Description of problem: Version-Release number of selected component: anaconda-18.6 Additional info: libreport version: 2.0.12 cmdline: initrd=initrd.img quiet rd.live.check BOOT_IMAGE=vmlinuz kernel: 3.6.0-0.rc1.git6.1.fc18.x86_64
Created attachment 606567 [details] File: program.log
Created attachment 606568 [details] File: environ
Created attachment 606569 [details] File: type
Created attachment 606570 [details] File: storage.log
Created attachment 606571 [details] File: release
Created attachment 606572 [details] File: product
Created attachment 606573 [details] File: anaconda.log
Created attachment 606574 [details] File: syslog
Created attachment 606575 [details] File: hashmarkername
Created attachment 606576 [details] File: version
Created attachment 606577 [details] File: anaconda-tb
Created attachment 606578 [details] File: description
I have seen this twice on a bare metal machine using Fedora 18 Alpha TC3 i386 DVD. It appears when entering the "hub" screen. I saw this bug twice and bug 851212 once, out of three attempts. That means I wasn't able to use the installer successfully even once. Proposing as an Alpha blocker.
(In reply to comment #13) > I have seen this twice on a bare metal machine using Fedora 18 Alpha TC3 > i386 DVD. ^^^^^ x86_64
Discussed at 2012-08-27 QA meeting, acting as a blocker review meeting. The traceback seems pretty unequivocal that some codepath in anaconda does a definitely wrong thing which will always result in a crash, so we considered this a clear enough crasher to be worthy of blocker status. We'll cite criterion "The installer must be able to complete package installation with the default package set for each supported installation method" for the record, but a crash this early in installation really infringes any number of criteria.
Kamil, do you have IPv6 configured in your testing environment? I.e. radvd running. This might be one of the causes that I can't reproduce the issue.
I have the machine prepared for you, stop by anytime and I can show it live.
This bug seems to be fixed in TC5, I'm unable to reproduce it now (using the same machine as Kamil).
I can confirm I can no longer hit this issue using the same machine with TC5 DVD. Closing as fixed.