Re: [PATCH] PCI PM: Restore standard config registers of all devices early (was: Re: EeePC resume failure - timers)

[Jesse Barnes <jbarnes@virtuousgeek.org>]

On Tuesday, January 20, 2009 12:15 pm Jesse Barnes wrote:
> On Tuesday, January 20, 2009 12:09 pm Jesse Barnes wrote:
> > On Tuesday, January 20, 2009 11:45 am Linus Torvalds wrote:
> > > On Tue, 20 Jan 2009, Jesse Barnes wrote:
> > > > Hm it's probably not display programming related then... Sounds like
> > > > the i830_update_dri_buffers() call is failing for some reason.  The
> > > > kernel driver does do some I/O mapping stuff at enter/leavevt time
> > > > (and therefore suspend/resume time), can you reproduce the issue just
> > > > by VT switching?  If so you might be able to see something useful in
> > > > the kernel log...
> > >
> > > I was going to say that I've been VT switching and it has worked, but I
> > > decided to double-check. And yes, I can reproduce it that way. And no,
> > > no kernel messages that are visible that way either.
> >
> > So the machine is alive enough for you to see the logs?  Can you get a
> > backtrace (both kernel & userspace)?
>
> Oops, just realized by "no messages visible" you probably meant that it
> really was hung rather than you saw no useful messages.  That'll make
> things more difficult.  Which 2D driver are you running?

And I assume you've tried keeping an ssh session open on the machine while VT 
switching (that's *usually* sufficient for me, in all but the worst of 
crashes)...