On Sat, Feb 7, 2009 at 10:56 AM, Thomas Steffen
<[email protected]> wrote:
> Well, I nailed it - although using oprofile is not as easy as it could
> be. As expected the driver got stuck in the kernel, or more precisely
> in via_dma.c:via_wait_idle(). Shortening the timeout makes it so quick
> that it is no longer a noticeable delay. I used:
>
>       int count = 500000;
>
> down from 10000000.
>
> Ok, so it works, but why does it go wrong in the first place? I don't
> know. Well, I have a guess, but for some reason I cannot figure out my
> trace statements (DRM_ERROR) have no effect - so that remains
> difficult to analyse. I still hope that I can get a bit further there.
>
> And does it have any negative influence? I have not noticed any.
>
Hey,

If your board is based off the K8M890 then you will need some
different register checks.  I will try to get a patch worked up for
you soon.

Jon

_______________________________________________
openchrome-users mailing list
[email protected]
http://wiki.openchrome.org/mailman/listinfo/openchrome-users
Main page:
http://www.openchrome.org
Wiki:
http://wiki.openchrome.org
User Forum:
http://wiki.openchrome.org/tikiwiki/tiki-view_forum.php?forumId=1


Reply via email to