diff --git a/gcc/tree-vect-slp.c b/gcc/tree-vect-slp.c
index 2a58e54fe51471df5f55ce4a524d0022744054b0..89e226ca3a25a6c77b86d46ba234ce54bd3cb83b 100644
--- a/gcc/tree-vect-slp.c
+++ b/gcc/tree-vect-slp.c
@@ -2228,6 +2228,114 @@ calculate_unrolling_factor (poly_uint64 nunits, unsigned int group_size)
   return exact_div (common_multiple (nunits, group_size), group_size);
 }
 
+/* Helper that checks to see if a node is a load node. This is done based on
+   two criterias:
+   1) The node is internal
+   2) The node has no childen.  */
+
+static inline bool
+vect_is_slp_load_node  (slp_tree root)
+{
+  return (SLP_TREE_DEF_TYPE (root) == vect_internal_def
+	  && !SLP_TREE_CHILDREN (root).exists ());
+}
+
+
+/* Helper function of optimize_load_redistribution that performs the operation
+   recursively.  */
+
+static slp_tree
+optimize_load_redistribution_1 (scalar_stmts_to_slp_tree_map_t *bst_map,
+				vec_info *vinfo, unsigned int group_size,
+				hash_set<slp_tree> *visited, slp_tree root)
+{
+  if (visited->add (root))
+    return NULL;
+
+  slp_tree node;
+  unsigned i;
+
+  /* For now, we don't know anything about externals so do not do anything.  */
+  if (SLP_TREE_DEF_TYPE (root) != vect_internal_def)
+    return NULL;
+  else if (SLP_TREE_CODE (root) == VEC_PERM_EXPR)
+    {
+      /* First convert this node into a load node and add it to the leaves
+         list and flatten the permute from a lane to a load one.  If it's
+         unneeded it will be elided later.  */
+      vec<stmt_vec_info> stmts;
+      stmts.create (SLP_TREE_LANES (root));
+      lane_permutation_t lane_perm = SLP_TREE_LANE_PERMUTATION (root);
+      for (unsigned j = 0; j < lane_perm.length (); j++)
+        {
+          std::pair<unsigned, unsigned> perm = lane_perm[j];
+          node = SLP_TREE_CHILDREN (root)[perm.first];
+
+	  if (!vect_is_slp_load_node (node))
+	   return NULL;
+
+	  stmts.quick_push (SLP_TREE_SCALAR_STMTS (node)[perm.second]);
+        }
+
+      if (dump_enabled_p ())
+	dump_printf_loc (MSG_NOTE, vect_location,
+			 "converting stmts on permute node %p\n", root);
+
+      bool *matches = XALLOCAVEC (bool, group_size);
+      poly_uint64 max_nunits = 1;
+      unsigned tree_size = 0, limit = 1;
+      node = vect_build_slp_tree (vinfo, stmts, group_size, &max_nunits,
+				  matches, &limit, &tree_size, bst_map);
+      if (!node)
+	stmts.release ();
+
+      return node;
+    }
+
+  FOR_EACH_VEC_ELT (SLP_TREE_CHILDREN (root), i , node)
+    {
+      slp_tree value
+	= optimize_load_redistribution_1 (bst_map, vinfo, group_size, visited,
+					  node);
+      if (value)
+	{
+          SLP_TREE_CHILDREN (root)[i] = value;
+          vect_free_slp_tree (node);
+	}
+    }
+
+  return NULL;
+}
+
+/* Temporary workaround for loads not being CSEd during SLP build.  This
+   function will traverse the SLP tree rooted in ROOT for INSTANCE and find
+   VEC_PERM nodes that blend vectors from multiple nodes that all read from the
+   same DR such that the final operation is equal to a permuted load.  Such
+   NODES are then directly converted into LOADS themselves.  The nodes are
+   CSEd using BST_MAP.  */
+
+static void
+optimize_load_redistribution (scalar_stmts_to_slp_tree_map_t *bst_map,
+			      vec_info *vinfo, unsigned int group_size,
+			      slp_tree root)
+{
+  slp_tree node;
+  unsigned i;
+  hash_set<slp_tree> visited;
+
+  FOR_EACH_VEC_ELT (SLP_TREE_CHILDREN (root), i , node)
+    {
+      slp_tree value
+	= optimize_load_redistribution_1 (bst_map, vinfo, group_size, &visited,
+					  node);
+      if (value)
+	{
+          SLP_TREE_CHILDREN (root)[i] = value;
+          vect_free_slp_tree (node);
+	}
+    }
+}
+
 /* Helper function of vect_match_slp_patterns.
 
    Attempts to match patterns against the slp tree rooted in REF_NODE using
@@ -2276,7 +2384,7 @@ static bool
 vect_match_slp_patterns (slp_instance instance, vec_info *vinfo,
 			 hash_set<slp_tree> *visited,
 			 slp_tree_to_load_perm_map_t *perm_cache,
-			 scalar_stmts_to_slp_tree_map_t * /* bst_map */)
+			 scalar_stmts_to_slp_tree_map_t *bst_map)
 {
   DUMP_VECT_SCOPE ("vect_match_slp_patterns");
   slp_tree *ref_node = &SLP_INSTANCE_TREE (instance);
@@ -2291,6 +2399,9 @@ vect_match_slp_patterns (slp_instance instance, vec_info *vinfo,
 
   if (found_p)
     {
+      optimize_load_redistribution (bst_map, vinfo, SLP_TREE_LANES (*ref_node),
+				    *ref_node);
+
       if (dump_enabled_p ())
 	{
 	  dump_printf_loc (MSG_NOTE, vect_location,