Computing the minimum size was easy, and done at compile-time for no
extra overhead here. Making the binding table smaller wastes less batch
space.
Adding the CACHE_NEW_WM_PROG dirty bit isn't strictly necessary, since
other atoms depend on it and flag BRW_NEW_SURFACES. However, it's best
to add it for clarity and safety. It shouldn't add any new overhead.
Signed-off-by: Kenneth Graunke <kenn...@whitecape.org>
---
src/mesa/drivers/dri/i965/brw_wm_surface_state.c | 11 +++++------
1 file changed, 5 insertions(+), 6 deletions(-)
diff --git a/src/mesa/drivers/dri/i965/brw_wm_surface_state.c
b/src/mesa/drivers/dri/i965/brw_wm_surface_state.c
index 8847f91..3384e76 100644
--- a/src/mesa/drivers/dri/i965/brw_wm_surface_state.c
+++ b/src/mesa/drivers/dri/i965/brw_wm_surface_state.c
@@ -873,15 +873,14 @@ brw_upload_wm_binding_table(struct brw_context *brw)
gen7_create_shader_time_surface(brw,
&brw->wm.surf_offset[SURF_INDEX_WM_SHADER_TIME]);
}
- /* Might want to calculate nr_surfaces first, to avoid taking up so much
- * space for the binding table.
- */
+ /* CACHE_NEW_WM_PROG */
+ unsigned entries = brw->wm.prog_data->max_surface_index + 1;
bind = brw_state_batch(brw, AUB_TRACE_BINDING_TABLE,
- sizeof(uint32_t) * BRW_MAX_WM_SURFACES,
+ sizeof(uint32_t) * entries,
32, &brw->wm.bind_bo_offset);
/* BRW_NEW_SURFACES */
- for (i = 0; i < BRW_MAX_WM_SURFACES; i++) {
+ for (i = 0; i < entries; i++) {
bind[i] = brw->wm.surf_offset[i];
}
@@ -893,7 +892,7 @@ const struct brw_tracked_state brw_wm_binding_table = {
.mesa = 0,
.brw = (BRW_NEW_BATCH |
BRW_NEW_SURFACES),
- .cache = 0
+ .cache = CACHE_NEW_WM_PROG
},
.emit = brw_upload_wm_binding_table,
};
--
1.8.3.4
_______________________________________________
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
