From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-GB; rv:1.4) Gecko/20030703
Description of problem:
At the time this happened my system wasn't entirely usable, and I was
upgrading (partly) in order to make the MBR point at linux again (a
windows xp setup had overwritten it), so I wasn't too concerned with
recording the error and this bug report is from memory, my apologies.
While upgrading from FC 0.96 x86_64 to FC1 x86_64 final, bootloader.py
failed on the following line, complaining that kernelLabel was undefined.
kernelList.append((kernelLabel, kernelLongLabel, version))
Dropping into the python debugger "print kernelLabel" and "print
kernelLongLabel" showed they were both undefined. IIRC version was
Before the upgrade to FC1 final I had installed the 2174 kernel
released as an update to the 0.96 test release, and _I_think_ my
grub.conf included both the original 2135 kernel from the 0.96 release
and the 2174 update kernel (although it might have only had the 2174
update). The MBR had been overwritten by WinXP though, so I couldn't
boot any linux kernel.
After rebooting from the CD in rescue mode I did a manual grub-install
and it worked fine. At this point grub.conf had only one item, the
Version-Release number of selected component (if applicable):
The version on the FC1 x86_64 CDs.
Steps to Reproduce:
This is what I did when the bug happened, I haven't tried to reproduce
1. Install Fedora Core 0.96 for x86_64
2. Install 2174 update kernel.
3. Upgrade to FC1 for x86_64.
Actual Results: After installing packages bootloader.py script fails
to setup bootloader.
Expected Results: Boot loader installs and installation continues.
I forgot to mention, this was during a text-mode install/upgrade (my
video card isn't recognised by FC1)
*** Bug 118232 has been marked as a duplicate of this bug. ***
Upgrades from a beta/test release to a final release are not
supported. If you attach the complete traceback, I can take a look at
it, though (without the complete traceback, there's not much I can
really do, though other than guessing randomly)
There's a complete traceback attached to bug 118232.
I realise the upgrade path I took was unsupported, but I thought I'd
log the bug anyway as I didn't know whether it was a direct result of
upgrading from the test release or not.
The traceback on bug 118232 does look like the same error as mine,
thanks mrsam. My package selection was different, of course. Now that
I've got the box playing nicely I'll check tonight whether the install
log has some more useful information.
*** Bug 118360 has been marked as a duplicate of this bug. ***
*** Bug 122153 has been marked as a duplicate of this bug. ***
Haven't reproduced but put a fix that should fix this in CVS