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;
From:   John Ogness <john.ogness@linutronix.de>
To:     Thomas Gleixner <tglx@linutronix.de>,
        LKML <linux-kernel@vger.kernel.org>
Cc:     Petr Mladek <pmladek@suse.com>,
        Sergey Senozhatsky <senozhatsky@chromium.org>,
        Steven Rostedt <rostedt@goodmis.org>,
        Linus Torvalds <torvalds@linuxfoundation.org>,
        Peter Zijlstra <peterz@infradead.org>,
        "Paul E. McKenney" <paulmck@kernel.org>,
        Daniel Vetter <daniel@ffwll.ch>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        Helge Deller <deller@gmx.de>,
        Jason Wessel <jason.wessel@windriver.com>,
        Daniel Thompson <daniel.thompson@linaro.org>
Subject: Re: [patch RFC 20/29] printk: Add non-BKL console acquire/release
 logic
In-Reply-To: <20220910222301.539158418@linutronix.de>
References: <20220910221947.171557773@linutronix.de>
 <20220910222301.539158418@linutronix.de>
Date:   Tue, 27 Sep 2022 15:55:24 +0206
Message-ID: <87tu4tkjvf.fsf@jogness.linutronix.de>
MIME-Version: 1.0
Content-Type: text/plain
Precedence: bulk

Below is a fix that was used for the LPC2022 demo so that a hostile
takeover upon timeout is possible.

On 2022-09-11, Thomas Gleixner <tglx@linutronix.de> wrote:
> --- a/kernel/printk/printk_nobkl.c
> +++ b/kernel/printk/printk_nobkl.c
> +/**
> + * __cons_try_acquire - Try to acquire the console for printk output
> + * @ctxt:	Pointer to an aquire context which contains
> + *		all information about the acquire mode
> + *
> + * Returns: True if the acquire was successful. False on fail.
> + *
> + * In case of success @ctxt->state contains the acquisition
> + * state.
> + *
> + * In case of fail @ctxt->old_state contains the state
> + * which was read from @con->state for analysis by the caller.
> + */
> +static bool __cons_try_acquire(struct cons_context *ctxt)
> +{
> +	struct console *con = ctxt->console;
> +	unsigned int cpu = smp_processor_id();
> +	struct cons_state old, new;
> +
> +	if (WARN_ON_ONCE(!(con->flags & CON_NO_BKL)))
> +		return false;
> +
> +	cons_state_read(con, STATE_REAL, &old);
> +
> +again:
> +	if (cons_check_panic())
> +		return false;
> +
> +	/* Preserve it for the caller and for spinwait */
> +	copy_full_state(ctxt->old_state, old);
> +
> +	if (!cons_state_ok(old))
> +		return false;
> +
> +	/* Set up the new state for takeover */
> +	copy_full_state(new, old);
> +	new.locked = 1;
> +	new.thread = ctxt->thread;
> +	new.cur_prio = ctxt->prio;
> +	new.req_prio = CONS_PRIO_NONE;
> +	new.cpu = cpu;
> +
> +	/* Attempt to acquire it directly if unlocked */
> +	if (!old.locked) {
> +		if (!cons_state_try_cmpxchg(con, STATE_REAL, &old, &new))
> +			goto again;
> +
> +		ctxt->hostile = false;
> +		copy_full_state(ctxt->state, new);
> +		goto success;
> +	}
> +
> +	/*
> +	 * Give up if the calling context is the printk thread. The
> +	 * atomic writer will wake the thread when it is done with
> +	 * the important output.
> +	 */
> +	if (ctxt->thread)
> +		return false;
> +
> +	/*
> +	 * If the active context is on the same CPU then there is
> +	 * obviously no handshake possible.
> +	 */
> +	if (old.cpu == cpu)
> +		goto check_hostile;
> +
> +	/*
> +	 * If the caller did not request spin-waiting or a request with the
> +	 * same or higher priority is pending then check whether a hostile
> +	 * takeover is due.
> +	 */
> +	if (!ctxt->spinwait || old.req_prio >= ctxt->prio)
> +		goto check_hostile;
> +
> +	/* Proceed further with spin acquire */
> +	if (!cons_setup_handover(ctxt))
> +		return false;
> +
> +	/*
> +	 * Setup the request in state[REAL]. Hand in the state, which was
> +	 * used to make the decision to spinwait above, for comparison.
> +	 */
> +	if (!cons_setup_request(ctxt, old))
> +		return false;
> +
> +	/* Now spin on it */
> +	if (!cons_try_acquire_spin(ctxt))
> +		return false;

Instead of returning false here, it needs to:

		goto check_hostile;

cons_try_acquire_spin() may have failed due to a timeout, in which case
a hostile takeover might be necessary.

> +success:
> +	/* Common updates on success */
> +	return true;
> +
> +check_hostile:
> +	if (!ctxt->hostile)
> +		return false;
> +
> +	if (!cons_state_try_cmpxchg(con, STATE_REAL, &old, &new))
> +		goto again;
> +
> +	ctxt->hostile = true;
> +	copy_full_state(ctxt->state, new);
> +	goto success;
> +}

John Ogness