On Fri, 2009-10-30 at 11:19 -0400, Kristian Høgsberg wrote: > On Thu, Oct 29, 2009 at 5:52 PM, Adam Jackson <a...@nwnk.net> wrote: > > On Thu, 2009-10-29 at 14:01 -0700, keithp wrote: > >> Excerpts from Adam Jackson's message of Thu Oct 29 11:01:29 -0700 2009: > >> > >> > + screen->base.GLXmajor = 1; > >> > + screen->base.GLXminor = 4; > >> > >> Should this define be in a header somewhere? > > > > Maybe? It ends up being variable per renderer atm, so you'd end up with > > > > #define GLX_SWRAST_PROTOCOL_MINOR 4 > > #define GLX_DRI_PROTOCOL_MINOR 2 > > #define GLX_DRI2_PROTOCOL_MINOR 4 > > > > That doesn't seem especially useful, even as documentation. > > > > The whole thing is broken, in reality. pbuffer support and separable > > read/draw surfaces in the context are features that require some level > > of hardware (well, renderer) support, and we're not querying the > > renderer for their presence. > > No, pbuffer support is entirely in GLX and all DRI2 drivers support > separate draw and read drawables.
I didn't say we weren't giving the right answer, although I did say "in reality" when I should have said "in theory". I was trying to say we were being chummy with the implementation. Not all DRI1 drivers necessarily support separate read/draw (although I think they do), and I'm reasonably sure hardware has existed where the only offscreen renderbuffer was the backbuffer. It wouldn't be impossible to implement pbuffers in that environment, but, the point remains, it requires renderer support. Still, swrast _can_ do 1.4, and I don't see a lot of value in #define'ing that bit of truth. - ajax
signature.asc
Description: This is a digitally signed message part
_______________________________________________ xorg-devel mailing list xorg-devel@lists.x.org http://lists.x.org/mailman/listinfo/xorg-devel