On Mon, 2010-01-11 at 11:14 -0800, Brian Paul wrote: > Keith Whitwell wrote: > > Module: Mesa > > Branch: gallium-fb-dimensions > > Commit: 609b043d442c99e48d5310244e648ea8a6cc2e8a > > URL: > > http://cgit.freedesktop.org/mesa/mesa/commit/?id=609b043d442c99e48d5310244e648ea8a6cc2e8a > > > > Author: Keith Whitwell <kei...@vmware.com> > > Date: Mon Jan 11 11:03:44 2010 +0000 > > > > gallium: doc tweaks for bypass_vs_clip_and_viewport > > > > --- > > > > src/gallium/docs/source/cso/rasterizer.rst | 10 +++++++--- > > 1 files changed, 7 insertions(+), 3 deletions(-) > > > > diff --git a/src/gallium/docs/source/cso/rasterizer.rst > > b/src/gallium/docs/source/cso/rasterizer.rst > > index 4d8e170..2ca9b64 100644 > > --- a/src/gallium/docs/source/cso/rasterizer.rst > > +++ b/src/gallium/docs/source/cso/rasterizer.rst > > @@ -133,9 +133,13 @@ implementation-dependent, but will usually be Gourard > > for most hardware. > > bypass_vs_clip_and_viewport > > ^^^^^^^^^^^^^^^^^^^^^^^^^^^ > > > > -When set, this implies that vertices are pre-transformed for the viewport, > > and > > -will not be run through the vertex shader. Note that implementations may > > still > > -clip away vertices that are not visible. > > +When set, this implies that vertices are pre-transformed for the > > +viewport, and will not be run through the vertex shader. Note that > > +these vertices will fall outside the conventional unit-cube clip > > +space. Implementations are expected to discard pixels which fall > > +outside the viewport (XXX: how is the viewport defined in this case), > > +and may if desired choose to subject these vertices to clipping after > > +appropriate transformation to the clipping cube. > > The viewport doesn't describe a clipping rect. It just happens that > tris clipped to the NDC cube lie fully inside the viewport. Wide > lines and points will generate vertices and fragments outside the > viewport.
OK, I didn't think of that case. > Do we really want to require that the viewport do fragment clipping too? No. What I was trying to do (fairly poorly) is figure out what actually are the semantics of the bypass_vs_clip_and_viewport flag, and word it in such a way that it could be implemented even on hardware that doesn't allow you to turn off the vertex shader or clipping -- ie permit an implementation where we squeeze the incoming vertices down to the clipping cube and then puff them back out with the viewport afterwards. > I'd say it's up to the user to also set up a scissor rect if he really > doesn't want to see any fragments outside the viewport. Yeah. I wasn't really trying to introduce new behaviours. I often feel life would be easier without bypass_vs_clip_viewport, but it is also quite convenient for state trackers and blitters to have this mode available. Keith ------------------------------------------------------------------------------ This SF.Net email is sponsored by the Verizon Developer Community Take advantage of Verizon's best-in-class app development support A streamlined, 14 day to market process makes app distribution fast and easy Join now and get one step closer to millions of Verizon customers http://p.sf.net/sfu/verizon-dev2dev _______________________________________________ Mesa3d-dev mailing list Mesa3d-dev@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/mesa3d-dev
