[Mesa-dev] [PATCH] i965: Rework the unlit centroid workaround.

[Kenneth Graunke]

Previously, for every input, we moved the dispatch mask to the flag
register, then emitted two predicated PLN instructions, one with
centroid barycentric coordinates (for normal pixels), and one with
pixel barycentric coordinates (for unlit helper pixels).

Instead, we can simply emit a set of predicated MOVs at the top of
the program which copy the pixel barycentric coordinates over the
centroid ones for unlit helper pixel channels.  Then, we can just
use normal PLNs.

On Sandybridge:

total instructions in shared programs: 7538470 -> 7534500 (-0.05%)
instructions in affected programs: 101268 -> 97298 (-3.92%)
helped: 705
HURT: 9 (all of which are SIMD16 programs)

Signed-off-by: Kenneth Graunke <kenneth at whitecape.org>
---
 src/mesa/drivers/dri/i965/brw_fs_nir.cpp     | 26 +---------------------
 src/mesa/drivers/dri/i965/brw_fs_visitor.cpp | 32 ++++++++++++++++++++++++++++
 2 files changed, 33 insertions(+), 25 deletions(-)

diff --git a/src/mesa/drivers/dri/i965/brw_fs_nir.cpp b/src/mesa/drivers/dri/i965/brw_fs_nir.cpp
index 5236d0e..6e94051 100644
--- a/src/mesa/drivers/dri/i965/brw_fs_nir.cpp
+++ b/src/mesa/drivers/dri/i965/brw_fs_nir.cpp
@@ -3422,31 +3422,7 @@ fs_visitor::nir_emit_fs_intrinsic(const fs_builder &bld,
          interp.type = BRW_REGISTER_TYPE_F;
          dest.type = BRW_REGISTER_TYPE_F;
 
-         if (devinfo->needs_unlit_centroid_workaround &&
-             bary_intrin == nir_intrinsic_load_barycentric_centroid) {
-
-            /* Get the pixel/sample mask into f0 so that we know which
-             * pixels are lit.  Then, for each channel that is unlit,
-             * replace the centroid data with non-centroid data.
-             */
-            bld.emit(FS_OPCODE_MOV_DISPATCH_TO_FLAGS);
-
-            fs_reg dest_i = offset(dest, bld, i);
-            fs_reg dst_xy_pixel =
-               delta_xy[brw_barycentric_mode(interp_mode,
-                  nir_intrinsic_load_barycentric_pixel)];
-
-            fs_inst *inst;
-            inst = bld.emit(FS_OPCODE_LINTERP, dest_i, dst_xy_pixel, interp);
-            inst->predicate = BRW_PREDICATE_NORMAL;
-            inst->predicate_inverse = true;
-            inst->no_dd_clear = true;
-
-            inst = bld.emit(FS_OPCODE_LINTERP, dest_i, dst_xy, interp);
-            inst->predicate = BRW_PREDICATE_NORMAL;
-            inst->predicate_inverse = false;
-            inst->no_dd_check = true;
-         } else if (devinfo->gen < 6 && interp_mode == INTERP_MODE_SMOOTH) {
+         if (devinfo->gen < 6 && interp_mode == INTERP_MODE_SMOOTH) {
             fs_reg tmp = vgrf(glsl_type::float_type);
             bld.emit(FS_OPCODE_LINTERP, tmp, dst_xy, interp);
             bld.MUL(offset(dest, bld, i), tmp, this->pixel_w);
diff --git a/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp b/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp
index 6d84374..4e0db06 100644
--- a/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp
+++ b/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp
@@ -308,9 +308,41 @@ fs_visitor::emit_interpolation_setup_gen6()
    this->wpos_w = vgrf(glsl_type::float_type);
    abld.emit(SHADER_OPCODE_RCP, this->wpos_w, this->pixel_w);
 
+   brw_wm_prog_data *wm_prog_data = (brw_wm_prog_data *) prog_data;
+   uint32_t centroid_modes = wm_prog_data->barycentric_interp_modes &
+      (1 << BRW_BARYCENTRIC_PERSPECTIVE_CENTROID |
+       1 << BRW_BARYCENTRIC_NONPERSPECTIVE_CENTROID);
+
    for (int i = 0; i < BRW_BARYCENTRIC_MODE_COUNT; ++i) {
       uint8_t reg = payload.barycentric_coord_reg[i];
       this->delta_xy[i] = fs_reg(brw_vec16_grf(reg, 0));
+
+      if (devinfo->needs_unlit_centroid_workaround &&
+          (centroid_modes & (1 << i))) {
+         /* Get the pixel/sample mask into f0 so that we know which
+          * pixels are lit.  Then, for each channel that is unlit,
+          * replace the centroid data with non-centroid data.
+          */
+         bld.emit(FS_OPCODE_MOV_DISPATCH_TO_FLAGS);
+
+         uint8_t pixel_reg = payload.barycentric_coord_reg[i - 1];
+
+         set_predicate_inv(BRW_PREDICATE_NORMAL, true,
+                           bld.half(0).MOV(brw_vec8_grf(reg, 0),
+                                           brw_vec8_grf(pixel_reg, 0)));
+         set_predicate_inv(BRW_PREDICATE_NORMAL, true,
+                           bld.half(0).MOV(brw_vec8_grf(reg + 1, 0),
+                                           brw_vec8_grf(pixel_reg + 1, 0)));
+         if (dispatch_width == 16) {
+            set_predicate_inv(BRW_PREDICATE_NORMAL, true,
+                              bld.half(1).MOV(brw_vec8_grf(reg + 2, 0),
+                                              brw_vec8_grf(pixel_reg + 2, 0)));
+            set_predicate_inv(BRW_PREDICATE_NORMAL, true,
+                              bld.half(1).MOV(brw_vec8_grf(reg + 3, 0),
+                                              brw_vec8_grf(pixel_reg + 3, 0)));
+         }
+         assert(dispatch_width != 32); /* not implemented yet */
+      }
    }
 }
 
-- 
2.9.0