Re: [lkp] [xfs] a45086e27d: -100.0% xfstests.xfs.033.seconds

From: Brian Foster
Date: Wed Nov 18 2015 - 09:32:13 EST


cc linux-xfs

On Wed, Nov 18, 2015 at 02:57:21PM +0800, kernel test robot wrote:
> FYI, we noticed the below changes on
>
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master
> commit a45086e27dfa21a4b39134f7505c8f60a3ecdec4 ("xfs: validate metadata LSNs against log on v5 superblocks")
>
> 2015-11-17 05:11:29 export TEST_DIR=/fs/sdd1
> 2015-11-17 05:11:29 export TEST_DEV=/dev/sdd1
> 2015-11-17 05:11:29 export FSTYP=xfs
> 2015-11-17 05:11:29 export SCRATCH_MNT=/fs/scratch
> 2015-11-17 05:11:29 mkdir /fs/scratch -p
> 2015-11-17 05:11:29 export SCRATCH_DEV=/dev/sdg1
> 2015-11-17 05:11:29 export SCRATCH_LOGDEV=/dev/sdh1
> 2015-11-17 05:11:29 ./check xfs/004 xfs/010 xfs/011 xfs/013 xfs/017 xfs/020 xfs/026 xfs/027 xfs/028 xfs/031 xfs/032 xfs/033 xfs/041 xfs/042 xfs/046 xfs/047 xfs/049 xfs/050 xfs/054 xfs/056 xfs/059 xfs/060 xfs/061 xfs/063 xfs/064 xfs/065 xfs/066 xfs/068 xfs/073 xfs/076
> FSTYP -- xfs (non-debug)
> PLATFORM -- Linux/x86_64 lkp-ws02 4.3.0-rc2-00002-ga45086e
> MKFS_OPTIONS -- -f -bsize=4096 /dev/sdg1
> MOUNT_OPTIONS -- /dev/sdg1 /fs/scratch
>
> xfs/004 3s
> xfs/010 13s
> xfs/011 19s
> xfs/013 318s
> xfs/017 15s
> xfs/020 82s
> xfs/026 14s
> xfs/027 15s
> xfs/028 25s
> xfs/031 58s
> xfs/032 14s
> xfs/033 - output mismatch (see /lkp/benchmarks/xfstests/results//xfs/033.out.bad)
> --- tests/xfs/033.out 2015-11-12 16:56:51.000000000 +0800
> +++ /lkp/benchmarks/xfstests/results//xfs/033.out.bad 2015-11-17 05:21:38.091589526 +0800
> @@ -38,7 +38,19 @@
> Phase 7 - verify and correct link counts...
> resetting inode INO nlinks from 1 to 2
> done
> +mount: wrong fs type, bad option, bad superblock on /dev/sdg1,
> + missing codepage or helper program, or other error
> +
> + In some cases useful info is found in syslog - try
> ...
> (Run 'diff -u tests/xfs/033.out /lkp/benchmarks/xfstests/results//xfs/033.out.bad' to see the entire diff)

I reproduced this running xfs/033 directly on a 4.3.0-rc2+ kernel
(linux-xfs tree) and xfsprogs-3.2.2 from fedora. The problem here is
that the mount complains about an invalid metadata LSN. Via dmesg:

...
XFS (dm-3): Mounting V5 Filesystem
XFS (dm-3): Corruption warning: Metadata has LSN (1:16) ahead of current LSN (1:2). Please unmount and run xfs_repair (>= v4.3) to resolve.
XFS (dm-3): log mount/recovery failed: error -22
XFS (dm-3): log mount failed
...

This occurs because older versions of xfs_repair unconditionally zero
out the log, which can lead to corruption after a crash on v5 superblock
(-m crc=1) filesystems. Therefore, the invalid log state is detected in
the kernel as of the commit referenced above and ultimately fixed in
xfsprogs v4.3. In other words, XFS on a v4.4 kernel requires xfsprogs
v4.3 or newer.

I'm not sure what the xfsprogs release status is (tot looks like it's at
4.3.0-rc2), but the fix is ultimately to upgrade to a supported
userspace.

Brian

> xfs/041 18s
> xfs/042 14s
> xfs/046 15s
> xfs/047 25s
> xfs/049 48s
> xfs/050 23s
> xfs/054 15s
> xfs/056 14s
> xfs/059 18s
> xfs/060 16s
> xfs/061 14s
> xfs/063 14s
> xfs/064 228s
> xfs/065 35s
> xfs/066 28s
> xfs/068 25s
> xfs/073 16s
> xfs/076 53s
> Ran: xfs/004 xfs/010 xfs/011 xfs/013 xfs/017 xfs/020 xfs/026 xfs/027 xfs/028 xfs/031 xfs/032 xfs/033 xfs/041 xfs/042 xfs/046 xfs/047 xfs/049 xfs/050 xfs/054 xfs/056 xfs/059 xfs/060 xfs/061 xfs/063 xfs/064 xfs/065 xfs/066 xfs/068 xfs/073 xfs/076
> Failures: xfs/033
> Failed 1 of 30 tests
>
>
> =========================================================================================
> tbox_group/testcase/rootfs/kconfig/compiler/disk/fs/test:
> lkp-ws02/xfstests/debian-x86_64-2015-02-07.cgz/x86_64-rhel/gcc-4.9/4HDD/xfs/xfs-mid1
>
> commit:
> b7cdc66be54b64daef593894d12ecc405f117829
> a45086e27dfa21a4b39134f7505c8f60a3ecdec4
>
> b7cdc66be54b64da a45086e27dfa21a4b39134f750
> ---------------- --------------------------
> fail:runs %reproduction fail:runs
> | | |
> :4 100% 4:4 last_state.xfstests.exit_code.1
> :4 100% 4:4 xfstests.nr_fail
> :4 100% 4:4 xfstests.xfs.033.fail
> :4 25% 1:4 kmsg.XFS(loop0):Filesystem_has_duplicate_UUID#-#-#f7b-b747-f26a6b2b589b-can't_mount
> :4 25% 1:4 kmsg.XFS(loop0):Filesystem_has_duplicate_UUID#-#c55-#af8-a93b-#c6382174ba-can't_mount
> 1:4 -25% :4 kmsg.XFS(loop0):Filesystem_has_duplicate_UUID#-#d5c-#dd8-#c4c-#b9758dd7946-can't_mount
> 1:4 -25% :4 kmsg.XFS(loop0):Filesystem_has_duplicate_UUID#-#d8b-#bd0-bd22-be703a050d73-can't_mount
> :4 25% 1:4 kmsg.XFS(loop0):Filesystem_has_duplicate_UUID#-#e49-#-#f14-c1f5b7a39914-can't_mount
> :4 25% 1:4 kmsg.XFS(loop0):Filesystem_has_duplicate_UUID#-#f2c-#ba-#-#a8309204098-can't_mount
> 1:4 -25% :4 kmsg.XFS(loop0):Filesystem_has_duplicate_UUID#-#f88-#-a30c-f7972224a2c2-can't_mount
> 1:4 -25% :4 kmsg.XFS(loop0):Filesystem_has_duplicate_UUID#-a0f8-#-b678-a0e21fca1aea-can't_mount
> :4 100% 4:4 kmsg.XFS(sdg1):log_mount/recovery_failed:error
>
> lkp-ws02: Westmere-EP
> Memory: 16G
>
> To reproduce:
>
> git clone git://git.kernel.org/pub/scm/linux/kernel/git/wfg/lkp-tests.git
> cd lkp-tests
> bin/lkp install job.yaml # job file is attached in this email
> bin/lkp run job.yaml
>
>
> Disclaimer:
> Results have been estimated based on internal Intel analysis and are provided
> for informational purposes only. Any difference in system hardware or software
> design or configuration may affect actual performance.
>
>
> Thanks,
> Ying Huang

> ---
> LKP_SERVER: inn
> LKP_CGI_PORT: 80
> LKP_CIFS_PORT: 139
> testcase: xfstests
> default-monitors:
> wait: activate-monitor
> kmsg:
> vmstat:
> interval: 10
> default-watchdogs:
> oom-killer:
> watchdog:
> commit: 96a2f2162deb97760a44ad8558e374342260cee6
> model: Westmere-EP
> memory: 16G
> nr_hdd_partitions: 11
> hdd_partitions: "/dev/disk/by-id/scsi-35000c500*-part1"
> swap_partitions:
> rootfs_partition: "/dev/disk/by-id/ata-WDC_WD1002FAEX-00Z3A0_WD-WCATR5408564-part3"
> cpufreq_governor:
> category: functional
> disk: 4HDD
> fs: xfs
> xfstests:
> test: xfs-mid1
> queue: cyclic
> testbox: lkp-ws02
> tbox_group: lkp-ws02
> kconfig: x86_64-rhel
> enqueue_time: 2015-11-14 09:35:40.999257430 +08:00
> id: 37df3a54bced8a69c8daf7c99bfff44d990622e7
> user: lkp
> compiler: gcc-4.9
> head_commit: 96a2f2162deb97760a44ad8558e374342260cee6
> base_commit: 6a13feb9c82803e2b815eca72fa7a9f5561d7861
> branch: linux-devel/devel-hourly-2015111410
> kernel: "/pkg/linux/x86_64-rhel/gcc-4.9/96a2f2162deb97760a44ad8558e374342260cee6/vmlinuz-4.3.0-bochs-virtio-gpu-wl-ath-16719-g96a2f21"
> rootfs: debian-x86_64-2015-02-07.cgz
> result_root: "/result/xfstests/4HDD-xfs-xfs-mid1/lkp-ws02/debian-x86_64-2015-02-07.cgz/x86_64-rhel/gcc-4.9/96a2f2162deb97760a44ad8558e374342260cee6/0"
> job_file: "/lkp/scheduled/lkp-ws02/cyclic_xfstests-4HDD-xfs-xfs-mid1-x86_64-rhel-CYCLIC_HEAD-96a2f2162deb97760a44ad8558e374342260cee6-20151114-75902-1fjjdep-0.yaml"
> dequeue_time: 2015-11-14 12:49:18.590161421 +08:00
> nr_cpu: "$(nproc)"
> max_uptime: 3600
> initrd: "/osimage/debian/debian-x86_64-2015-02-07.cgz"
> bootloader_append:
> - root=/dev/ram0
> - user=lkp
> - job=/lkp/scheduled/lkp-ws02/cyclic_xfstests-4HDD-xfs-xfs-mid1-x86_64-rhel-CYCLIC_HEAD-96a2f2162deb97760a44ad8558e374342260cee6-20151114-75902-1fjjdep-0.yaml
> - ARCH=x86_64
> - kconfig=x86_64-rhel
> - branch=linux-devel/devel-hourly-2015111410
> - commit=96a2f2162deb97760a44ad8558e374342260cee6
> - BOOT_IMAGE=/pkg/linux/x86_64-rhel/gcc-4.9/96a2f2162deb97760a44ad8558e374342260cee6/vmlinuz-4.3.0-bochs-virtio-gpu-wl-ath-16719-g96a2f21
> - max_uptime=3600
> - RESULT_ROOT=/result/xfstests/4HDD-xfs-xfs-mid1/lkp-ws02/debian-x86_64-2015-02-07.cgz/x86_64-rhel/gcc-4.9/96a2f2162deb97760a44ad8558e374342260cee6/0
> - LKP_SERVER=inn
> - |-
> ipmi_watchdog.start_now=1
>
> earlyprintk=ttyS0,115200 systemd.log_level=err
> debug apic=debug sysrq_always_enabled rcupdate.rcu_cpu_stall_timeout=100
> panic=-1 softlockup_panic=1 nmi_watchdog=panic oops=panic load_ramdisk=2 prompt_ramdisk=0
> console=ttyS0,115200 console=tty0 vga=normal
>
> rw
> lkp_initrd: "/lkp/lkp/lkp-x86_64.cgz"
> modules_initrd: "/pkg/linux/x86_64-rhel/gcc-4.9/96a2f2162deb97760a44ad8558e374342260cee6/modules.cgz"
> bm_initrd: "/osimage/deps/debian-x86_64-2015-02-07.cgz/lkp.cgz,/osimage/deps/debian-x86_64-2015-02-07.cgz/run-ipconfig.cgz,/osimage/deps/debian-x86_64-2015-02-07.cgz/fs.cgz,/osimage/deps/debian-x86_64-2015-02-07.cgz/xfstests.cgz,/lkp/benchmarks/xfstests.cgz"
> job_state: finished
> loadavg: 0.66 0.72 0.56 1/350 26465
> start_time: '1447476647'
> end_time: '1447477486'
> version: "/lkp/lkp/.src-20151113-180526"

> mkfs -t xfs /dev/sdd1
> mkfs -t xfs /dev/sdk1
> mkfs -t xfs /dev/sdg1
> mkfs -t xfs /dev/sdh1
> mount -t xfs -o nobarrier,inode64 /dev/sdd1 /fs/sdd1
> mount -t xfs -o nobarrier,inode64 /dev/sdh1 /fs/sdh1
> mount -t xfs -o nobarrier,inode64 /dev/sdk1 /fs/sdk1
> mount -t xfs -o nobarrier,inode64 /dev/sdg1 /fs/sdg1
> export TEST_DIR=/fs/sdd1
> export TEST_DEV=/dev/sdd1
> export FSTYP=xfs
> export SCRATCH_MNT=/fs/scratch
> mkdir /fs/scratch -p
> export SCRATCH_DEV=/dev/sdg1
> export SCRATCH_LOGDEV=/dev/sdh1
> ./check xfs/004 xfs/010 xfs/011 xfs/013 xfs/017 xfs/020 xfs/026 xfs/027 xfs/028 xfs/031 xfs/032 xfs/033 xfs/041 xfs/042 xfs/046 xfs/047 xfs/049 xfs/050 xfs/054 xfs/056 xfs/059 xfs/060 xfs/061 xfs/063 xfs/064 xfs/065 xfs/066 xfs/068 xfs/073 xfs/076


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/