r600g: Implement sm5 geometry shader instancing

Requires Evergreen or later hardware.

Signed-off-by: Glenn Kennard <glenn.kennard@gmail.com>

diff --git a/docs/GL3.txt b/docs/GL3.txt
index 3d7a450..f5d5e72 100644 (file)
--- a/docs/GL3.txt
+++ b/docs/GL3.txt
@@ -104,7 +104,7 @@ GL 4.0, GLSL 4.00:
   - Fused multiply-add                                 DONE ()
   - Packing/bitfield/conversion functions              DONE (r600)
   - Enhanced textureGather                             DONE (r600, radeonsi)
-  - Geometry shader instancing                         DONE ()
+  - Geometry shader instancing                         DONE (r600)
   - Geometry shader multiple streams                   DONE ()
   - Enhanced per-sample shading                        DONE (r600)
   - Interpolation functions                            DONE ()

diff --git a/src/gallium/drivers/r600/evergreen_state.c b/src/gallium/drivers/r600/evergreen_state.c
index 9f0e82d..8ce31dc 100644 (file)
--- a/src/gallium/drivers/r600/evergreen_state.c
+++ b/src/gallium/drivers/r600/evergreen_state.c
@@ -2963,8 +2963,8 @@ void evergreen_update_gs_state(struct pipe_context *ctx, struct r600_pipe_shader
 
        if (rctx->screen->b.info.drm_minor >= 35) {
                r600_store_context_reg(cb, R_028B90_VGT_GS_INSTANCE_CNT,
-                               S_028B90_CNT(0) |
-                               S_028B90_ENABLE(0));
+                               S_028B90_CNT(MIN2(rshader->gs_num_invocations, 127)) |
+                               S_028B90_ENABLE(rshader->gs_num_invocations > 0));
        }
        r600_store_context_reg_seq(cb, R_02891C_SQ_GS_VERT_ITEMSIZE, 4);
        r600_store_value(cb, cp_shader->ring_item_size >> 2);

diff --git a/src/gallium/drivers/r600/r600_shader.c b/src/gallium/drivers/r600/r600_shader.c
index 3f089b4..9f10c20 100644 (file)
--- a/src/gallium/drivers/r600/r600_shader.c
+++ b/src/gallium/drivers/r600/r600_shader.c
@@ -706,6 +706,8 @@ static int tgsi_declaration(struct r600_shader_ctx *ctx)
                        break;
                } else if (d->Semantic.Name == TGSI_SEMANTIC_VERTEXID)
                        break;
+               else if (d->Semantic.Name == TGSI_SEMANTIC_INVOCATIONID)
+                       break;
        default:
                R600_ERR("unsupported file %d declaration\n", d->Declaration.File);
                return -EINVAL;
@@ -811,6 +813,12 @@ static void tgsi_src(struct r600_shader_ctx *ctx,
                        r600_src->swizzle[2] = 0;
                        r600_src->swizzle[3] = 0;
                        r600_src->sel = 0;
+               } else if (ctx->info.system_value_semantic_name[tgsi_src->Register.Index] == TGSI_SEMANTIC_INVOCATIONID) {
+                       r600_src->swizzle[0] = 3;
+                       r600_src->swizzle[1] = 3;
+                       r600_src->swizzle[2] = 3;
+                       r600_src->swizzle[3] = 3;
+                       r600_src->sel = 1;
                }
        } else {
                if (tgsi_src->Register.Indirect)
@@ -1753,6 +1761,9 @@ static int r600_shader_from_tgsi(struct r600_context *rctx,
                        case TGSI_PROPERTY_GS_MAX_OUTPUT_VERTICES:
                                shader->gs_max_out_vertices = property->u[0].Data;
                                break;
+                       case TGSI_PROPERTY_GS_INVOCATIONS:
+                               shader->gs_num_invocations = property->u[0].Data;
+                               break;
                        }
                        break;
                default:

diff --git a/src/gallium/drivers/r600/r600_shader.h b/src/gallium/drivers/r600/r600_shader.h
index d6db8f0..4b27ede 100644 (file)
--- a/src/gallium/drivers/r600/r600_shader.h
+++ b/src/gallium/drivers/r600/r600_shader.h
@@ -74,6 +74,7 @@ struct r600_shader {
        unsigned                gs_input_prim;
        unsigned                gs_output_prim;
        unsigned                gs_max_out_vertices;
+       unsigned                gs_num_invocations;
        /* size in bytes of a data item in the ring (single vertex data) */
        unsigned                ring_item_size;