Re: [PATCH] rcu/tree: consider time a VM was suspended

[Paul E. McKenney]

On Fri, May 21, 2021 at 07:34:41AM +0900, Sergey Senozhatsky wrote:
> On (21/05/20 07:57), Paul E. McKenney wrote:
> > > 
> > > Sounds good. I can cook a patch and run some tests.
> > > Or do you want to send a patch?
> > 
> > Given that you have the test setup, things might go faster if you do
> > the patch, especially taking timezones into consideration.  Of course,
> > if you run into difficulties, you know where to find me.
> 
> OK. Sounds good to me.
> 
> > > While VCPU-2 has PVCLOCK_GUEST_STOPPED set (resuming) and is in
> > > check_cpu_stall(), the VCPU-3 is executing:
> > > 
> > > 	apic_timer_interrupt()
> > > 	 tick_irq_enter()
> > > 	  tick_do_update_jiffies64()
> > > 	   do_timer()
> > 
> > OK, but the normal grace period time is way less than one second, and
> > the stall timeout in mainline is 21 seconds, so that would be a -lot-
> > of jiffies of skew.  Or does the restarting really take that long a time?
> 
> That's a good question. I see huge jiffies spike in the logs.
> I suspect that resuming a VM can take some time, especially on a "not
> powerful at all" overcommitted host (more virtual CPUs than physical
> ones).

I really am just asking the question.  ;-)

After all, if restarting a VM can take that long, then it can take
that long.

							Thanx, Paul