On Saturday, March 14, 2015 09:54:18 PM Jordan Justen wrote:
> Signed-off-by: Jordan Justen <jordan.l.jus...@intel.com>
> ---
> src/mesa/drivers/dri/i965/brw_fs_reg_allocate.cpp | 4 ++++
> 1 file changed, 4 insertions(+)
>
> diff --git a/src/mesa/drivers/dri/i965/brw_fs_reg_allocate.cpp
> b/src/mesa/drivers/dri/i965/brw_fs_reg_allocate.cpp
> index 72c490b..1f30a19 100644
> --- a/src/mesa/drivers/dri/i965/brw_fs_reg_allocate.cpp
> +++ b/src/mesa/drivers/dri/i965/brw_fs_reg_allocate.cpp
> @@ -405,6 +405,10 @@ fs_visitor::setup_payload_interference(struct ra_graph
> *g,
> }
> break;
>
> + case CS_OPCODE_CS_TERMINATE:
> + payload_last_use_ip[0] = use_ip;
> + break;
> +
> default:
> if (inst->eot) {
> /* We could omit this for the !inst->header_present case, except
> I was going to ask why we didn't just fall through to the inst->eot case below, but...it looks like the VS, GS, and FS all want g0/g1 reserved, while compute only needs g0. So this looks like a sensible solution. Reviewed-by: Kenneth Graunke <kenn...@whitecape.org>
signature.asc
Description: This is a digitally signed message part.
_______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
