search for: gl_culldistance

...ar *name, unsigned size, > _mesa_glsl_error(&loc, state, "`gl_ClipDistance' array size cannot " > "be larger than gl_MaxClipDistances (%u)", > state->Const.MaxClipPlanes); > + } else if (strcmp("gl_CullDistance", name) == 0 > + && size > state->Const.MaxClipPlanes) { > + /* From the ARB_cull_distance spec: > + * > + * "The gl_CullDistance array is predeclared as unsized and > + * must be sized by the shader either redeclaring...

This patch series adds the needed support for this extension to the various parts of mesa to finally enable it for nvc0. Dave Airlie (1): glsl: lower cull_distance into cull_distance_mesa Tobias Klausmann (10): glapi: add GL_ARB_cull_distance mesa/main: add support for GL_ARB_cull_distance mesa/prog: Add varyings for arb_cull_distance mesa/st: add support for GL_ARB_cull_distance

...ht > nvidia passed that test just fine. My take (and note that I last read the extension many months ago) is that you're supposed to figure out the max gl_ClipDistance[] written, and then write all your cull distances above that. So if you, e.g., have something like gl_ClipDistance[5] = 1; gl_CullDistance[0] = 1; Then it would decide that there are 6 clip distances (or if there's an explicit out float gl_ClipDistance[n], then use that), and 1 cull distance. In the TGSI, I'm thinking this might look approximately like PROPERTY CULL_MASK (1<<6) DCL OUT[0], CLIPDIST[0] DCL OUT[1], CLIPD...

On 24.05.2015 20:25, Ilia Mirkin wrote: > I'm having a bit of trouble tracing through this. What happens if I > have a shader that just does: > > gl_ClipDistance[0] = 1; > gl_CullDistance[0] = 1; > > what does the resulting TGSI look like? (Assuming that clip plane 0 is > enabled.) What about the generated nvc0 code (for the vertex shader)? (hack up a patch for this, run it without DRI_PRIME=1, see i pass and forget to check it again) yeah those are equal, sorry for wasti...

...tobias.johannes.klausmann at mni.thm.de> wrote: >> >> On 24.05.2015 20:25, Ilia Mirkin wrote: >>> I'm having a bit of trouble tracing through this. What happens if I >>> have a shader that just does: >>> >>> gl_ClipDistance[0] = 1; >>> gl_CullDistance[0] = 1; >>> >>> what does the resulting TGSI look like? (Assuming that clip plane 0 is >>> enabled.) What about the generated nvc0 code (for the vertex shader)? >> >> (hack up a patch for this, run it without DRI_PRIME=1, see i pass and forget >> to chec...

On 25 May 2015 at 08:11, Marek Olšák <maraeo at gmail.com> wrote: > It's the same on Radeon. There are 2x ClipOrCullDistance output > vectors and a mask saying it should clip or cull or do nothing. > > Marek > My thinking was gallium should have a single semantic and a mask in the shader definition maybe. though it doesn't solve the does nvidia do the right thing with

...and note that I last read the extension many months ago) is >> that you're supposed to figure out the max gl_ClipDistance[] written, >> and then write all your cull distances above that. So if you, e.g., >> have something like >> >> gl_ClipDistance[5] = 1; >> gl_CullDistance[0] = 1; >> >> Then it would decide that there are 6 clip distances (or if there's an >> explicit out float gl_ClipDistance[n], then use that), and 1 cull >> distance. In the TGSI, I'm thinking this might look approximately like >> >> PROPERTY CULL_MASK (1...