Description of problem: This is a problem similar to the problem that I encountered and earlier reported a bug regarding. FC7 wants to initialize a disk which has XP and FC6 successfully installed on it. The problem was earlier re-assigned to parted as the culpret. The bug was pushed around awhile and some ideas were tried and I tested some versions of parted on my working FC6 installation and still get this dump. It is still here for me and probably will see more episodes since FC7 is now released. Version-Release number of selected component (if applicable): Anaconda on the i386 DVD for FC7 How reproducible: Start installer and do not allow anaconda to initialize the working disk. Get error when continuing on with the installation. Steps to Reproduce: 1. Start installer 2. Say no to initializing /dev/sda 3. Continue installation 4. Get exception 5. Save exception to floppy 6. Get prompt to reboot system. Actual results: FC7 thinks the geometry on /dev/sda is wrong. The detection of the disk is correct for FC6 and XP and shows no errors. Expected results: FC7 to realize that the disk is alright and to continue without the exception. Additional info:
Created attachment 155960 [details] This is what anaconda wanted to reboot because of.
To clarify what each lvm is in dump file. This is the / partition for FC6 22:10:00 INFO : lv is hda5vol/LogVol00, size of 13056 This is the swap partition for FC6 22:10:00 INFO : lv is hda5vol/LogVol01, size of 544 There should be no problem mounting the LVM for / (FC6) 22:10:00 DEBUG : isys.py:mount()- going to mount /dev/hda5vol/LogVol00 on /mnt/sysimage
Is this the same as: bug 241249 ? referenced at: http://fedoraproject.org/wiki/Bugs/F7Common#head-65f561a07fdf2889f310f3aac0ab0f984faf3e9c
no it is bug 229745 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=229745
Thanks David for pointing out the reason for this error. I resorted to a yum localupdate method to upgrade this system so I will post the excerpt into this bug report for future reference. I saw this error spit out from a rawhide installation for a long time now. boot: linux libata.ignore_hpa=1 Those upgrading from FC6 or earlier via yum on impacted systems can add a line to modprobe.conf along these same lines to allow an F7 kernel to boot their system. Ex: options libata ignore_hpa=1
Created attachment 155964 [details] modprobe.conf - added suggested line to file Since I am using yum because of the installer problem, I added the line so I hopefully can boot the system after the upgrade. Is this correct?
Yep, that's correct. Unfortunately, given the kernel interface, there's no way for us to really probe and do this automatically :(
I ended up getting a kernel panic with the entry in modprobe.conf and ended up removing the last partition while booted into development using fdisk. I then recreated an extended partition and added two partitions, one for / and the other for swap. I could then run the installer with success without getting an exception. Now the disk does not show errors in either development or F7. I could not confirm whether adding to modprobe.conf works or not. The disk no longer needs the entry to work.