Re: PROBLEM: XFS on ARM corruption 'Structure needs cleaning'

[Brian Foster <bfoster@redhat.com>]

On Thu, Jun 11, 2015 at 07:32:04PM +0300, Török Edwin wrote:
> On 06/11/2015 06:58 PM, Eric Sandeen wrote:
> > On 6/11/15 10:51 AM, Eric Sandeen wrote:
> >> On 6/11/15 10:28 AM, Török Edwin wrote:
> >>> On 06/11/2015 06:16 PM, Brian Foster wrote:
> >>>> On Thu, Jun 11, 2015 at 09:23:38AM +0300, Török Edwin wrote:
> >>>>> [1.] XFS on ARM corruption 'Structure needs cleaning'
> >>>>> [2.] Full description of the problem/report:
> >>>>>
> >>>>> I have been running XFS sucessfully on x86-64 for years, however I'm having trouble running it on ARM.
> >>>>>
> >>>>> Running the testcase below [7.] reliably reproduces the filesystem corruption starting from a freshly
> >>>>> created XFS filesystem: running ls after 'sxadm node --new --batch /export/dfs/a/b' shows a 'Structure needs cleaning' error,
> >>>>> and dmesg shows a corruption error [6.].
> >>>>> xfs_repair 3.1.9 is not able to repair the corruption: after mounting the repair filesystem
> >>>>> I still get the 'Structure needs cleaning' error.
> >>>>>
> >>>>> Note: using /export/dfs/a/b is important for reproducing the problem: if I only use one level of directories in /export/dfs then the problem
> >>>>> doesn't reproduce. Also if I use a tuned version of sxadm that creates fewer database files then the problem doesn't reproduce either.
> >>>>>
> >>>>> [3.] Keywords: filesystems, XFS corruption, ARM
> >>>>> [4.] Kernel information
> >>>>> [4.1.] Kernel version (from /proc/version):
> >>>>> Linux hornet34 3.14.3-00088-g7651c68 #24 Thu Apr 9 16:13:46 MDT 2015 armv7l GNU/Linux
> >>>>>
> >>>> ...
> >>>>> [5.] Most recent kernel version which did not have the bug: Unknown, first kernel I try on ARM
> >>>>>
> >>>>> [6.] dmesg stacktrace
> >>>>>
> >>>>> [4627578.440000] XFS (sda4): Mounting Filesystem
> >>>>> [4627578.510000] XFS (sda4): Ending clean mount
> >>>>> [4627621.470000] dd6ee000: 58 46 53 42 00 00 10 00 00 00 00 00 37 40 21 00  XFSB........7@!.
> >>>>> [4627621.480000] dd6ee010: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
> >>>>> [4627621.490000] dd6ee020: 5b 08 7f 79 0e 3a 46 3d 9b ea 26 ad 9d 62 17 8d  [..y.:F=..&..b..
> >>>>> [4627621.490000] dd6ee030: 00 00 00 00 20 00 00 04 00 00 00 00 00 00 00 80  .... ...........
> >>>>
> >>>> Just a data point... the magic number here looks like a superblock magic
> >>>> (XFSB) rather than one of the directory magic numbers. I'm wondering if
> >>>> a buffer disk address has gone bad somehow or another.
> >>>>
> >>>> Does this happen to be a large block device? I don't see any partition
> >>>> or xfs_info data below. If so, it would be interesting to see if this
> >>>> reproduces on a smaller device. It does appear that the large block
> >>>> device option is enabled in the kernel config above, however, so maybe
> >>>> that's unrelated.
> >>>
> >>> This is mkfs.xfs /dev/sda4:
> >>> meta-data=/dev/sda4              isize=256    agcount=4, agsize=231737408 blks
> >>>          =                       sectsz=512   attr=2, projid32bit=0
> >>> data     =                       bsize=4096   blocks=926949632, imaxpct=5
> >>>          =                       sunit=0      swidth=0 blks
> >>> naming   =version 2              bsize=4096   ascii-ci=0
> >>> log      =internal log           bsize=4096   blocks=452612, version=2
> >>>          =                       sectsz=512   sunit=0 blks, lazy-count=1
> >>> realtime =none                   extsz=4096   blocks=0, rtextents=0
> >>>
> >>> But it also reproduces with this small loopback file:
> >>> meta-data=/tmp/xfs.test          isize=256    agcount=2, agsize=5120 blks
> >>>          =                       sectsz=512   attr=2, projid32bit=0
> >>> data     =                       bsize=4096   blocks=10240, imaxpct=25
> >>>          =                       sunit=0      swidth=0 blks
> >>> naming   =version 2              bsize=4096   ascii-ci=0
> >>> log      =internal log           bsize=4096   blocks=1200, version=2
> >>>          =                       sectsz=512   sunit=0 blks, lazy-count=1
> >>> realtime =none                   extsz=4096   blocks=0, rtextents=0
> >>
> >> ok so not a block number overflow issue, thanks.
> >>
> >>> You can have a look at xfs.test here: http://vol-public.s3.indian.skylable.com:8008/armel/testcase/xfs.test.gz
> >>>
> >>> If I loopback mount that on an x86-64 box it doesn't show the corruption message though ...
> >>
> >> FWIW, this is the 2nd report we've had of something similar, both on Armv7, both ok on x86_64.
> >>
> >> I'll take a look at your xfs.test; that's presumably copied after it reported the error, and you unmounted it before uploading, correct?  And it was mkfs'd on armv7, never mounted or manipulated in any way on x86_64?
> 
> Thanks, yes it was mkfs.xfs on ARMv7 and unmounted.
> 
> > 
> > Oh, and what were the kernel messages when you produced the corruption with xfs.txt?
> 
> Takes only a couple of minutes to reproduce the issue so I've prepared a fresh set of xfs2.test and corresponding kernel messages to make sure its all consistent.
> Freshly created XFS by mkfs.xfs: http://vol-public.s3.indian.skylable.com:8008/armel/testcase/xfs2.test.orig.gz
> The corrupted XFS: http://vol-public.s3.indian.skylable.com:8008/armel/testcase/xfs2.test.corrupted.gz
> 

I managed to get an updated kernel on a beaglebone I had sitting around,
but I don't reproduce any errors with the "corrupted" image (I think
we've established that the image is fine on-disk and something is going
awry at runtime):

root@beaglebone:~# uname -a
Linux beaglebone 3.14.1+ #5 SMP Thu Jun 11 20:58:02 EDT 2015 armv7l GNU/Linux
root@beaglebone:~# mount ./xfs2.test.corrupted /mnt/
root@beaglebone:~# ls -al /mnt/a/
total 12
drwxr-xr-x 3 root root   14 Jun 11 16:11 .
drwxr-xr-x 3 root root   14 Jun 11 16:11 ..
drwxr-x--- 2 root root 8192 Jun 11 16:11 b
root@beaglebone:~# ls -al /mnt/a/b/
total 17996
drwxr-x--- 2 root root    8192 Jun 11 16:11 .
drwxr-xr-x 3 root root      14 Jun 11 16:11 ..
-rw-r--r-- 1 root root   12288 Jun 11 16:11 events.db
-rw-r--r-- 1 root root   15360 Jun 11 16:11 f00000000.db
-rw-r--r-- 1 root root   15360 Jun 11 16:11 f00000001.db
-rw-r--r-- 1 root root   15360 Jun 11 16:11 f00000002.db
-rw-r--r-- 1 root root   15360 Jun 11 16:11 f00000003.db
...
root@beaglebone:~#

I echo Dave's suggestion down thread with regard to toolchain. This
kernel was compiled with the following cross-gcc (installed via Fedora
package):

	gcc version 4.9.2 20150212 (Red Hat Cross 4.9.2-5) (GCC) 

Are you using something different?

Brian

> All commands below were run on armv7, and unmounted, the files from /tmp copied over to x86-64, gzipped and uploaded, they were never mounted on x86-64:
> 
> # dd if=/dev/zero of=/tmp/xfs2.test bs=1M count=40
> 40+0 records in
> 40+0 records out
> 41943040 bytes (42 MB) copied, 0.419997 s, 99.9 MB/s
> # mkfs.xfs /tmp/xfs2.test
> meta-data=/tmp/xfs2.test         isize=256    agcount=2, agsize=5120 blks
>          =                       sectsz=512   attr=2, projid32bit=0
> data     =                       bsize=4096   blocks=10240, imaxpct=25
>          =                       sunit=0      swidth=0 blks
> naming   =version 2              bsize=4096   ascii-ci=0
> log      =internal log           bsize=4096   blocks=1200, version=2
>          =                       sectsz=512   sunit=0 blks, lazy-count=1
> realtime =none                   extsz=4096   blocks=0, rtextents=0
> # cp /tmp/xfs2.test /tmp/xfs2.test.orig
> # umount /export/dfs
> # mount -o loop -t xfs /tmp/xfs2.test /export/dfs
> # mkdir /export/dfs/a
> # sxadm node --new --batch /export/dfs/a/b
> # ls /export/dfs/a/b
> ls: reading directory /export/dfs/a/b: Structure needs cleaning
> # umount /export/dfs
> # cp /tmp/xfs2.test /tmp/xfs2.test.corrupted
> # dmesg >/tmp/dmesg
> # exit
> 
> the latest corruption message from dmesg:
> [4744604.870000] XFS (loop0): Mounting Filesystem
> [4744604.900000] XFS (loop0): Ending clean mount
> [4745016.610000] dc61e000: 58 46 53 42 00 00 10 00 00 00 00 00 00 00 28 00  XFSB..........(.
> [4745016.620000] dc61e010: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
> [4745016.630000] dc61e020: 64 23 d2 06 32 2e 4c 20 82 6e f0 36 a7 d9 54 f9  d#..2.L .n.6..T.
> [4745016.640000] dc61e030: 00 00 00 00 00 00 20 04 00 00 00 00 00 00 00 80  ...... .........
> [4745016.640000] XFS (loop0): Internal error xfs_dir3_data_read_verify at line 274 of file fs/xfs/xfs_dir2_data.c.  Caller 0xc01c1528
> [4745016.650000] CPU: 0 PID: 37 Comm: kworker/0:1H Not tainted 3.14.3-00088-g7651c68 #24
> [4745016.650000] Workqueue: xfslogd xfs_buf_iodone_work
> [4745016.650000] [<c0013948>] (unwind_backtrace) from [<c0011058>] (show_stack+0x10/0x14)
> [4745016.650000] [<c0011058>] (show_stack) from [<c01c3dc4>] (xfs_corruption_error+0x54/0x70)
> [4745016.650000] [<c01c3dc4>] (xfs_corruption_error) from [<c01f7854>] (xfs_dir3_data_read_verify+0x60/0xd0)
> [4745016.650000] [<c01f7854>] (xfs_dir3_data_read_verify) from [<c01c1528>] (xfs_buf_iodone_work+0x7c/0x94)
> [4745016.650000] [<c01c1528>] (xfs_buf_iodone_work) from [<c00309f0>] (process_one_work+0xf4/0x32c)
> [4745016.650000] [<c00309f0>] (process_one_work) from [<c0030fb4>] (worker_thread+0x10c/0x388)
> [4745016.650000] [<c0030fb4>] (worker_thread) from [<c0035e10>] (kthread+0xbc/0xd8)
> [4745016.650000] [<c0035e10>] (kthread) from [<c000e8f8>] (ret_from_fork+0x14/0x3c)
> [4745016.650000] XFS (loop0): Corruption detected. Unmount and run xfs_repair
> [4745016.650000] XFS (loop0): metadata I/O error: block 0xa000 ("xfs_trans_read_buf_map") error 117 numblks 8
> 
> Best regards,
> --Edwin
> 
> _______________________________________________
> xfs mailing list
> xfs@oss.sgi.com
> http://oss.sgi.com/mailman/listinfo/xfs

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs