Version-Release number of selected component: anaconda-18.13 Additional info: libreport version: 2.0.12 cmdline: /usr/bin/python /sbin/anaconda kernel: 3.6.0-2.fc18.ppc64
Created attachment 623670 [details] File: anaconda-tb
Created attachment 623671 [details] File: product
Created attachment 623672 [details] File: type
Created attachment 623673 [details] File: environ
Created attachment 623674 [details] File: storage.log
Created attachment 623675 [details] File: version
Created attachment 623676 [details] File: program.log
Created attachment 623677 [details] File: anaconda.log
Created attachment 623678 [details] File: syslog
Created attachment 623679 [details] File: hashmarkername
Created attachment 623680 [details] File: packaging.log
Created attachment 623681 [details] File: cmdline_file
Created attachment 623682 [details] File: release
Created attachment 623683 [details] File: description
1) Select "Software Selection," choose "Minimal install," select Done. 2) Select "Installation Destination," deselect disk, select Done. 3) Select "Installation Destination," select disk, select Done. 4) Select "Installation Destination," select Continue. 5) Select "Reclaim Space," select "/ (Fedora Linux 18 for ppc64)," change action to Delete, select "Reclaim Space." 6) Select "Reclaim Space," anaconda traps The anaconda updates.img file contains the following because I am unable to change the execution line number back to the dev = getDevicesByPath() line [root@bluebill anaconda-18.13-1]# git diff diff --git a/pyanaconda/ui/gui/spokes/lib/resize.py b/pyanaconda/ui/gui/spokes/lib/resize.py index fcccbeb..14b8982 100644 --- a/pyanaconda/ui/gui/spokes/lib/resize.py +++ b/pyanaconda/ui/gui/spokes/lib/resize.py @@ -120,6 +120,11 @@ class ResizeDialog(GUIObject): for part in disk.format.partitions: dev = self.storage.devicetree.getDeviceByPath(part.path) + if dev is None: + import pdb + pdb.set_trace() + self.storage.devicetree.getDeviceByPath(part.path) + if dev.isExtended and disk.format.logicalPartitions: continue > /tmp/updates/pyanaconda/storage/devicetree.py(2062)getDeviceByPath() -> if not path: (Pdb) print path /dev/sda5 (Pdb) print preferLeaves True (Pdb) path in [d.path for d in self._devices] False (Pdb) [d.path for d in self._devices] ['/dev/sda', '/dev/sda1', '/dev/sda2', '/dev/sda3', '/dev/sda4', '/dev/sr0', '/run/install/repo/LiveOS/squashfs.img', '/dev/loop0', '/LiveOS/rootfs.img', '/dev/loop1', '/overlay (deleted)', '/dev/loop2', '/dev/mapper/live-rw'] So it seems that the extended partition is no longer in _devices. [anaconda root@sharpie /]# fdisk -l /dev/sda Disk /dev/sda: 32.2 GB, 32212254720 bytes, 62914560 sectors Units = sectors of 1 * 512 = 512 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes Disk identifier: 0x90570900 Device Boot Start End Blocks Id System /dev/sda1 * 2048 10239 4096 41 PPC PReP Boot /dev/sda2 10240 1034239 512000 83 Linux /dev/sda3 1034240 9324543 4145152 82 Linux swap / Solaris /dev/sda4 9324544 62914559 26795008 5 Extended /dev/sda5 9326592 62914559 26793984 83 Linux
Different traceback, but same underlying cause. *** This bug has been marked as a duplicate of bug 862972 ***
Different traceback, different fix.
With a multipath disk only and anaconda-18.14-1.fc18, I immediately hit this exception. Traceback (most recent call first): File "/usr/lib64/python2.7/site-packages/pyanaconda/ui/gui/spokes/lib/resize.py", line 123, in populate if dev.isExtended and disk.format.logicalPartitions: File "/usr/lib64/python2.7/site-packages/pyanaconda/ui/gui/spokes/lib/resize.py", line 167, in refresh self.populate(disks) File "/usr/lib64/python2.7/site-packages/pyanaconda/ui/gui/spokes/storage.py", line 665, in on_continue_clicked resizeDialog.refresh(disks) AttributeError: 'NoneType' object has no attribute 'isExtended' > /usr/lib64/python2.7/site-packages/pyanaconda/ui/gui/spokes/lib/resize.py(123)populate() -> if dev.isExtended and disk.format.logicalPartitions: (Pdb) print part.path /dev/mapper/mpatha1 (Pdb) [d.path for d in self.storage.devicetree._devices] ['/dev/sr0', '/dev/sda', '/dev/sdb', '/dev/sdc', '/dev/sdd', '/run/install/repo/LiveOS/squashfs.img', '/dev/loop0', '/LiveOS/rootfs.img', '/dev/loop1', '/overlay (deleted)', '/dev/loop2', '/dev/mapper/mpatha', '/dev/mapper/live-rw'] (Pdb) [part.path for part in disk.format.partitions] ['/dev/mapper/mpatha1', '/dev/mapper/mpatha2', '/dev/mapper/mpatha3', '/dev/mapper/mpatha4', '/dev/mapper/mpatha5']
It still traps with: https://lists.fedorahosted.org/pipermail/anaconda-patches/2012-October/001512.html https://lists.fedorahosted.org/pipermail/anaconda-patches/2012-October/001513.html applied
Please double check that you are, in fact, using an updates image with the correct files. And if that is the case, please attach the traceback to this bug.
I have verified that I am using an updates image with: https://lists.fedorahosted.org/pipermail/anaconda-patches/2012-October/001512.html https://lists.fedorahosted.org/pipermail/anaconda-patches/2012-October/001513.html which reference one file and has been applied. /tmp/updates/pyanaconda/ui/gui/spokes/storage.py does have the patches applied.
Created attachment 624278 [details] current traceback
I tried to reclaim space by deleting existing partitions Package: anaconda-18.14 Architecture: i686 OS Release: Fedora release 18-Beta-TC3
Proposing as Beta blocker, I hit this with a standard harddisk, no fancy stuff.
Mark - multipath is completely untested territory as far as I know, so there could very well be more multipath-specific things going on there.
*** Bug 864718 has been marked as a duplicate of this bug. ***
I have tried to reclaim space by using reclaim dialog, "deleting" partitions. I returned me into "Choose local disk", so I clicked "continue" and boom goes the dynamite. Package: anaconda-18.14 Architecture: x86_64 OS Release: Fedora release 18-Beta-TC3
anaconda-18.15-1.fc18 has been submitted as an update for Fedora 18. https://admin.fedoraproject.org/updates/anaconda-18.15-1.fc18
Went through 'guided partitioning' to clear space on a system into which I'd done a default autopart F18 beta tc3 install to an empty 15GB disk. I was able to set action to 'delete' for all partitions and 'finish' the dialog, then it went back to the 'you don't have enough space but we can help you reclaim it' screen. I think I hit 'reclaim space' again on that screen, and it crashed. Package: anaconda-18.14 OS Release: Fedora release 18-Beta-TC3
Package anaconda-18.15-1.fc18: * should fix your issue, * was pushed to the Fedora 18 testing repository, * should be available at your local mirror within two days. Update it with: # su -c 'yum update --enablerepo=updates-testing anaconda-18.15-1.fc18' as soon as you are able to. Please go to the following url: https://admin.fedoraproject.org/updates/FEDORA-2012-15903/anaconda-18.15-1.fc18 then log in and leave karma (feedback).
Discussed at 2012-10-11 blocker review meeting: http://meetbot.fedoraproject.org/fedora-qa/2012-10-11/f18beta-blocker-review-3.1.2012-10-11-16.04.log.txt . Accepted as a blocker per Alpha criterion "The installer must be able to complete an installation using automatic partitioning to any sufficiently large target disk, whether unformatted, empty, or containing any kind of existing data" - it's possible to hit this with a pretty simple attempt to just install to an existing populated disk by removing all the partitions on the populated disk (that's what I was doing in comment #29).
anaconda-18.16-1.fc18 has been submitted as an update for Fedora 18. https://admin.fedoraproject.org/updates/anaconda-18.16-1.fc18
*** Bug 865737 has been marked as a duplicate of this bug. ***
The bug as described in bug marked as duplicate 864718 still persists in smoke9. if getattr(dev, 'isExtended', False) and disk.format.logicalPartitions: would probably axe the rest of the issues but I suppose the key questions is: must a dev have isExtended attribute? BTRFSVolumeDevice seems not to. Not always, anyway
Chris, does #34 sound like this fix didn't work, or rather that we should re-opne 864718?
I need to see data to be able to tell.
I am unable to test the fix because of bug 865009, I'll test with TC5.
I've some technical difficulties providing the stack traces since report bug button crashes Anaconda. However, on painful looking, appears that in fact my stack trace doesn't mention BTRFSVolumeDevice but BTRFSSubvolumeDevice. Otherwise the stack trace is identical. The partitioning sda1 Ext4, sda2 Btrfs, sda3 swap, all primary, no extended partitions. BTRFSSubvolumeDevice will not get attribute isExtended. Reproducible always on my system. I ended up in this situation in the first place by installing with Fedora TC3 with Btrfs enabled.
I just read some source code: StorageDevice->PartitionDevice (PartitionDevice defines isExtended) StorageDevice->BTRFSDevice->BTRFSVolumeDevice->BTRFSSubvolumeDevice (no isExtended exists because we inherit StorageDevice, not PartitionDevice) This would explain the really botched behaviour. Source code is from anaconda 18.16. Could someone confirm? Other devices should be checked for incorrect inheritance as well.
anaconda-18.17-1.fc18 has been submitted as an update for Fedora 18. https://admin.fedoraproject.org/updates/anaconda-18.17-1.fc18
anaconda-18.18-1.fc18 has been submitted as an update for Fedora 18. https://admin.fedoraproject.org/updates/anaconda-18.18-1.fc18
Confirming these problems no longer occur for me in smoke11.
setting VERIFIED then, thanks.
anaconda-18.19-1.fc18 has been submitted as an update for Fedora 18. https://admin.fedoraproject.org/updates/anaconda-18.19-1.fc18
Package anaconda-18.19-1.fc18: * should fix your issue, * was pushed to the Fedora 18 testing repository, * should be available at your local mirror within two days. Update it with: # su -c 'yum update --enablerepo=updates-testing anaconda-18.19-1.fc18' as soon as you are able to. Please go to the following url: https://admin.fedoraproject.org/updates/FEDORA-2012-16528/anaconda-18.19-1.fc18 then log in and leave karma (feedback).
I haven't seen this with TC6. Seppo confirmed fix as well, closing this bug. If someone else still hits this, please reopen.