Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Mon, 27 Jan 2020 10:29:52 +0100
From:   Oleg Nesterov <oleg@redhat.com>
To:     madhuparnabhowmik10@gmail.com
Cc:     mingo@redhat.com, peterz@infradead.org, juri.lelli@redhat.com,
        vincent.guittot@linaro.org, ebiederm@xmission.com,
        christian.brauner@ubuntu.com, paulmck@kernel.org,
        joel@joelfernandes.org, linux-kernel@vger.kernel.org,
        linux-kernel-mentees@lists.linuxfoundation.org,
        frextrite@gmail.com, rcu@vger.kernel.org
Subject: Re: [PATCH] sched.h: Annotate sighand_struct with __rcu
Message-ID: <20200127092951.GA1116@redhat.com>
References: <20200124045908.26389-1-madhuparnabhowmik10@gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20200124045908.26389-1-madhuparnabhowmik10@gmail.com>
User-Agent: Mutt/1.5.24 (2015-08-30)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 01/24, madhuparnabhowmik10@gmail.com wrote:
>
> --- a/include/linux/sched.h
> +++ b/include/linux/sched.h
> @@ -918,7 +918,7 @@ struct task_struct {
>  
>  	/* Signal handlers: */
>  	struct signal_struct		*signal;
> -	struct sighand_struct		*sighand;
> +	struct sighand_struct __rcu		*sighand;
>  	sigset_t			blocked;
>  	sigset_t			real_blocked;
>  	/* Restored if set_restore_sigmask() was used: */
> diff --git a/kernel/signal.c b/kernel/signal.c
> index bcd46f547db3..9ad8dea93dbb 100644
> --- a/kernel/signal.c
> +++ b/kernel/signal.c
> @@ -1383,7 +1383,7 @@ struct sighand_struct *__lock_task_sighand(struct task_struct *tsk,
>  		 * must see ->sighand == NULL.
>  		 */
>  		spin_lock_irqsave(&sighand->siglock, *flags);
> -		if (likely(sighand == tsk->sighand))
> +		if (likely(sighand == rcu_access_pointer(tsk->sighand)))
>  			break;
>  		spin_unlock_irqrestore(&sighand->siglock, *flags);
>  	}

ACK,

perhaps you can also cleanup copy_sighand(). rcu_assign_pointer() makes no
sense, we should either move it down or simply use RCU_INIT_POINTER().

Oleg.