Date: Sat, 21 Jan 2006 23:14:27 -0500
From: Kurt Wall <kwall@kurtwerks.com>
To: Dave Jones <davej@redhat.com>, Erwin Rol <mailinglists@erwinrol.com>,
       Andrew Morton <akpm@osdl.org>, "Carlo E. Prelz" <fluido@fluido.as>,
       linux-kernel@vger.kernel.org
Subject: Re: ATI RS480-based motherboard: stuck while booting with kernel >= 2.6.15 rc1
Message-ID: <20060122041427.GA6183@kurtwerks.com>
Mail-Followup-To: Dave Jones <davej@redhat.com>,
	Erwin Rol <mailinglists@erwinrol.com>, Andrew Morton <akpm@osdl.org>,
	"Carlo E. Prelz" <fluido@fluido.as>, linux-kernel@vger.kernel.org
References: <20060120123202.GA1138@epio.fluido.as> <20060121010932.5d731f90.akpm@osdl.org> <20060121125741.GA13470@epio.fluido.as> <20060121125822.570b0d99.akpm@osdl.org> <1137878926.2976.28.camel@xpc.home.erwinrol.com> <20060121220402.GC28051@redhat.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20060121220402.GC28051@redhat.com>
User-Agent: Mutt/1.4.2.1i
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1508
Lines: 32

On Sat, Jan 21, 2006 at 05:04:02PM -0500, Dave Jones took 21 lines to write:
> On Sat, Jan 21, 2006 at 10:28:46PM +0100, Erwin Rol wrote:
>  > I had also had the problem that my Shuttle ST20G5 (a RS480 IGP based
>  > system) hung in pci_init. This was with one of the Fedora Rawhide
>  > kernels, after reporting it  Dave Jones fixed it cause the next rawhide
>  > kernel worked again, maybe he could explain what it was, and where the
>  > fix is (if it is the same thing, but it really looks like it).
> 
> That was due to us carrying one of the 'make the clock not tick
> at twice the speed on ati chipsets' patches. Matthew Garrett's variant iirc.
> It worked fine in .14, but caused havoc in .15+
> 
> I put it down to the problem being fixed in other ways upstream.

Heh. I think you mean s/fixed/worked around/.

The bugme tracking this, 3927, is a joyous muddle of patches, 
patches to patches, workarounds, and dueling command line options. 
I ported this patch,
http://www.ussg.iu.edu/hypermail/linux/kernel/0504.0/1625.html,
to 2.6.16-rc1 to fix my pain, but 1) it didn't work without the
disable_timer_pin_1 kernel option and, 2) I don't think it's a fix 
so much as a workaround.

Kurt
-- 
I can read your mind, and you should be ashamed of yourself.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/