Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <f0b1f33108fec467689b52568b5ad7c906eb5435.camel@sipsolutions.net>
Subject: Re: [PATCH] um: protect find_task_by_pid_ns() with rcu lock
From:   Johannes Berg <johannes@sipsolutions.net>
To:     Dawei Li <set_pte_at@outlook.com>, richard@nod.at,
        anton.ivanov@cambridgegreys.com
Cc:     linux-um@lists.infradead.org, linux-kernel@vger.kernel.org
Date:   Sun, 16 Oct 2022 19:52:48 +0200
In-Reply-To: <TYCP286MB2323CA02DF3A741B5F58B297CA269@TYCP286MB2323.JPNP286.PROD.OUTLOOK.COM>
References: <TYCP286MB2323CA02DF3A741B5F58B297CA269@TYCP286MB2323.JPNP286.PROD.OUTLOOK.COM>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.44.4 (3.44.4-2.fc36) 
MIME-Version: 1.0
Precedence: bulk

On Sun, 2022-10-16 at 23:10 +0800, Dawei Li wrote:
> find_task_by_pid_ns() is safe if and only if it's under protection
> from rcu read lock, implements the protection.
>=20
> base-commit: 193cb8372424184dde28088a4230a5fed0afb0ad
>=20
> Signed-off-by: Dawei Li <set_pte_at@outlook.com>
> ---
>  arch/um/drivers/mconsole_kern.c | 3 +++
>  1 file changed, 3 insertions(+)
>=20
> diff --git a/arch/um/drivers/mconsole_kern.c b/arch/um/drivers/mconsole_k=
ern.c
> index 5026e7b9adfe..6b80e766ae80 100644
> --- a/arch/um/drivers/mconsole_kern.c
> +++ b/arch/um/drivers/mconsole_kern.c
> @@ -683,7 +683,10 @@ void mconsole_stack(struct mc_request *req)
>  		return;
>  	}
> =20
> +	rcu_read_lock();
>  	to =3D find_task_by_pid_ns(pid_requested, &init_pid_ns);
> +	rcu_read_unlock();
> +
>  	if ((to =3D=3D NULL) || (pid_requested =3D=3D 0)) {
>  		mconsole_reply(req, "Couldn't find that pid", 1, 0);
>  		return;

This can't be right - the 'to' pointer is used later in the code (not in
the existing context), so the protection, if needed, would have to
include the later "with_console()" call?

johannes