I'm currently running a 2.6.20, but i experienced this bug since quite some versions.
This is while using lvm2/dm.
It happens even on a freshly mkfs-ed fs. Certain workloads trigger it. Best way i found is installing World of Warcraft with all it's updates under wine, which results first in corruption, then in an ro filesystem during the installing. another one is using mldonkey, a p2p file sharing application (especially used it's bittorrent module), but this only shows after days of usage.
EXT3-fs error (device dm-0): ext3_free_blocks_sb: bit already cleared for block 6858329
Aborting journal on device dm-0.
Remounting filesystem read-only
EXT3-fs error (device dm-0): ext3_free_blocks_sb: bit already cleared for block 6858330
EXT3-fs error (device dm-0): ext3_free_blocks_sb: bit already cleared for block 6858331
EXT3-fs error (device dm-0): ext3_free_blocks_sb: bit already cleared for block 6858332
EXT3-fs error (device dm-0): ext3_free_blocks_sb: bit already cleared for block 6858333
EXT3-fs error (device dm-0): ext3_free_blocks_sb: bit already cleared for block 6858334
EXT3-fs error (device dm-0): ext3_free_blocks_sb: bit already cleared for block 6858335
EXT3-fs error (device dm-0): ext3_free_blocks_sb: bit already cleared for block 6858336
EXT3-fs error (device dm-0): ext3_free_blocks_sb: bit already cleared for block 6858337
EXT3-fs error (device dm-0): ext3_free_blocks_sb: bit already cleared for block 6858338
EXT3-fs error (device dm-0): ext3_free_blocks_sb: bit already cleared for block 6858339
EXT3-fs error (device dm-0): ext3_free_blocks_sb: bit already cleared for block 6858340
EXT3-fs error (device dm-0) in ext3_free_blocks_sb: Journal has aborted
EXT3-fs error (device dm-0) in ext3_reserve_inode_write: Journal has aborted
EXT3-fs error (device dm-0) in ext3_truncate: Journal has aborted
EXT3-fs error (device dm-0) in ext3_reserve_inode_write: Journal has aborted
EXT3-fs error (device dm-0) in ext3_orphan_del: Journal has aborted
EXT3-fs error (device dm-0) in ext3_reserve_inode_write: Journal has aborted
__journal_remove_journal_head: freeing b_committed_data
[last line repeated quite some time]