From: Tobias Burnus <tobias@codesourcery.com>
To: Jakub Jelinek <jakub@redhat.com>
Cc: Mikael Morin <morin-mikael@orange.fr>, <gscfq@t-online.de>,
Kwok Cheung Yeung <kcy@codesourcery.com>,
gcc-patches <gcc-patches@gcc.gnu.org>,
fortran <fortran@gcc.gnu.org>
Subject: Re: [PATCH] openmp, fortran: Check that event handles passed to detach clauses are not arrays [PR104131]
Date: Tue, 1 Mar 2022 10:17:54 +0100 [thread overview]
Message-ID: <14be1163-51e6-19bd-0032-c18a53ffe254@codesourcery.com> (raw)
In-Reply-To: <Yh3WcFjhygX+Q7Fd@tucnak>
First, thanks for looking into the standard. I think the information is
too much spread through the standard. I keep searching for restrictions,
find them at 5 places and miss another 5. With OpenMP 5.2, it became
even worse.
On 01.03.22 09:16, Jakub Jelinek wrote:
> As there is no explicit allowing of array sections, array sections aren't
> allowed in detach, and it must be scalar integer.
> The question is if a non-coindexed coarray is a scalar integer variable or
> not.
I think the Fortran sense, yes. That only coindexed vars are disallowed
also implies this in the OpenMP spec. In terms of the implementation,
a local part of a coarray is really not much different from any other
variable, except that all coarrays have to be either in static memory
or allocatables/pointers (which are collectively allocated).
For a non-descriptor variable like:
program main
integer, save :: a[*]
call foo(a)
contains
subroutine foo(x)
integer :: x[*]
end
end
That's also the case for the internal representation:
static void * restrict caf_token.2;
static integer(kind=4) * restrict a;
void foo (integer(kind=4) * restrict x,
void * restrict caf_token.0,
integer(kind=8) caf_offset.1)
'x' is a pretty normal variable (admittedly, internally now a pointer)
which can be used like a noncoarray. (The pointer is there to permit
to put the var into some special, remote accessible memory.)
Tobias
-----------------
Siemens Electronic Design Automation GmbH; Anschrift: Arnulfstraße 201, 80634 München; Gesellschaft mit beschränkter Haftung; Geschäftsführer: Thomas Heurung, Frank Thürauf; Sitz der Gesellschaft: München; Registergericht München, HRB 106955
next prev parent reply other threads:[~2022-03-01 9:18 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-02-28 14:01 Kwok Cheung Yeung
2022-02-28 14:07 ` Jakub Jelinek
2022-02-28 14:27 ` Kwok Cheung Yeung
2022-02-28 15:54 ` Mikael Morin
2022-02-28 16:00 ` Jakub Jelinek
2022-02-28 17:33 ` Mikael Morin
2022-02-28 17:37 ` Jakub Jelinek
2022-02-28 18:38 ` Kwok Cheung Yeung
2022-02-28 20:37 ` Mikael Morin
2022-02-28 20:45 ` Mikael Morin
2022-02-28 21:37 ` Jakub Jelinek
2022-02-28 22:36 ` Mikael Morin
2022-03-01 7:58 ` Tobias Burnus
2022-03-01 8:16 ` Jakub Jelinek
2022-03-01 9:17 ` Tobias Burnus [this message]
2022-03-01 15:37 ` Mikael Morin
2022-03-02 17:22 ` [PATCH][v2] openmp, fortran: Check that the type of an event handle in a detach clause is suitable [PR104131] Kwok Cheung Yeung
2022-03-02 17:31 ` Jakub Jelinek
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=14be1163-51e6-19bd-0032-c18a53ffe254@codesourcery.com \
--to=tobias@codesourcery.com \
--cc=fortran@gcc.gnu.org \
--cc=gcc-patches@gcc.gnu.org \
--cc=gscfq@t-online.de \
--cc=jakub@redhat.com \
--cc=kcy@codesourcery.com \
--cc=morin-mikael@orange.fr \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).