Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1629134 > unrolled thread

TREE_SRCU slows hotplug by factor ~16

Started byMike Galbraith <efault@gmx.de>
First post2017-04-24 05:00 +0200
Last post2017-04-26 20:00 +0200
Articles 4 on this page of 24 — 2 participants

Back to article view | Back to linux.kernel

Contents

  TREE_SRCU slows hotplug by factor ~16 Mike Galbraith <efault@gmx.de> - 2017-04-24 05:00 +0200
    Re: TREE_SRCU slows hotplug by factor ~16 "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2017-04-24 05:40 +0200
      Re: TREE_SRCU slows hotplug by factor ~16 Mike Galbraith <efault@gmx.de> - 2017-04-24 07:30 +0200
        Re: TREE_SRCU slows hotplug by factor ~16 "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2017-04-24 08:30 +0200
          Re: TREE_SRCU slows hotplug by factor ~16 Mike Galbraith <efault@gmx.de> - 2017-04-24 09:40 +0200
            Re: TREE_SRCU slows hotplug by factor ~16 Mike Galbraith <efault@gmx.de> - 2017-04-24 10:50 +0200
            Re: TREE_SRCU slows hotplug by factor ~16 "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2017-04-24 18:30 +0200
              Re: TREE_SRCU slows hotplug by factor ~16 "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2017-04-26 00:40 +0200
                Re: TREE_SRCU slows hotplug by factor ~16 "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2017-04-26 16:40 +0200
                  Re: TREE_SRCU slows hotplug by factor ~16 Mike Galbraith <efault@gmx.de> - 2017-04-26 17:30 +0200
                    Re: TREE_SRCU slows hotplug by factor ~16 "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2017-04-26 17:50 +0200
                      Re: TREE_SRCU slows hotplug by factor ~16 Mike Galbraith <efault@gmx.de> - 2017-04-26 18:00 +0200
                        Re: TREE_SRCU slows hotplug by factor ~16 "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2017-04-26 18:10 +0200
                        Re: TREE_SRCU slows hotplug by factor ~16 Mike Galbraith <efault@gmx.de> - 2017-04-26 19:50 +0200
                          Re: TREE_SRCU slows hotplug by factor ~16 "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2017-04-26 20:00 +0200
                            Re: TREE_SRCU slows hotplug by factor ~16 Mike Galbraith <efault@gmx.de> - 2017-04-26 20:20 +0200
                              Re: TREE_SRCU slows hotplug by factor ~16 "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2017-04-26 20:30 +0200
                              Re: TREE_SRCU slows hotplug by factor ~16 Mike Galbraith <efault@gmx.de> - 2017-04-27 05:50 +0200
                                Re: TREE_SRCU slows hotplug by factor ~16 "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2017-04-27 06:20 +0200
                                  Re: TREE_SRCU slows hotplug by factor ~16 Mike Galbraith <efault@gmx.de> - 2017-04-27 06:20 +0200
                                    Re: TREE_SRCU slows hotplug by factor ~16 "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2017-04-27 07:40 +0200
                                      Re: TREE_SRCU slows hotplug by factor ~16 Mike Galbraith <efault@gmx.de> - 2017-04-27 07:50 +0200
                                        Re: TREE_SRCU slows hotplug by factor ~16 "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2017-04-27 14:40 +0200
                          Re: TREE_SRCU slows hotplug by factor ~16 "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2017-04-26 20:00 +0200

Page 2 of 2 — ← Prev page 1 [2]

#1631869

From"Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Date2017-04-27 07:40 +0200
Message-ID<tAKjL-4vc-7@gated-at.bofh.it>
In reply to#1631853
On Thu, Apr 27, 2017 at 06:15:56AM +0200, Mike Galbraith wrote:
> On Wed, 2017-04-26 at 21:11 -0700, Paul E. McKenney wrote:
> 
> > This is with srcutree.exp_holdoff set to 25*1000?
> 
> Yup.

And please see below for the trivial patch, just for confirmation.
May I add your Tested-by?

							Thanx, Paul

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

diff --git a/kernel/rcu/srcutree.c b/kernel/rcu/srcutree.c
index ccbe1fb84165..65f03bca5b32 100644
--- a/kernel/rcu/srcutree.c
+++ b/kernel/rcu/srcutree.c
@@ -39,7 +39,7 @@
 
 #include "rcu.h"
 
-ulong exp_holdoff = 50 * 1000; /* Holdoff (ns) for auto-expediting. */
+ulong exp_holdoff = 25 * 1000; /* Holdoff (ns) for auto-expediting. */
 module_param(exp_holdoff, ulong, 0444);
 
 static void srcu_invoke_callbacks(struct work_struct *work);

[toc] | [prev] | [next] | [standalone]

#1631874

FromMike Galbraith <efault@gmx.de>
Date2017-04-27 07:50 +0200
Message-ID<tAKts-4z4-13@gated-at.bofh.it>
In reply to#1631869
On Wed, 2017-04-26 at 22:32 -0700, Paul E. McKenney wrote:
> On Thu, Apr 27, 2017 at 06:15:56AM +0200, Mike Galbraith wrote:
> > On Wed, 2017-04-26 at 21:11 -0700, Paul E. McKenney wrote:
> > 
> > > This is with srcutree.exp_holdoff set to 25*1000?
> > 
> > Yup.
> 
> And please see below for the trivial patch, just for confirmation.
> May I add your Tested-by?

Sure.

> diff --git a/kernel/rcu/srcutree.c b/kernel/rcu/srcutree.c
> index ccbe1fb84165..65f03bca5b32 100644
> --- a/kernel/rcu/srcutree.c
> +++ b/kernel/rcu/srcutree.c
> @@ -39,7 +39,7 @@
>  
>  #include "rcu.h"
>  
> -ulong exp_holdoff = 50 * 1000; /* Holdoff (ns) for auto-expediting. */
> +ulong exp_holdoff = 25 * 1000; /* Holdoff (ns) for auto-expediting. */

Yup, exactly what I did.

[toc] | [prev] | [next] | [standalone]

#1632074

From"Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Date2017-04-27 14:40 +0200
Message-ID<tAQSe-Bq-9@gated-at.bofh.it>
In reply to#1631874
On Thu, Apr 27, 2017 at 07:44:00AM +0200, Mike Galbraith wrote:
> On Wed, 2017-04-26 at 22:32 -0700, Paul E. McKenney wrote:
> > On Thu, Apr 27, 2017 at 06:15:56AM +0200, Mike Galbraith wrote:
> > > On Wed, 2017-04-26 at 21:11 -0700, Paul E. McKenney wrote:
> > > 
> > > > This is with srcutree.exp_holdoff set to 25*1000?
> > > 
> > > Yup.
> > 
> > And please see below for the trivial patch, just for confirmation.
> > May I add your Tested-by?
> 
> Sure.
> 
> > diff --git a/kernel/rcu/srcutree.c b/kernel/rcu/srcutree.c
> > index ccbe1fb84165..65f03bca5b32 100644
> > --- a/kernel/rcu/srcutree.c
> > +++ b/kernel/rcu/srcutree.c
> > @@ -39,7 +39,7 @@
> >  
> >  #include "rcu.h"
> >  
> > -ulong exp_holdoff = 50 * 1000; /* Holdoff (ns) for auto-expediting. */
> > +ulong exp_holdoff = 25 * 1000; /* Holdoff (ns) for auto-expediting. */
> 
> Yup, exactly what I did.

Woo-hoo!!!

I will beat up on this a bit and send a pull request to Ingo.

And thank you very much for finding this, gathering the trace data
(which was critically important to my figuring out what was going
on), and for testing the patches!

							Thanx, Paul

[toc] | [prev] | [next] | [standalone]

#1631630

From"Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Date2017-04-26 20:00 +0200
Message-ID<tAzol-5vz-5@gated-at.bofh.it>
In reply to#1631625
On Wed, Apr 26, 2017 at 07:45:57PM +0200, Mike Galbraith wrote:
> On Wed, 2017-04-26 at 17:49 +0200, Mike Galbraith wrote:
> > On Wed, 2017-04-26 at 08:44 -0700, Paul E. McKenney wrote:
> 
> > > Should I be comparing this with the 55s number from your initial email,
> > > or to the 39s number?
> > 
> > Should be the 39...
> 
> And 39 it is.

Thank you again for your testing efforts!

OK, I do need to do more work.  My current guess is that I should have
set the default for srcutree.exp_holdoff to 25*1000 instead of 50*1000.
But I am sure that further data will show me the error of my ways.  ;-)

							Thanx, Paul

[toc] | [prev] | [standalone]

Page 2 of 2 — ← Prev page 1 [2]

Back to top | Article view | linux.kernel

csiph-web