[3183] in linux-scsi channel archive
Re: AHA1542CF suddenly stopped working! Help!
daemon@ATHENA.MIT.EDU (Scott Kozicki)
Tue Feb 3 12:13:41 1998
Date: Tue, 03 Feb 1998 10:57:23 -0600
From: Scott Kozicki <SKozicki@bluestar.net>
Reply-To: SKozicki@bluestar.net
To: Jay Vassos-Libove <libove@felines.org>
CC: redhat-list@redhat.com, linux-scsi@vger.rutgers.edu
Jay Vassos-Libove wrote:
>
> I have recently added a Pentium 133 to my (old) collection. I plugged an
> ISA bus Adaptec AHA1542CF in to it, added a couple of drives, and away I
> went. It was working great, until a few hours ago.
>
> I did not do anything to it. I shut the system down, added some more RAM,
> turned it back on, and all was great. Then I shut it down to take out the
> video board so that I could read the chipset numbers off of it, put the
> board back in, turned it back on, and since then (despite a dozen power
> cycles and board pulls) all I can get it to do is to load the Linux
> kernel, and get as far as probing the SCSI bus and it always gets the
> error:
>
> scsi : aborting command due to timeout : pid 0, scsi0, channel 0, id 0,
> lun 0 Test Unit Ready 00 00 00 00 00
>
> I have removed the second device from the bus (Hitachi DK516C) - no
> effect. I have removed the first device from the bus (Seagate Medalist
> Pro ST52160N) - no effect. I have removed both devices from the bus and
> put a different device on the bus as device zero (HP 307M old thing) - no
> effect.
>
> Clearly, it isn't the specific device(s) on the bus that is the cause of
> this problem. Software did not change. The data on the drives is okay --
> I moved one of them to another system with a compatible controller
> (AHA2842VL) and it mounts happily and is in use presently.
>
> What could have happenned which
>
> - allows the hardware self test, including disk surface verification, to
> work find
>
> - allows the SCSI Controller DMA test to run loops and loops without error
>
> + yet results in the Linux kernel booting only as far as the AHA1542
> driver, and then wiping out as described above
>
> + given that I have not changed anything!
>
> Heelllppppp!!!!!!!!
I'm by no means an expert here, but I've had similar problems. After
about a week solid of trying to get a Linux box up using the AHA1542, I
just gave up. I had tried everything, even installing on an IDE drive,
building a custom kernal with support for the AHA, and trying to boot
off of it. It didn't work. I got the same error you did. I think that
this is a timing issue. The board is already notorious for being
extremely delicate. While it's amazing that it would stop working the
way you described, it's not all that surprising given it's reputation.
I've never had a problem with the board in other OS's. It's worked
admirably. But Linux obviously exacerbates something bad about the
board.
So in the meantime, I'm looking for used SCSI controllers. =)
entertaining all offers,
~sk
--
"Do not meddle in the affairs of wizards,
for you are crunchy and taste good with ketchup."
-- Joe Klemmer, Network Wizard
