Re: XFS internal error

[Jeffrey Hundstad <jeffrey.hundstad@mnsu.edu>]

The use of linux-2.6.17.6 almost certainly caused your problems.  You'll 
need a xfs_repair version 2.8.10 or later to repair the damage.

See the following for complete info:
http://oss.sgi.com/projects/xfs/faq.html#dir2

-- 
Jeffrey Hundstad

Gerardo Exequiel Pozzi wrote:
> Hi,
>
> (sorry for my english)
> (please cc to my email i am not suscribe to this list)
> Linux version 2.6.17.11 (root@djgera) (gcc version 3.4.6) #1 PREEMPT 
> Thu Aug 24 00:27:47 ART 2006
> and previusly used, .7 .6 .2, and .16.X ...
>
> This problem can be related with previus kernels that write on fs, and 
> now when delete files apears the errors?
>
> I delete a big tree directory with many files (39G) from my filesystem,
> then appears this error:
>
> djgera@djgera:/mnt/sdb1$ time rm -rf frugalware
> rm: cannot remove 
> `frugalware/frugalware-stable/extra/frugalware-x86_64/gnet-2.0.7-1-x86_64.fpm': 
> Unknown error 990
> rm: cannot lstat 
> `frugalware/frugalware-stable/extra/frugalware-x86_64/firefox-hu-1.5-2-x86_64.fpm': 
> Input/output error
>
> real    0m54.650s
> user    0m0.059s
> sys     0m2.582s
>
> the dmesg shows:
> xfs_da_do_buf: bno 16777216
> dir: inode 168328424
> Filesystem "sdb1": XFS internal error xfs_da_do_buf(1) at line 2119 of 
> file fs/xfs/xfs_da_btree.c.  Caller 0xc01e6597
> <c01e6276> xfs_da_do_buf+0x5f6/0x880  <c01e6597> 
> xfs_da_read_buf+0x47/0x50
> <c01e43bd> xfs_da_node_lookup_int+0xcd/0x3a0  <c01ebcbb> 
> xfs_dir2_data_log_unused+0x6b/0x80
> <c01e6597> xfs_da_read_buf+0x47/0x50  <c01f03b3> 
> xfs_dir2_leafn_remove+0x2a3/0x440
> <c01f03b3> xfs_dir2_leafn_remove+0x2a3/0x440  <c01f1824> 
> xfs_dir2_node_removename+0xb4/0x100
> <c01e8c92> xfs_dir2_removename+0x122/0x130  <c0229e46> 
> kmem_zone_alloc+0x56/0xe0
> <c021d205> xfs_trans_ijoin+0x35/0xa0  <c0225050> xfs_remove+0x280/0x560
> <c01ff3e0> xfs_iget+0xd0/0x140  <c0231cd8> xfs_vn_unlink+0x48/0x90
> <c021bfa1> xfs_trans_unlocked_item+0x41/0x60  <c0174b10> 
> d_rehash+0x50/0x80
> <c0174648> d_splice_alias+0xa8/0x110  <c01687e5> permission+0x85/0xb0
> <c016a7a1> may_delete+0x41/0x120  <c016bc7f> vfs_unlink+0xaf/0xc0
> <c016bd3c> do_unlinkat+0xac/0x130  <c016e67c> sys_getdents64+0xcc/0xf0
> <c016e4c0> filldir64+0x0/0xf0  <c016be17> sys_unlink+0x17/0x20
> <c01031e3> syscall_call+0x7/0xb
> Filesystem "sdb1": XFS internal error xfs_trans_cancel at line 1150 of 
> file fs/xfs/xfs_trans.c.  Caller 0xc02250be
> <c021b998> xfs_trans_cancel+0x108/0x150  <c02250be> 
> xfs_remove+0x2ee/0x560
> <c02250be> xfs_remove+0x2ee/0x560  <c01ff3e0> xfs_iget+0xd0/0x140
> <c0231cd8> xfs_vn_unlink+0x48/0x90  <c021bfa1> 
> xfs_trans_unlocked_item+0x41/0x60
> <c0174b10> d_rehash+0x50/0x80  <c0174648> d_splice_alias+0xa8/0x110
> <c01687e5> permission+0x85/0xb0  <c016a7a1> may_delete+0x41/0x120
> <c016bc7f> vfs_unlink+0xaf/0xc0  <c016bd3c> do_unlinkat+0xac/0x130
> <c016e67c> sys_getdents64+0xcc/0xf0  <c016e4c0> filldir64+0x0/0xf0
> <c016be17> sys_unlink+0x17/0x20  <c01031e3> syscall_call+0x7/0xb
> xfs_force_shutdown(sdb1,0x8) called from line 1151 of file 
> fs/xfs/xfs_trans.c.  Return address = 0xc021b9be
> Filesystem "sdb1": Corruption of in-memory data detected.  Shutting 
> down filesystem: sdb1
> Please umount the filesystem, and rectify the problem(s)
>
> when umount appears this at dmesg:
> xfs_force_shutdown(sdb1,0x1) called from line 338 of file 
> fs/xfs/xfs_rw.c.  Return address = 0xc02299e7
>
> now mount and: (to replay log for check with xfs_check)
> XFS mounting filesystem sdb1
> Starting XFS recovery on filesystem: sdb1 (logdev: internal)
> Ending XFS recovery on filesystem: sdb1 (logdev: internal)
>
> umount and xfs_check:
> root@djgera:~# xfs_check /dev/sdb1
> bad free block nused 34 should be 43 for dir ino 78944224 block 16777216
> missing free index for data block 0 in dir ino 168328424
> missing free index for data block 2 in dir ino 168328424
> missing free index for data block 3 in dir ino 168328424
> missing free index for data block 4 in dir ino 168328424
> missing free index for data block 5 in dir ino 168328424
> missing free index for data block 6 in dir ino 168328424
> missing free index for data block 7 in dir ino 168328424
> missing free index for data block 8 in dir ino 168328424
> missing free index for data block 9 in dir ino 168328424
> bad free block nused 4 should be 21 for dir ino 168328959 block 16777216
>
> xfs_repair:
> root@djgera:~# xfs_repair /dev/sdb1
> Phase 1 - find and verify superblock...
> Phase 2 - using internal log
>        - zero log...
>        - scan filesystem freespace and inode maps...
>        - found root inode chunk
> Phase 3 - for each AG...
>        - scan and clear agi unlinked lists...
>        - process known inodes and perform inode discovery...
>        - agno = 0
>        - agno = 1
>        - agno = 2
>        - agno = 3
>        - agno = 4
>        - agno = 5
>        - agno = 6
>        - agno = 7
>        - agno = 8
>        - agno = 9
>        - agno = 10
>        - agno = 11
>        - agno = 12
>        - agno = 13
>        - agno = 14
>        - agno = 15
>        - process newly discovered inodes...
> Phase 4 - check for duplicate blocks...
>        - setting up duplicate extent list...
>        - clear lost+found (if it exists) ...
>        - check for inodes claiming duplicate blocks...
>        - agno = 0
>        - agno = 1
>        - agno = 2
>        - agno = 3
>        - agno = 4
>        - agno = 5
>        - agno = 6
>        - agno = 7
>        - agno = 8
>        - agno = 9
>        - agno = 10
>        - agno = 11
>        - agno = 12
>        - agno = 13
>        - agno = 14
>        - agno = 15
> Phase 5 - rebuild AG headers and trees...
>        - reset superblock...
> Phase 6 - check inode connectivity...
>        - resetting contents of realtime bitmap and summary inodes
>        - ensuring existence of lost+found directory
>        - traversing filesystem starting at / ...
> free block 16777216 for directory inode 78944224 bad nused
> rebuilding directory inode 78944224
> free block 16777216 for directory inode 168328959 bad nused
> rebuilding directory inode 168328959
> can't read freespace block 16777216 for directory inode 168328424
> rebuilding directory inode 168328424
>        - traversal finished ...
>        - traversing all unattached subtrees ...
>        - traversals finished ...
>        - moving disconnected inodes to lost+found ...
> Phase 7 - verify and correct link counts...
> cache_purge: shake on cache 0x80daa78 left 3 nodes!?
> cache_purge: shake on cache 0x80daa78 left 3 nodes!?
> done
>
> root@djgera:~# xfs_info /dev/sdb1
> meta-data=/dev/sdb1              isize=256    agcount=16, 
> agsize=1525923 blks
>         =                       sectsz=512   attr=0
> data     =                       bsize=4096   blocks=24414768, imaxpct=25
>         =                       sunit=0      swidth=0 blks, unwritten=1
> naming   =version 2              bsize=4096
> log      =internal               bsize=4096   blocks=11921, version=1
>         =                       sectsz=512   sunit=0 blks
> realtime =none                   extsz=65536  blocks=0, rtextents=0
>
>