MIME-Version: 1.0
References: <20230228155121.3416-1-ubizjak@gmail.com> <Y/5mguXPPqdP3MZF@google.com>
 <20230228160324.2a7c1012@gandalf.local.home> <20230228212911.GX2948950@paulmck-ThinkPad-P17-Gen-1>
 <20230228164124.77c126d2@gandalf.local.home> <CAEXW_YQ515_DOLVUm48GvDADuaY2mSrYTaKa7u6jYDNqBncJww@mail.gmail.com>
 <20230228190846.79b06089@gandalf.local.home> <CAFULd4aY3Y8tyLN70oebZDagBebvH0erwRxKDaEX8L83Xo8gYw@mail.gmail.com>
 <20230301113813.4f16a689@gandalf.local.home> <CAFULd4aWZ+fvVZ+MEt6JD1rFaQZeO5DDWAHtPKOrx8R8DETFBw@mail.gmail.com>
 <20230301135201.59771b73@gandalf.local.home>
In-Reply-To: <20230301135201.59771b73@gandalf.local.home>
From:   Uros Bizjak <ubizjak@gmail.com>
Date:   Wed, 1 Mar 2023 20:18:11 +0100
Message-ID: <CAFULd4YPP_5koKN5ZbdgOueXcpUgap97UkaxxyzzGcQ=GEV__Q@mail.gmail.com>
Subject: Re: [PATCH] rcu: use try_cmpxchg in check_cpu_stall
To:     Steven Rostedt <rostedt@goodmis.org>
Cc:     Joel Fernandes <joel@joelfernandes.org>,
        "Paul E. McKenney" <paulmck@kernel.org>, rcu@vger.kernel.org,
        linux-kernel@vger.kernel.org,
        Frederic Weisbecker <frederic@kernel.org>,
        Neeraj Upadhyay <quic_neeraju@quicinc.com>,
        Josh Triplett <josh@joshtriplett.org>,
        Mathieu Desnoyers <mathieu.desnoyers@efficios.com>,
        Lai Jiangshan <jiangshanlai@gmail.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Wed, Mar 1, 2023 at 7:52=E2=80=AFPM Steven Rostedt <rostedt@goodmis.org>=
 wrote:
>
> On Wed, 1 Mar 2023 19:43:34 +0100
> Uros Bizjak <ubizjak@gmail.com> wrote:
>
> > On Wed, Mar 1, 2023 at 5:38=E2=80=AFPM Steven Rostedt <rostedt@goodmis.=
org> wrote:
> > >
> > > On Wed, 1 Mar 2023 11:28:47 +0100
> > > Uros Bizjak <ubizjak@gmail.com> wrote:
> > >
> > > > These benefits are the reason the change to try_cmpxchg was accepte=
d
> > > > also in the linear code elsewhere in the linux kernel, e.g. [2,3] t=
o
> > > > name a few commits, with a thumbs-up and a claim that the new code =
is
> > > > actually *clearer* at the merge commit [4].
> > >
> > > I'll say it here too. I really like Joel's suggestion of having a
> > > cmpxchg_success() that does not have the added side effect of modifyi=
ng the
> > > old variable.
> > >
> > > I think that would allow for the arch optimizations that you are tryi=
ng to
> > > achieve, as well as remove the side effect that might cause issues do=
wn the
> > > road.
> >
> > Attached patch implements this suggestion.
>
> I like it!

Thanks!

> Anyway to make this more generic?

If we want to put the definition in generic headers, then we also need
to define acquire/release/relaxed and 64bit variants. ATM, we have two
sites that require this definition and I think that for now we could
live with two instances of the same definition in two separate
subsystems. But this would definitely be a good future addition. There
is some code in the form of

if (cmpxchg (ptr, 0, 1) =3D=3D 0)

that can not be converted to try_cmpxchg, but can use cmpxchg_success.

Uros.