public inbox for gcc@gcc.gnu.org
 help / color / mirror / Atom feed
From: Jason Merrill <jason@redhat.com>
To: Richard Biener <richard.guenther@gmail.com>
Cc: Andrew Pinski <pinskia@gmail.com>,
	Eric Gallager <egall@gwmail.gwu.edu>,
	 GCC Mailing List <gcc@gcc.gnu.org>
Subject: Re: Deprecating -fgnu-tm support for GCC 14 and removing it for GCC 15
Date: Wed, 20 Dec 2023 11:40:50 -0500	[thread overview]
Message-ID: <CADzB+2m7UpWwUN5rTxHDoghvv-G5GFk89VoUiOtVJz+aVSYw9Q@mail.gmail.com> (raw)
In-Reply-To: <CAFiYyc3BGJDzv+FjZ1FsJ_nbrNAPtzAoAMPFfsTqxozzK6899A@mail.gmail.com>

[-- Attachment #1: Type: text/plain, Size: 1986 bytes --]

On Mon, Dec 18, 2023 at 3:04 AM Richard Biener via Gcc <gcc@gcc.gnu.org>
wrote:

> On Mon, Dec 18, 2023 at 2:35 AM Andrew Pinski via Gcc <gcc@gcc.gnu.org>
> wrote:
> >
> > On Sun, Dec 17, 2023 at 1:20 PM Eric Gallager <egall@gwmail.gwu.edu>
> wrote:
> > >
> > > On Sat, Dec 16, 2023 at 3:16 PM Andrew Pinski via Gcc <gcc@gcc.gnu.org>
> wrote:
> > > >
> > > > -fgnu-tm support has not been improved since GCC 5 or earlier. It is
> > > > not even supported with LTO. Does it make sense to deprecate the
> > > > support for GCC 14 and remove it in GCC 15?
> > > >
> > > > Thanks,
> > > > Andrew Pinski
> > >
> > > Personally, since GCC is in stage 3 now, I would push that schedule
> > > back a release and move deprecation to GCC 15, and then only remove it
> > > for GCC 16 if no one objects, but then again I don't actually use
> > > -fgnu-tm myself, so I wouldn't be too upset if the faster schedule is
> > > chosen instead.
> >
> > Considering -fgnu-tm has been broken for LTO ever since LTO was
> > introduced, and broken with -fsanitize=undefined and broken with many
> > code that might use internal functions (known since 2015), I suspect
> > nobody is using this option in production nor even trying it out. If
> > this was stage1, I might even just recommend removing the support. But
> > deprecating it during stage 3 seems like a fair compromise.
>
> Btw, I'm OK with deprecating it for GCC 14.  Can you please propose a
> patch for changes.html and add a diagnostic message when -fgnu-tm is used
> (disabled with -Wno-deprecated)?
>

Deprecation makes sense to me.

But keep in mind that transactional memory is still the subject of research
and standardization efforts, though the current proposal (wg21.link/n4923)
is significantly simpler than the earlier TS that GCC implemented.  I don't
know how much of the current implementation would carry over, but I'd be
cautious about tearing everything out just yet.

Jason

      reply	other threads:[~2023-12-20 16:41 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-12-16 20:15 Andrew Pinski
2023-12-17 16:26 ` Florian Weimer
2023-12-17 18:05   ` Andrew Pinski
2023-12-17 21:20 ` Eric Gallager
2023-12-18  1:33   ` Andrew Pinski
2023-12-18  8:01     ` Richard Biener
2023-12-20 16:40       ` Jason Merrill [this message]

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=CADzB+2m7UpWwUN5rTxHDoghvv-G5GFk89VoUiOtVJz+aVSYw9Q@mail.gmail.com \
    --to=jason@redhat.com \
    --cc=egall@gwmail.gwu.edu \
    --cc=gcc@gcc.gnu.org \
    --cc=pinskia@gmail.com \
    --cc=richard.guenther@gmail.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).