From owner-freebsd-current  Thu Jan  6 23:46:17 2000
Delivered-To: freebsd-current@freebsd.org
Received: from freebsd.dk (freebsd.dk [212.242.42.178])
	by hub.freebsd.org (Postfix) with ESMTP id 12C5D1522B
	for <current@FreeBSD.ORG>; Thu,  6 Jan 2000 23:46:14 -0800 (PST)
	(envelope-from sos@freebsd.dk)
Received: (from sos@localhost)
	by freebsd.dk (8.9.3/8.9.1) id IAA20488;
	Fri, 7 Jan 2000 08:46:09 +0100 (CET)
	(envelope-from sos)
From: Soren Schmidt <sos@freebsd.dk>
Message-Id: <200001070746.IAA20488@freebsd.dk>
Subject: Re: Found it --- 'ata' driver screwing up the machine (was Re: New kernel no longer boots on one of my machines... ata, other proble
In-Reply-To: <200001070655.WAA82937@apollo.backplane.com> from Matthew Dillon at "Jan 6, 2000 10:55:12 pm"
To: dillon@apollo.backplane.com (Matthew Dillon)
Date: Fri, 7 Jan 2000 08:46:09 +0100 (CET)
Cc: current@FreeBSD.ORG
X-Mailer: ELM [version 2.4ME+ PL54 (25)]
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
Sender: owner-freebsd-current@FreeBSD.ORG
Precedence: bulk
X-Loop: FreeBSD.ORG

It seems Matthew Dillon wrote:
> 
> :    From what I can tell so far, something during the kernel boot is
> :    disabling the timer interrupt.  The ATA probe does a tsleep() which
> :    never times out.  Output is queued to the console during the boot
> :    sequence which is never printed (until I CTL-ALT-ESC)... 
> :
> :    has someone messed with spl*()'s anywhere recently?
> 
>     Another datapoint:  When I DDB the machine and 'print *ticks',
>     it comes up 0, and when I panic it it says "uptime 0s".
> 
>     If I remove the 'ata' drivers I get the bootconf screen
>     since there is no root - but *ticks is incrementing!
> 
>     If I add 'ata' back in, buckus (DDB> print *ticks yields 0, machine
>     is screwed up).
> 
>     When I remove 'ata' and enable the 'wd' driver, everything works 
>     normally.  The machine boots up fine.

Thats to be expected, the wd driver doesn't use config_intrhook_establish()
to hook in the probes when interrupts should be running.
From your previous mail it looked like you had both ATA & SCSI in there,
then tried without SCSI and it kindof worked. The error 6 problem has
been up here before, but I dont remember the solution off hand, but
you should use /dev/ad* entries in you fstab, and have MAKEDEV'd 
them fairly resently, new bootblocks and loader are also a must.
My best guess is that config_intrhook_establish() screws up something
when more than one consumer uses it to hook into the boot process...

-Søren


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