Mesa (main): radeonsi: remove incorrect comment about PA

[GitLab Mirror]

Module: Mesa
Branch: main
Commit: 901f4aa5d62f8a3e7b90233b6dd1b6e7c0dfde5a
URL:    http://cgit.freedesktop.org/mesa/mesa/commit/?id=901f4aa5d62f8a3e7b90233b6dd1b6e7c0dfde5a

Author: Marek Olšák <marek.olsak at amd.com>
Date:   Mon May 24 19:14:18 2021 -0400

radeonsi: remove incorrect comment about PA

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer at amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11102>

---

 src/gallium/drivers/radeonsi/gfx10_shader_ngg.c | 7 -------
 1 file changed, 7 deletions(-)

diff --git a/src/gallium/drivers/radeonsi/gfx10_shader_ngg.c b/src/gallium/drivers/radeonsi/gfx10_shader_ngg.c
index 5f3763bcd83..09590878e5e 100644
--- a/src/gallium/drivers/radeonsi/gfx10_shader_ngg.c
+++ b/src/gallium/drivers/radeonsi/gfx10_shader_ngg.c
@@ -1881,13 +1881,6 @@ void gfx10_ngg_gs_emit_epilogue(struct si_shader_context *ctx)
 
    /* Allocate export space. Send this message as early as possible, to
     * hide the latency of the SQ <-> SPI roundtrip.
-    *
-    * Note: We could consider compacting primitives for export as well.
-    *       PA processes 1 non-null prim / clock, but it fetches 4 DW of
-    *       prim data per clock and skips null primitives at no additional
-    *       cost. So compacting primitives can only be beneficial when
-    *       there are 4 or more contiguous null primitives in the export
-    *       (in the common case of single-dword prim exports).
     */
    ac_build_sendmsg_gs_alloc_req(&ctx->ac, get_wave_id_in_tg(ctx), vertlive_scan.result_reduce,
                                  num_emit_threads);