Re: [PATCH 2/3] gl/msaa: Ignore the antialiasing property of the clip.
Chris Wilson <[email protected]>
| Newsgroups | gmane.comp.lib.cairo |
|---|---|
| Message-ID | <[email protected]> |
On Thu, 15 Dec 2011 10:49:27 -0800, Martin Robinson <[email protected]> wrote: > If the clip suggests anti-aliasing and we don't support it, > draw the clip anyway. Currently disabling anti-aliasing does > not disable anti-aliasing on certain clips. I added a couple of sentences to explain why this is justified, or rather why I think we can get away with this: /* We ignore the antialias mode of the clip here, since the user requested * unantialiased rendering of their path and we expect that this stencil * based rendering of the clip to be a reasonable approximation to * the intersection between that clip and the path. * * In other words, what the user expects when they try to perform * a geometric intersection between an unantialiased polygon and an * antialiased polygon is open to interpretation. And we choose the * fast option. */ Can anyone poke a hole in that argument? Or come up with a strict definition of that the intersection should be? In the mean time I've pushed both of the patches along with a switch back to spans by default. Use CAIRO_GL_COMPOSITOR=msaa to continue testing and improving the msaa compositor. (The reason is simply I don't have enough memory to keep the driver happy whilst running cairo-perf-trace.) Thanks, -Chris -- Chris Wilson, Intel Open Source Technology Centre -- cairo mailing list [email protected] http://lists.cairographics.org/mailman/listinfo/cairo