portelli/Grid
1
0
Fork 0
mirror of https://github.com/paboyle/Grid.git synced 2025-04-27 14:15:55 +01:00
Code Releases Activity

stream synchronise

Browse Source
This commit is contained in:
Peter Boyle 2023-03-24 15:40:30 -04:00
parent d0bb033ea2
commit d8a9a745d8
1 changed files with 3 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
Grid/lattice/Lattice_reduction_gpu.h
View File

@ -222,14 +222,14 @@ inline typename vobj::scalar_objectD sumD_gpu_small(const vobj *lat, Integer osi
commVector<sobj> buffer(numBlocks); commVector<sobj> buffer(numBlocks);
sobj *buffer_v = &buffer[0]; sobj *buffer_v = &buffer[0];
sobj result; sobj result;
reduceKernel<<< numBlocks, numThreads, smemSize >>>(lat, buffer_v, size); reduceKernel<<< numBlocks, numThreads, smemSize, computeStream >>>(lat, buffer_v, size);
accelerator_barrier(); accelerator_barrier();
acceleratorCopyFromDevice(buffer_v,&result,sizeof(result)); acceleratorCopyFromDevice(buffer_v,&result,sizeof(result));
#else #else
Vector<sobj> buffer(numBlocks); Vector<sobj> buffer(numBlocks);
sobj *buffer_v = &buffer[0]; sobj *buffer_v = &buffer[0];
sobj result; sobj result;
reduceKernel<<< numBlocks, numThreads, smemSize >>>(lat, buffer_v, size); reduceKernel<<< numBlocks, numThreads, smemSize, computeStream >>>(lat, buffer_v, size);
accelerator_barrier(); accelerator_barrier();
result = *buffer_v; result = *buffer_v;
#endif #endif