Created attachment 474386 [details] output of yum upgrade Description of problem: env LANG=C yum -y upgrade Thunar was killed with SIGFPE as attached Version-Release number of selected component (if applicable): rpm-4.9.0-0.beta1.1.fc15.i686 python-2.7.1-4.fc15.i686 How reproducible: 100% Steps to Reproduce: 1. See attached 2. 3.
Created attachment 474387 [details] gdb log from generated core file
Interesting. It's obvious that it crashes on division by zero but how it comes to that... What does 'mount' say on this system?
Created attachment 474410 [details] result of mount You mean this?
Yup, thats the one. Can you also post the output of the reproducer case with rpm debugging enabled? Ie "LANG=C yum --rpmverbosity=debug -y upgrade Thunar" Do you get this crash on all rpm operations (I suppose yes) or is it somehow specific to Thunar?
(In reply to comment #4) > Can you also post the output of the reproducer case with > rpm debugging enabled? Ie "LANG=C yum --rpmverbosity=debug -y upgrade Thunar" It seems that this does not show any different outputs compared to my comment 0. Perhaps "--rpmverbosity=debug" works when actual transaction is run. By the way # find /var/cache/yum/ -name Thunar\*rpm -or -name exo\*rpm -or -name xfce4-panel\*rpm | xargs rpm -Uvh --test succeeds (i.e. without killed with SIGFPE) > Do you get this crash on all rpm operations (I suppose yes) or is it somehow > specific to Thunar? For this case (i.e. yum -y upgrade Thunar), currently 100% reproducible, however - As written above # rpm -Uvh --test succeeds - And actually after this issue occured I already installed python-debuginfo, rpm-debuginfo with # rpm -ivh without any error.
Yum behaving differently from rpm here seems highly peculiar and I haven't been able to reproduce it. I plugged the division by zero for now so the crash should be gone in rpm-4.9.0-0.beta1.3.fc15, but this wants further investigation (the current fix feels more like treating the symptom than the disease).
With rpm-4.9.0-0.beta1.3.fc15.i686 this issue seems no longer reproducible, thank you (with rpm-4.9.0-0.beta1.1.fc15.i686 it was still reproducible).
Hmm, I thought I'd already closed this one (been fixed for quite a while now)