[Intel-gfx] [patch] Fix XV scan line calculation when rotated

Barry Scott barry.scott at onelan.co.uk
Fri Jun 19 14:21:08 CEST 2009


Łukasz Kuryło wrote:
> Friday 19 of June 2009 13:07:43 Łukasz Kuryło napisał(a):
> > On czwartek 18 czerwca 2009 18:04:35 Barry Scott wrote:
> > > Łukasz Kuryło wrote:
> > > > On środa 17 czerwca 2009 18:52:48 you wrote:
> > > > > Łukasz Kuryło wrote:
> > > > > > Wednesday 17 of June 2009 18:18:33 Barry Scott napisał(a):
> > > > > > > The attached patch fixes the problems I was seeing with XV 
> when
> > > > rotated.
> > > > > > >
> > > > > > > The box is in screen coordinates and must be transformed 
> into crtc
> > > > > > > coordinates
> > > > > > > to calculate the correct scan line numbers.
> > > > > > >
> > > > > > > Barry
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > > > Hi Barry,
> > > > > > another issue is multiple monitor configuration.
> > > > > > What if box is not entirely in one crtc.
> > > > > > Are those
> > > > > > y1 = box_in_crtc_coordinates.y1 - crtc->y;
> > > > > > y2 = box_in_crtc_coordinates.y2 - crtc->y;
> > > > > > correct then?
> > > > > If it is a problem then the original and patched versions have 
> the same
> > > > > problem.
> > > > >
> > > > > Barry
> > > > >
> > > > >
> > > > Barry have you checked that patch?
> > > Yes I'm running with it applied.
> > > > Does it work when screen is inverted (upside down),
> > > Yes it works for all four rotations.
> > > > also isn't crtc->y in screen coordinates as well?
> > > I don't know, one of the intel developers would be best to comment 
> on that.
> > > The value while I was testing was always 0.
> > >
> > > Did you mean to reply off list?
> > >
> > > Barry
> > >
> > >
> >
> > Hi Barry,
> > in commit
> > e5bfa2702a31566fa94fa75f7289d7cbe9825420
> >
> > I've added clipping of scanlines I think in i830_video it would be 
> useful as well. But what bothers me is crtc->y value. As long as you 
> have one screen crtc->y is 0 and thats not a problem but please have a 
> look at this odg file.
> >
> > I hit reply instead of repry to list.
> >
>
>
> Or maybe after transformation to crtc coordinates, crtc->y is unnecesary?
We need someone that knows what crtc->y is used for to comment.
Then fixing the code to match will be simple.

Barry




More information about the Intel-gfx mailing list