The weird bug around changing partitions is back, reproducible in autopilot (Cannot create JFS partition)

[Dan Chapman ]

Found while running daily autopilot ubiquity iso tests.

https://jenkins.qa.ubuntu.com/view/Ubiquity/view/Lubuntu/job/ubiquity_ap-lubuntu_devel_daily-test_custom_install/ARCH=amd64,label=rabisu/lastCompletedBuild/testReport/ubiquity_autopilot_tests.tests.test_custom_install/CustomInstallTestCase/test_custom_install/

When running the tests locally it turns out this is actually a error when creating a JFS partition (see attached screenshot)

Steps to reproduce

1) branch lp:ubiquity
2) make sure qemu and bsdtar are installed and have a daily lubuntu iso
3) from dir ./ubiquity/autopilot/ubiquity-autopilot-runner execute using the configfile attached to this bug

    ./run-ubiquity-test -T ~/path/to/testrunner.cfg --test ubiquity_autopilot_tests.tests.test_custom_install ~/path/to/trusty-desktop.iso

4) test artifacts are stored in /tmp/ubiquity.tests

If you want additional artifacts or additional autopilot options or even stop the VM shutting down for debugging you can change the config file accordingly

You can also use the --sdl param if you want to view the desktop. All other test runner info can be found in the README if needed

If any other info is needed contact myself or jibel

ProblemType: Bug
DistroRelease: Ubuntu 14.04
Package: ubiquity 2.17.1 [modified: lib/partman/automatically_partition/question]
ProcVersionSignature: Ubuntu 3.12.0-7.15-generic 3.12.4
Uname: Linux 3.12.0-7-generic i686
ApportVersion: 2.12.7-0ubuntu2
Architecture: i386
CasperVersion: 1.336ubuntu1
CurrentDesktop: LXDE
Date: Tue Dec 17 13:07:08 2013
InstallCmdLine: file=/cdrom/preseed/lubuntu.seed boot=casper initrd=/casper/initrd.lz quiet splash --
LiveMediaBuild: Lubuntu 14.04 "Trusty Tahr" - Alpha i386 (20131216)
SourcePackage: ubiquity
UpgradeStatus: No upgrade log present (probably fresh install)

[Dan Chapman  (dpniel)]

Screenshot of error

[Dan Chapman  (dpniel)]

The debug config file to reproduce

[Brian Murray (brian-murray)]

There is actually a kernel issue found in the syslog:

Dec 17 13:05:57 lubuntu kernel: [ 153.359697] FAT-fs (sda5): Can't find a valid FAT filesystem
Dec 17 13:05:57 lubuntu kernel: [ 153.366565] ISOFS: Unable to identify CD-ROM format.
Dec 17 13:05:57 lubuntu kernel: [ 153.366818] SQUASHFS error: Can't find a SQUASHFS superblock on sda5
Dec 17 13:05:57 lubuntu kernel: [ 153.367158] EXT2-fs (sda5): error: can't find an ext2 filesystem on dev sda5.
Dec 17 13:05:57 lubuntu kernel: [ 153.368768] XFS (sda5): bad magic number
Dec 17 13:05:57 lubuntu kernel: [ 153.368775] da204000: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
Dec 17 13:05:57 lubuntu kernel: [ 153.368778] da204010: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
Dec 17 13:05:57 lubuntu kernel: [ 153.368780] da204020: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
Dec 17 13:05:57 lubuntu kernel: [ 153.368782] da204030: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
Dec 17 13:05:57 lubuntu kernel: [ 153.368785] XFS (sda5): Internal error xfs_sb_read_verify at line 637 of file /build/buildd/linux-3.12.0/fs/xfs/xfs_sb.c. Caller 0xf923520c
Dec 17 13:05:57 lubuntu kernel: [ 153.368785]
Dec 17 13:05:57 lubuntu kernel: [ 153.368790] CPU: 0 PID: 5 Comm: kworker/0:0H Tainted: G C 3.12.0-7-generic #15-Ubuntu
Dec 17 13:05:57 lubuntu kernel: [ 153.368793] Hardware name: innotek GmbH VirtualBox/VirtualBox, BIOS VirtualBox 12/01/2006
Dec 17 13:05:57 lubuntu kernel: [ 153.368838] Workqueue: xfslogd xfs_buf_iodone_work [xfs]
Dec 17 13:05:57 lubuntu kernel: [ 153.368842] 00000000 00000000 f549de80 c164cc9c e62bf800 f549dea4 f92383cb e62bf800
Dec 17 13:05:57 lubuntu kernel: [ 153.368846] 00000008 f92af524 f92a9f00 0000027d f92b1b20 f923520c f549dec4 f9238418
Dec 17 13:05:57 lubuntu kernel: [ 153.368849] f92b1b20 0000027d f923520c eb020c00 00000016 e62bf800 f549def0 f928b7f6
Dec 17 13:05:57 lubuntu kernel: [ 153.368852] Call Trace:
Dec 17 13:05:57 lubuntu kernel: [ 153.368863] [<c164cc9c>] dump_stack+0x41/0x52
Dec 17 13:05:57 lubuntu kernel: [ 153.368881] [<f92383cb>] xfs_error_report+0x4b/0x50 [xfs]
Dec 17 13:05:57 lubuntu kernel: [ 153.368897] [<f923520c>] ? xfs_buf_iodone_work+0x6c/0xc0 [xfs]
Dec 17 13:05:57 lubuntu kernel: [ 153.368914] [<f9238418>] xfs_corruption_error+0x48/0x60 [xfs]
Dec 17 13:05:57 lubuntu kernel: [ 153.368930] [<f923520c>] ? xfs_buf_iodone_work+0x6c/0xc0 [xfs]
Dec 17 13:05:57 lubuntu kernel: [ 153.368956] [<f928b7f6>] xfs_sb_read_verify+0x106/0x120 [xfs]
Dec 17 13:05:57 lubuntu kernel: [ 153.368972] [<f923520c>] ? xfs_buf_iodone_work+0x6c/0xc0 [xfs]
Dec 17 13:05:57 lubuntu kernel: [ 153.368976] [<c10812da>] ? try_to_wake_up+0x14a/0x240
Dec 17 13:05:57 lubuntu kernel: [ 153.368992] [<f923520c>] xfs_buf_iodone_work+0x6c/0xc0 [xfs]
Dec 17 13:05:57 lubuntu kernel: [ 153.368996] [<c106bd1b>] process_one_work+0x11b/0x3a0
Dec 17 13:05:57 lubuntu kernel: [ 153.368998] [<c1069ce5>] ? start_worker+0x25/0x30
Dec 17 13:05:57 lubuntu kernel: [ 153.369001] [<c106c45d>] ? manage_workers.isra.28+0x1ad/0x270
Dec 17 13:05:57 lubuntu kernel: [ 153.369003] [<c106c...

[Dimitri John Ledkov (xnox)]

Thanks a lot for hunting this down.

Last time around it was discovered that is it "normal" for xfs module to fail like this when trying to mount a non-xfs partition. Usually modules/mounting is tried in order and therefore it's very rare when xfs module is presented with !xfs filesystem.

This has happened before, during one of the previous releases, and at the time it was attributed somewhere in partman not waiting long enough for partition table / kernel to settle whilst doing the partitioning. I'll assign this to Colin, as I think he ended up patching this last time around.

[Marcus Tomlinson (marcustomlinson)]

This release of Ubuntu is no longer receiving maintenance updates. If this is still an issue on a maintained version of Ubuntu please let us know.

[Launchpad Janitor (janitor)]

[Expired for ubiquity (Ubuntu) because there has been no activity for 60 days.]