v3d: Drop a perf note about merging unpack_half_*, which has been implemented.

This is handled with copy-propagation now.

diff --git a/src/broadcom/compiler/nir_to_vir.c b/src/broadcom/compiler/nir_to_vir.c
index 0a3275de7dcb88535cedfa38e2cb014c9bdeea6f..82da32ac6be1cb14b00124d1eaa710fb41652dd6 100644 (file)
--- a/src/broadcom/compiler/nir_to_vir.c
+++ b/src/broadcom/compiler/nir_to_vir.c
@@ -1048,9 +1048,6 @@ ntq_emit_alu(struct v3d_compile *c, nir_alu_instr *instr)
                 break;
 
         case nir_op_unpack_half_2x16_split_x:
-                /* XXX perf: It would be good to be able to merge this unpack
-                 * with whatever uses our result.
-                 */
                 result = vir_FMOV(c, src[0]);
                 vir_set_unpack(c->defs[result.index], 0, V3D_QPU_UNPACK_L);
                 break;