Hi!
As mentioned in the PR, this testcase is not vectorized e.g. on power7
(32-bit as well as 64-bit), because there is no misalign support.
Tested on x86_64-linux (-m32/-m64) where the scan-tree-dump-times still
passes and on powerpc64-linux (-m32/-m64) where it previously FAILed and now
isn't tested.
Ok for trunk?
2019-01-02 Jakub Jelinek <ja...@redhat.com>
PR testsuite/87304
* gcc.dg/vect/bb-slp-over-widen-1.c: Expect basic block vectorized
messages only on vect_hw_misalign targets.
--- gcc/testsuite/gcc.dg/vect/bb-slp-over-widen-1.c.jj 2018-08-03
17:05:27.067196476 +0200
+++ gcc/testsuite/gcc.dg/vect/bb-slp-over-widen-1.c 2019-01-02
19:18:50.205495553 +0100
@@ -64,4 +64,4 @@ main (void)
/* { dg-final { scan-tree-dump "demoting int to signed short" "slp2" { target
{ ! vect_widen_shift } } } } */
/* { dg-final { scan-tree-dump "demoting int to unsigned short" "slp2" {
target { ! vect_widen_shift } } } } */
/* { dg-final { scan-tree-dump {\.AVG_FLOOR} "slp2" { target vect_avg_qi } } }
*/
-/* { dg-final { scan-tree-dump-times "basic block vectorized" 2 "slp2" } } */
+/* { dg-final { scan-tree-dump-times "basic block vectorized" 2 "slp2" {
target vect_hw_misalign } } } */
Jakub
