More BB vectorization tweaks

This tweaks the op build from splats to allow loads marked as not
vectorizable.  It also amends some dump prints with the address of
the SLP node or the instance to better be able to debug things.

2020-10-29  Richard Biener  <rguenther@suse.de>

	* tree-vect-slp.c (vect_build_slp_tree_2): Allow splatting
	not vectorizable loads.
	(vect_build_slp_instance): Amend dumping with address.
	(vect_slp_convert_to_external): Likewise.

	* gcc.dg/vect/bb-slp-pr65935.c: Adjust.

diff --git a/gcc/testsuite/gcc.dg/vect/bb-slp-pr65935.c b/gcc/testsuite/gcc.dg/vect/bb-slp-pr65935.c
index ea37e4e..c262d73 100644 (file)
--- a/gcc/testsuite/gcc.dg/vect/bb-slp-pr65935.c
+++ b/gcc/testsuite/gcc.dg/vect/bb-slp-pr65935.c
@@ -60,6 +60,7 @@ int main()
 /* We should also be able to use 2-lane SLP to initialize the real and
    imaginary components in the first loop of main.  */
 /* { dg-final { scan-tree-dump-times "optimized: basic block" 10 "slp1" } } */
-/* We should see the s->phase[dir] operand and only that operand built
+/* We should see the s->phase[dir] operand splatted and no other operand built
    from scalars.  See PR97334.  */
-/* { dg-final { scan-tree-dump-times "Building vector operands from scalars" 1 "slp1" } } */
+/* { dg-final { scan-tree-dump-times "Using a splat" 1 "slp1" } } */
+/* { dg-final { scan-tree-dump-times "Building vector operands from scalars" 0 "slp1" } } */

diff --git a/gcc/tree-vect-slp.c b/gcc/tree-vect-slp.c
index ff3a0c2..0a7b8e6 100644 (file)
--- a/gcc/tree-vect-slp.c
+++ b/gcc/tree-vect-slp.c
@@ -1627,8 +1627,10 @@ vect_build_slp_tree_2 (vec_info *vinfo, slp_tree node,
              break;
          if (j == group_size
              /* But avoid doing this for loads where we may be
-                able to CSE things.  */
-             && !gimple_vuse (first_def->stmt))
+                able to CSE things, unless the stmt is not
+                vectorizable.  */
+             && (!STMT_VINFO_VECTORIZABLE (first_def)
+                 || !gimple_vuse (first_def->stmt)))
            {
              if (dump_enabled_p ())
                dump_printf_loc (MSG_NOTE, vect_location,
@@ -2379,7 +2381,7 @@ vect_build_slp_instance (vec_info *vinfo,
          if (dump_enabled_p ())
            {
              dump_printf_loc (MSG_NOTE, vect_location,
-                              "Final SLP tree for instance:\n");
+                              "Final SLP tree for instance %p:\n", new_instance);
              vect_print_slp_graph (MSG_NOTE, vect_location,
                                    SLP_INSTANCE_TREE (new_instance));
            }
@@ -3402,7 +3404,7 @@ vect_slp_convert_to_external (vec_info *vinfo, slp_tree node,
 
   if (dump_enabled_p ())
     dump_printf_loc (MSG_NOTE, vect_location,
-                    "Building vector operands from scalars instead\n");
+                    "Building vector operands of %p from scalars instead\n", node);
 
   /* Don't remove and free the child nodes here, since they could be
      referenced by other structures.  The analysis and scheduling phases