(j3.2006) Suggest addition from interp F08/0061

[Malcolm Cohen]

This note is suggesting that the processor might do a bad job (from the point of 
view of people who want good performance).  I don't think we want to encourage 
that.

Besides which, unless both the dummy and actual have the TARGET attribute, the 
processor is free to use copy-in/copy-out anyway.  Which makes the note not just 
"recommending bad practice" but "seriously misleading to boot".

This note does not belong in the standard.

Cheers,

-----Original Message----- 
From: Van Snyder
Date: ?? 26?1?28? 8:29
To: j3
Subject: (j3.2006) Suggest addition from interp F08/0061

In 12-006Ar1, There was a suggested edit for a future revision:

  "NOTE 5.10a
   If an actual argument is not simply contiguous and the
   corresponding dummy argument is an assumed-shape array with the
   CONTIGUOUS attribute, the processor might use the so-called
   copy-in/copy-out argument passing mechanism to assure the
   contiguity of the dummy argument."

Is it out of bounds to consider this, with whatever revised wordsmithing
is appropriate?

This was gone from 13-006A.  The whole interp was gone from 13-006Ar1
(because it had been incorporated into Corrigendum 2).


_______________________________________________
J3 mailing list
J3 at mailman.j3-fortran.org
http://mailman.j3-fortran.org/mailman/listinfo/j3

________________________________________________________________________
This e-mail has been scanned for all viruses by Star.
________________________________________________________________________

-- 
................................Malcolm Cohen, Nihon NAG, Tokyo.