Issue 601 in fds-smv: Fatal error in MPI_Waitall

34 views
Skip to first unread message

codesite...@google.com

unread,
Jan 7, 2009, 11:18:16 PM1/7/09
to fds-smv...@googlegroups.com
Status: New
Owner: ----
Labels: Type-Defect Priority-Medium

New issue 601 by kevin.c...@wfra.com.au: Fatal error in MPI_Waitall
http://code.google.com/p/fds-smv/issues/detail?id=601

Please complete the following lines...

Application Version: 5.2.4 Parallel
SVN Revision Number: 2651
Compile Date: 11 Nov 2008
Operating System: Windows XP 64 bit

Describe details of the issue below:
I tried to run the attached input file over 2 machines and get the error
message below. I have tried running using 1 computer and over 2 computers
with 32 bit Windows XP, both getting similar message.

Fatal error in MPI_Waitall: Other MPI error, error stack:
MPI_Waitall(257)..........................: MPI_Waitall(count=20,
req_array=0000
0000139ABEB0, status_array=00000000139AD2C0) failed
MPIDI_CH3i_Progress_wait(215).............: an error occurred while
handling an
event returned by MPIDU_Sock_Wait()
MPIDI_CH3I_Progress_handle_sock_event(660):
MPIDU_Sock_wait(2275).....................: An invalid argument was
supplied. (e
rrno 10022)
job aborted:
rank: node: exit code[: error message]
0: FDS1: 1
1: FDS1: 1
2: FDS1: 1
3: FDS1: 1
4: FDS2: 1
5: FDS2: 1
6: FDS2: 1: Fatal error in MPI_Waitall: Other MPI error, error stack:
MPI_Waitall(257)..........................: MPI_Waitall(count=20,
req_array=0000
0000139ABEB0, status_array=00000000139AD2C0) failed
MPIDI_CH3i_Progress_wait(215).............: an error occurred while
handling an
event returned by MPIDU_Sock_Wait()
MPIDI_CH3I_Progress_handle_sock_event(660):
MPIDU_Sock_wait(2275).....................: An invalid argument was
supplied. (e
rrno 10022)
7: FDS2: 1

Attachments:
kev3.fds 9.6 KB

--
You received this message because you are listed in the owner
or CC fields of this issue, or because you starred this issue.
You may adjust your issue notification preferences at:
http://code.google.com/hosting/settings

codesite...@google.com

unread,
Jan 8, 2009, 8:24:51 AM1/8/09
to fds-smv...@googlegroups.com
Updates:
Status: Accepted
Owner: mcgratta

Comment #1 on issue 601 by mcgratta: Fatal error in MPI_Waitall
http://code.google.com/p/fds-smv/issues/detail?id=601

I'll take a look at it.

codesite...@google.com

unread,
Jan 8, 2009, 11:02:38 AM1/8/09
to fds-smv...@googlegroups.com
Updates:
Status: MoreInfo

Comment #2 on issue 601 by mcgratta: Fatal error in MPI_Waitall
http://code.google.com/p/fds-smv/issues/detail?id=601

Some of your meshes contain over 6 million cells. This cannot be run on a
32 bit
Windows machine. If your OS is 64 bit, are you using the 64 bit executable?

codesite...@google.com

unread,
Jan 8, 2009, 7:38:16 PM1/8/09
to fds-smv...@googlegroups.com

Comment #3 on issue 601 by kevin.c...@wfra.com.au: Fatal error in
MPI_Waitall
http://code.google.com/p/fds-smv/issues/detail?id=601

Yes. I am using the 64 bit executable. We have 2 machines running 64 bit
Windows
and we have done a number of other simulations over the 2 machines
successfully
using the 64 bit executable.

codesite...@google.com

unread,
Jan 9, 2009, 8:10:08 AM1/9/09
to fds-smv...@googlegroups.com
Updates:
Status: Accepted
Owner: randy.mcdermott
Cc: bryanwklein

Comment #4 on issue 601 by mcgratta: Fatal error in MPI_Waitall
http://code.google.com/p/fds-smv/issues/detail?id=601

OK, we have 64 bit linux machines and we'll try the case there. We don't
have 64 bit
Windows machines with enough RAM to run it. If we find a bug that is not
specific to
Windows, we'll fix it.

Reply all
Reply to author
Forward
0 new messages