i965: Implement gl_PatchVerticesIn by baking it into brw_tcs_prog_key.

The hardware provides us no decent way of getting at the number of input
vertices in the patch topology from the tessellation control shader.
It's actually very surprising - normally this sort of information would
be available in the thread payload.

For the precompile, we guess that the number of vertices will be the
same for both the input and output patches.  This usually seems to be
the case.

On Gen8+, we could pass in an extra push constant containing this value.
We may be able to do that on Haswell too.  It's quite a bit trickier on
Ivybridge, however.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>

src/mesa/drivers/dri/i965/brw_compiler.h

@@ -198,6 +198,8 @@ struct brw_tcs_prog_key
 
    GLenum tes_primitive_mode;
 
+   unsigned input_vertices;
+
    struct brw_sampler_prog_key_data tex;
 };
 

src/mesa/drivers/dri/i965/brw_tcs.c

@@ -65,6 +65,8 @@ brw_tcs_debug_recompile(struct brw_context *brw,
       return;
    }
 
+   found |= key_debug(brw, "input vertices", old_key->input_vertices,
+                      key->input_vertices);
    found |= key_debug(brw, "TES primitive mode", old_key->tes_primitive_mode,
                       key->tes_primitive_mode);
    found |= brw_debug_recompile_sampler_key(brw, &old_key->tex, &key->tex);
@@ -188,6 +190,7 @@ brw_upload_tcs_prog(struct brw_context *brw)
 
    if (!brw_state_dirty(brw,
                         _NEW_TEXTURE,
+                        BRW_NEW_PATCH_PRIMITIVE |
                         BRW_NEW_TESS_CTRL_PROGRAM |
                         BRW_NEW_TESS_EVAL_PROGRAM))
       return;
@@ -207,6 +210,8 @@ brw_upload_tcs_prog(struct brw_context *brw)
 
    key.program_string_id = tcp->id;
 
+   key.input_vertices = ctx->TessCtrlProgram.patch_vertices;
+
    /* _NEW_TEXTURE */
    brw_populate_sampler_prog_key_data(ctx, prog, stage_state->sampler_count,
                                       &key.tex);
@@ -251,6 +256,9 @@ brw_tcs_precompile(struct gl_context *ctx,
    key.program_string_id = btcp->id;
    brw_setup_tex_for_precompile(brw, &key.tex, prog);
 
+   /* Guess that the input and output patches have the same dimensionality. */
+   key.input_vertices = shader_prog->TessCtrl.VerticesOut;
+
    key.tes_primitive_mode = GL_TRIANGLES;
 
    success = brw_codegen_tcs_prog(brw, shader_prog, btcp, &key);

src/mesa/drivers/dri/i965/brw_vec4_tcs.cpp

@@ -257,7 +257,8 @@ vec4_tcs_visitor::nir_emit_intrinsic(nir_intrinsic_instr *instr)
            get_nir_dest(instr->dest, BRW_REGISTER_TYPE_UD));
       break;
    case nir_intrinsic_load_patch_vertices_in:
-      unreachable("XXX: gl_PatchVerticesIn not implemented yet.");
+      emit(MOV(get_nir_dest(instr->dest, BRW_REGISTER_TYPE_D),
+               brw_imm_d(key->input_vertices)));
       break;
    case nir_intrinsic_load_per_vertex_input: {
       src_reg indirect_offset = get_indirect_offset(instr);