[MPICH] Bug in mpiexec_rsh.c builds wrong command line
Jayesh Krishna
jayesh at mcs.anl.gov
Mon Jan 28 15:55:01 CST 2008
Hi,
Great. We will be including the patch in the upcoming release, 1.0.7, and
will update you when the release is available (in couple of weeks).
Regards,
Jayesh
-----Original Message-----
From: Reuti [mailto:reuti at staff.uni-marburg.de]
Sent: Monday, January 28, 2008 3:47 PM
To: Jayesh Krishna
Cc: mpich-discuss at mcs.anl.gov
Subject: Re: [MPICH] Bug in mpiexec_rsh.c builds wrong command line
Hi,
Am 28.01.2008 um 21:13 schrieb Jayesh Krishna:
> After fixing the bug (mentioned below) does SGE work with the latest
> version of MPICH2 (Is there any other problems that you see with
> MPICH2+SGE)
> ?
yes, it's working after the patch. For now I'm not aware of additional
problems to integrate in in SGE.
-- Reuti
> Regards,
> Jayesh
>
> -----Original Message-----
> From: owner-mpich-discuss at mcs.anl.gov
> [mailto:owner-mpich-discuss at mcs.anl.gov] On Behalf Of Jayesh Krishna
> Sent: Friday, January 25, 2008 5:19 PM
> To: 'Reuti'
> Cc: mpich-discuss at mcs.anl.gov
> Subject: RE: [MPICH] Bug in mpiexec_rsh.c builds wrong command line
>
> Hi,
> Thanks for reporting the bug.
> This bug with be fixed in the next release.
>
> Regards,
> Jayesh
> -----Original Message-----
> From: owner-mpich-discuss at mcs.anl.gov
> [mailto:owner-mpich-discuss at mcs.anl.gov] On Behalf Of Reuti
> Sent: Friday, January 25, 2008 4:33 PM
> To: mpich-discuss at mcs.anl.gov
> Subject: [MPICH] Bug in mpiexec_rsh.c builds wrong command line
>
> Hi,
>
> one user of SGE noticed, that the daemonless smpd startup isn't
> working any longer in MPICH2 1.0.6p1 (http://gridengine.sunsource.net/
> howto/mpich2-integration/mpich2-integration.html). After he and I
> digged in the code, we found in mpich2-1.0.6p1/src/pm/smpd/
> mpiexec_rsh.c a bug. Line
> 472 must be changed from:
>
> pExe += curLen;
>
> to
>
> pExe = process->exe + curLen;
>
> Otherwise the string will not correctly be built, as curLen is added
> to pExe again, which is already pointing to a location somewhere in
> the middle of the string.
>
>
> -- Reuti
>
>
More information about the mpich-discuss
mailing list