Sounds good. I'll try to do it this week. I'll let you know if I run into anything.
Brian ----- Original Message ----- > Oh an upon closer inspection, the patches I put into trunk a little > while back that fixed my segfaulting problems actually include a check > similar to yours. It covers all 4 major arrays in that function. CRE, > JRE, ORE and NRE. I also changed the mallocs for those arrays to > callocs so that they were properly initialized to zero. So your > suggested changes are already in trunk. Thanks again for the patch. > > Brian: Any chance we can get trunk rolled to a new release on the > site? > There has been a handful of people with these segfault issues and a > new point release might be a good idea. > > -- > Jason > > Alexander Oltu wrote: > > We tried version from trunk. It is not segfaulting without applying > > JRE > > not null check. > > > > Best regards, > > Alex. > > > > On Tue, 22 Feb 2011 14:57:43 -0500 > > Jason Williams wrote: > > > >> Alexander, > >> > >> This looks like it might help catch problems when JRE somehow comes > >> up as NULL. That's definitely good. I'll try to get to merging this > >> into trunk within the next day or so. Have you tried using the code > >> from SVN trunk? It also has a few patches of mine to fix some > >> segfaults as well. I'm also curious what your .cfg file looks like. > >> Any chance you can copy that to the list, with the proper bits for > >> your site scrubbed (ie. ip addresses and user/group names) > >> > >> -- > >> Jason > >> > >> On 2/22/2011 11:55 AM, Alexander Oltu wrote: > >>> Brian, > >>> > >>> Thanks! Below is what I send to bugzilla, patched is attached: > >>> > >>> With our maui.cfg file we get: > >>> > >>> Program terminated with signal 11, Segmentation fault. > >>> #0 MResAdjustDRes (JName=0x0, ForceEval=0) at MRes.c:7447 > >>> 7447 JRE[jreindex].Type = mreNONE; > >>> (gdb) bt > >>> #0 MResAdjustDRes (JName=0x0, ForceEval=0) at MRes.c:7447 > >>> #1 0x000000000042a45e in MSchedProcessJobs (OldDay=0x7fff6e7c9d50 > >>> "", GlobalSQ=0x7fff6e7c3d40, GlobalHQ=0x7fff6e7bfd40) at > >>> MSched.c:6955 #2 0x000000000040596d in main (ArgC=1, ArgV=<value > >>> optimized out>) at Server.c:192 > >>> (gdb) print JName > >>> $1 = 0x0 > >>> (gdb) print JRE > >>> $2 = (mre_t *) 0x0 > >>> > >>> It is not reproducible with default cfg file. > >>> > >>> The patch attached we are using to workaround this issue. > >>> > >>> What this segfault is related to and is our solution OK? > >>> > >>> PS. We can provide coredump and binary. > >>> > >>> Thanks, > >>> Alex. > >>> > >>> On Tue, 22 Feb 2011 09:16:54 -0700 (MST) > >>> Brian Christiansen wrote: > >>> > >>>> Alex, > >>>> > >>>> You can send your patch to the mailing list and either I or one > >>>> of > >>>> the others that have check-in rights can review it and check it > >>>> in. > >>>> Or you can request check-in rights for yourself. Bugzilla is > >>>> currently only being used for Torque. > >>>> > >>>> http://www.clusterresources.com/pipermail/mauiusers/2010-February/004145.html > >>>> > >>>> Thanks, > >>>> Brian > >>>> > >>>> ----- Original Message ----- > >>>>> Hello list, > >>>>> > >>>>> I am trying to fill a MAUI bug with a patch, but can't find the > >>>>> proper place. I found http://www.clusterresources.com/bugzilla/ > >>>>> . > >>>>> Looks like it has only Torque there. I already opened a bug > >>>>> there, > >>>>> as #114 and got > >>>>> no response. > >>>>> > >>>>> Does anybody know where we submit patches and bugs for MAUI? > >>>>> > >>>>> Thanks, > >>>>> Alex. > >>>>> _______________________________________________ > >>>>> mauiusers mailing list > >>>>> [email protected] > >>>>> http://www.supercluster.org/mailman/listinfo/mauiusers > >>> > >>> _______________________________________________ > >>> mauiusers mailing list > >>> [email protected] > >>> http://www.supercluster.org/mailman/listinfo/mauiusers > > > > _______________________________________________ mauiusers mailing list [email protected] http://www.supercluster.org/mailman/listinfo/mauiusers
