[2234] in linux-scsi channel archive
Re: BusLogic BT-958 + st
daemon@ATHENA.MIT.EDU (Leonard N. Zubkoff)
Sun Aug 3 00:17:01 1997
Date: Sat, 2 Aug 1997 21:13:01 -0700
From: "Leonard N. Zubkoff" <lnz@dandelion.com>
To: maillard@vt.edu
CC: linux-scsi@vger.rutgers.edu
In-reply-to: <199708011235.IAA09285@blue.me.vt.edu> (message from Julien
Maillard on Fri, 01 Aug 1997 08:35:49 -0400)
Date: Fri, 01 Aug 1997 08:35:49 -0400
From: Julien Maillard <maillard@blue.me.vt.edu>
Has anyone encountered the kernel messages below? ...
I've seen messages like that on occasion. They are consistent with a couple of
possibilities:
(1) A SCSI bus hang that can only be cleared by a SCSI Bus Reset.
(2) A device staying connected on the SCSI bus long enough that it
appears as (1).
If you increase the timeouts in "linux/drivers/scsi/sd.c" by a large enough
factor that no device could be holding the bus that long (e.g. 100), it should
be apparent which is the case.
--
Jul 31 03:00:55 blue kernel: scsi : aborting command due to timeout : pid
1618337, scsi0, channel 0, id 0, lun 0 Write (6) 0f 32 ff 04 00
Jul 31 03:00:55 blue kernel: scsi0: Aborting CCB #1618410 to Target 0
Jul 31 03:00:55 blue kernel: scsi : aborting command due to timeout : pid
1618342, scsi0, channel 0, id 0, lun 0 Write (6) 11 35 01 02 00
Jul 31 03:00:55 blue kernel: scsi0: Aborting CCB #1618415 to Target 0
Jul 31 03:00:55 blue kernel: SCSI host 0 abort (pid 1618337) timed out -
resetting
Jul 31 03:00:55 blue kernel: SCSI bus is being reset for host 0 channel 0.
Jul 31 03:00:55 blue kernel: scsi0: Sending Bus Device Reset CCB #1618426
to Target 0
Jul 31 03:00:55 blue kernel: SCSI host 0 abort (pid 1618342) timed out -
resetting
Jul 31 03:00:55 blue kernel: SCSI bus is being reset for host 0 channel 0.
Jul 31 03:00:55 blue kernel: scsi0: Unable to Reset Command to Target 0 -
Reset Pending
Jul 31 03:00:57 blue kernel: SCSI host 0 channel 0 reset (pid 1618337)
timed out - trying harder
Jul 31 03:00:57 blue kernel: SCSI bus is being reset for host 0 channel 0.
Jul 31 03:00:57 blue kernel: scsi0: Resetting BusLogic BT-958 due to
Target 0
Jul 31 03:00:57 blue kernel: scsi0: *** BusLogic BT-958 Initialized
Successfully ***
Jul 31 03:00:59 blue kernel: st0: Error with sense data: extra data not
valid Current error st09:00: sense key Unit Attention
Jul 31 03:00:59 blue kernel: Additional sense indicates Power on, reset,
or bus device reset occurred
Jul 31 03:01:01 blue kernel: st: Unloaded.
Jul 31 03:01:56 blue kernel: scsi0: Tagged Queuing now active for Target 0
--
I an running kernel 2.0.30 with the BusLogic BT-958 driver (as a module)...
my backup cron job loads st.o and then removes it after the backup program
(bru) excited... the above messages occur about 5 minutes after removing
st.o ... i tend to disregard them but if anyone has some explanation as to
what they mean, please let me know!
Does this also happen if st.o is compiled into the kernel?
BTW, i insert and remove the module manually rather than using kerneld (i
have 'alias char-major-9 off' in /etc/conf.modules) to make sure kerneld
does not remove the module while the tape is left unrewound for some time
(more than a minute) due to pre-backup file searches for instance... I
heard removing the module of a non-rewind tape device could possibly lead
to data loss... at least for ftape... is this still true with st? any
precision would be appreciated. thank you.
No idea about this.
Leonard
