Relax boolean processing in vect_maybe_update_slp_op_vectype

The following makes VECTOR_BOOLEAN_TYPE_P processing consistent with
what we do without SLP.  The original motivation for rejecting of
VECTOR_BOOLEAN_TYPE_P extern defs was bad code generation.  But
the non-SLP codepath happily goes along - but always hits the
case of an uniform vector and this case specifically we can now
code-generate optimally.  So the following allows single-lane
externs as well.

Requiring patterns to code-generate can have bad influence on
the vectorization factor though a prototype patch of mine shows
that generating vector compares externally isn't always trivial.

The patch fixes the gcc.dg/vect/vect-early-break_82.c FAIL on x86_64
when --param vect-force-slp=1 is in effect.

	PR tree-optimization/117171
	* tree-vect-stmts.cc (vect_maybe_update_slp_op_vectype):
	Relax vect_external_def VECTOR_BOOLEAN_TYPE_P constraint.

diff --git a/gcc/tree-vect-stmts.cc b/gcc/tree-vect-stmts.cc
index 6967d50288e9e37de40a0e3e7afab30aaa49c247..e7f14c3144c3289f5b84b879055ea6c3d36f0c3f 100644 (file)
--- a/gcc/tree-vect-stmts.cc
+++ b/gcc/tree-vect-stmts.cc
@@ -14290,9 +14290,12 @@ vect_maybe_update_slp_op_vectype (slp_tree op, tree vectype)
   if (SLP_TREE_VECTYPE (op))
     return types_compatible_p (SLP_TREE_VECTYPE (op), vectype);
   /* For external defs refuse to produce VECTOR_BOOLEAN_TYPE_P, those
-     should be handled by patters.  Allow vect_constant_def for now.  */
+     should be handled by patters.  Allow vect_constant_def for now
+     as well as the trivial single-lane uniform vect_external_def case
+     both of which we code-generate reasonably.  */
   if (VECTOR_BOOLEAN_TYPE_P (vectype)
-      && SLP_TREE_DEF_TYPE (op) == vect_external_def)
+      && SLP_TREE_DEF_TYPE (op) == vect_external_def
+      && SLP_TREE_LANES (op) > 1)
     return false;
   SLP_TREE_VECTYPE (op) = vectype;
   return true;