From: Waiman Long <longman@redhat.com>
To: paulmck@kernel.org, linux-kernel@vger.kernel.org
Cc: clm@meta.com, jstultz@google.com, tglx@linutronix.de,
sboyd@kernel.org, feng.tang@intel.com
Subject: Re: [PATCH clocksource] Reject bogus watchdog clocksource measurements
Date: Wed, 19 Oct 2022 23:01:50 -0400 [thread overview]
Message-ID: <df0960dc-86d4-984f-8dd8-c9d118aae449@redhat.com> (raw)
In-Reply-To: <20221019230904.GA2502730@paulmck-ThinkPad-P17-Gen-1>
On 10/19/22 19:09, Paul E. McKenney wrote:
> One remaining clocksource-skew issue involves extreme CPU overcommit,
> which can cause the clocksource watchdog measurements to be delayed by
> tens of seconds. This in turn means that a clock-skew criterion that
> is appropriate for a 500-millisecond interval will instead give lots of
> false positives.
CPU overcommit means it is running in a VM. Right? Unfortunately, there
is not a consistent cross-arch way to check for running under a
hypervisor or we may want to add such a test if available. However,
CLOCKSOURCE_WATCHDOG is only enabled in x86 and mips. Maybe we can add a
helper function to do that.
>
> Therefore, check for the watchdog clocksource reporting much larger or
> much less than the time specified by WATCHDOG_INTERVAL. In these cases,
> print a pr_warn() warning and refrain from marking the clocksource under
> test as being unstable.
>
> Reported-by: Chris Mason <clm@meta.com>
> Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
> Cc: John Stultz <jstultz@google.com>
> Cc: Thomas Gleixner <tglx@linutronix.de>
> Cc: Stephen Boyd <sboyd@kernel.org>
> Cc: Feng Tang <feng.tang@intel.com>
> Cc: Waiman Long <longman@redhat.com>
>
> diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c
> index 8058bec87acee..dcaf38c062161 100644
> --- a/kernel/time/clocksource.c
> +++ b/kernel/time/clocksource.c
> @@ -386,7 +386,7 @@ EXPORT_SYMBOL_GPL(clocksource_verify_percpu);
>
> static void clocksource_watchdog(struct timer_list *unused)
> {
> - u64 csnow, wdnow, cslast, wdlast, delta;
> + u64 csnow, wdnow, cslast, wdlast, delta, wdi;
> int next_cpu, reset_pending;
> int64_t wd_nsec, cs_nsec;
> struct clocksource *cs;
> @@ -440,6 +440,17 @@ static void clocksource_watchdog(struct timer_list *unused)
> if (atomic_read(&watchdog_reset_pending))
> continue;
>
> + /* Check for bogus measurements. */
> + wdi = jiffies_to_nsecs(WATCHDOG_INTERVAL);
> + if (wd_nsec < (wdi >> 2)) {
> + pr_warn("timekeeping watchdog on CPU%d: Watchdog clocksource '%s' advanced only %lld ns during %d-jiffy time interval, skipping watchdog check.\n", smp_processor_id(), watchdog->name, wd_nsec, WATCHDOG_INTERVAL);
> + continue;
> + }
> + if (wd_nsec > (wdi << 2)) {
> + pr_warn("timekeeping watchdog on CPU%d: Watchdog clocksource '%s' advanced an excessive %lld ns during %d-jiffy time interval, probable CPU overutilization, skipping watchdog check.\n", smp_processor_id(), watchdog->name, wd_nsec, WATCHDOG_INTERVAL);
> + continue;
> + }
> +
> /* Check the deviation from the watchdog clocksource. */
> md = cs->uncertainty_margin + watchdog->uncertainty_margin;
> if (abs(cs_nsec - wd_nsec) > md) {
In the worst case, there will be a warning every half second or so.
Should we rate limit the number of these warnings in some way?
Cheers,
Longman
next prev parent reply other threads:[~2022-10-20 3:02 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-10-19 23:09 [PATCH clocksource] Reject bogus watchdog clocksource measurements Paul E. McKenney
2022-10-20 3:01 ` Waiman Long [this message]
2022-10-20 8:09 ` Feng Tang
2022-10-20 14:09 ` Paul E. McKenney
2022-10-20 17:29 ` Waiman Long
2022-10-21 0:46 ` Feng Tang
2022-10-28 17:52 ` Paul E. McKenney
2022-10-28 17:52 ` Paul E. McKenney
2022-10-31 5:59 ` Feng Tang
2022-10-31 17:42 ` Paul E. McKenney
2022-11-01 5:43 ` Feng Tang
2022-11-01 19:06 ` Paul E. McKenney
2022-11-02 2:58 ` Feng Tang
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=df0960dc-86d4-984f-8dd8-c9d118aae449@redhat.com \
--to=longman@redhat.com \
--cc=clm@meta.com \
--cc=feng.tang@intel.com \
--cc=jstultz@google.com \
--cc=linux-kernel@vger.kernel.org \
--cc=paulmck@kernel.org \
--cc=sboyd@kernel.org \
--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).