Re: [PATCH clocksource] Reject bogus watchdog clocksource measurements

["Paul E. McKenney" <paulmck@kernel.org>]

On Fri, Oct 21, 2022 at 08:46:32AM +0800, Feng Tang wrote:
> On Thu, Oct 20, 2022 at 07:09:44AM -0700, Paul E. McKenney wrote:
> [...]
> > > >  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 this happens (500ms timer happens only after less than 125ms),
> > > there is some severe problem with timer/interrupt system. 
> > 
> > Should I add ", suspect timer/interrupt bug" just after "jiffy time
> > interval"?  Or would a comment before that pr_warn() work better for you?
> 
> Both are fine for me.

Here is the patch, which just adds comments.  (The exponential-backoff
patch is on its way.)

Thoughts?

							Thanx, Paul

------------------------------------------------------------------------

commit 59d9db36dc15b3b40a30d7a3d733dbd412c8557a
Author: Paul E. McKenney <paulmck@kernel.org>
Date:   Thu Oct 27 11:58:27 2022 -0700

    clocksource: Add comments to classify bogus measurements
    
    An extremely busy system can delay the clocksource watchdog, so that
    the corresponding too-long bogus-measurement error does not necessarily
    imply an error in the system.  However, a too-short bogus-measurement
    error likely indicates a bug in hardware, firmware or software.
    
    Therefore, add comments clarifying these bogus-measurement pr_warn()s.
    
    Reported-by: Feng Tang <feng.tang@intel.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 dcaf38c062161..3f5317faf891f 100644
--- a/kernel/time/clocksource.c
+++ b/kernel/time/clocksource.c
@@ -443,10 +443,12 @@ static void clocksource_watchdog(struct timer_list *unused)
 		/* Check for bogus measurements. */
 		wdi = jiffies_to_nsecs(WATCHDOG_INTERVAL);
 		if (wd_nsec < (wdi >> 2)) {
+			/* This usually indicates broken timer code or hardware. */
 			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)) {
+			/* This can happen on busy systems, which can delay the watchdog. */
 			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;
 		}