On Sat, Aug 31, 2019 at 09:56:52AM -0700, Steve Kargl wrote:
> On Sat, Aug 31, 2019 at 04:59:03PM +0100, Paul Richard Thomas wrote:
> > +
> > + /* Match a SELECT RANK statement. */
> > +
> > + match
> > + gfc_match_select_rank (void)
> > + {
> > + gfc_expr *expr1, *expr2 = NULL;
> > + match m;
> > + char name[GFC_MAX_SYMBOL_LEN];
> > + gfc_symbol *sym, *sym2;
> > + gfc_namespace *ns = gfc_current_ns;
> > + gfc_array_spec *as;
>
> (snip)
>
> > + if (expr2->expr_type != EXPR_VARIABLE
> > + || !(as && as->type == AS_ASSUMED_RANK))
> > + gfc_error_now ("The SELECT RANK selector at %C must be an assumed "
> > + "rank variable");
>
> If an error occurs, should this set m = MATCH_ERROR and goto cleanup?
> OR, set m = MATCH_ERROR, free expr1 and expr2 and return m?
> IOW, if an error occurs, why should gfortran continue to match select rank?
>
> Still looking at the remaining part of patch.
>
There's another gfc_error_now several lines down. Does the
same early return apply there as well.
Also, found
Index: gcc/fortran/parse.c
===================================================================
*** gcc/fortran/parse.c (revision 275242)
--- gcc/fortran/parse.c (working copy)
+
+ /* At this point, we're got a nonempty select block. */
s/we're/we've
--
Steve
