[PATCH] Fix PR66110

Richard Biener rguenther@suse.de
Tue May 12 13:50:00 GMT 2015


The following patch fixes PR66110 - there is no reason why two alias-sets
conflict just because one or the other has a child with alias-set zero.

We need to paper over strict-aliasing bugs in dfp.c (re-using
struct real_value unsigned long sig[] as decimal128) by adding
-Wno-strict-aliasing, otherwise bootstrap with -Werror fails.

Bootstrapped and tested on x86_64-unknown-linux-gnu (with 
--disable-werror, now re-trying with just the dfp.c issue workaround).

I intent to commit this if that succeeds.

Thanks,
Richard.

2015-05-12  Richard Biener  <rguenther@suse.de>

	PR middle-end/66110
	* alias.c (alias_sets_conflict_p): Do not treat has_zero_child
	specially.
	* Makefile.in (dfp.o-warn): Add -Wno-strict-aliasing.

	* gcc.dg/alias-2.c: Adjust.

Index: gcc/alias.c
===================================================================
--- gcc/alias.c	(revision 223032)
+++ gcc/alias.c	(working copy)
@@ -470,15 +470,13 @@ alias_sets_conflict_p (alias_set_type se
   /* See if the first alias set is a subset of the second.  */
   ase = get_alias_set_entry (set1);
   if (ase != 0
-      && (ase->has_zero_child
-	  || ase->children->get (set2)))
+      && ase->children->get (set2))
     return 1;
 
   /* Now do the same, but with the alias sets reversed.  */
   ase = get_alias_set_entry (set2);
   if (ase != 0
-      && (ase->has_zero_child
-	  || ase->children->get (set1)))
+      && ase->children->get (set1))
     return 1;
 
   /* The two alias sets are distinct and neither one is the
Index: gcc/Makefile.in
===================================================================
--- gcc/Makefile.in	(revision 223032)
+++ gcc/Makefile.in	(working copy)
@@ -211,6 +211,7 @@ libgcov-driver-tool.o-warn = -Wno-error
 libgcov-merge-tool.o-warn = -Wno-error
 gimple-match.o-warn = -Wno-unused
 generic-match.o-warn = -Wno-unused
+dfp.o-warn = -Wno-strict-aliasing
 
 # All warnings have to be shut off in stage1 if the compiler used then
 # isn't gcc; configure determines that.  WARN_CFLAGS will be either
Index: gcc/testsuite/gcc.dg/alias-2.c
===================================================================
--- gcc/testsuite/gcc.dg/alias-2.c	(revision 223032)
+++ gcc/testsuite/gcc.dg/alias-2.c	(working copy)
@@ -11,6 +11,6 @@ struct foo {
 int
 sub1 (long long int foobar)
 {
-  struct foo *tmp = (struct foo *) &foobar; // { dg-warning "type-punned pointer might" "" }
+  struct foo *tmp = (struct foo *) &foobar; // { dg-warning "type-punned pointer will" "" }
   return tmp->i;
 }



More information about the Gcc-patches mailing list