Yes I go through the source there are many many sprintf that try to do the same:-(
Sent from my iPad On Feb 18, 2011, at 2:05 AM, Ake Sandgren <[email protected]> wrote: > On Thu, 2011-02-17 at 16:10 -0800, Sergey Blagodurov wrote: >> Hello, >> >> I've been trying to resolve the problem that is described in the >> following threads: >> http://www.supercluster.org/pipermail/torqueusers/2010-October/011518.html >> http://comments.gmane.org/gmane.comp.clustering.torque.user/9112 >> >> Basically, PBS_NODEFILE was incomplete: it contained only one node >> when job was submitted to run on several nodes. None of the solutions >> proposed there worked for my installation. I've ended up looking into >> maui source and found out that the following piece of code: >> >> sprintf(TSBuf,"%s%s:ppn=%d", >> TSBuf, >> tmpHostName, >> NL[tindex].TC); >> >> in the file /src/moab/MPBSI.c from Maui 3.3 distribution does not work >> as it supposed to: instead of appending to TSBuf it rewrote TSBuf with >> the new value. This thread: > > The behaviour of this sprintf is not defined in the C spec. The compiler > is allowed to handle it anyway it feels like. > > The line is simply incorrect. Look at the maui wrappers for sprintf > (can't remember the name at the moment), or simply rewrite it as > sprintf(TSBuf[strlen(TSBuf)], "%s:ppn=%d", tmpHostName... > > I've sent patches for LOTS of these in the past but i gave up since > there are too many and people insist of writing incorrect code like that > all the time for some reason. > > -- > Ake Sandgren, HPC2N, Umea University, S-90187 Umea, Sweden > Internet: [email protected] Phone: +46 90 7866134 Fax: +46 90 7866126 > Mobile: +46 70 7716134 WWW: http://www.hpc2n.umu.se > > _______________________________________________ > mauiusers mailing list > [email protected] > http://www.supercluster.org/mailman/listinfo/mauiusers _______________________________________________ mauiusers mailing list [email protected] http://www.supercluster.org/mailman/listinfo/mauiusers
