[PATCH][PR tree-optimization/68844] Fix testcase expected output
authorJeff Law <law@redhat.com>
Fri, 11 Dec 2015 23:18:22 +0000 (16:18 -0700)
committerJeff Law <law@gcc.gnu.org>
Fri, 11 Dec 2015 23:18:22 +0000 (16:18 -0700)
PR tree-optimization/68844
* gcc.dg/tree-ssa/ssa-dom-thread-4.c: Update expected output.

2015-12-11  Jan Beulich  <jbeulich@suse.com>

From-SVN: r231577

gcc/testsuite/ChangeLog
gcc/testsuite/gcc.dg/tree-ssa/ssa-dom-thread-4.c

index af682a95af025405241686ee7bd3bbea61c9f6e2..3877b19c95ea5f15a661dd926c623ea0d1aaee81 100644 (file)
@@ -1,3 +1,8 @@
+2015-12-11  Jeff Law  <law@redhat.com>
+
+       PR tree-optimization/68844
+       * gcc.dg/tree-ssa/ssa-dom-thread-4.c: Update expected output.
+
 2015-12-11  Nathan Sidwell  <nathan@acm.org>
 
        * gcc.dg/pr59605-1.c: Reduce iterations for nvptx.
index 77ba74c1352a534f2e27cc9a4fe23c81b866b236..4258fb5acbaa2a49df51df8c68831f654c405005 100644 (file)
@@ -74,6 +74,10 @@ bitmap_ior_and_compl (bitmap dst, const_bitmap a, const_bitmap b,
    2x "kill_elt->indx >= b_elt->indx" in the first "while" loop
       -> "kill_elt->indx == b_elt->indx" in the second condition,
         skipping the known-true "b_elt && kill_elt" in the second
-        condition.  */
-/* { dg-final { scan-tree-dump-times "Threaded" 4 "dom2" { target logical_op_short_circuit } } } */
+        condition.
+
+   However, 3 of those 4 opportunities are ultimately eliminated by
+   DOM optimizing away conditionals.  So there's only one jump threading
+   opportunity left.  */
+/* { dg-final { scan-tree-dump-times "Threaded" 1 "dom2" { target logical_op_short_circuit } } } */