Re: RCU Kconfig help text

[Paul E. McKenney]

On Sun, Sep 06, 2009 at 08:24:41AM +0200, Pavel Machek wrote:
> On Sat 2009-09-05 13:01:46, Paul E. McKenney wrote:
> > On Sat, Sep 05, 2009 at 09:27:10AM -0400, Valdis.Kletnieks@xxxxxx wrote:
> > > init/Kconfig says:
> > > 
> > > "choice
> > >         prompt "RCU Implementation"
> > >         default TREE_RCU
> > > 
> > > config TREE_RCU
> > >         bool "Tree-based hierarchical RCU"
> > >         help
> > >           This option selects the RCU implementation that is
> > >           designed for very large SMP system with hundreds or
> > >           thousands of CPUs.  It also scales down nicely to
> > >           smaller systems.
> > > 
> > > config TREE_PREEMPT_RCU
> > >         bool "Preemptable tree-based hierarchical RCU"
> > >         depends on PREEMPT
> > >         help
> > >           This option selects the RCU implementation that is
> > >           designed for very large SMP systems with hundreds or
> > >           thousands of CPUs, but for which real-time response
> > >           is also required.
> > > 
> > > endchoice"
> > > 
> > > This leaves somebody who has a laptop wondering which choice is best for
> > > a system with only one or two cores that has CONFIG_PREEMPT defined. One
> > > choice says it scales down nicely, the other explicitly has a 'depends on
> > > PREEMPT' attached to it...
> > > 
> > > (Yes, I realize in practice, the RCU sections on a laptop are probably usually
> > > so short they don't matter in practice.  I finally concluded TREE_PREEMPT was
> > > apparently a rename of CONFIG_PREEMPT_RCU and went with that since that was
> > > working for me before...)
> > 
> > Good point -- I will add the "It also scales down nicely to smaller
> > systems" to TREE_PREEMPT_RCU.
> > 
> > For -really- small systems, TINY_RCU will hopefully be there at some
> > point, but it can only handle single-CPU systems.
> 
> If so, call it 'UP_RCU' or 'UNIPROCESSOR_RCU'?

"UP_RCU".  Cute!!!  ;-)

My hope is to drive the RCU selection directly from CONFIG_SMP and
CONFIG_PREEMPT, so that the choice of RCU would be implicit (aside from
things like tracing and fanout).

Seem reasonable, or would you expect anyone to want to hand-select
the RCU implementation?

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