From: Tony Lindgren <tony@atomide.com>
To: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: Petr Mladek <pmladek@suse.com>,
Steven Rostedt <rostedt@goodmis.org>,
Peter Zijlstra <peterz@infradead.org>,
Thomas Gleixner <tglx@linutronix.de>,
linux-kernel@vger.kernel.org,
Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com>,
"Rafael J. Wysocki" <rjw@rjwysocki.net>,
linux-pm@vger.kernel.org
Subject: Re: Regression in next with use printk_safe buffers in printk
Date: Tue, 14 Feb 2017 08:54:52 -0800 [thread overview]
Message-ID: <20170214165451.GI21809@atomide.com> (raw)
In-Reply-To: <20170214160140.GA401@tigerII.localdomain>
* Sergey Senozhatsky <sergey.senozhatsky@gmail.com> [170214 08:03]:
> Hello,
>
> Cc Rafael, just in case
>
> On (02/13/17 10:59), Tony Lindgren wrote:
> > Looks like commit f975237b7682 ("printk: use printk_safe buffers in
> > printk") causes "possible circular locking dependency detected " for
> > me on the first suspend.
>
> thanks for the report.
>
> > Reverting the following four patches in next makes it go away:
> >
> > d9c23523ed98 ("printk: drop call_console_drivers() unused param")
> > de6fcbdb68b2 ("printk: convert the rest to printk-safe")
> > 8b1742c9c207 ("printk: remove zap_locks() function")
> > f975237b7682 ("printk: use printk_safe buffers in printk")
>
>
> these patches basically just enable locked where it previously was
> forcibly turned off. no timekeeping/pm/sched/etc code was modified.
> can you share the link where Peter pointed out that this might be
> caused by printk() changes?
Oh sorry I should have been more specific. Not much there to share,
I got redirected over to Peter's department in a private email thread
while chasing this issue. So that was just Peter's comment looking at
the log output.
> timekeeping_resume()
> lock timekeeper_lock
> lock tk_core
> tk_debug_account_sleep_time()
> printk() << lockdep was disabled here before
> try_to_wake_up()
> lock_hrtimer_base() ##hrtimer_bases.lock
Yeah above seems describe what changed, so your patch makes sense.
> shouldn't tk_debug_account_sleep_time() do printk_deferred() instead of
> 'normal' printk()?
> printk() calls from under timekeeping seqlock are not safe, aren't they?
> and tk_debug_account_sleep_time() is under tk_core seq lock.
>
> IOW, replace pr_info() in tk_debug_account_sleep_time() with something
> like this
>
> printk_deferred(KERN_INFO "Suspended for %lld.%03lu seconds\n",
> (s64)t->tv_sec, t->tv_nsec / NSEC_PER_MSEC);
Your patch below fixes the issue for me thanks. I had to apply it manually
though as tabs got replaced by spaces probably by some mail daemons.
Regards,
Tony
> ---
>
> diff --git a/kernel/time/timekeeping_debug.c b/kernel/time/timekeeping_debug.c
> index ca9fb800336b..b8f7146c3538 100644
> --- a/kernel/time/timekeeping_debug.c
> +++ b/kernel/time/timekeeping_debug.c
> @@ -75,7 +75,8 @@ void tk_debug_account_sleep_time(struct timespec64 *t)
> int bin = min(fls(t->tv_sec), NUM_BINS-1);
>
> sleep_time_bin[bin]++;
> - pr_info("Suspended for %lld.%03lu seconds\n", (s64)t->tv_sec,
> + printk_deferred(KERN_INFO "Suspended for %lld.%03lu seconds\n",
> + (s64)t->tv_sec,
> t->tv_nsec / NSEC_PER_MSEC);
> }
>
>
next prev parent reply other threads:[~2017-02-14 16:55 UTC|newest]
Thread overview: 16+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-02-13 18:59 Regression in next with use printk_safe buffers in printk Tony Lindgren
2017-02-14 16:01 ` Sergey Senozhatsky
2017-02-14 16:18 ` Peter Zijlstra
2017-02-14 16:56 ` Sergey Senozhatsky
2017-02-14 17:03 ` Tony Lindgren
2017-02-15 4:44 ` Sergey Senozhatsky
2017-02-14 18:29 ` Peter Zijlstra
2017-02-15 4:49 ` Sergey Senozhatsky
2017-02-14 16:54 ` Tony Lindgren [this message]
2017-02-15 18:01 ` Tony Lindgren
2017-02-16 1:31 ` Sergey Senozhatsky
2017-02-16 4:03 ` Tony Lindgren
2017-02-16 4:25 ` Sergey Senozhatsky
2017-02-16 15:10 ` Tony Lindgren
2017-02-16 16:31 ` Sergey Senozhatsky
2017-02-16 19:13 ` Tony Lindgren
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=20170214165451.GI21809@atomide.com \
--to=tony@atomide.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pm@vger.kernel.org \
--cc=peterz@infradead.org \
--cc=pmladek@suse.com \
--cc=rjw@rjwysocki.net \
--cc=rostedt@goodmis.org \
--cc=sergey.senozhatsky.work@gmail.com \
--cc=sergey.senozhatsky@gmail.com \
--cc=tglx@linutronix.de \
/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).