RFC: Two possible fixes for g++.dg/torture/pr32950.C failure

[Richard Guenther]

On Tue, Jun 16, 2009 at 5:57 PM, Steve Ellcey<sje@cup.hp.com> wrote:
> On Tue, 2009-06-16 at 11:31 +0200, Richard Guenther wrote:
>
>> You can refer to TYPE_MODE (TREE_TYPE (from)) instead, that
>> should work here.
>>
>> Richard.
>
> Yes, that works.  I have done a bootstrap and test with no regressions
> using this patch and it fixes pr32950.C, is it OK to check it in?

This is ok if the fix for PR40446 didn't already fix this as a side-effect.

Thanks,
Richard.

> Steve Ellcey
> sje@cup.hp.com
>
>
> 2009-06-16  Steve Ellcey  <sje@cup.hp.com>
>
>        * expr.c (expand_assignment): Change complex type check.
>
>
>
> Index: expr.c
> ===================================================================
> --- expr.c      (revision 148510)
> +++ expr.c      (working copy)
> @@ -4250,7 +4250,7 @@ expand_assignment (tree to, tree from, b
>       /* Handle expand_expr of a complex value returning a CONCAT.  */
>       if (GET_CODE (to_rtx) == CONCAT)
>        {
> -         if (TREE_CODE (TREE_TYPE (from)) == COMPLEX_TYPE)
> +         if (COMPLEX_MODE_P (TYPE_MODE (TREE_TYPE (from))))
>            {
>              gcc_assert (bitpos == 0);
>              result = store_expr (from, to_rtx, false, nontemporal);
>
>