On Monday 24 March 2008, Michael Schwendt wrote: >On Mon, 24 Mar 2008 00:01:52 -0400, Gene Heskett wrote: >> I've had several hard lockups over the last day, possibly related to a >> tickless kernel2.6.24.1 I could keep running long enough to recompile it >> with the usual kilohertz tick, so ATM I'm on one of the older 2.6.23 >> kernels from the F8 repo's, and kde is 3.5.9 >> >> Anyway, kmail (and kontact but I don't use it) is now doing a nearly >> instant Signal 11. The kmail.kcrash report sort of points a finger at >> libkmailprivate.so, but I've stopped X, had yum rip out kdepim and >> kdepim-libs, and then re-install them with no effect on the crash. >> >> So here is the kmail.kcrash text: >> Using host libthread_db library "/lib/libthread_db.so.1". >> (no debugging symbols found) >> [snip many lines oof this] >> (no debugging symbols found) >> [Thread debugging using libthread_db enabled] >> [New Thread -1208654128 (LWP 25523)] >> (no debugging symbols found) >> [snip many more lines] >> (no debugging symbols found) > >This means that you should install the corresponding -debuginfo rpms to >fill in the missing details in the backtrace. > Where might those be found?, yum shows nothing. I have found a bad set of .index files for the debian-users mailing list folder, and it appears kmail crashed while regenerating it. Now its a name only, not even an inode attached that I can find, and un-deletable. HOWEVER, kmail is now running again after I manually deleted all other references to the debian-user folder except this non-existent .debian-user.index file. I had tried to delete all the *.index files after kmail first crashed because that was the trigger 2 or 3 times before, so when I ram kmail this time, it was several minutes, showing as high as 50% cpu in an htop display, before it ever opened its operating screens. I tried booting the Dec-2007 respin in rescue mode, but then cannot umount /mnt/sysimage so that I can safely run a session of e2fsck against /dev/sda1, but its busy and cannot be unmounted, so: How can I force an fsck on /dev/VolGroupVol00 on the next boot? In the meantime I have built, and rebooted to, a 2.6.24.1 kernel with that 'tickless' BS turned off, and while its a bit early to state it without reservation, the system does feel a heck of a lot more responsive, I was having pauses in the whole system for 30 seconds at a time like some IRQ wasn't being serviced in a timely manner. I am purposely making this message longer to see if it hangs while I'm typing, and so far in this messages editing it has not hung once like it was before. Now its off to go read up on tune2fs & see if it can set that check me flag on /dev/Vol* or /dev/sda1 Thanks for the reply, this has been an interesting trip indeed... >Do you realise that it could be a segfault caused by corrupted input data >and not by corrupted packages? Depending on what libkmailprivate does in >this area of the code, it could be that it doesn't verify the input data >enough as would be necessary to prevent it from crashing. Should I, when things settle, file a bugzilla against libkmailprivate.so? Or whatever it is that manages the .index files? -- Cheers, Gene "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) It has been said that man is a rational animal. All my life I have been searching for evidence which could support this. -- Bertrand Russell
