Mar 24 06:39:05 syn-proc4 kernel: [65682.636659] Call Trace:
Mar 24 06:39:05 syn-proc4 kernel: [65682.636690]  [<f8d17478>] diWrite+0x16d/0x4a2 [jfs]
Mar 24 06:39:05 syn-proc4 kernel: [65682.636721]  [<c0121355>] __wake_up+0x29/0x39
Mar 24 06:39:05 syn-proc4 kernel: [65682.636744]  [<f8d260a6>] txCommit+0x18d/0xecc [jfs]
Mar 24 06:39:05 syn-proc4 kernel: [65682.636770]  [<c016b2c7>] write_cache_pages+0x231/0x279
Mar 24 06:39:05 syn-proc4 kernel: [65682.636790]  [<c016ad5d>] __writepage+0x0/0x21
Mar 24 06:39:05 syn-proc4 kernel: [65682.636808]  [<c01f3ea2>] __next_cpu+0x12/0x21
Mar 24 06:39:05 syn-proc4 kernel: [65682.636830]  [<f8d25d5c>] txBegin+0x25/0x1e2 [jfs]
Mar 24 06:39:05 syn-proc4 kernel: [65682.636858]  [<f8d0d0d9>] jfs_commit_inode+0x9e/0xde [jfs]
Mar 24 06:39:05 syn-proc4 kernel: [65682.636883]  [<f8d0d328>] jfs_write_inode+0x2d/0x3b [jfs]
Mar 24 06:39:05 syn-proc4 kernel: [65682.636905]  [<c01a1120>] __writeback_single_inode+0x15a/0x251
 
Ok. Even more problems. I tried to remount this fs in read-only but 'mount -o remount,ro /my/mountpoint'
command hadnged up. I rebooted machine, upgrade jfsutils to the latest version 1.1.4 and tried to fsck it.
Here is the output:

# sudo fsck.jfs -f /dev/sdb3
fsck.jfs version 1.1.14, 06-Apr-2009
processing started: 3/25/2010 19.17.30
The current device is:  /dev/sdb3
Block size in bytes:  4096
Filesystem size in blocks:  363779876
**Phase 0 - Replay Journal Log
**Phase 1 - Check Blocks, Files/Directories, and  Directory Entries
Insufficient dynamic storage available for required workspace (1,6).  CANNOT CONTINUE
.......|.
#

This files system contains huge amount of files, hundreds of thousands of them, maybe millions,
and huge amount of hard links. Can this be a problem for JFS and fsck.jfs?

--
Best regards, Aleksey Cheusov.