RFC: Bug #4633 (Initial mouse pointer incorrect with EXA)

[Alan Hourihane]

On Fri, 2005-11-11 at 17:21 +0100, Michel Dänzer wrote:
> On Fri, 2005-11-11 at 11:05 +0000, Alan Hourihane wrote:
> > On Fri, 2005-11-11 at 11:51 +0100, Michel Dänzer wrote:
> > > 
> > >       * xf86RandRSetMode() might be able to take care of this by calling
> > >         something like UnrealizeCursor() before disabling FB access and
> > >         RealizeCursor() after re-enabling it.
> > 
> > I think I like this approach best. It'll also cure Thomas' problem with
> > the SiS rotation and hw cursor upload.
> 
> Thanks for reminding me of that, and thank you, Ben and Thomas for your
> feedback in general. You've convinced me that this is the best of the
> three approaches I listed. However, on second though, I think it would
> be even better if EnableDisableFBAccess() itself took care of it
> implicitly rather than xf86RandRSetMode() explicitly. That should also
> remove the need for kludges such as the fix for bug #4951, e.g.
> 
> Do you guys agree? If so, any preferences for the level at which
> EnableDisableFBAccess() should do this?

I agree that EnableDisableFBAccess is probably the better place to do
this too.

But there isn't much in that function, so UnrealizeCursor should get
called before the SetRootClip when disabling, and RealizeCursor should
come after the SetRootClip when enabling.

Alan.