MIME-Version: 1.0
References: <20230213015506.778246-1-joel@joelfernandes.org>
 <Y/JS5SYKPeeDQErL@rowland.harvard.edu> <CAEXW_YQrFSiDEM9cuhkTT2_1+CZoGbg7vC9oL-D-Wd5OQ2mm2w@mail.gmail.com>
In-Reply-To: <CAEXW_YQrFSiDEM9cuhkTT2_1+CZoGbg7vC9oL-D-Wd5OQ2mm2w@mail.gmail.com>
From:   Joel Fernandes <joel@joelfernandes.org>
Date:   Sun, 19 Feb 2023 12:13:14 -0500
Message-ID: <CAEXW_YR6eKDCv+E8Xv2aX=Eo=H0667cqrXkMqKhc_QMZ4Vf59A@mail.gmail.com>
Subject: Re: [PATCH] tools/memory-model: Add details about SRCU read-side
 critical sections
To:     Alan Stern <stern@rowland.harvard.edu>
Cc:     linux-kernel@vger.kernel.org, Boqun Feng <boqun.feng@gmail.com>,
        Jade Alglave <j.alglave@ucl.ac.uk>,
        Luc Maranget <luc.maranget@inria.fr>,
        "Paul E. McKenney" <paulmck@linux.ibm.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Will Deacon <will.deacon@arm.com>,
        Jonas Oberhauser <jonas.oberhauser@huaweicloud.com>,
        Akira Yokosawa <akiyks@gmail.com>,
        Andrea Parri <parri.andrea@gmail.com>,
        Daniel Lustig <dlustig@nvidia.com>,
        David Howells <dhowells@redhat.com>,
        Jonas Oberhauser <jonas.oberhauser@huawei.com>,
        linux-arch@vger.kernel.org, Nicholas Piggin <npiggin@gmail.com>,
        "Paul E. McKenney" <paulmck@kernel.org>,
        =?UTF-8?Q?Paul_Heidekr=C3=BCger?= <paul.heidekrueger@in.tum.de>,
        Will Deacon <will@kernel.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Sun, Feb 19, 2023 at 12:11 PM Joel Fernandes <joel@joelfernandes.org> wrote:
>
> On Sun, Feb 19, 2023 at 11:48 AM Alan Stern <stern@rowland.harvard.edu> wrote:
> >
> > On Mon, Feb 13, 2023 at 01:55:06AM +0000, Joel Fernandes (Google) wrote:
[...]
> > +       A: idx1 = srcu_read_lock(&s);
> > +       B: srcu_read_unlock(&s, idx1);
> > +       C: idx2 = srcu_read_lock(&s);
> > +       D: srcu_read_unlock(&s, idx2);
> > +
> > +it would appear that B was a store to a temporary variable (i.e., s)
> > +and C was a load from that variable, thereby allowing carry-srcu-data
> > +to extend a data dependency from A to D and giving the impression
> > +that D was the srcu-unlock event matching A's srcu-lock.
>
> Even though it may be redundant: would it be possible to also mention
> (after this paragraph) that this case forms an undesirable "->rf" link
> between B and C, which then causes us to link A and D as a result?
>
> A[srcu-lock] ->data B[once] ->rf C[once] ->data D[srcu-unlock].

Apologies, I meant here, care must be taken to avoid:

A[srcu-lock] ->data B[srcu-unlock] ->rf C[srcu-lock] ->data D[srcu-unlock].

Thanks,

  - Joel