[4982] in linux-scsi channel archive
AIC-7880 Ultra parity error w/Linux 2.1.12x
daemon@ATHENA.MIT.EDU (Roy C Bixler)
Fri Oct 23 13:11:00 1998
Date: Fri, 23 Oct 1998 12:03:52 -0500 (CDT)
From: Roy C Bixler <rcb@press-gopher.uchicago.edu>
To: linux-scsi@vger.rutgers.edu
Hi:
I have been using the 2.1 series or kernels because I have generally found
them to be stable and they do seem to perform better than 2.0.xx. However,
with 2.1.12x revisions, I have started seeing unexplained SCSI errors.
The machine is a Dell Poweredge 2200 with Pentium II 233 MHz UP CPU,
Adaptec AIC-7880 Ultra controller and a couple of Seagate 4 Gig. drives
attached. I also have a tape drive and CD-ROM on the same SCSI chain, but
they aren't used when the problem happens. In fact, the problem generally
occurs during periods of low usage.
The following sequence was under 2.1.125 UP, but I also saw something
similar under 2.1.122.
Oct 21 01:37:51 press kernel: (scsi0:0:0:0) Parity error during Message-In phase.
Oct 21 01:37:51 press kernel: (scsi0:0:0:-1) Unexpected busfree, LASTPHASE = 0xa0, SEQADDR = 0x155
Oct 21 01:39:06 press kernel: (scsi0:0:0:0) Parity error during Data-In phase.
Oct 21 01:39:06 press kernel: (scsi0:0:0:0) Synchronous at 40.0 Mbyte/sec, offset 8.
Oct 21 01:39:36 press kernel: (scsi0:0:0:0) Parity error during Message-In phase.
Oct 21 01:39:36 press kernel: (scsi0:0:0:0) Parity error during Data-In phase.
Oct 21 01:40:36 press kernel: (scsi0:0:0:0) Parity error during Message-In phase.
Oct 21 01:40:36 press kernel: (scsi0:0:0:0) Parity error during Data-In phase.
Oct 21 01:40:36 press kernel: (scsi0:0:0:0) Parity error during Message-In phase.
Oct 21 01:40:36 press kernel: (scsi0:0:0:-1) Unexpected busfree, LASTPHASE = 0xa0, SEQADDR = 0x155
Oct 21 01:40:40 press kernel: (scsi0:0:0:0) Parity error during Message-In phase.
Oct 21 01:40:40 press kernel: (scsi0:0:0:0) Parity error during Data-In phase.
Oct 21 01:40:40 press kernel: (scsi0:0:0:0) Parity error during Data-In phase.
Oct 21 01:40:42 press kernel: scsi0 channel 0 : resetting for second half of retries.
Oct 21 01:40:42 press kernel: SCSI bus is being reset for host 0 channel 0.
Oct 21 01:40:42 press kernel: scsi : aborting command due to timeout : pid 2562122, scsi0, channel 0, id 0, lun 0 Read (10) 00 00 43 1d 31 00 00 02 00
Oct 21 01:40:42 press kernel: (scsi0:0:0:0) Synchronous at 40.0 Mbyte/sec, offset 8.
Oct 21 01:40:42 press kernel: (scsi0:0:0:0) Parity error during Command phase.
Oct 21 01:40:42 press kernel: (scsi0:-1:-1:-1) CMDCMPLT without command for SCB 9, SCB flags 0x0, cmd 0x0
Oct 21 01:40:59 press kernel: (scsi0:0:1:0) Synchronous at 40.0 Mbyte/sec, offset 8.
Later on, I saw that the 'updatedb' job that runs at 6:40 each morning was
stuck. If I did a 'find' command manually, it got wedged in a 'D' state
and was unkillable. Subsequent to this, I upgraded to 2.1.126pre2 since I
saw there was an update to the AIC7xxx driver (v. 5.1.2). I got the
following message then
Oct 23 07:31:15 press kernel: (scsi0:0:0:0) Parity error during Message-In phase.
Oct 23 07:31:15 press kernel: (scsi0:0:0:-1) Unexpected busfree, LASTPHASE = 0xa0, SEQADDR = 0x155
About an hour later, the system locked up hard.
I don't think the problem is with the drive, although it's a possibility.
When I searched all the newsgroups and mailing list archives I could find,
I only turned up a suggestion in 'comp.os.linux.hardware' that this
problem began with 2.1.12x kernels. This seems to match my experience.
Are there any new fixes I should try or should I just use an earlier
version of the kernel?
Thanks ...
---
Roy Bixler
The University of Chicago Press
rcb@press-gopher.uchicago.edu
-
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@vger.rutgers.edu
