[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: i915 DRI broken after suspend/resume



On Wed, Jul 27, 2005 at 10:45:20AM +0200, David Martínez Moreno wrote:
> El Martes, 26 de Julio de 2005 05:20, Drew Parsons escribió:
> > Intel video driver (855GM etc) support is broken in Debian
> > (linux-tree-2.6  v2.6.12-1, xserver-xorg v6.8.2.dfsg.1-4) at the moment.
> > That's the i810 Xserver driver using the i915 kernel drm module.
> >
> > After a suspend/resume cycle, any OpenGL programs fail, e.g.
> >
> > $ glxgears
> > intelWaitIrq: drmI830IrqWait: -16
> >
> > A message appears in dmesg:
> >
> > [drm:i915_wait_irq] *ERROR* i915_wait_irq: EBUSY -- rec: 4229 emitted:
> > 4239
> >
> >
> > This is a known bug, reported at
> > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=159960
> > and
> > https://bugzilla.ubuntu.com/show_bug.cgi?id=7787
> > and
> > http://www.mail-archive.com/dri-devel@lists.sourceforge.net/msg21138.html
> >
> > A patch has apparently been made to X.org CVS, so we'd need to check it
> > has been applied in Debian's copy of 6.8.2 (Daniel Stone has apparently
> > applied it to Ubuntu).
> >
> > A kernel patch to i915 is also needed (provided by dri.freedesktop.org,
> > exact urls mentioned in the Redhat bug report).

They seem more like tarballs with new drivers than patches.
Which makes it somewhat difficult to isolate the fix.

> > I don't know if it's Debian's X.org or 2.6.12 kernel, or both, which
> > needs the patches, so I'm not filing a bug till we know.
> 
> 	Hello, Drew. We know that the i810 Xserver included in X.Org 6.8.2 is old and 
> does not support newer chips, so probably the problem will be in the X.Org 
> packages. We are currently working in the upcoming 6.9 X.Org release that we 
> hope to solve your problem (and several others).

Is there any value in the patches been added to the debian kernel
at this stage? And more to the point, is there any information
on the status of this change in the upstream (linus) kernel.

I heard on the grapevine that the dri/xorg thing is a bit of a mess
at the moment, can someone confirm or deny this?

-- 
Horms



Reply to: