From: Richard Biener <richard.guenther@gmail.com>
To: Mikael Morin <morin-mikael@orange.fr>
Cc: Thomas Koenig <tkoenig@netcologne.de>,
Mikael Morin <mikael@gcc.gnu.org>,
gcc-patches@gcc.gnu.org, fortran@gcc.gnu.org
Subject: Re: [PATCH 09/10] fortran: Support clobbering of variable subreferences [PR88364]
Date: Mon, 19 Sep 2022 09:58:56 +0200 [thread overview]
Message-ID: <CAFiYyc1Zg-SgADG9QUNZmwU--NmP6YsddQ3ngcbHQhJmZzyoRQ@mail.gmail.com> (raw)
In-Reply-To: <62e5ef00-b739-37b2-269d-67b9089dfc94@orange.fr>
On Mon, Sep 19, 2022 at 9:31 AM Mikael Morin <morin-mikael@orange.fr> wrote:
>
> Le 18/09/2022 à 12:48, Richard Biener a écrit :
> >
> >> Does *(&a[1]) count as a pointer dereference?
> >
> > Yes, technically.
> >
> >> Even in the original dump it is already simplified to a straight a[1].
> >
> > But this not anymore. The check can probably be relaxed, it stems from the dual purpose of CLOBBER.
> >
> So the following makes the frontend-emitted IL valid, by handing the
> simplification over to the middle-end, but I can't help thinking that
> behavior won't be more reliable.
I think that will just delay the folding. We are basically relying on
the frontends to
restrict what they produce, the *ptr case was specifically for C++
this in CTOR/DTORs
and during inlining we throw away all clobbers that end up "wrong" but
I guess we
might have latent issues when we obfuscate the address in the caller enough and
get undesired simplification ...
>
> diff --git a/gcc/fortran/trans-expr.cc b/gcc/fortran/trans-expr.cc
> index f8fcd2d97d9..5fb9a3a536d 100644
> --- a/gcc/fortran/trans-expr.cc
> +++ b/gcc/fortran/trans-expr.cc
> @@ -6544,8 +6544,9 @@ gfc_conv_procedure_call (gfc_se * se, gfc_symbol *
> sym,
> && !sym->attr.elemental)
> {
> tree var;
> - var = build_fold_indirect_ref_loc (input_location,
> - parmse.expr);
> + var = build1_loc (input_location, INDIRECT_REF,
> + TREE_TYPE (TREE_TYPE
> (parmse.expr)),
> + parmse.expr);
> tree clobber = build_clobber (TREE_TYPE (var));
> gfc_add_modify (&clobbers, var, clobber);
> }
>
next prev parent reply other threads:[~2022-09-19 7:59 UTC|newest]
Thread overview: 36+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-09-16 20:24 [PATCH 00/10] fortran: clobber fixes [PR41453] Mikael Morin
2022-09-16 20:24 ` [PATCH 01/10] fortran: Move the clobber generation code Mikael Morin
2022-09-16 20:24 ` [PATCH 02/10] fortran: Fix invalid function decl clobber ICE [PR105012] Mikael Morin
2022-09-16 20:24 ` [PATCH 03/10] fortran: Move clobbers after evaluation of all arguments [PR106817] Mikael Morin
2022-09-16 20:24 ` [PATCH 04/10] fortran: Support clobbering with implicit interfaces [PR105012] Mikael Morin
2022-09-16 20:24 ` [PATCH 05/10] fortran: Support clobbering of reference variables [PR41453] Mikael Morin
2022-09-16 20:24 ` [PATCH 06/10] fortran: Support clobbering of SAVE variables [PR87395] Mikael Morin
2022-09-16 20:24 ` [PATCH 07/10] fortran: Support clobbering of ASSOCIATE variables [PR87397] Mikael Morin
2022-09-16 20:24 ` [PATCH 08/10] fortran: Support clobbering of allocatables and pointers [PR41453] Mikael Morin
2022-09-16 20:24 ` [PATCH 09/10] fortran: Support clobbering of variable subreferences [PR88364] Mikael Morin
2022-09-17 17:03 ` Thomas Koenig
2022-09-17 19:33 ` Mikael Morin
2022-09-17 19:49 ` Bernhard Reutner-Fischer
2022-09-17 19:50 ` Mikael Morin
2022-09-17 21:24 ` Bernhard Reutner-Fischer
2022-09-18 6:12 ` Richard Biener
2022-09-18 9:10 ` Mikael Morin
2022-09-18 10:23 ` Thomas Koenig
2022-09-18 18:32 ` Harald Anlauf
2022-09-18 20:55 ` Mikael Morin
2022-09-19 7:11 ` Mikael Morin
2022-09-19 19:46 ` Harald Anlauf
2022-09-19 20:50 ` Mikael Morin
2022-09-20 6:54 ` Thomas Koenig
2022-09-20 8:46 ` Mikael Morin
2022-09-20 21:08 ` Harald Anlauf
2022-09-20 21:08 ` Harald Anlauf
2022-09-21 9:57 ` Thomas Koenig
2022-09-21 18:56 ` Mikael Morin
2022-09-21 19:12 ` Harald Anlauf
2022-09-21 19:12 ` Harald Anlauf
2022-09-18 20:43 ` Mikael Morin
2022-09-18 10:48 ` Richard Biener
2022-09-19 7:31 ` Mikael Morin
2022-09-19 7:58 ` Richard Biener [this message]
2022-09-16 20:24 ` [PATCH 10/10] fortran: Support clobbering of derived types [PR41453] Mikael Morin
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=CAFiYyc1Zg-SgADG9QUNZmwU--NmP6YsddQ3ngcbHQhJmZzyoRQ@mail.gmail.com \
--to=richard.guenther@gmail.com \
--cc=fortran@gcc.gnu.org \
--cc=gcc-patches@gcc.gnu.org \
--cc=mikael@gcc.gnu.org \
--cc=morin-mikael@orange.fr \
--cc=tkoenig@netcologne.de \
/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).