[Call for testers] DRM device-independent code update to Linux 3.8

[J.R. Oldroyd]

On Tue, 24 Feb 2015 20:45:48 +0100 Jean-Sébastien Pédron <dumbbell at FreeBSD.org> wrote:
>
> On 21.02.2015 04:12, J.R. Oldroyd wrote:
> > But, booting does not go well:
> > 
> > Feb 20 21:58:22 xx kernel: panic: mtx_lock() of destroyed mutex @ /usr/src/sys/modules/drm2/drm2/../../../dev/drm2/drm_irq.c:1036
> 
> Could you please try the following patch?
> https://people.freebsd.org/~dumbbell/graphics/drm_fb_helper-mtx-destroyed-panic.1.patch
> 
> Apply it over your already patched source tree (not in place of
> drm-update-38.*.patch).
> 

With that patch, I am now able to boot without the mtx_lock() panic.
Thanks.

Other observations:

On a cold boot, the CP init now succeeds with no CAFEDEAD - wow!
However, on a reboot (without power off) the CP init is still *mostly*
failing with the CAFEDEAD - I've rebooted about 6 times and 4 times
got the CAFEDEAD.  This is still pointing at there being something not
quite right with the initialization sequence.  Chip is RS690 X1270 IGP,
and the CAFEDEAD is from r100_ring_test().

There are 10 LORs being logged on booting this revision (r279092),
but none related to drm code.

	-jr