radeonsi/gfx9: proper workaround for LS/HS VGPR initialization bug

[mesa.git] / src / gallium / drivers / radeonsi / si_state_draw.c

diff --git a/src/gallium/drivers/radeonsi/si_state_draw.c b/src/gallium/drivers/radeonsi/si_state_draw.c
index 7ee6cf88e8822906289739306dd3419c3eff21d0..051dfea8f7c6747d191329e5655298957958a3c6 100644 (file)
--- a/src/gallium/drivers/radeonsi/si_state_draw.c
+++ b/src/gallium/drivers/radeonsi/si_state_draw.c
@@ -195,11 +195,7 @@ static void si_emit_derived_tess_state(struct si_context *sctx,
         */
        *num_patches = MIN2(*num_patches, 40);
 
-       if (sctx->b.chip_class == SI ||
-           /* TODO: fix GFX9 where a threadgroup contains more than 1 wave and
-            * LS vertices per patch > HS vertices per patch. Piglit: 16in-1out */
-           (sctx->b.chip_class == GFX9 &&
-            num_tcs_input_cp > num_tcs_output_cp)) {
+       if (sctx->b.chip_class == SI) {
                /* SI bug workaround, related to power management. Limit LS-HS
                 * threadgroups to only one wave.
                 */
@@ -1264,6 +1260,27 @@ void si_draw_vbo(struct pipe_context *ctx, const struct pipe_draw_info *info)
                sctx->do_update_shaders = true;
        }
 
+       if (sctx->tes_shader.cso &&
+           (sctx->b.family == CHIP_VEGA10 || sctx->b.family == CHIP_RAVEN)) {
+               /* Determine whether the LS VGPR fix should be applied.
+                *
+                * It is only required when num input CPs > num output CPs,
+                * which cannot happen with the fixed function TCS. We should
+                * also update this bit when switching from TCS to fixed
+                * function TCS.
+                */
+               struct si_shader_selector *tcs = sctx->tcs_shader.cso;
+               bool ls_vgpr_fix =
+                       tcs &&
+                       info->vertices_per_patch >
+                       tcs->info.properties[TGSI_PROPERTY_TCS_VERTICES_OUT];
+
+               if (ls_vgpr_fix != sctx->ls_vgpr_fix) {
+                       sctx->ls_vgpr_fix = ls_vgpr_fix;
+                       sctx->do_update_shaders = true;
+               }
+       }
+
        if (sctx->gs_shader.cso) {
                /* Determine whether the GS triangle strip adjacency fix should
                 * be applied. Rotate every other triangle if