Date:   Sun, 19 Dec 2021 10:48:12 +0100
From:   Sebastian Andrzej Siewior <bigeasy@linutronix.de>
To:     Herbert Xu <herbert@gondor.apana.org.au>
Cc:     Jason@zx2c4.com, linux-kernel@vger.kernel.org, tytso@mit.edu,
        tglx@linutronix.de, peterz@infradead.org
Subject: Re: [PATCH 5/5] random: Defer processing of randomness on PREEMPT_RT.
Message-ID: <Yb7/3JUcDENyYySN@linutronix.de>
References: <20211207201037.h46573oa5nfj33xq@linutronix.de>
 <20211217022338.GA19411@gondor.apana.org.au>
 <YbxRxPdBDMu8KIy6@linutronix.de>
 <20211218032409.GA11425@gondor.apana.org.au>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
In-Reply-To: <20211218032409.GA11425@gondor.apana.org.au>
Precedence: bulk

On 2021-12-18 14:24:09 [+1100], Herbert Xu wrote:
> On Fri, Dec 17, 2021 at 10:00:52AM +0100, Sebastian Andrzej Siewior wrote:
> >
> > I'm sorry, I can't connect the dots here. I was trying to explain that
> > for the lock in question it is not possible to spin-wait without
> > disabling interrupts first.
>=20
> That's precisely the problem the socket lock was designed to
> solve.
=E2=80=A6
> Take a look at lock_sock/release_sock and bh_lock_sock/bh_unlock_sock
> in net/core/sock.c.

That one. I find the semantic difficult to understand. Some BH-user
check sock_owned_by_user(), some don't.
However, a semaphore should work. The atomic user would do
down_trylock() while preemptible caller would invoke down() and sleep
until the lock is available. Let me check if that works here=E2=80=A6

> Cheers,

Sebastian