http://oss.sgi.com/bugzilla/show_bug.cgi?id=284
------- Additional Comments From xfs@xxxxxxxxxxxx 2003-12-12 11:15 PDT -------
I'm using RH7.2 (XFS release)
I'm using rsync to backup 1 partition to another (hdg is the target)
These are the problems that have occured since upgrading from
2.4.18-4SGI_XFS_1.1
to
2.6.0-test9
I have not applied the patch eric suggested. The drives are not mounted with the
ikeep option. Here's additional info asked of others:
kernel CONFIG_XFS:
CONFIG_XFS_FS=y
# CONFIG_XFS_RT is not set
# CONFIG_XFS_QUOTA is not set
# CONFIG_XFS_POSIX_ACL is not set
This is after the xfs_force_shutdown() and before any remount
# xfs_info /share/backup/mirror
meta-data=/share/backup/mirror isize=256 agcount=19, agsize=1048576 blks
= sectsz=512
data = bsize=4096 blocks=19537040, imaxpct=25
= sunit=0 swidth=0 blks, unwritten=0
naming =version 2 bsize=4096
log =internal bsize=4096 blocks=2384, version=1
= sectsz=512 sunit=0 blks
realtime =none extsz=65536 blocks=0, rtextents=0
# mount -oremount /share/backup/mirror
root@willow /usr/src/linux-2.6.0-test11
# ll /share/backup/mirror
?--------- 2053 3367557920 12288 577991275711802121 Jan 1 1970
/share/backup/mirror
Eeek
root@willow /usr/src/linux-2.6.0-test11
# umount /share/backup/mirror
root@willow /usr/src/linux-2.6.0-test11
# mount /share/backup/mirror
root@willow /usr/src/linux-2.6.0-test11
# ll /share/backup/mirror/
total 0
drwxr-xr-x 10 root root 148 Dec 11 01:56 ./
drwxr-xr-x 4 root root 33 Nov 6 18:46 ../
drwxr-xr-x 10 root root 140 Nov 7 01:52 yesterday/
drwxr-xr-x 10 root root 140 Nov 7 01:52 yesterday-1/
drwxr-xr-x 10 root root 140 Nov 7 01:52 yesterday-2/
drwxr-xr-x 10 root root 140 Nov 7 01:52 yesterday-3/
drwxr-xr-x 10 root root 140 Nov 7 01:52 yesterday-4/
drwxr-xr-x 10 root root 140 Nov 7 01:52 yesterday-5/
drwxr-xr-x 10 root root 140 Nov 7 01:52 yesterday-6/
drwxr-xr-x 3 root root 19 Dec 12 01:51 yesterday-7/
phew
anyway, logs:
Dec 2 01:55:12 willow kernel: Filesystem "hdg1": xfs_log_write: reservation ran
out. Need to up reservation
Dec 2 01:55:12 willow kernel: xfs_force_shutdown(hdg1,0x8) called from line 173
9 of file fs/xfs/xfs_log.c. Return address = 0xc01f023d
Dec 2 01:55:12 willow kernel: Filesystem "hdg1": Corruption of in-memory data d
etected. Shutting down filesystem: hdg1
Dec 2 01:55:12 willow kernel: Please umount the filesystem, and rectify the pro
blem(s)
Dec 2 01:55:12 willow kernel: xfs_force_shutdown(hdg1,0x2) called from line 132
1 of file fs/xfs/xfs_log.c. Return address = 0xc01f023d
Dec 5 12:30:34 willow kernel: Unable to handle kernel NULL pointer dereference
at virtual address 00000000
Dec 5 12:30:34 willow kernel: printing eip:
Dec 5 12:30:34 willow kernel: c01c212b
Dec 5 12:30:34 willow kernel: *pde = 00000000
Dec 5 12:30:34 willow kernel: Oops: 0002 [#1]
Dec 5 12:30:34 willow kernel: CPU: 0
Dec 5 12:30:34 willow kernel: EIP: 0060:[<c01c212b>] Not tainted
Dec 5 12:30:34 willow kernel: EFLAGS: 00010296
Dec 5 12:30:34 willow kernel: EIP is at xfs_inobt_lookup+0x24b/0x350
Dec 5 12:30:34 willow kernel: eax: ffffffff ebx: 0000000f ecx: 0000000e e
dx: 00000000
Dec 5 12:30:34 willow kernel: esi: cd0d1000 edi: 00000000 ebp: 0000000e e
sp: c5bb9b5c
Dec 5 12:30:34 willow kernel: ds: 007b es: 007b ss: 0068
Dec 5 12:30:34 willow kernel: Process nmbd (pid: 5550, threadinfo=c5bb8000 task
=ce2ac6d0)
Dec 5 12:30:34 willow kernel: Stack: c13eb6f8 c01aab86 c65cef6c 00000000 000000
06 d0dfc800 00000000 ffffffe4
Dec 5 12:30:34 willow kernel: ffffffff 00000006 00000003 00000001 d05241
60 00000000 0000049c 00000006
Dec 5 12:30:34 willow kernel: d0dfc800 c01bfce7 c65cef30 00000001 c5bb9c
00 00000000 00000000 c5bb9c00
Dec 5 12:30:34 willow kernel: Call Trace:
Dec 5 12:30:34 willow kernel: [<c01aab86>] xfs_btree_init_cursor+0x36/0x1d0
Dec 5 12:30:34 willow kernel: [<c01bfce7>] xfs_dialloc+0x2d7/0x990
Dec 5 12:30:34 willow kernel: [<c01343b0>] cache_alloc_refill+0x170/0x1c0
Dec 5 12:30:34 willow kernel: [<c01c63f6>] xfs_ialloc+0x56/0x420
Dec 5 12:30:34 willow kernel: [<c01af0b6>] xfs_da_buf_make+0x36/0x1f0
Dec 5 12:30:34 willow kernel: [<c01dc241>] xfs_dir_ialloc+0x71/0x280
Dec 5 12:30:34 willow kernel: [<c01d9504>] xfs_trans_reserve+0xa4/0x180
Dec 5 12:30:34 willow kernel: [<c01e108f>] xfs_create+0x30f/0x5b0
Dec 5 12:30:34 willow kernel: [<c01ed078>] linvfs_mknod+0x1a8/0x250
Dec 5 12:30:34 willow kernel: [<c01b270a>] xfs_dir2_block_lookup+0x1a/0xa0
Dec 5 12:30:34 willow kernel: [<c01b0e24>] xfs_dir2_lookup+0xc4/0x130
Dec 5 12:30:34 willow kernel: [<c0131699>] __alloc_pages+0xb9/0x320
Dec 5 12:30:34 willow kernel: [<c0150ec1>] vfs_create+0x71/0xa0
Dec 5 12:30:34 willow kernel: [<c0150b65>] __lookup_hash+0x75/0xa0
Dec 5 12:30:34 willow kernel: [<c01511fc>] open_namei+0x18c/0x420
Dec 5 12:30:34 willow kernel: [<c0115f55>] schedule+0x295/0x4b0
Dec 5 12:30:34 willow kernel: [<c0143dd2>] filp_open+0x32/0x50
Dec 5 12:30:34 willow kernel: [<c0144172>] sys_open+0x32/0x70
Dec 5 12:30:34 willow kernel: [<c0108f47>] syscall_call+0x7/0xb
Dec 5 12:30:34 willow kernel:
Dec 5 12:30:34 willow kernel: Code: 6c 24 08 04 ff 4c 24 18 0f 89 0e fe ff ff 8
3 7c 24 4c 01 74
Dec 12 01:54:27 willow kernel: Filesystem "hdg1": xfs_log_write: reservation ran
out. Need to up reservation
Dec 12 01:54:27 willow kernel: xfs_force_shutdown(hdg1,0x8) called from line 173
9 of file fs/xfs/xfs_log.c. Return address = 0xc01f023d
Dec 12 01:54:27 willow kernel: Filesystem "hdg1": Corruption of in-memory data d
etected. Shutting down filesystem: hdg1
Dec 12 01:54:27 willow kernel: Please umount the filesystem, and rectify the pro
blem(s)
Dec 12 01:54:27 willow kernel: xfs_force_shutdown(hdg1,0x2) called from line 132
1 of file fs/xfs/xfs_log.c. Return address = 0xc01f023d
Since this is occuring reasonably often on a non-critical system I am more than
happy to apply changes, report back and try to reproduce crashes *if requested*.
For now I'll :
a) upgrade to 2.6.0-test11
b) apply the "+ xfs_stack_trace();" patch
c) echo 5 > /proc/sys/fs/xfs/error_level
d) echo 2 > /proc/sys/fs/xfs/panic_mask
But I'm remounting with the ikeep option unless someone mails me to ask me to
leave it off :)
David
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
|