Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 18 Feb 1999 00:06:47 -0700
From:      Brett Glass <brett@lariat.org>
To:        Mike Smith <mike@smith.net.au>
Cc:        chat@freebsd.org
Subject:   Re: Walnut Creek, Where Are You? 
Message-ID:  <4.1.19990217234718.0401de40@mail.lariat.org>
In-Reply-To: <199902180609.WAA03463@dingo.cdrom.com>
References:  <Your message of "Wed, 17 Feb 1999 23:06:37 MST."             <4.1.19990217225925.0401f9d0@mail.lariat.org>

next in thread | previous in thread | raw e-mail | index | archive | help
At 10:09 PM 2/17/99 -0800, Mike Smith wrote:
 
>Brett, I take back everything I said about offering you hardware to 
>write drivers with.  If you're that far behind the ball with the way 
>that PCI works, it wouldn't be even vaguely economical.

It has nothing to do with PCI. A probe for something else ENTIRELY
could be mucking things up. No, it's not "supposed" to happen, but
it does. Take it from an EE who's designed chips and motherboards:
a VGA driver or a sound card driver can mess up the works in subtle
ways. Heck, you might even find that the machine never hangs when
booted with, say, a serial console and no video card.

>Just accept that courtesy of the marvels of modern PnP architectures, 
>what you are fretting about is effectively impossible.

There's a reason why they call it "Plug 'n Pray." Scans for various
sorts of legacy hardware *can* mess up a motherboard. And IBM's
motherboards are sometimes weird.

This is why one of the first things you should try -- if you haven't
already -- is disabling nonessential, seemingly unrelated drivers in 
the kernel. If you're lucky, you'll find that removing one of them
mysteriously makes the problem go away. If you're not, you'll have 
to check for other problems such as race conditions, etc. I'd have 
to observe the problem occurring, and perhaps single-step the
machine through the problem code while looking at the spec sheet, to 
develop other ideas.

--Brett Glass



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




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4.1.19990217234718.0401de40>