[Pm-utils] Problems with 98smart-kernel-video and intel graphics chipsets

[Victor Lowther]

On Thu, 2008-09-25 at 14:34 +0200, Michael Biebl wrote:
> 2008/9/21 Dan Nicholson <dbn.lists at gmail.com>:
> 
> >
> > Hang on, Michael. I don't think we're talking about kernel modesetting
> > here. Correct me if I'm wrong, but none of the modesetting patches
> > have landed in upstream linux, much less in 2.6.26. I don't think
> > Debian has backported the modesetting patches, right?
> 
> It might very well be, that I'm mixing things up here, that's why I
> asked for clarification.
> 
> I don't know, if Debian has any modesetting patches backported, but I doubt it.
> 
> >
> > What we're concerned with right now is just if the 915 drm driver in
> > recent vanilla kernels (2.6.26 and 2.6.27-rc*) is saving and restoring
> > all the state necessary around a suspend/resume cycle to get the
> > display up. Having all the modesetting done in the drm driver is a new
> > feature around the corner. Right now, you need some coordination
> > between the X video driver and the DRM driver to get the display
> > restored.
> 
> Going back to the original bug report I referenced, you see that the
> bug reporter was running a 2.6.26 kernel and has a 915 hardware.
> Apparently he still requires the vbe* quirks to successfully suspend and resume.
> 
> So, is there something within pm-utils that has to be done about this
> or is this a kernel bug?

Good question, and unless I hear that it is a kernel bug I will release
1.1.2.5 and 1.2.1 with a 98-smart-kernel-video that does not filter out
all the vbe* quirks for Intel hardware.

ETA for these releases right now is Friday, October 3.

> Michael
> 
> 
> 
-- 
Victor Lowther
Ubuntu Certified Professional