Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 20 Oct 1998 21:48:11 -0500 (CDT)
From:      Vijay Ramasubramanian <ramasubr@ews.uiuc.edu>
To:        "Justin T. Gibbs" <gibbs@narnia.plutotech.com>
Cc:        current@FreeBSD.ORG
Subject:   Re: Installation probs w/3.0
Message-ID:  <Pine.GSO.3.96.981020211418.4544B-100000@eesn17.ews.uiuc.edu>
In-Reply-To: <199810201932.NAA29092@narnia.plutotech.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, 20 Oct 1998, Justin T. Gibbs wrote:

> Is this an ABP-940, an ABP-940U, an ABP-940UA, or an ABP-940UW?
> I have all three here, and I want to ensure that I use the same
> board you have to reproduce the bug.

This is an ABP-940UA.  BTW, what is the difference between U and UA?
BIOS version 2.8M.

AdvanSys BIOS settings: 

Adapter config: SCSI ID 7
                Spin Up Delay 1
                Termination Enabled (last device on SCSI chain, so this is
correct)
                PCI Burst Mode Enabled
                Multiple Drive Support Enabled
                Removable Media Support Disabled (no removable disks
are/will be attached)
                Bootable CD Support Disabled
                SCSI Parity Enabled (all devices on this bus support
parity and have it turned on)
                Ultra SCSI Enabled
                Host Queue Size 240 (this was the default value - can
I/should I set this any higher or lower?)
                Device Queue Size 32

Device Config: All SCSI IDs are set identically with the following
settings:

Start Unit Command - Yes
SCSI Disconnection - Yes
Command Queuing - Yes
Synchronous Transfer - Yes
BIOS Target Control - Yes

> Which of these drives is being accessed at the time of the failure?

The failure occurs while data is being written to the / filesystem, which
means that it is the 1st disk, the IBM drive at ID 0, being accessed.

> Does the Fujitsu claim to support tagged queuing?  Are there any
> SCSI error messages printed on the console?  You can get this
> information by hitting <scroll lock> and <Page Up> once you hit
> the install screen to see the kernel messages for the drives.

SCSI info from the device probe at boot is as follows:

adv0: <AdvanSYs ASC3050 Ultra SCSI controller> rev 0x03 int a irq 15 on
pci0.12.0

adv0: AdvanSys Ultra SCSI Host Adapter, SCSI ID 7, queue depth 240

da2 at adv0 bus 0 target 2 lun 0
da2: <FUJITSU M2623F-512 0403> Fixed Direct Access SCSI2 device
da2: 10.0MB/s transfers (10.0 MHz, offset 15), Tagged Queuing Enabled
da2: 405MB (830430 512 byte sectors: 64H 32S/T 405C)

da0 at adv0 bus 0 target 0 lun 0
da0: <IBM DCAS-32160 S61A> Fixed Direct Access SCSI2 device
da0: 20.0 MB/s transfers (20.0 MHz, offset 15), Tagged Queing Enabled
da0: 2063 MB (4226725 512 byte sectors: 255H 63S/T 263C)

da1's output is exactly the same as da0's, except that it is target 1

When the failures occur, I do not see any messages indicating SCSI errors
with the exception of the Fujitsu error messages which Mr. Smith indicated
might be due to the Synchronize Cache command not being supported but the
error being harmless.  Even if I take the Fujitsu off the SCSI bus (and
designate a partition on the 2nd IBM drive as Swap) these installation
failures occur.

I installed RedHat Linux 5.1 on the machine yesterday (also via FTP) to
try to test the hardware and make sure that it was not a problem with my
hardware.  The install of approximately 125 MB succeeded.  It did seem to
pause in places, but it continued fine all the way through.  (I've never
installed RedHat before so I don't know if there's anything unusual about
the pausing - I suspect not since the whole installation went through
without a hitch) The Linux install used the same partitioning scheme as I
was trying to use with FreeBSD.  It did not detect any bad blocks during
the formatting of the partitions (and I had the check for bad blocks
option enabled and monitored the debug consoles).  I will attempt a kernel
build and let you know the results of that. 

I can also install NT 4.0 onto the machine if you think that might serve
as a reasonable test of the hardware. 

What are your suspicions?

Thanks.

.______
| Vijay N. Ramasubramanian
   mailto:ramasubr@ews.uiuc.edu
    http://www.ews.uiuc.edu/~ramasubr/



To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.GSO.3.96.981020211418.4544B-100000>