broadcom/vc5: Simplify separate stencil surface setup.

If we just make another gallium surface for the separate stencil, it's a
lot easier to keep track of which set of fields we're using in RCL setup.

This also incidentally fixes a little bug in setting up the surface's
padded height for separate stencil when the UIF-ness changes at different
levels of Z versus stencil.

3 files changed, 77 insertions, 99 deletions

diff --git a/src/gallium/drivers/vc5/vc5_rcl.c b/src/gallium/drivers/vc5/vc5_rcl.c
index 8018bfdcba..86ea22628d 100644
--- a/src/gallium/drivers/vc5/vc5_rcl.c
+++ b/src/gallium/drivers/vc5/vc5_rcl.c
@@ -38,39 +38,29 @@ static void
 load_general(struct vc5_cl *cl, struct pipe_surface *psurf, int buffer)
 {
         struct vc5_surface *surf = vc5_surface(psurf);
-        struct vc5_resource *rsc = vc5_resource(psurf->texture);
+        bool separate_stencil = surf->separate_stencil && buffer == STENCIL;
+        if (separate_stencil) {
+                psurf = surf->separate_stencil;
+                surf = vc5_surface(psurf);
+        }
 
-        struct vc5_resource *separate_stencil = NULL;
-        if (rsc->separate_stencil && buffer == STENCIL)
-                separate_stencil = rsc->separate_stencil;
+        struct vc5_resource *rsc = vc5_resource(psurf->texture);
 
         cl_emit(cl, LOAD_TILE_BUFFER_GENERAL, load) {
                 load.buffer_to_load = buffer;
-                if (separate_stencil) {
-                        load.address = cl_address(separate_stencil->bo,
-                                                  surf->separate_stencil_offset);
-                } else {
-                        load.address = cl_address(rsc->bo, surf->offset);
-                }
+                load.address = cl_address(rsc->bo, surf->offset);
 
 #if V3D_VERSION >= 40
-                if (separate_stencil) {
+                load.memory_format = surf->tiling;
+                if (separate_stencil)
                         load.input_image_format = V3D_OUTPUT_IMAGE_FORMAT_S8;
-                        load.memory_format = surf->separate_stencil_tiling;
-                }  else {
+                else
                         load.input_image_format = surf->format;
-                        load.memory_format = surf->tiling;
-                }
 
                 if (surf->tiling == VC5_TILING_UIF_NO_XOR ||
                     surf->tiling == VC5_TILING_UIF_XOR) {
-                        if (separate_stencil) {
-                                load.height_in_ub_or_stride =
-                                        surf->separate_stencil_padded_height_of_output_image_in_uif_blocks;
-                        } else {
-                                load.height_in_ub_or_stride =
-                                        surf->padded_height_of_output_image_in_uif_blocks;
-                        }
+                        load.height_in_ub_or_stride =
+                                surf->padded_height_of_output_image_in_uif_blocks;
                 } else if (surf->tiling == VC5_TILING_RASTER) {
                         struct vc5_resource_slice *slice =
                                 &rsc->slices[psurf->u.tex.level];
@@ -92,24 +82,19 @@ store_general(struct vc5_job *job,
               int pipe_bit, bool last_store, bool general_color_clear)
 {
         struct vc5_surface *surf = vc5_surface(psurf);
+        bool separate_stencil = surf->separate_stencil && buffer == STENCIL;
+        if (separate_stencil) {
+                psurf = surf->separate_stencil;
+                surf = vc5_surface(psurf);
+        }
+
         struct vc5_resource *rsc = vc5_resource(psurf->texture);
 
-        struct vc5_resource *separate_stencil = NULL;
-        if (rsc->separate_stencil && buffer == STENCIL) {
-                separate_stencil = rsc->separate_stencil;
-                separate_stencil->writes++;
-        } else {
-                rsc->writes++;
-        }
+        rsc->writes++;
 
         cl_emit(cl, STORE_TILE_BUFFER_GENERAL, store) {
                 store.buffer_to_store = buffer;
-                if (separate_stencil) {
-                        store.address = cl_address(separate_stencil->bo,
-                                                   surf->separate_stencil_offset);
-                } else {
-                        store.address = cl_address(rsc->bo, surf->offset);
-                }
+                store.address = cl_address(rsc->bo, surf->offset);
 
 #if V3D_VERSION >= 40
                 store.clear_buffer_being_stored =
@@ -117,23 +102,17 @@ store_general(struct vc5_job *job,
                          (general_color_clear ||
                           !(pipe_bit & PIPE_CLEAR_COLOR_BUFFERS)));
 
-                if (separate_stencil) {
+                if (separate_stencil)
                         store.output_image_format = V3D_OUTPUT_IMAGE_FORMAT_S8;
-                        store.memory_format = surf->separate_stencil_tiling;
-                }  else {
+                else
                         store.output_image_format = surf->format;
-                        store.memory_format = surf->tiling;
-                }
+
+                store.memory_format = surf->tiling;
 
                 if (surf->tiling == VC5_TILING_UIF_NO_XOR ||
                     surf->tiling == VC5_TILING_UIF_XOR) {
-                        if (separate_stencil) {
-                                store.height_in_ub_or_stride =
-                                        surf->separate_stencil_padded_height_of_output_image_in_uif_blocks;
-                        } else {
-                                store.height_in_ub_or_stride =
-                                        surf->padded_height_of_output_image_in_uif_blocks;
-                        }
+                        store.height_in_ub_or_stride =
+                                surf->padded_height_of_output_image_in_uif_blocks;
                 } else if (surf->tiling == VC5_TILING_RASTER) {
                         struct vc5_resource_slice *slice =
                                 &rsc->slices[psurf->u.tex.level];
@@ -433,7 +412,37 @@ v3d_setup_render_target(struct vc5_job *job, int cbuf,
         *rt_type = surf->internal_type;
         *rt_clamp = V3D_RENDER_TARGET_CLAMP_NONE;
 }
-#endif /* V3D_VERSION >= 40 */
+
+#else /* V3D_VERSION < 40 */
+
+static void
+v3d_emit_z_stencil_config(struct vc5_job *job, struct vc5_surface *surf,
+                          struct vc5_resource *rsc, bool is_separate_stencil)
+{
+        cl_emit(&job->rcl, TILE_RENDERING_MODE_CONFIGURATION_Z_STENCIL_CONFIG, zs) {
+                zs.address = cl_address(rsc->bo, surf->offset);
+
+                if (!is_separate_stencil) {
+                        zs.internal_type = surf->internal_type;
+                        zs.output_image_format = surf->format;
+                } else {
+                        zs.z_stencil_id = 1; /* Separate stencil */
+                }
+
+                zs.padded_height_of_output_image_in_uif_blocks =
+                        surf->padded_height_of_output_image_in_uif_blocks;
+
+                assert(surf->tiling != VC5_TILING_RASTER);
+                zs.memory_format = surf->tiling;
+        }
+
+        if (job->resolve & (is_separate_stencil ?
+                            PIPE_CLEAR_STENCIL :
+                            PIPE_CLEAR_DEPTHSTENCIL)) {
+                rsc->writes++;
+        }
+}
+#endif /* V3D_VERSION < 40 */
 
 #define div_round_up(a, b) (((a) + (b) - 1) / b)
 
@@ -584,41 +593,16 @@ v3dX(emit_rcl)(struct vc5_job *job)
                 struct vc5_surface *surf = vc5_surface(psurf);
                 struct vc5_resource *rsc = vc5_resource(psurf->texture);
 
-                cl_emit(&job->rcl, TILE_RENDERING_MODE_CONFIGURATION_Z_STENCIL_CONFIG, zs) {
-                        zs.address = cl_address(rsc->bo, surf->offset);
-
-                        zs.internal_type = surf->internal_type;
-                        zs.output_image_format = surf->format;
-                        zs.padded_height_of_output_image_in_uif_blocks =
-                                surf->padded_height_of_output_image_in_uif_blocks;
-
-                        assert(surf->tiling != VC5_TILING_RASTER);
-                        zs.memory_format = surf->tiling;
-                }
-
-                if (job->resolve & PIPE_CLEAR_DEPTHSTENCIL)
-                        rsc->writes++;
+                v3d_emit_z_stencil_config(job, surf, rsc, false);
 
                 /* Emit the separate stencil packet if we have a resource for
                  * it.  The HW will only load/store this buffer if the
                  * Z/Stencil config doesn't have stencil in its format.
                  */
-                if (rsc->separate_stencil) {
-                        cl_emit(&job->rcl,
-                                TILE_RENDERING_MODE_CONFIGURATION_Z_STENCIL_CONFIG,
-                                zs) {
-                                zs.address =
-                                        cl_address(rsc->separate_stencil->bo,
-                                                   surf->separate_stencil_offset);
-
-                                zs.z_stencil_id = 1; /* Separate stencil */
-
-                                zs.padded_height_of_output_image_in_uif_blocks =
-                                        surf->separate_stencil_padded_height_of_output_image_in_uif_blocks;
-
-                                assert(surf->tiling != VC5_TILING_RASTER);
-                                zs.memory_format = surf->separate_stencil_tiling;
-                        }
+                if (surf->separate_stencil) {
+                        v3d_emit_z_stencil_config(job,
+                                                  vc5_surface(surf->separate_stencil),
+                                                  rsc->separate_stencil, true);
                 }
         }
 #endif /* V3D_VERSION < 40 */
diff --git a/src/gallium/drivers/vc5/vc5_resource.c b/src/gallium/drivers/vc5/vc5_resource.c
index a9cc27127f..9ecd083b05 100644
--- a/src/gallium/drivers/vc5/vc5_resource.c
+++ b/src/gallium/drivers/vc5/vc5_resource.c
@@ -654,10 +654,6 @@ vc5_create_surface(struct pipe_context *pctx,
         unsigned level = surf_tmpl->u.tex.level;
         struct vc5_resource_slice *slice = &rsc->slices[level];
 
-        struct vc5_resource_slice *separate_stencil_slice = NULL;
-        if (rsc->separate_stencil)
-                separate_stencil_slice = &rsc->separate_stencil->slices[level];
-
         pipe_reference_init(&psurf->reference, 1);
         pipe_resource_reference(&psurf->texture, ptex);
 
@@ -672,14 +668,6 @@ vc5_create_surface(struct pipe_context *pctx,
         surface->offset = (slice->offset +
                            psurf->u.tex.first_layer * rsc->cube_map_stride);
         surface->tiling = slice->tiling;
-        if (separate_stencil_slice) {
-                surface->separate_stencil_offset =
-                        (separate_stencil_slice->offset +
-                         psurf->u.tex.first_layer *
-                         rsc->separate_stencil->cube_map_stride);
-                surface->separate_stencil_tiling =
-                        separate_stencil_slice->tiling;
-        }
 
         surface->format = vc5_get_rt_format(&screen->devinfo, psurf->format);
 
@@ -709,13 +697,12 @@ vc5_create_surface(struct pipe_context *pctx,
                 surface->padded_height_of_output_image_in_uif_blocks =
                         ((slice->size / slice->stride) /
                          (2 * vc5_utile_height(rsc->cpp)));
+        }
 
-                if (separate_stencil_slice) {
-                        surface->separate_stencil_padded_height_of_output_image_in_uif_blocks =
-                        ((separate_stencil_slice->size /
-                          separate_stencil_slice->stride) /
-                         (2 * vc5_utile_height(rsc->separate_stencil->cpp)));
-                }
+        if (rsc->separate_stencil) {
+                surface->separate_stencil =
+                        vc5_create_surface(pctx, &rsc->separate_stencil->base,
+                                           surf_tmpl);
         }
 
         return &surface->base;
@@ -724,6 +711,11 @@ vc5_create_surface(struct pipe_context *pctx,
 static void
 vc5_surface_destroy(struct pipe_context *pctx, struct pipe_surface *psurf)
 {
+        struct vc5_surface *surf = vc5_surface(psurf);
+
+        if (surf->separate_stencil)
+                pipe_surface_reference(&surf->separate_stencil, NULL);
+
         pipe_resource_reference(&psurf->texture, NULL);
         FREE(psurf);
 }
diff --git a/src/gallium/drivers/vc5/vc5_resource.h b/src/gallium/drivers/vc5/vc5_resource.h
index 688f03d53f..631a7b8fba 100644
--- a/src/gallium/drivers/vc5/vc5_resource.h
+++ b/src/gallium/drivers/vc5/vc5_resource.h
@@ -80,9 +80,7 @@ struct vc5_resource_slice {
 struct vc5_surface {
         struct pipe_surface base;
         uint32_t offset;
-        uint32_t separate_stencil_offset;
         enum vc5_tiling_mode tiling;
-        enum vc5_tiling_mode separate_stencil_tiling;
         /**
          * Output image format for TILE_RENDERING_MODE_CONFIGURATION
          */
@@ -101,7 +99,11 @@ struct vc5_surface {
         uint8_t internal_bpp;
 
         uint32_t padded_height_of_output_image_in_uif_blocks;
-        uint32_t separate_stencil_padded_height_of_output_image_in_uif_blocks;
+
+        /* If the resource being referenced is separate stencil, then this is
+         * the surface to use when reading/writing stencil.
+         */
+        struct pipe_surface *separate_stencil;
 };
 
 struct vc5_resource {