This removes the XFAIL by fixing the testcase - it cannot be ever
optimized in the desired way because the write to type.alias_set
_does_ kill the previous load. But that's because the testcase
bogously simplified the tree inheritance.
Fixed. Also disabled FRE so this really tests DOM. (all the
20030814-x testcases have the same issue, but this one is exposed
by an artificial XPASS I was seeing with a local patch).
Installed.
Richard.
2016-04-05 Richard Biener <rguent...@suse.de>
* gcc.dg/tree-ssa/20030814-6.c: Fix testcase, disable FRE,
remove XFAIL.
Index: gcc/testsuite/gcc.dg/tree-ssa/20030814-6.c
===================================================================
--- gcc/testsuite/gcc.dg/tree-ssa/20030814-6.c (revision 234705)
+++ gcc/testsuite/gcc.dg/tree-ssa/20030814-6.c (working copy)
@@ -1,5 +1,5 @@
/* { dg-do compile } */
-/* { dg-options "-O1 -fdump-tree-dom2" } */
+/* { dg-options "-O1 -fno-tree-fre -fdump-tree-dom2" } */
extern void abort (void);
union tree_node;
@@ -15,6 +15,7 @@ struct tree_common
};
struct tree_type
{
+ struct tree_common common;
double alias_set;
};
union tree_node
@@ -39,6 +40,4 @@ foo (t, set)
}
/* There should be precisely one load of common.code. If there is
more than one, then the dominator optimizations failed. */
-/* ??? Will fail until we properly distinguish member stores. At
- present the write to type.alias_set kills the previous load. */
-/* { dg-final { scan-tree-dump-times "common.code" 1 "dom2" { xfail *-*-* } }
} */
+/* { dg-final { scan-tree-dump-times "common.code" 1 "dom2" } } */
