Bug 84067
Summary: | rhn-applet-gui segfault immediately | ||||||
---|---|---|---|---|---|---|---|
Product: | [Retired] Red Hat Public Beta | Reporter: | Harris Landgarten <harrisl> | ||||
Component: | rhn-applet | Assignee: | Daniel Veillard <veillard> | ||||
Status: | CLOSED NOTABUG | QA Contact: | |||||
Severity: | medium | Docs Contact: | |||||
Priority: | medium | ||||||
Version: | phoebe | CC: | cturner | ||||
Target Milestone: | --- | ||||||
Target Release: | --- | ||||||
Hardware: | i686 | ||||||
OS: | Linux | ||||||
Whiteboard: | |||||||
Fixed In Version: | Doc Type: | Bug Fix | |||||
Doc Text: | Story Points: | --- | |||||
Clone Of: | Environment: | ||||||
Last Closed: | 2003-02-13 14:23:02 UTC | Type: | --- | ||||
Regression: | --- | Mount Type: | --- | ||||
Documentation: | --- | CRM: | |||||
Verified Versions: | Category: | --- | |||||
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||
Cloudforms Team: | --- | Target Upstream Version: | |||||
Embargoed: | |||||||
Bug Depends On: | |||||||
Bug Blocks: | 79578 | ||||||
Attachments: |
|
Description
Harris Landgarten
2003-02-11 20:15:34 UTC
I tried updating to : rhn-applet-2.0.8-1 rhnlib-1.0-3 from rawhide with same results Is /usr/lib/python2.2/site-packages/bonobo empty? If not, this could be bug 79652 again (graphical up2date would also fail in this case, but text mode up2date could work). Created attachment 90031 [details]
strace of rhn-applet-gui with segfault
strace of rhn-applet-gui with segfault
[harrisl@d8100 harrisl]$ ls /usr/lib/python2.2/site-packages/bonobo activationmodule.so _bonobomodule.so __init__.py __init__.pyc __init__.pyo uimodule.so Graphical up2date works perfectly. I uploaded an strace. Even so it is still probably bug 79652 , as that directory should be empty. Read the instructions in that bug or upgrade gnome-python2-bonobo to 1.99.14-5 or later. I removed all files from /usr/lib/python2.2/site-packages/bonobo and upgraded to gnome-python2-bonobo-1.99.14-5 but am still getting the same segfault. When I upgraded to phoebe, Anaconda messed up and ignored the /usr is fstab, created a new /usr on / and wrote all content meant for /usr in the wrong place. Inorder to rescue the installation, I had to copy the wrong /usr on top of the old /usr so everything that wasn't overwritten is still there. Are there other left over files that could be causing this problem? There are certainly other files that are not part of RPMs but I don't want to do a wholesale rm -rf on everything rpm doesn't lay claim to. Are there specific directories that should not have any unpackaged files? This is getting way beyond my area of expertise, but if you have anything in /usr/lib/python2.2/site-packages/gtk (empty on my machine) that isn't claimed by a package (rpm -qf filename) try removing it or renaming it - I am guessing the stuff that was in that directory in 8.0 has been moved to .../gtk-2.0/gtk in phoebe. Hum, considering that 1/ you're the only one reporting the problem, 2/ the /usr copy do mess with timestamps 3/ that as a result you may have python compiled object more recent than their source counterparts then the fact that you may have serious troubles is not that surprizing, I highly suggest you restart a clean installation and report if you still see the problem. Obviously the state of your system is nearly impossible to reproduce, it is not the normal state generated by a normal install or upgrade path, and I think it's nearly impossible to fix just by getting symptoms reports, Daniel I removed all non owned packages from /usr/lib/python2.2 and that resolved the problem. rhn-applet-gui is running normally now. Thanks for the help. BTW the /usr problem was submitted by me as 83816. As far as I know it has not been resolved and given the problems you have pointed out, if not fixed it can really hose an upgrade and kill alot of time. Okay 83816 has a life of its own, we don't need to keep this bug open since apparently it was solved by trying to fix the correction measure taken against 83816. I think we can resonnably close this bug as not being a rhn-applet bug at this point, thanks for the quick feedback, Daniel |