public inbox for libc-alpha@sourceware.org
 help / color / mirror / Atom feed
From: Noah Goldstein <goldstein.w.n@gmail.com>
To: Sunil Pandey <skpgkp2@gmail.com>
Cc: "H.J. Lu" <hjl.tools@gmail.com>,
	 Libc-stable Mailing List <libc-stable@sourceware.org>,
	Florian Weimer <fweimer@redhat.com>,
	 Jangwoong Kim <6812skiii@gmail.com>,
	GNU C Library <libc-alpha@sourceware.org>
Subject: Re: [PATCH v4] nptl: Effectively skip CAS in spinlock loop
Date: Wed, 28 Sep 2022 20:11:20 -0400	[thread overview]
Message-ID: <CAFUsyfJKvaP1NQju+3TWbnmwB9N-4m2kDNgqoZOcQamDGHX_pQ@mail.gmail.com> (raw)
In-Reply-To: <CAMAf5_c9fHwcAMr-QmnpFBeghXifrZxTDDQ0Bv21K3X_GFSAyA@mail.gmail.com>

On Sun, Sep 11, 2022 at 4:24 PM Sunil Pandey via Libc-stable
<libc-stable@sourceware.org> wrote:
>
> On Thu, Jan 20, 2022 at 5:08 AM H.J. Lu via Libc-alpha
> <libc-alpha@sourceware.org> wrote:
> >
> > On Thu, Jan 20, 2022 at 04:53:27AM -0800, H.J. Lu wrote:
> > > On Tue, Dec 14, 2021 at 4:31 AM Jangwoong Kim via Libc-alpha
> > > <libc-alpha@sourceware.org> wrote:
> > > >
> > > > The commit:
> > > > "Add LLL_MUTEX_READ_LOCK [BZ #28537]"
> > > > SHA1: d672a98a1af106bd68deb15576710cd61363f7a6
> > > >
> > > > introduced LLL_MUTEX_READ_LOCK, to skip CAS in spinlock loop
> > > > if atomic load fails. But, "continue" inside of do-while loop
> > > > does not skip the evaluation of escape expression, thus CAS
> > > > is not skipped.
> > > >
> > > > Replace do-while with while and skip LLL_MUTEX_TRYLOCK if
> > > > LLL_MUTEX_READ_LOCK fails.
> > > > ---
> > > >  nptl/pthread_mutex_lock.c | 5 +----
> > > >  1 file changed, 1 insertion(+), 4 deletions(-)
> > > >
> > > > diff --git a/nptl/pthread_mutex_lock.c b/nptl/pthread_mutex_lock.c
> > > > index 47b88a6b5b..06274c68e4 100644
> > > > --- a/nptl/pthread_mutex_lock.c
> > > > +++ b/nptl/pthread_mutex_lock.c
> > > > @@ -138,7 +138,7 @@ PTHREAD_MUTEX_LOCK (pthread_mutex_t *mutex)
> > > >           int cnt = 0;
> > > >           int max_cnt = MIN (max_adaptive_count (),
> > > >                              mutex->__data.__spins * 2 + 10);
> > > > -         do
> > > > +         while (LLL_MUTEX_READ_LOCK (mutex) != 0 || LLL_MUTEX_TRYLOCK (mutex) != 0)
> > > >             {
> > > >               if (cnt++ >= max_cnt)
> > > >                 {
> > > > @@ -146,10 +146,7 @@ PTHREAD_MUTEX_LOCK (pthread_mutex_t *mutex)
> > > >                   break;
> > > >                 }
> > > >               atomic_spin_nop ();
> > > > -             if (LLL_MUTEX_READ_LOCK (mutex) != 0)
> > > > -               continue;
> > > >             }
> > > > -         while (LLL_MUTEX_TRYLOCK (mutex) != 0);
> > > >
> > > >           mutex->__data.__spins += (cnt - mutex->__data.__spins) / 8;
> > > >         }
> > > > --
> > > > 2.25.1
> > > >
> > >
> > > LGTM.
> > >
> > > Reviewed-by: H.J. Lu <hjl.tools@gmail.com>
> > >
> > > Thanks.
> > >
> >
> > This is the v4 patch I am checkinng in.
> >
> > H.J.
> > ---
> > From: Jangwoong Kim <6812skiii@gmail.com>
> > Date: Tue, 14 Dec 2021 21:30:51 +0900
> >
> > The commit:
> > "Add LLL_MUTEX_READ_LOCK [BZ #28537]"
> > SHA1: d672a98a1af106bd68deb15576710cd61363f7a6
> >
> > introduced LLL_MUTEX_READ_LOCK, to skip CAS in spinlock loop
> > if atomic load fails. But, "continue" inside of do-while loop
> > does not skip the evaluation of escape expression, thus CAS
> > is not skipped.
> >
> > Replace do-while with while and skip LLL_MUTEX_TRYLOCK if
> > LLL_MUTEX_READ_LOCK fails.
> >
> > Reviewed-by: H.J. Lu <hjl.tools@gmail.com>
> > ---
> >  nptl/pthread_mutex_lock.c | 5 ++---
> >  1 file changed, 2 insertions(+), 3 deletions(-)
> >
> > diff --git a/nptl/pthread_mutex_lock.c b/nptl/pthread_mutex_lock.c
> > index 31dfc7f1d7..d2e652d151 100644
> > --- a/nptl/pthread_mutex_lock.c
> > +++ b/nptl/pthread_mutex_lock.c
> > @@ -146,10 +146,9 @@ PTHREAD_MUTEX_LOCK (pthread_mutex_t *mutex)
> >                   break;
> >                 }
> >               atomic_spin_nop ();
> > -             if (LLL_MUTEX_READ_LOCK (mutex) != 0)
> > -               continue;
> >             }
> > -         while (LLL_MUTEX_TRYLOCK (mutex) != 0);
> > +         while (LLL_MUTEX_READ_LOCK (mutex) != 0
> > +                || LLL_MUTEX_TRYLOCK (mutex) != 0);
> >
> >           mutex->__data.__spins += (cnt - mutex->__data.__spins) / 8;
> >         }
> > --
> > 2.34.1
> >
>
> I would like to backport this patch to release branch 2.33 and 2.34

Fine by me.
>
> Any comments/suggestions or objections on this.
>
> commit 6b8dbbd03ac88f169b65b5c7d7278576a11d2e44
> Author: Jangwoong Kim <6812skiii@gmail.com>
> Date:   Tue Dec 14 21:30:51 2021 +0900
>
>     nptl: Effectively skip CAS in spinlock loop
>
>     The commit:
>     "Add LLL_MUTEX_READ_LOCK [BZ #28537]"
>     SHA1: d672a98a1af106bd68deb15576710cd61363f7a6

      reply	other threads:[~2022-09-29  0:11 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-12-14 12:30 [PATCH v3] " Jangwoong Kim
2022-01-20 12:53 ` H.J. Lu
2022-01-20 13:08   ` [PATCH v4] " H.J. Lu
2022-09-11 20:23     ` Sunil Pandey
2022-09-29  0:11       ` Noah Goldstein [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=CAFUsyfJKvaP1NQju+3TWbnmwB9N-4m2kDNgqoZOcQamDGHX_pQ@mail.gmail.com \
    --to=goldstein.w.n@gmail.com \
    --cc=6812skiii@gmail.com \
    --cc=fweimer@redhat.com \
    --cc=hjl.tools@gmail.com \
    --cc=libc-alpha@sourceware.org \
    --cc=libc-stable@sourceware.org \
    --cc=skpgkp2@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).