[5663] in Release_7.7_team
Re: ATA abnormal status error
daemon@ATHENA.MIT.EDU (Mitchell E Berger)
Mon Nov 27 14:39:29 2006
Message-Id: <200611271939.kARJdLc8028776@byte-me.mit.edu>
To: Matthew Goldstein <austein@MIT.EDU>
cc: release-team@MIT.EDU
In-Reply-To: Your message of "Mon, 27 Nov 2006 12:42:20 EST."
<Pine.LNX.4.62L.0611271234090.24575@dewey.mit.edu>
Date: Mon, 27 Nov 2006 14:39:21 -0500
From: Mitchell E Berger <mitchb@MIT.EDU>
X-Spam-Flag: NO
X-Spam-Score: 0.00
Hi,
> OLC has a question from somebody from the libraries who has been trying to
> install Athena on a Dell GX270 (unsupported configuration).
> I was told you might be interested in it, so I've copied the error message
> below. The question in OLCR is from natkins in case you want to take a
> look.
Unless you're 100% sure you've got all the useful information covered,
it saves work if you just give us everything the user sent. Important
things not said here are that the actual install succeeded, and this problem
only showed up on rebooting into Athena, and that this isn't verbatim
because the user left out things that "did not appear to be important."
> ATA: abnormal status 0xFF on port 0xFE07
> Mount: error 6 mounting ext3
> Mount: error 2 mounting none
> Switching to new root
> Switchroot: mount failed:22
> Moun /initrd/dev failed:22
> Kernel Panic
If not for the first line of that error message (the ATA abnormal status)
line, I'd say the symptoms sound similar to one of the problems described
by Garrett Barter in release-team[5358] where the initrd image didn't
include the SATA drivers... I wish he'd given us the actual error messages
he saw at the time. There's a "solution" in there that involves doing a
custom install, getting a prompt before the reboot, and manually running
mkinitrd again forcing the sd_mod, ata_piix, and ide-scsi modules into the
initrd. However,
1) Even if that were the solution, it may be too pointy-hat to give a user,
because I think any future kernel update would mess up the initrd again
unless they remembered to manually fix it before the reboot.
2) I think (hope) the reason Garrett was seeing that problem was that he was
using a hack to make the SATA drives look like IDE drives during the
install... if Noel isn't playing with legacy mode or anything for the
install to recognize his SATA drives, mkinitrd really ought to be doing
the right thing on its own. It might be interesting to know what the
SATA settings in his BIOS are.
3) If we were trying this, the 2.6.9-11.EL would need to be changed to
2.6.9-32.0.2.EL for 9.4.32.
4) The "ATA: abnormal status" line sort of sounds like his machine has the
drivers and is genuinely having a hardware problem to me. Again, I wish
we'd seen the exact error messages that Garrett got.
Hopefully Andrew will have more insight.
Mitch
