A better trade-off for ext-partition-types search

Changes a pruning criterion that seems to give a little better
compression efficiency at a little faster speed.

Change-Id: I8e3f9aa552b093c4af4ba615bb6ce29587bc8c36

av1/encoder/encodeframe.c

diff --git a/av1/encoder/encodeframe.c b/av1/encoder/encodeframe.c
index 8a6bc09..bf5fc93 100644
--- a/av1/encoder/encodeframe.c
+++ b/av1/encoder/encodeframe.c
@@ -3077,7 +3077,7 @@
     horz4_partition_allowed &= (pc_tree->partitioning == PARTITION_HORZ ||
                                 pc_tree->partitioning == PARTITION_HORZ_A ||
                                 pc_tree->partitioning == PARTITION_HORZ_B ||
-                                pc_tree->partitioning == PARTITION_SPLIT);
+                                pc_tree->partitioning == PARTITION_NONE);
   }
   if (horz4_partition_allowed && !force_horz_split &&
       (do_rectangular_split || av1_active_h_edge(cpi, mi_row, mi_step))) {
@@ -3116,7 +3116,7 @@
     vert4_partition_allowed &= (pc_tree->partitioning == PARTITION_VERT ||
                                 pc_tree->partitioning == PARTITION_VERT_A ||
                                 pc_tree->partitioning == PARTITION_VERT_B ||
-                                pc_tree->partitioning == PARTITION_SPLIT);
+                                pc_tree->partitioning == PARTITION_NONE);
   }
   if (vert4_partition_allowed && !force_vert_split &&
       (do_rectangular_split || av1_active_v_edge(cpi, mi_row, mi_step))) {