[Mesa-dev] [PATCH] i965: Skip recalculating URB allocations if the entry size didn't change.

[Eric Anholt]

We only get here if the VS/GS compiled programs change, but we can even
skip it if the VS/GS size didn't change.

Affects cairo runtime on glamor by -1.26471% +/- 0.674335% (n=234)
---
 src/mesa/drivers/dri/i965/brw_context.c |  2 +-
 src/mesa/drivers/dri/i965/brw_context.h |  5 +++--
 src/mesa/drivers/dri/i965/gen6_urb.c    |  4 ++--
 src/mesa/drivers/dri/i965/gen7_urb.c    | 13 +++++++++++++
 4 files changed, 19 insertions(+), 5 deletions(-)

diff --git a/src/mesa/drivers/dri/i965/brw_context.c b/src/mesa/drivers/dri/i965/brw_context.c
index 17ae685..2a4620b 100644
--- a/src/mesa/drivers/dri/i965/brw_context.c
+++ b/src/mesa/drivers/dri/i965/brw_context.c
@@ -745,7 +745,7 @@ brwCreateContext(gl_api api,
    brw->max_gtt_map_object_size = gtt_size / 4;
 
    if (brw->gen == 6)
-      brw->urb.gen6_gs_previously_active = false;
+      brw->urb.gs_present = false;
 
    brw->prim_restart.in_progress = false;
    brw->prim_restart.enable_cut_index = false;
diff --git a/src/mesa/drivers/dri/i965/brw_context.h b/src/mesa/drivers/dri/i965/brw_context.h
index 92e1592..b450777 100644
--- a/src/mesa/drivers/dri/i965/brw_context.h
+++ b/src/mesa/drivers/dri/i965/brw_context.h
@@ -1183,6 +1183,7 @@ struct brw_context
     */
    struct {
       GLuint vsize;		/* vertex size plus header in urb registers */
+      GLuint gsize;	        /* GS output size in urb registers */
       GLuint csize;		/* constant buffer size in urb registers */
       GLuint sfsize;		/* setup data size in urb registers */
 
@@ -1205,10 +1206,10 @@ struct brw_context
       GLuint cs_start;
       GLuint size; /* Hardware URB size, in KB. */
 
-      /* gen6: True if the most recently sent _3DSTATE_URB message allocated
+      /* True if the most recently sent _3DSTATE_URB message allocated
        * URB space for the GS.
        */
-      bool gen6_gs_previously_active;
+      bool gs_present;
    } urb;
 
 
diff --git a/src/mesa/drivers/dri/i965/gen6_urb.c b/src/mesa/drivers/dri/i965/gen6_urb.c
index b694f5d..9197bcf 100644
--- a/src/mesa/drivers/dri/i965/gen6_urb.c
+++ b/src/mesa/drivers/dri/i965/gen6_urb.c
@@ -109,9 +109,9 @@ gen6_upload_urb( struct brw_context *brw )
     * doesn't exist on Gen6).  So for now we just do a full pipeline flush as
     * a workaround.
     */
-   if (brw->urb.gen6_gs_previously_active && !brw->ff_gs.prog_active)
+   if (brw->urb.gs_present && !brw->ff_gs.prog_active)
       intel_batchbuffer_emit_mi_flush(brw);
-   brw->urb.gen6_gs_previously_active = brw->ff_gs.prog_active;
+   brw->urb.gs_present = brw->ff_gs.prog_active;
 }
 
 const struct brw_tracked_state gen6_urb = {
diff --git a/src/mesa/drivers/dri/i965/gen7_urb.c b/src/mesa/drivers/dri/i965/gen7_urb.c
index 2653e9c..3c21e1e 100644
--- a/src/mesa/drivers/dri/i965/gen7_urb.c
+++ b/src/mesa/drivers/dri/i965/gen7_urb.c
@@ -150,6 +150,19 @@ gen7_upload_urb(struct brw_context *brw)
    unsigned gs_size = gs_present ? brw->gs.prog_data->base.urb_entry_size : 1;
    unsigned gs_entry_size_bytes = gs_size * 64;
 
+   /* If we're just switching between programs with the same URB requirements,
+    * skip the rest of the logic.
+    */
+   if (!(brw->state.dirty.brw & BRW_NEW_CONTEXT) &&
+       brw->urb.vsize == vs_size &&
+       brw->urb.gs_present == gs_present &&
+       brw->urb.gsize == gs_size) {
+      return;
+   }
+   brw->urb.vsize = vs_size;
+   brw->urb.gs_present = gs_present;
+   brw->urb.gsize = gs_size;
+
    /* From p35 of the Ivy Bridge PRM (section 1.7.1: 3DSTATE_URB_GS):
     *
     *     VS Number of URB Entries must be divisible by 8 if the VS URB Entry
-- 
1.9.2