[J3] Assumed-size Argument to Collective Subroutine

[Malcolm Cohen]

Hi Brad,

The bit you quoted is already sufficient. No interp. No edit. Already correct.

I note that virtually all the intrinsics that operate on arrays rely on that requirement, they do not have specific cases called out.

Cheers,
-- 
..............Malcolm Cohen, NAG Oxford/Tokyo.

-----Original Message-----
From: J3 <j3-bounces at mailman.j3-fortran.org> On Behalf Of Brad Richardson via J3
Sent: Tuesday, October 22, 2024 3:53 AM
To: General J3 interest list <j3 at mailman.j3-fortran.org>
Cc: Brad Richardson <everythingfunctional at protonmail.com>; fortran at lbl.gov
Subject: [J3] Assumed-size Argument to Collective Subroutine

Hi all,

The group at LBL recently encountered this question.

Does the standard allow or prohibit assumed-size arrays as actual arguments to the collective subroutines?

The only bit of the standard I've found that seems like it might have some bearing is the very generic statement in 8.5.8.5 Assumed-size
array:

    An assumed-size array shall not appear in a context that requires its shape.

We believe it would be impossible to implement the collective subroutines for arguments that are assumed size (i.e. their size is not known), and so I have an interp I could submit to add that requirement to the their descriptions. Have I missed something and this is already covered somewhere? Or do others believe it should be allowed somehow?

Regards,
Brad Richardson