linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Petr Mladek <pmladek@suse.com>
To: John Ogness <john.ogness@linutronix.de>
Cc: linux-kernel@vger.kernel.org,
	Peter Zijlstra <peterz@infradead.org>,
	Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com>,
	Steven Rostedt <rostedt@goodmis.org>,
	Linus Torvalds <torvalds@linux-foundation.org>,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	Andrea Parri <andrea.parri@amarulasolutions.com>,
	Thomas Gleixner <tglx@linutronix.de>,
	Sergey Senozhatsky <sergey.senozhatsky@gmail.com>,
	Brendan Higgins <brendanhiggins@google.com>,
	kexec@lists.infradead.org
Subject: Re: [RFC PATCH v5 1/3] printk-rb: new printk ringbuffer implementation (writer)
Date: Mon, 2 Dec 2019 16:59:55 +0100	[thread overview]
Message-ID: <20191202155955.meawljmduiciw5t2@pathway.suse.cz> (raw)
In-Reply-To: <20191202154841.qikvuvqt4btudxzg@pathway.suse.cz>

On Mon 2019-12-02 16:48:41, Petr Mladek wrote:
> > +/* Reserve a new descriptor, invalidating the oldest if necessary. */
> > +static bool desc_reserve(struct printk_ringbuffer *rb, u32 *id_out)
> > +{
> > +	struct prb_desc_ring *desc_ring = &rb->desc_ring;
> > +	struct prb_desc *desc;
> > +	u32 id_prev_wrap;
> > +	u32 head_id;
> > +	u32 id;
> > +
> > +	head_id = atomic_read(&desc_ring->head_id);
> > +
> > +	do {
> > +		desc = to_desc(desc_ring, head_id);
> > +
> > +		id = DESC_ID(head_id + 1);
> > +		id_prev_wrap = DESC_ID_PREV_WRAP(desc_ring, id);
> > +
> > +		if (id_prev_wrap == atomic_read(&desc_ring->tail_id)) {
> > +			if (!desc_push_tail(rb, id_prev_wrap))
> > +				return false;
> > +		}
> > +	} while (!atomic_try_cmpxchg(&desc_ring->head_id, &head_id, id));
> 
> Hmm, in theory, ABA problem might cause that we successfully
> move desc_ring->head_id when tail has not been pushed yet.
> 
> As a result we would never call desc_push_tail() until
> it overflows again.
> 
> I am not sure if we need to take care of it. The code is called with
> interrupts disabled. IMHO, only NMI could cause ABA problem
> in reality. But the game (debugging) is lost anyway when NMI ovewrites
> the buffer several times.

BTW: If I am counting correctly. The ABA problem would happen when
exactly 2^30 (1G) messages is written in the mean time.

> Also it should not be a complete failure. We still could bail out when
> the previous state was not reusable. We are on the safe side
> when it was reusable.
> 
> Also we could probably detect when desc_ring->tail_id is not
> updated any longer and do something about it.
> 
> > +
> > +	desc = to_desc(desc_ring, id);
> > +
> > +	/* Set the descriptor's ID and also set its state to reserved. */
> > +	atomic_set(&desc->state_var, id | 0);
> 
> We should check here that the original state id from the previous
> wrap in reusable state (or 0 for bootstrap). On error, we know that
> there was the ABA problem and would need to do something about it.

I believe that it should be enough to add this check and
bail out in case of error.

Best Regards,
Petr

  reply	other threads:[~2019-12-02 16:00 UTC|newest]

Thread overview: 34+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-11-28  1:52 [RFC PATCH v5 0/3] printk: new ringbuffer implementation John Ogness
2019-11-28  1:52 ` [RFC PATCH v5 1/3] printk-rb: new printk ringbuffer implementation (writer) John Ogness
2019-12-02 15:48   ` Petr Mladek
2019-12-02 15:59     ` Petr Mladek [this message]
2019-12-02 16:37       ` John Ogness
2019-12-03  1:17         ` Sergey Senozhatsky
2019-12-03 14:18           ` Steven Rostedt
2019-12-05 12:01             ` John Ogness
2019-12-03  8:54         ` Petr Mladek
2019-12-03 14:13     ` John Ogness
2019-12-03 14:36       ` Petr Mladek
2019-12-09  9:19     ` Sergey Senozhatsky
2019-12-09  7:42   ` Sergey Senozhatsky
2019-12-09  9:00     ` John Ogness
2019-12-09  9:27   ` Sergey Senozhatsky
2019-12-09  9:34     ` John Ogness
2019-12-21 14:22   ` Andrea Parri
2019-12-23 16:01     ` John Ogness
2020-01-03 10:24       ` Petr Mladek
2020-01-04 14:33         ` Andrea Parri
2019-11-28  1:52 ` [RFC PATCH v5 2/3] printk-rb: new printk ringbuffer implementation (reader) John Ogness
2019-12-03 12:06   ` Petr Mladek
2019-12-03 13:46     ` John Ogness
2019-12-04 12:54       ` Petr Mladek
2019-12-04 13:28         ` John Ogness
2019-12-09  8:43   ` Sergey Senozhatsky
2019-12-09  9:03     ` Sergey Senozhatsky
2019-12-09  9:09     ` John Ogness
2019-11-28  1:52 ` [RFC PATCH v5 3/3] printk-rb: add test module John Ogness
2019-12-09  8:44   ` Sergey Senozhatsky
2019-12-05 13:46 ` [RFC PATCH v5 0/3] printk: new ringbuffer implementation Prarit Bhargava
2019-12-05 14:05   ` John Ogness
2019-12-06 14:16     ` Prarit Bhargava
2020-01-27 12:20 ` Eugeniu Rosca

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20191202155955.meawljmduiciw5t2@pathway.suse.cz \
    --to=pmladek@suse.com \
    --cc=andrea.parri@amarulasolutions.com \
    --cc=brendanhiggins@google.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=john.ogness@linutronix.de \
    --cc=kexec@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=peterz@infradead.org \
    --cc=rostedt@goodmis.org \
    --cc=sergey.senozhatsky.work@gmail.com \
    --cc=sergey.senozhatsky@gmail.com \
    --cc=tglx@linutronix.de \
    --cc=torvalds@linux-foundation.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).