Skip some mode SAD calculation in non-RD mode

This commit checks if the motion vector associated with the current
mode has been computed in previous mode tests. If possible, skip the
redundant reference block generation and SAD calculation in the
non-RD mode decision process.

For test sequence pedestrian_area 1080p, the runtime goes from
24261 ms to 23770 ms. This does not change compression performance.
The speed-up is mostly around places with consistent motion.

Change-Id: I97be63c6a2d07c57be26b3c600fbda3803adddda

diff --git a/vp9/encoder/vp9_pickmode.c b/vp9/encoder/vp9_pickmode.c
index 87f20fa1c68d25e2ab194f406f866b9b30d6519b..0f2e3034f5eb2f8d3719479d89076ec25e7aecb6 100644 (file)
--- a/vp9/encoder/vp9_pickmode.c
+++ b/vp9/encoder/vp9_pickmode.c
@@ -270,12 +270,21 @@ int64_t vp9_pick_inter_mode(VP9_COMP *cpi, MACROBLOCK *x,
                                 &frame_mv[NEWMV][ref_frame]);
       }
 
-      mbmi->mode = this_mode;
-      mbmi->mv[0].as_int = frame_mv[this_mode][ref_frame].as_int;
-      vp9_build_inter_predictors_sby(xd, mi_row, mi_col, bsize);
+      if (frame_mv[this_mode][ref_frame].as_int == 0) {
+        dist = x->mode_sad[ref_frame][INTER_OFFSET(ZEROMV)];
+      } else if (this_mode != NEARESTMV &&
+                 frame_mv[NEARESTMV][ref_frame].as_int ==
+                     frame_mv[this_mode][ref_frame].as_int) {
+        dist = x->mode_sad[ref_frame][INTER_OFFSET(NEARESTMV)];
+      } else {
+        mbmi->mode = this_mode;
+        mbmi->mv[0].as_int = frame_mv[this_mode][ref_frame].as_int;
+        vp9_build_inter_predictors_sby(xd, mi_row, mi_col, bsize);
+        dist = x->mode_sad[ref_frame][INTER_OFFSET(this_mode)] =
+            cpi->fn_ptr[bsize].sdf(p->src.buf, p->src.stride,
+                                   pd->dst.buf, pd->dst.stride, INT_MAX);
+      }
 
-      dist = cpi->fn_ptr[bsize].sdf(p->src.buf, p->src.stride,
-                                    pd->dst.buf, pd->dst.stride, INT_MAX);
       this_rd = rate + dist;
 
       if (this_rd < best_rd) {