From: Augustin Degomme <adegomme@gmail.com>
Date: Wed, 10 Jul 2019 12:59:57 +0000 (+0200)
Subject: Try to avoid rare bug on one ci node.
X-Git-Tag: v3.24~346
X-Git-Url: http://info.iut-bm.univ-fcomte.fr/pub/gitweb/simgrid.git/commitdiff_plain/89a7204b5cc355432e0fddd01a54f7a1975b49e0?hp=b5c0093dd3fb7d66ec6e1f72963ba7b9faa45fe7;ds=sidebyside

Try to avoid rare bug on one ci node.
A guess is that the win_lock call by a process may be finishing the comms posted by the raccumulate call from the other process.
And this would mean having max(0,undefined) as a result if memory is not initialized before win_lock_all.
I didn't find anything saying lock_all should be blocking/collective, so I'm not modifying it for so little.
Other guess: something's rotten with locks, as I think we don't check properly for them before posting rma calls.
---

diff --git a/teshsuite/smpi/mpich3-test/rma/racc_local_comp.c b/teshsuite/smpi/mpich3-test/rma/racc_local_comp.c
index a86c766d1a..35c228a0df 100644
--- a/teshsuite/smpi/mpich3-test/rma/racc_local_comp.c
+++ b/teshsuite/smpi/mpich3-test/rma/racc_local_comp.c
@@ -33,10 +33,9 @@ int main(int argc, char *argv[])
     MPI_Alloc_mem(MAX_SIZE * sizeof(int), MPI_INFO_NULL, &winbuf);
     MPI_Win_create(winbuf, MAX_SIZE * sizeof(int), sizeof(int), MPI_INFO_NULL,
                    MPI_COMM_WORLD, &window);
-
-    MPI_Win_lock_all(0, window);
     buf[0] = 0;
     winbuf[0]=0;
+    MPI_Win_lock_all(0, window);
     /* Test Raccumulate local completion with small data.
      * Small data is always copied to header packet as immediate data. */
     if (rank == 1) {