lkml.org 
[lkml]   [2016]   [Jul]   [25]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH] rcu_sync: simplify the state machine, introduce __rcu_sync_enter()
On Mon, Jul 25, 2016 at 07:01:17PM +0200, Oleg Nesterov wrote:
> Paul, sorry for delay.
>
> On 07/21, Paul E. McKenney wrote:
> >
> > On Thu, Jul 21, 2016 at 07:34:36PM +0200, Oleg Nesterov wrote:
> > > On 07/20, Paul E. McKenney wrote:
> > > >
> > > > On Wed, Jul 20, 2016 at 07:16:03PM +0200, Oleg Nesterov wrote:
> > > >
> > > > > Now, suppose we add the additional enter/exit's:
> > > > >
> > > > > freeze_super(sb)
> > > > > {
> > > > > // this doesn't block
> > > > > __rcu_sync_enter(SEM3);
> > > > > __rcu_sync_enter(SEM2);
> > > > > __rcu_sync_enter(SEM1);
> > > > >
> > > > > down_write(&sb->s_umount);
> > > > > if (NEED_TO_FREEZE) {
> > > > > percpu_down_write(SEM1);
> > > >
> > > > The above waits for the grace period initiated by __rcu_sync_enter(),
> > > > correct? Presumably "yes", because it will invoke rcu_sync_enter(), which
> > > > will see the state as GP_ENTER, and will thus wait.
> > >
> > > But if down_write() blocks and/or NEED_TO_FREEZE takes some time it
> > > could already see the GP_PASSED state, or at least it can sleep less.
> > >
> > > > But your point is that if !NEED_TO_FREEZE, we will get here without
> > > > waiting for a grace period.
> > > >
> > > > But why aren't the __rcu_sync_enter() and rcu_sync_exit() calls inside
> > > > the "if" statement?
> > >
> > > Yes, if we do __rcu_sync_enter() inside "if", then rcu_sync_exit() can't
> > > hit GP_ENTER.
> > >
> > > But why we should disallow this use-case? It does not complicate the code
> > > at all.
> >
> > I do agree that it doesn't complicate the current implementation.
> > But it relies on a global lock, so I am not at all confident that this
> > implementation is the final word.
>
> Hmm. which global lock? Or did you mean freeze_super(), not rcu_sync?

OK, you are right, they are per-superblock locks rather than being global
locks. Still, given that workloads that hammer a single filesystem hard
are quite common, it might still eventually be of some concern.

> > And speaking of global locks, failing to discourage the pattern above
> > means that the code is unnecessarily acquiring three global locks,
> > which doesn't seem like a good thing to me.
>
> Well, I do not agree, but this wasn't written by me. Just in case, all these
> locks above are not really global, they are per-sb, but this is minor.

Agreed.

> And the patches which changed sb->s_writers to use percpu_rw_semaphore/rcu_sync
> didn't change this logic.
>
> Except the old implementation was buggy, and the readers were slower than now.
>
> > I agree that there are use cases for beginning-of-time __rcu_sync_enter()
> > or whatever we end up naming it.
>
> OK, at least iiuc you agree that cgroup_init() can use __rcu_sync_enter().
> As for other potential use-cases, we will disccuss this later. I will have
> to CC you anyway ;)
>
> So I'll send v2 with renames after I test it. Thanks again.

Sounds good!

Thanx, Paul

\
 
 \ /
  Last update: 2016-07-25 20:21    [W:0.559 / U:0.088 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site