeviceID);
printf("AcceleratorCudaInit[%d]: maxGridSize (%d,%d,%d)\n",rank,prop.maxGridSize[0],prop.maxGridSize[1],prop.maxGridSize[2]);
} }
#endif
// GPU_PROP(unifiedAddressing); // GPU_PROP(unifiedAddressing);
// GPU_PROP(l2CacheSize); // GPU_PROP(l2CacheSize);
// GPU_PROP(singleToDoublePrecisionPerfRatio); // GPU_PROP(singleToDoublePrecisionPerfRatio);

7
Grid/threads/Accelerator.h
View File

@ -178,9 +178,10 @@ void Lambda6Apply(uint64_t num1, uint64_t num2, uint64_t num3,
cudaDeviceSynchronize(); \ cudaDeviceSynchronize(); \
cudaError err = cudaGetLastError(); \ cudaError err = cudaGetLastError(); \
if ( cudaSuccess != err ) { \ if ( cudaSuccess != err ) { \
printf("Cuda error %s \n", cudaGetErrorString( err )); \ printf("accelerator_barrier(): Cuda error %s \n", \
puts(__FILE__); \ cudaGetErrorString( err )); \
printf("Line %d\n",__LINE__); \ printf("File %s Line %d\n",__FILE__,__LINE__); \
fflush(stdout); \
if (acceleratorAbortOnGpuError) assert(err==cudaSuccess); \ if (acceleratorAbortOnGpuError) assert(err==cudaSuccess); \
} \ } \
} }