Re: [PATCH tip/core/rcu 02/15] rcu: Use timer as backstop for NOCB deferred wakeups

2017-07-27 Thread Paul E. McKenney
On Wed, Jul 26, 2017 at 07:09:34PM -0400, Steven Rostedt wrote: > On Wed, 26 Jul 2017 14:47:41 -0700 > "Paul E. McKenney" wrote: > > > > > It is much lighter weight than a timer setup. > > > > How much lighter weight? In other words, what fraction of the > > timers have to avoid being cancel

Re: [PATCH tip/core/rcu 02/15] rcu: Use timer as backstop for NOCB deferred wakeups

2017-07-26 Thread Steven Rostedt
On Wed, 26 Jul 2017 14:47:41 -0700 "Paul E. McKenney" wrote: > > It is much lighter weight than a timer setup. > > How much lighter weight? In other words, what fraction of the > timers have to avoid being cancelled for irq_work to break even? No idea. I guess that would be a nice academic

Re: [PATCH tip/core/rcu 02/15] rcu: Use timer as backstop for NOCB deferred wakeups

2017-07-26 Thread Paul E. McKenney
On Wed, Jul 26, 2017 at 05:18:01PM -0400, Steven Rostedt wrote: > On Tue, 25 Jul 2017 17:05:40 -0700 > "Paul E. McKenney" wrote: > > > On Tue, Jul 25, 2017 at 06:17:10PM -0400, Steven Rostedt wrote: > > > On Tue, 25 Jul 2017 12:18:14 -0700 > > > "Paul E. McKenney" wrote: > > > > > > > On Tue,

Re: [PATCH tip/core/rcu 02/15] rcu: Use timer as backstop for NOCB deferred wakeups

2017-07-26 Thread Steven Rostedt
On Tue, 25 Jul 2017 17:05:40 -0700 "Paul E. McKenney" wrote: > On Tue, Jul 25, 2017 at 06:17:10PM -0400, Steven Rostedt wrote: > > On Tue, 25 Jul 2017 12:18:14 -0700 > > "Paul E. McKenney" wrote: > > > > > On Tue, Jul 25, 2017 at 02:12:20PM -0400, Steven Rostedt wrote: > > > > On Mon, 24 Ju

Re: [PATCH tip/core/rcu 02/15] rcu: Use timer as backstop for NOCB deferred wakeups

2017-07-25 Thread Paul E. McKenney
On Tue, Jul 25, 2017 at 06:17:10PM -0400, Steven Rostedt wrote: > On Tue, 25 Jul 2017 12:18:14 -0700 > "Paul E. McKenney" wrote: > > > On Tue, Jul 25, 2017 at 02:12:20PM -0400, Steven Rostedt wrote: > > > On Mon, 24 Jul 2017 14:44:31 -0700 > > > "Paul E. McKenney" wrote: > > > > > > > The han

Re: [PATCH tip/core/rcu 02/15] rcu: Use timer as backstop for NOCB deferred wakeups

2017-07-25 Thread Steven Rostedt
On Tue, 25 Jul 2017 12:18:14 -0700 "Paul E. McKenney" wrote: > On Tue, Jul 25, 2017 at 02:12:20PM -0400, Steven Rostedt wrote: > > On Mon, 24 Jul 2017 14:44:31 -0700 > > "Paul E. McKenney" wrote: > > > > > The handling of RCU's no-CBs CPUs has a maintenance headache, namely > > > that if call

Re: [PATCH tip/core/rcu 02/15] rcu: Use timer as backstop for NOCB deferred wakeups

2017-07-25 Thread Paul E. McKenney
On Tue, Jul 25, 2017 at 02:12:20PM -0400, Steven Rostedt wrote: > On Mon, 24 Jul 2017 14:44:31 -0700 > "Paul E. McKenney" wrote: > > > The handling of RCU's no-CBs CPUs has a maintenance headache, namely > > that if call_rcu() is invoked with interrupts disabled, the rcuo kthread > > wakeup must

Re: [PATCH tip/core/rcu 02/15] rcu: Use timer as backstop for NOCB deferred wakeups

2017-07-25 Thread Steven Rostedt
On Mon, 24 Jul 2017 14:44:31 -0700 "Paul E. McKenney" wrote: > The handling of RCU's no-CBs CPUs has a maintenance headache, namely > that if call_rcu() is invoked with interrupts disabled, the rcuo kthread > wakeup must be defered to a point where we can be sure that scheduler > locks are not he

[PATCH tip/core/rcu 02/15] rcu: Use timer as backstop for NOCB deferred wakeups

2017-07-24 Thread Paul E. McKenney
The handling of RCU's no-CBs CPUs has a maintenance headache, namely that if call_rcu() is invoked with interrupts disabled, the rcuo kthread wakeup must be defered to a point where we can be sure that scheduler locks are not held. Of course, there are a lot of code paths leading from an interrupt