Re: 2.4.24 may be bug in prints.c:341

From: Oleg Drokin
Date: Sat Jan 17 2004 - 16:26:29 EST


Hello!

"Condor" <condor@xxxxxxxxxx> wrote:

C> Jan 15 19:59:20 heaven kernel: sd(8,1):Using r5 hash to sort names
C> Jan 15 21:34:03 heaven kernel: 3w-xxxx: scsi0: Command failed: status =
C> 0xc7, flags = 0x1b, unit #2.
C> Jan 15 21:34:03 heaven last message repeated 27 times
C> Jan 15 21:34:07 heaven kernel: 3w-xxxx: scsi0: Command failed: status =
C> 0xc7, flags = 0xd0, unit #2.
C> Jan 15 21:34:31 heaven kernel: 3w-xxxx: scsi0: Reset succeeded.
C> Jan 16 01:53:39 heaven kernel: Device busy for revalidation (usage=2)

These indicate problems with your hard drive or disk controller.

C> Jan 16 01:55:17 heaven kernel: is_tree_node: node level 0 does not match to
C> the expected one -1
C> Jan 16 01:55:17 heaven kernel: sd(8,1):vs-5150: search_by_key: invalid
C> format found in block 0. Fsck?
C> Jan 16 01:55:17 heaven kernel: sd(8,1):vs-13050: reiserfs_update_sd: i/o
C> failure occurred trying to update [403 8975 0x0 SD] stat data
C> Jan 16 01:55:18 heaven kernel: is_tree_node: node level 0 does not match to
C> the expected one -1
C> Jan 16 01:55:18 heaven kernel: sd(8,1):vs-5150: search_by_key: invalid
C> format found in block 0. Fsck?
C> Jan 16 01:55:18 heaven kernel: sd(8,1):vs-13050: reiserfs_update_sd: i/o
C> failure occurred trying to update [403 8975 0x0 SD] stat data

Some garbage was returned for READ requests and reiserfs was not able
to find metadata it expected to find.

C> Jan 16 01:55:19 heaven kernel: journal-003: journal_end: j_start (4757) is
C> too high

Now somehow superblock data that is pinned in memory also got corrupted
and reiserfs panicked because of that.

So you have a hardware problem.

Bye,
Oleg
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/