Date: Wed, 1 Jul 2015 14:49:14 -0700
From: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
To: josh@joshtriplett.org
Cc: Peter Zijlstra <peterz@infradead.org>, linux-kernel@vger.kernel.org,
        mingo@kernel.org, laijs@cn.fujitsu.com, dipankar@in.ibm.com,
        akpm@linux-foundation.org, mathieu.desnoyers@efficios.com,
        tglx@linutronix.de, rostedt@goodmis.org, dhowells@redhat.com,
        edumazet@google.com, dvhart@linux.intel.com, fweisbec@gmail.com,
        oleg@redhat.com, bobby.prani@gmail.com
Subject: Re: [PATCH RFC tip/core/rcu 0/5] Expedited grace periods encouraging
 normal ones
Message-ID: <20150701214914.GV3717@linux.vnet.ibm.com>
Reply-To: paulmck@linux.vnet.ibm.com
References: <20150630221224.GQ3717@linux.vnet.ibm.com>
 <20150630234633.GA11450@cloud>
 <20150701100939.GR19282@twins.programming.kicks-ass.net>
 <20150701105511.GN18673@twins.programming.kicks-ass.net>
 <20150701140031.GB3717@linux.vnet.ibm.com>
 <20150701141710.GG25159@twins.programming.kicks-ass.net>
 <20150701161705.GK3717@linux.vnet.ibm.com>
 <20150701170242.GL3644@twins.programming.kicks-ass.net>
 <20150701200936.GP3717@linux.vnet.ibm.com>
 <20150701212001.GA18023@cloud>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20150701212001.GA18023@cloud>
User-Agent: Mutt/1.5.21 (2010-09-15)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 2380
Lines: 48

On Wed, Jul 01, 2015 at 02:20:01PM -0700, josh@joshtriplett.org wrote:
> On Wed, Jul 01, 2015 at 01:09:36PM -0700, Paul E. McKenney wrote:
> > On Wed, Jul 01, 2015 at 07:02:42PM +0200, Peter Zijlstra wrote:
> > > USB sure, but a backing dev is involved in nfs clients, loopback and all
> > > sorts of block/filesystem like setups.
> > > 
> > > unmount an NFS mount and voila expedited rcu, unmount a loopback, tada.
> > > 
> > > All you need is a regular server workload triggering any of that on a
> > > semi regular basis and even !rt people might start to notice something
> > > is up.
> > 
> > I don't believe that latency-sensitive systems are going to be messing
> > with remapping their storage at runtime, let alone on a regular basis.
> > If they are not latency sensitive, and assuming that the rate of
> > storage remapping is at all sane, I bet that they won't notice the
> > synchronize_rcu_expedited() overhead.  The overhead of the actual
> > remapping will very likely leave the synchronize_rcu_expedited() overhead
> > way down in the noise.
> > 
> > And if they are doing completely insane rates of storage remapping,
> > I suspect that the batching in the synchronize_rcu_expedited()
> > implementation will reduce the expedited-grace-period overhead still
> > further as a fraction of the total.
> 
> Consider the case of container-based systems, calling mount as part of
> container setup and umount as part of container teardown.
> 
> And those workloads are often sensitive to latency, not throughput.

So people are really seeing a synchronize_rcu_expedited() on each
container setup/teardown right now?  Or is this something that could
happen if they were mounting block devices rather than rebind mounts?

And when you say that these workloads are sensitive to latency, I am
guessing that you mean to the millisecond-level latencies seen from
synchronize_rcu() as opposed to the microsecond-level OS jitter from
synchronize_rcu_expedited().  Or are there really containers workloads
that care about the few microseconds of OS jitter that would be incurred
due to expedited grace periods?

							Thanx, Paul

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/