lkml.org 
[lkml]   [2013]   [Oct]   [10]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH 0/6] Optimize the cpu hotplug locking -v2
On Thu, 10 Oct 2013 18:53:37 +0200
Peter Zijlstra <peterz@infradead.org> wrote:

> On Thu, Oct 10, 2013 at 09:43:55AM -0700, Andrew Morton wrote:
> > > But we would like to remove stomp machine from
> > > CPU hotplug.
> >
> > We do? That's news. It wasn't mentioned in the changelog and should
> > have been. Why?
>
> It would be an unrelated change to this and unrelated to the reasons as
> to why I want a faster get_online_cpus().

Yeah, sorry for the confusion. My comment wasn't really about this
change set but about stop machine and hotplug in general. Needing stop
machine for hotplug has been a complaint by many, but off topic for
this particular change set.

>
> > > Why prevent all CPUs from running when we want to remove
> > > one?
> >
> > So get_online_cpus() goes away. Nothing is more scalable than nothing!
>
> Very much agreed; now stop_machine() wouldn't actually work for hotplug
> because it will instantly preempt everybody, including someone who might
> be in the middle of using per-cpu state of the cpu we're about to
> remove.

Well, stop machine doesn't instantly preempt everybody. Only those that
don't have preemption disabled. Using per_cpu without preemption
disabled can be dangerous. Except for the migrate disable we want to
add for -rt. Then we can't rely on migrate disable and stop machine
making sure per_cpu data isn't being used. Oh, and for threads that use
per_cpu that are bound to a CPU. But then they need to be taken care of
too for their CPU going off line.

But this would also require the "get_online_cpu()s" to disable
preemption as well. Not quite a "nothing" we are looking for.


-- Steve


\
 
 \ /
  Last update: 2013-10-10 19:21    [W:0.061 / U:0.112 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site