Re: pseudo-fix for PR/38984

[Richard Guenther <richard dot guenther at gmail dot com>]

On Tue, Jan 27, 2009 at 5:48 PM, Paolo Bonzini <bonzini@gnu.org> wrote:
> This is only a pseudo-fix because the testcase also exhibits PR/38985.
> The testcase will be fixed in 4.5 thanks to the alias-improvements branch.
>
> The patch however is trivial enough that Richard agreed to put it in 4.4
> too.  Fixing PR38985 in time for 4.4.0 also seems not impossible to me.
>
> Bootstrapped/regtested i686-pc-linux-gnu (regtested C only so far; but
> there is no testcase outside gcc.dg that tries
> -fno-delete-null-pointer-checks).  Ok?

Ok.

Thanks,
Richard.

> Paolo
>
> 2009-01-27  Paolo Bonzini  <bonzini@gnu.org>
>
>        * tree-ssa-structalias.c (get_constraint_for_1): Do not
>        use nothing_id if -fno-delete-null-pointer-checks.
>
> Index: tree-ssa-structalias.c
> ===================================================================
> --- tree-ssa-structalias.c      (revision 142960)
> +++ tree-ssa-structalias.c      (working copy)
> @@ -3030,8 +3030,14 @@ get_constraint_for_1 (tree t, VEC (ce_s,
>      happens below, since it will fall into the default case. The only
>      case we know something about an integer treated like a pointer is
>      when it is the NULL pointer, and then we just say it points to
> -     NULL.  */
> -  if (TREE_CODE (t) == INTEGER_CST
> +     NULL.
> +
> +     Do not do that if -fno-delete-null-pointer-checks though, because
> +     in that case *NULL does not fail, so it _should_ alias *anything.
> +     It is not worth adding a new option or renaming the existing one,
> +     since this case is relatively obscure.  */
> +  if (flag_delete_null_pointer_checks
> +      && TREE_CODE (t) == INTEGER_CST
>       && integer_zerop (t))
>     {
>       temp.var = nothing_id;
>