Message-Id: <20181121024107.GI4170@linux.ibm.com>
Date:   Tue, 20 Nov 2018 18:41:07 -0800
From:   "Paul E. McKenney" <paulmck@...ux.ibm.com>
To:     Joel Fernandes <joel@...lfernandes.org>
Cc:     linux-kernel@...r.kernel.org, josh@...htriplett.org,
        rostedt@...dmis.org, mathieu.desnoyers@...icios.com,
        jiangshanlai@...il.com
Subject: Re: dyntick-idle CPU and node's qsmask

On Tue, Nov 20, 2018 at 06:06:12PM -0800, Joel Fernandes wrote:
> On Tue, Nov 20, 2018 at 02:28:14PM -0800, Paul E. McKenney wrote:
> > On Tue, Nov 20, 2018 at 12:42:43PM -0800, Joel Fernandes wrote:
> > > On Sun, Nov 11, 2018 at 10:36:18AM -0800, Paul E. McKenney wrote:
> > > > On Sun, Nov 11, 2018 at 10:09:16AM -0800, Joel Fernandes wrote:
> > > > > On Sat, Nov 10, 2018 at 08:22:10PM -0800, Paul E. McKenney wrote:
> > > > > > On Sat, Nov 10, 2018 at 07:09:25PM -0800, Joel Fernandes wrote:
> > > > > > > On Sat, Nov 10, 2018 at 03:04:36PM -0800, Paul E. McKenney wrote:
> > > > > > > > On Sat, Nov 10, 2018 at 01:46:59PM -0800, Joel Fernandes wrote:
> > > > > > > > > Hi Paul and everyone,
> > > > > > > > > 
> > > > > > > > > I was tracing/studying the RCU code today in paul/dev branch and noticed that
> > > > > > > > > for dyntick-idle CPUs, the RCU GP thread is clearing the rnp->qsmask
> > > > > > > > > corresponding to the leaf node for the idle CPU, and reporting a QS on their
> > > > > > > > > behalf.
> > > > > > > > > 
> > > > > > > > > rcu_sched-10    [003]    40.008039: rcu_fqs:              rcu_sched 792 0 dti
> > > > > > > > > rcu_sched-10    [003]    40.008039: rcu_fqs:              rcu_sched 801 2 dti
> > > > > > > > > rcu_sched-10    [003]    40.008041: rcu_quiescent_state_report: rcu_sched 805 5>0 0 0 3 0
> > > > > > > > > 
> > > > > > > > > That's all good but I was wondering if we can do better for the idle CPUs if
> > > > > > > > > we can some how not set the qsmask of the node in the first place. Then no
> > > > > > > > > reporting would be needed of quiescent state is needed for idle CPUs right?
> > > > > > > > > And we would also not need to acquire the rnp lock I think.
> > > > > > > > > 
> > > > > > > > > At least for a single node tree RCU system, it seems that would avoid needing
> > > > > > > > > to acquire the lock without complications. Anyway let me know your thoughts
> > > > > > > > > and happy to discuss this at the hallways of the LPC as well for folks
> > > > > > > > > attending :)
> > > > > > > > 
> > > > > > > > We could, but that would require consulting the rcu_data structure for
> > > > > > > > each CPU while initializing the grace period, thus increasing the number
> > > > > > > > of cache misses during grace-period initialization and also shortly after
> > > > > > > > for any non-idle CPUs.  This seems backwards on busy systems where each
> > > > > > > 
> > > > > > > When I traced, it appears to me that rcu_data structure of a remote CPU was
> > > > > > > being consulted anyway by the rcu_sched thread. So it seems like such cache
> > > > > > > miss would happen anyway whether it is during grace-period initialization or
> > > > > > > during the fqs stage? I guess I'm trying to say, the consultation of remote
> > > > > > > CPU's rcu_data happens anyway.
> > > > > > 
> > > > > > Hmmm...
> > > > > > 
> > > > > > The rcu_gp_init() function does access an rcu_data structure, but it is
> > > > > > that of the current CPU, so shouldn't involve a communications cache miss,
> > > > > > at least not in the common case.
> > > > > > 
> > > > > > Or are you seeing these cross-CPU rcu_data accesses in rcu_gp_fqs() or
> > > > > > functions that it calls?  In that case, please see below.
> > > > > 
> > > > > Yes, it was rcu_implicit_dynticks_qs called from rcu_gp_fqs.
> > > > > 
> > > > > > > > CPU will with high probability report its own quiescent state before three
> > > > > > > > jiffies pass, in which case the cache misses on the rcu_data structures
> > > > > > > > would be wasted motion.
> > > > > > > 
> > > > > > > If all the CPUs are busy and reporting their QS themselves, then I think the
> > > > > > > qsmask is likely 0 so then rcu_implicit_dynticks_qs (called from
> > > > > > > force_qs_rnp) wouldn't be called and so there would no cache misses on
> > > > > > > rcu_data right?
> > > > > > 
> > > > > > Yes, but assuming that all CPUs report their quiescent states before
> > > > > > the first call to rcu_gp_fqs().  One exception is when some CPU is
> > > > > > looping in the kernel for many milliseconds without passing through a
> > > > > > quiescent state.  This is because for recent kernels, cond_resched()
> > > > > > is not a quiescent state until the grace period is something like 100
> > > > > > milliseconds old.  (For older kernels, cond_resched() was never an RCU
> > > > > > quiescent state unless it actually scheduled.)
> > > > > > 
> > > > > > Why wait 100 milliseconds?  Because otherwise the increase in
> > > > > > cond_resched() overhead shows up all too well, causing 0day test robot
> > > > > > to complain bitterly.  Besides, I would expect that in the common case,
> > > > > > CPUs would be executing usermode code.
> > > > > 
> > > > > Makes sense. I was also wondering about this other thing you mentioned about
> > > > > waiting for 3 jiffies before reporting the idle CPU's quiescent state. Does
> > > > > that mean that even if a single CPU is dyntick-idle for a long period of
> > > > > time, then the minimum grace period duration would be atleast 3 jiffies? In
> > > > > our mobile embedded devices, jiffies is set to 3.33ms (HZ=300) to keep power
> > > > > consumption low. Not that I'm saying its an issue or anything (since IIUC if
> > > > > someone wants shorter grace periods, they should just use expedited GPs), but
> > > > > it sounds like it would be shorter GP if we just set the qsmask early on some
> > > > > how and we can manage the overhead of doing so.
> > > > 
> > > > First, there is some autotuning of the delay based on HZ:
> > > > 
> > > > #define RCU_JIFFIES_TILL_FORCE_QS (1 + (HZ > 250) + (HZ > 500))
> > > > 
> > > > So at HZ=300, you should be seeing a two-jiffy delay rather than the
> > > > usual HZ=1000 three-jiffy delay.  Of course, this means that the delay
> > > > is 6.67ms rather than the usual 3ms, but the theory is that lower HZ
> > > > rates often mean slower instruction execution and thus a desire for
> > > > lower RCU overhead.  There is further autotuning based on number of
> > > > CPUs, but this does not kick in until you have 256 CPUs on your system,
> > > > and I bet that smartphones aren't there yet.  Nevertheless, check out
> > > > RCU_JIFFIES_FQS_DIV for more info on this.
> > > > 
> > > > But you can always override this autotuning using the following kernel
> > > > boot paramters:
> > > > 
> > > > rcutree.jiffies_till_first_fqs
> > > > rcutree.jiffies_till_next_fqs
> > > 
> > > Slightly related, I was just going through your patch in the dev branch "doc:
> > > Now jiffies_till_sched_qs solicits from cond_resched()".
> > > 
> > > If I understand correctly, what you're trying to do is set
> > > rcu_data.rcu_urgent_qs if you've not heard from the CPU long enough from
> > > rcu_implicit_dynticks_qs.
> > > 
> > > Then in the other paths, you are reading this value and similuating a dyntick
> > > idle transition even though you may not be really going into dyntick-idle.
> > > Actually in the scheduler-tick, you are also using it to set NEED_RESCHED
> > > appropriately.
> > > 
> > > Did I get it right so far?
> > 
> > Partially.
> > 
> > The simulated dyntick-idle transition happens if the grace period extends
> > for even longer, so that ->rcu_need_heavy_qs gets set.  Up to that point,
> > all that is asked for is a local-to-the-CPU report of a quiescent state.
> 
> Right, that's true. My feeling was the whole "fake a dyntick idle transition"
> seems to me a bit of a hack and I was thinking if not depending on that would
> simplify the code so we don't need the rcu_momentary_dyntick_idle.

It is there in case scheduling-clock interrupts are for whatever reason
not happening on that CPU.

> > > I was thinking if we could simplify rcu_note_context_switch (the parts that
> > > call rcu_momentary_dyntick_idle), if we did the following in
> > > rcu_implicit_dynticks_qs.
> > > 
> > > Since we already call rcu_qs in rcu_note_context_switch, that would clear the
> > > rdp->cpu_no_qs flag. Then there should be no need to call
> > > rcu_momentary_dyntick_idle from rcu_note_context switch.
> > 
> > But does this also work for the rcu_all_qs() code path?
> 
> Could we not do something like this in rcu_all_qs? as some over-simplified
> pseudo code:
> 
> rcu_all_qs() {
>   if (!urgent_qs || !heavy_qs)
>      return;
> 
>   rcu_qs();   // This clears the rdp->cpu_no_qs flags which we can monitor in
>               //  the diff in my last email (from rcu_implicit_dynticks_qs)
> }

Except that rcu_qs() doesn't necessarily report the quiescent state to
the RCU core.  Keeping down context-switch overhead and all that.

> > > I think this would simplify cond_resched as well.  Could this avoid the need
> > > for having an rcu_all_qs at all? Hopefully I didn't some Tasks-RCU corner cases..
> > 
> > There is also the code path from cond_resched() in PREEMPT=n kernels.
> > This needs rcu_all_qs().  Though it is quite possible that some additional
> > code collapsing is possible.
> > 
> > > Basically for some background, I was thinking can we simplify the code that
> > > calls "rcu_momentary_dyntick_idle" since we already register a qs in other
> > > ways (like by resetting cpu_no_qs).
> > 
> > One complication is that rcu_all_qs() is invoked with interrupts
> > and preemption enabled, while rcu_note_context_switch() is
> > invoked with interrupts disabled.  Also, as you say, Tasks RCU.
> > Plus rcu_all_qs() wants to exit immediately if there is nothing to
> > do, while rcu_note_context_switch() must unconditionally do rcu_qs()
> > -- yes, it could check, but that would be redundant with the checks
> 
> This immediate exit is taken care off in the above psuedo code, would that
> help the cond_resched performance?

It look like you are cautiously edging towards the two wrapper functions
calling common code, relying on inlining and simplification.  Why not just
try doing it?  ;-)

> > > diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c
> > > index c818e0c91a81..5aa0259c014d 100644
> > > --- a/kernel/rcu/tree.c
> > > +++ b/kernel/rcu/tree.c
> > > @@ -1063,7 +1063,7 @@ static int rcu_implicit_dynticks_qs(struct rcu_data *rdp)
> > >  	 * read-side critical section that started before the beginning
> > >  	 * of the current RCU grace period.
> > >  	 */
> > > -	if (rcu_dynticks_in_eqs_since(rdp, rdp->dynticks_snap)) {
> > > +	if (rcu_dynticks_in_eqs_since(rdp, rdp->dynticks_snap) || !rdp->cpu_no_qs.b.norm) {
> > 
> > If I am not too confused, this change could cause trouble for
> > nohz_full CPUs looping in the kernel.  Such CPUs don't necessarily take
> > scheduler-clock interrupts, last I checked, and this could prevent the
> > CPU from reporting its quiescent state to core RCU.
> 
> Would that still be a problem if rcu_all_qs called rcu_qs? Also the above
> diff is an OR condition so it is more relaxed than before.

Yes, because rcu_qs() is only guaranteed to capture the quiescent
state on the current CPU, not necessarily report it to the RCU core.

> Assuming the NOHZ_FULL CPUs call cond_resched during their looping, that
> would trigger the rcu_all_qs -> rcu_qs path which would clear cpu_no_qs flag
> for that CPU right? This would result in the above diff causing a return of 1
> for that CPU (from rcu_implicit_dynticks_qs).

One problem is that cond_resched() is a complete no-op in PREEMPT=y
kernels.

> > Or am I missing something here?
> 
> I think I might be the one missing something but I'm glad we are almost on
> the same page ;-)

There are a lot of moving parts in this area, to be sure.

							Thanx, Paul

Powered by blists - more mailing lists