Red Hat Bugzilla – Bug 32939
rpm 4.0.2-6x causes Seg fault on all operations
Last modified: 2007-04-18 12:32:20 EDT
From Bugzilla Helper:
User-Agent: Mozilla/4.72 [en] (X11; U; Linux 2.4.1 i686)
When I try to use rpm or up2date to do anything it returns with
Segmentation fault (core dumped)
Steps to Reproduce:
1.ran up2date and got the new rpm packages
2.submitted a bug report when 4.0.2 wouldn't run
3.downloaded the rpm-python package when it became available
4.rpm -Uvh on rpm-python
5.rpm now fails all commands
Actual Results: rpm will execute at all !!
it simply responds with:
Segmentation fault (core dumped)
(I don't know how to debug a core.)
Expected Results: I read the bug reports and installed rpm-python as
instructed from the ftp site but now nothing works !!
This would be fine if it was any other package but now I can't do anything
with packages since rpm is hosed. I consider this high priority since rpm
What exactly is failing?
Meanwhile, try doing "rpm --rebuilddb"
Thank you for your quick response. I should have been more explicit.
rpm commands that work and give expected results:
rpm -qi (on any rpm package)
rpm commands that produce a Segmentation fault
rpm -i (on any package)
rpm -Uvh (on anything)
rpm -Uvv (on any package)
the last command prints out all the dependency checks, opens the db index
Should I go and grab rpm-3.0.4 and then do a --force? I am afraid this won't
work bacause inbetween installing rpm-4.0.2 and rpm-python-4.0.2 a few days
later I ran rpm --rebuilddb successfully.
and then prints out "Preparing packages for installation" and then fails
I'm assuming that you are using rpm-3.0.x, rpm-4.0.x, on Red Hat 6.x.
OK, if --rebuilddb is segfaulting, then the file /var/lib/rpm/packages.rpm is
Hmm, you didn't perchance try Ximian and/or Nautilus recently did you?
Packages installed by those systems trigger a bug in rpm (yes the bug is in
rpm --rebuilddb -vv
Just before the segfault you may see the package that is causing the problem.
Try erasing that package and doing another --rebuilddb. Did that work?
Hmmm, that should have been "not rpm-4.0.x", but I see from the SUmmary
that you're trying to do that.
Are you using db1 (you have var/lib/rpm/packages.rpm) or db3 (you have
If you are already using db3, there's little reason to try to go back to
If you are using db3, what does
have to say?
Thanks for responding again. I had enough with my box, I had to do some
and couldn't wait to use rpm again. I had installed oracle on it and that
corrupted some files at
some point so it was due for a rebuild anyways. Yes, I use helix-gnome or
whatever it is
called, so maybe that caused the problems with rpm. Anyways, I have stayed up
last night and rebuilt and everything seems fine. Oracle is working, rpm is
working, and I
have Ximian on as well. By the way when is 7.1 coming out?
Thanks so much for you help !!
I had this same problem. I upgraded to 4.0.2 and rpm was segfaulting all over
the place. Even --rebuilddb failed.
The fix was to copy an old rpm 3.x binary to the machine and run --rebuilddb
with it. This is what happened:
[root@tmp4 lib]# /root/rpm --rebuilddb
record number 9194776 in database is bad -- skipping.
I immediately did a --rebuilddb with the 4.0.2 rpm. I was then able to use the
4.0.2 rpm without a problem.
So the new 4.0.2 is not as robust as the old 3.x one in handling database
problems in the db1 dbs
is what it looks like to me.