summary |
shortlog |
log |
commit | commitdiff |
tree
raw |
patch |
inline | side by side (from parent 1:
e35bbeb)
The test seems buggy, as the initialization part on one process may not be started when the other one reaches the second part, due to the way the locks are done.
Add a barrier to separate initialization from computation.
baseptr[0] = 1;
baseptr[1] = 2;
MPI_Win_unlock(rank, win);
baseptr[0] = 1;
baseptr[1] = 2;
MPI_Win_unlock(rank, win);
+ MPI_Barrier(MPI_COMM_WORLD);
/* Issue request-based get with testall. */
MPI_Win_lock_all(0, win);
MPI_Rget(&val1, 1, MPI_INT, 0, 0, 1, MPI_INT, win, &reqs[0]);
/* Issue request-based get with testall. */
MPI_Win_lock_all(0, win);
MPI_Rget(&val1, 1, MPI_INT, 0, 0, 1, MPI_INT, win, &reqs[0]);