[PATCH][alias-improvements] Throttle down TBAA for Fortran

[Richard Guenther <rguenther at suse dot de>]

Fortran has a type merging problem (like it has a function decl merging
problem).  This causes us to miscompile protein on the alias-improvements
branch.

Fixed with the following ugl^Wnice hack.

Richard.

2009-01-19  Richard Guenther  <rguenther@suse.de>

	* tree-ssa-alias.c (same_type_for_tbaa): Fortran cannot reliably
	say that two types are not equal, so say don't know for it.

	* gcc.dg/tree-ssa/pr38895.c: Adjust testcase to test what it
	was supposed to test.

Index: gcc/tree-ssa-alias.c
===================================================================
*** gcc/tree-ssa-alias.c	(revision 143496)
--- gcc/tree-ssa-alias.c	(working copy)
*************** same_type_for_tbaa (tree type1, tree typ
*** 373,379 ****
      return -1;
  
    /* Compare the canonical types.  */
!   return (TYPE_CANONICAL (type1) == TYPE_CANONICAL (type2)) ? 1 : 0;
  }
  
  /* Return true, if the two memory references REF1 and REF2 may alias.  */
--- 373,388 ----
      return -1;
  
    /* Compare the canonical types.  */
!   if (TYPE_CANONICAL (type1) == TYPE_CANONICAL (type2))
!     return 1;
! 
!   /* ???  The Fortran frontend does not properly merge equivalent types,
!      so we cannot say 0 for it.  */
!   if (strcmp (lang_hooks.name, "GNU Fortran") == 0)
!     return -1;
! 
!   /* The types are known to be not equal.  */
!   return 0;
  }
  
  /* Return true, if the two memory references REF1 and REF2 may alias.  */
Index: gcc/testsuite/gcc.dg/tree-ssa/pr38895.c
===================================================================
*** gcc/testsuite/gcc.dg/tree-ssa/pr38895.c	(revision 143496)
--- gcc/testsuite/gcc.dg/tree-ssa/pr38895.c	(working copy)
*************** struct C {
*** 16,22 ****
  int foo(struct C *c, struct B *b)
  {
    c->a1.i = 1;
!   b->a1.j = 0;
    return c->a1.i;
  }
  
--- 16,22 ----
  int foo(struct C *c, struct B *b)
  {
    c->a1.i = 1;
!   b->a1.i = 0;
    return c->a1.i;
  }