Red Hat Bugzilla – Bug 59726
up2date segfaults on e2fs progs
Last modified: 2015-01-07 18:54:56 EST
Description of Problem:
I'm using Red Hat 7.0, with new up2date, rhn, rpm, glibc, db3, popt,
python-xmlrpc, python-popt packages. When I try to use either up2date -u or
up2date e2fsprogs, up2date exits with a segmentation fault and a core dump.
Here's a few lines from the output
Getting headers for available packages...
Segmentation fault (core dumped)
I've attached a stack trace acquired by doing 'gdb -c core /usr/bin/python',
then 'thread apply all bt'. This seems quite consistent, and is on a fresh
(installed tonight) Red Hat 7.0 system.
Version-Release number of selected component (if applicable):
[root@basinstall /root]# rpm -q up2date
Created attachment 45361 [details]
My guess is a corrupt header in /var/spool/up2date.
Try removing all the .hdr files from there and trying again.
Also, the next version of the client attempts to do more
sanity checking on a header before attempting to load it,
so this shouldnt be an issue anymore.
Think issues similar to this should be cleared up
in all currently supported releases (aka, up2date
does more sanity checks, and rpm is much more
robust in the face of wacky headers, so closing