From: Jakub Jelinek <jakub@redhat.com>
To: Julian Brown <julian@codesourcery.com>
Cc: gcc-patches@gcc.gnu.org,
Thomas Schwinge <thomas@codesourcery.com>,
Tobias Burnus <tobias@codesourcery.com>,
Fortran List <fortran@gcc.gnu.org>
Subject: Re: [PATCH v2 05/11] OpenMP: Handle reference-typed struct members
Date: Tue, 24 May 2022 15:39:47 +0200 [thread overview]
Message-ID: <YozgIyNhEYGL4nnA@tucnak> (raw)
In-Reply-To: <5084d165c33b382d71b44eaf9e5f3ffdb87bbcfc.1647619144.git.julian@codesourcery.com>
On Fri, Mar 18, 2022 at 09:26:46AM -0700, Julian Brown wrote:
> This patch relates to OpenMP mapping clauses containing struct members of
> reference type, e.g. "mystruct.myref.myptr[:N]". To be able to access
> the array slice through the reference in the middle, we need to perform
> an attach action for that reference, since it is represented internally
> as a pointer.
>
> I don't think the spec allows for this case explicitly. The closest
> clause is (OpenMP 5.0, "2.19.7.1 map Clause"):
>
> "If the type of a list item is a reference to a type T then the
> reference in the device data environment is initialized to refer to
> the object in the device data environment that corresponds to the
> object referenced by the list item. If mapping occurs, it occurs as
> though the object were mapped through a pointer with an array section
> of type T and length one."
Plus the general rule that aggregates are mapped as mapping of all its
individual members/elements.
> --- a/gcc/gimplify.cc
> +++ b/gcc/gimplify.cc
> @@ -9813,7 +9813,10 @@ accumulate_sibling_list (enum omp_region_type region_type, enum tree_code code,
> /* FIXME: If we're not mapping the base pointer in some other clause on this
> directive, I think we want to create ALLOC/RELEASE here -- i.e. not
> early-exit. */
> - if (openmp && attach_detach)
> + if (openmp
> + && attach_detach
> + && !(TREE_CODE (TREE_TYPE (ocd)) == REFERENCE_TYPE
> + && TREE_CODE (TREE_TYPE (TREE_TYPE (ocd))) != POINTER_TYPE))
> return NULL;
Why isn't a reference to pointer handled that way too?
Jakub
next prev parent reply other threads:[~2022-05-24 13:39 UTC|newest]
Thread overview: 30+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-03-18 16:24 [PATCH v2 00/11] OpenMP 5.0: C & C++ "declare mapper" support (plus struct rework, etc.) Julian Brown
2022-03-18 16:24 ` [PATCH v2 01/11] OpenMP 5.0: Clause ordering for OpenMP 5.0 (topological sorting by base pointer) Julian Brown
2022-05-24 13:03 ` Jakub Jelinek
2022-06-08 15:00 ` Julian Brown
2022-06-09 14:45 ` Jakub Jelinek
2022-03-18 16:24 ` [PATCH v2 02/11] Remove omp_target_reorder_clauses Julian Brown
2022-05-24 13:05 ` Jakub Jelinek
2022-03-18 16:24 ` [PATCH v2 03/11] OpenMP/OpenACC struct sibling list gimplification extension and rework Julian Brown
2022-05-24 13:17 ` Jakub Jelinek
2022-03-18 16:24 ` [PATCH v2 04/11] OpenMP/OpenACC: Add inspector class to unify mapped address analysis Julian Brown
2022-05-24 13:32 ` Jakub Jelinek
2022-03-18 16:26 ` [PATCH v2 05/11] OpenMP: Handle reference-typed struct members Julian Brown
2022-05-24 13:39 ` Jakub Jelinek [this message]
2022-03-18 16:26 ` [PATCH v2 06/11] OpenMP: lvalue parsing for map clauses (C++) Julian Brown
2022-05-24 14:15 ` Jakub Jelinek
2022-11-01 21:50 ` Julian Brown
2022-11-01 21:54 ` [PATCH 2/2] OpenMP: Use OMP_ARRAY_SECTION instead of TREE_LIST in C++ FE Julian Brown
2022-11-02 11:58 ` [PATCH v2 06/11] OpenMP: lvalue parsing for map clauses (C++) Jakub Jelinek
2022-11-02 12:20 ` Julian Brown
2022-11-02 12:35 ` Jakub Jelinek
2022-11-08 14:36 ` Julian Brown
2022-11-25 13:22 ` Jakub Jelinek
2022-03-18 16:26 ` [PATCH v2 07/11] OpenMP: lvalue parsing for map clauses (C) Julian Brown
2022-03-18 16:26 ` [PATCH v2 08/11] Use OMP_ARRAY_SECTION instead of TREE_LIST in C++ FE Julian Brown
2022-05-24 14:19 ` Jakub Jelinek
2022-03-18 16:26 ` [PATCH v2 09/11] OpenMP 5.0 "declare mapper" support for C++ Julian Brown
2022-05-24 14:48 ` Jakub Jelinek
2022-05-25 13:37 ` Jakub Jelinek
2022-03-18 16:28 ` [PATCH v2 10/11] OpenMP: Use OMP_ARRAY_SECTION instead of TREE_LIST for array sections in C FE Julian Brown
2022-03-18 16:28 ` [PATCH v2 11/11] OpenMP: Support OpenMP 5.0 "declare mapper" directives for C Julian Brown
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=YozgIyNhEYGL4nnA@tucnak \
--to=jakub@redhat.com \
--cc=fortran@gcc.gnu.org \
--cc=gcc-patches@gcc.gnu.org \
--cc=julian@codesourcery.com \
--cc=thomas@codesourcery.com \
--cc=tobias@codesourcery.com \
/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).