Isn't this incredibly at odds with our previous discussion, in which we generally agreed to not advertise support for unaccelerated things? Geometry shaders are simply not going to have any kind of speed on i915, nv30, nv40, or r300.
I can't speak for Jakob or the Nouveau guys, but at least r300g is trying to keep Draw usage to a minimum. I know i915 has to use Draw, so that won't be a problem, but nv30 and nv40 are Draw-free IIRC. ~ C. On Thu, Dec 24, 2009 at 6:09 AM, Keith Whitwell <kei...@vmware.com> wrote: > Zack, > > In terms of advertising support for this - I wonder if this isn't something > we should just turn on for all drivers, given that in the worst case it is > just a vertex path fallback, and a lot of drivers will be hitting those for > various reasons anyway. > > Keith > ________________________________________ > From: Zack Rusin [za...@vmware.com] > Sent: Thursday, December 24, 2009 5:24 AM > To: mesa3d-dev > Cc: Keith Whitwell; Michal Krol > Subject: geometry shading patches > > yo, > > after our discussions i hacked a new version of geometry shading support in > gallium. the new geometry shading syntax looks as follows: > GEOM > PROPERTY GS_INPUT_PRIMITIVE TRIANGLES > PROPERTY GS_OUTPUT_PRIMITIVE TRIANGLE_STRIP > DCL IN[][0], POSITION, CONSTANT > DCL IN[][1], COLOR, CONSTANT > DCL OUT[0], POSITION, CONSTANT > DCL OUT[1], COLOR, CONSTANT > 0:MOV OUT[0], IN[0][0] > 1:MOV OUT[1], IN[0][1] > 2:EMIT > 3:MOV OUT[0], IN[1][0] > 4:MOV OUT[1], IN[1][1] > 5:EMIT > 6:MOV OUT[0], IN[2][0] > 7:MOV OUT[1], IN[2][1] > 8:EMIT > 9:ENDPRIM > 10:END > the attached patch series adds support for all of it throughout gallium > (including support for 2dimensional arrays in tgsi_sanity checking, tgsi_dump > and tgsi_text). > Right now don't support switching of primitives in the draw module (e.g. input > primitive being points and output primitive being a triangle_strip) but i'd > like to do that after committing the attached changes (also the semantic for > vertices which is a leftover from an old patch should be removed then if it's > ok with you). > > Keith, most relevant patches for you are 0001 and 0005. > Michal the most relevant to you are 0007, 0010, 0011 and 0012. > > z > > ------------------------------------------------------------------------------ > 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 > -- Only fools are easily impressed by what is only barely beyond their reach. ~ Unknown Corbin Simpson <mostawesomed...@gmail.com> ------------------------------------------------------------------------------ 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
