[Bug fortran/79492] LHS reallocation, intrinsic function RHS, and -fcheck=all

[sgk at troutmask dot apl.washington.edu]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=79492

--- Comment #2 from Steve Kargl <sgk at troutmask dot apl.washington.edu> ---
On Mon, Feb 13, 2017 at 06:54:16PM +0000, kargl at gcc dot gnu.org wrote:
> 
> The suggested back is

s/back/fix

> Index: trans-expr.c
> ===================================================================
> --- trans-expr.c        (revision 245254)
> +++ trans-expr.c        (working copy)
> @@ -9811,7 +9811,7 @@ gfc_trans_assignment_1 (gfc_expr * expr1
>    /* Walk the lhs.  */
>    lss = gfc_walk_expr (expr1);
>    if (gfc_is_reallocatable_lhs (expr1)
> -       && !(expr2->expr_type == EXPR_FUNCTION
> +       && (expr2->expr_type == EXPR_FUNCTION
>              && expr2->value.function.isym != NULL))
>      lss->is_alloc_lhs = 1;
>    rss = NULL;

This seems to cause regressions.