portelli/Grid
1
0
Fork 0
mirror of https://github.com/paboyle/Grid.git synced 2025-06-17 15:27:06 +01:00
Code Releases Activity

Overlap cudamemcpy - didn't set up stream right

Browse Source
This commit is contained in:
Peter Boyle
2021-10-11 13:31:26 -07:00
parent cda915a345
commit 16c2a99965
3 changed files with 3 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
Grid/communicator/Communicator_mpi3.cc
View File

@ -389,7 +389,6 @@ double CartesianCommunicator::StencilSendToRecvFromBegin(std::vector<CommsReques
void *shm = (void *) this->ShmBufferTranslate(dest,recv);
assert(shm!=NULL);
acceleratorCopyDeviceToDeviceAsynch(xmit,shm,bytes);
acceleratorCopySynchronise(); // MPI prob slower
}
if ( CommunicatorPolicy == CommunicatorPolicySequential ) {
@ -405,6 +404,7 @@ void CartesianCommunicator::StencilSendToRecvFromComplete(std::vector<CommsReque
if (nreq==0) return;
std::vector<MPI_Status> status(nreq);
acceleratorCopySynchronise();
int ierr = MPI_Waitall(nreq,&list[0],&status[0]);
assert(ierr==0);
list.resize(0);