[Mesa-dev] [PATCH] radv: remove unnecessary code around CACHE_FLUSH_AND_INV_TS_EVENT

[Bas Nieuwenhuizen]

Reviewed-by: Bas Nieuwenhuizen <bas at basnieuwenhuizen.nl>

On Tue, Jun 19, 2018 at 3:24 PM, Samuel Pitoiset
<samuel.pitoiset at gmail.com> wrote:
> AMDVLK also always uses CACHE_FLUSH_AND_INV_TS_EVENT. The other
> workaround is to flush DB metadata after emitting the framebuffer,
> but that seems slower.
>
> Signed-off-by: Samuel Pitoiset <samuel.pitoiset at gmail.com>
> ---
>  src/amd/vulkan/si_cmd_buffer.c | 19 +------------------
>  1 file changed, 1 insertion(+), 18 deletions(-)
>
> diff --git a/src/amd/vulkan/si_cmd_buffer.c b/src/amd/vulkan/si_cmd_buffer.c
> index 4c3a888f3c..671624cc0e 100644
> --- a/src/amd/vulkan/si_cmd_buffer.c
> +++ b/src/amd/vulkan/si_cmd_buffer.c
> @@ -834,26 +834,9 @@ si_cs_emit_cache_flush(struct radeon_winsys_cs *cs,
>         if (chip_class >= GFX9 && flush_cb_db) {
>                 unsigned cb_db_event, tc_flags;
>
> -#if 0
> -               /* This breaks a bunch of:
> -                  dEQP-VK.renderpass.dedicated_allocation.formats.d32_sfloat_s8_uint.input*.
> -                  use the big hammer always.
> -               */
>                 /* Set the CB/DB flush event. */
> -               switch (flush_cb_db) {
> -               case RADV_CMD_FLAG_FLUSH_AND_INV_CB:
> -                       cb_db_event = V_028A90_FLUSH_AND_INV_CB_DATA_TS;
> -                       break;
> -               case RADV_CMD_FLAG_FLUSH_AND_INV_DB:
> -                       cb_db_event = V_028A90_FLUSH_AND_INV_DB_DATA_TS;
> -                       break;
> -               default:
> -                       /* both CB & DB */
> -                       cb_db_event = V_028A90_CACHE_FLUSH_AND_INV_TS_EVENT;
> -               }
> -#else
>                 cb_db_event = V_028A90_CACHE_FLUSH_AND_INV_TS_EVENT;
> -#endif
> +
>                 /* These are the only allowed combinations. If you need to
>                  * do multiple operations at once, do them separately.
>                  * All operations that invalidate L2 also seem to invalidate
> --
> 2.17.1
>
> _______________________________________________
> mesa-dev mailing list
> mesa-dev at lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/mesa-dev