containers.lists.linux.dev archive mirror
 help / color / mirror / Atom feed
* [PATCH 00/11] cpuset: separate configured masks and effective masks
@ 2013-08-21  9:58 Li Zefan
  0 siblings, 0 replies; 2+ messages in thread
From: Li Zefan @ 2013-08-21  9:58 UTC (permalink / raw)
  To: Tejun Heo; +Cc: Cgroups, Containers, LKML

This patcheset introduces behavior changes, but only if you mount cgroupfs
with sane_behavior option:

- We introduce new interfaces cpuset.effective_cpus and cpuset.effective_mems,
  while cpuset.cpus and cpuset.mems will be configured masks.

- The configured masks can be changed by writing cpuset.cpus/mems only. They
  won't be changed when hotplug happens.

- Users can config cpus and mems without restrictions from the parent cpuset.
  effective masks will enforce the hierarchical behavior.

- Users can also config cpus and mems to have already offlined CPU/nodes.

- When a CPU/node is onlined, it will be brought back to the effective masks
  if it's in the configured masks.

- We build sched domains based on effective cpumask but not configured cpumask.

Li Zefan (11):
  cgroup: allow subsystems to create files for sane_behavior only
  cpuset: add cs->real_cpus_allowed and cs->real_mems_allowed
  cpuset: update cpuset->real_{cpus,mems}_allowed at hotplug
  cpuset: update cs->real_{cpus,mems}_allowed when config changes
  cpuset: inherite ancestor's masks if real_{cpus,mems}_allowed become empty
  cpuset: apply cs->real_{cpus,mems}_allowed
  cpuset: use effective cpumask to build sched domains
  cpuset: separate configured masks and efffective masks
  cpuset: enable onlined cpu/node in effective masks
  cpuset: allow writing offlined masks to cpuset.cpus/mems
  cpuset: export effective masks to userspace

 include/linux/cgroup.h |   1 +
 kernel/cgroup.c        |   2 +
 kernel/cpuset.c        | 466 ++++++++++++++++++++++++++++---------------------
 3 files changed, 271 insertions(+), 198 deletions(-)

-- 
1.8.0.2

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [PATCH 00/11] cpuset: separate configured masks and effective masks
       [not found] ` <52148F52.0-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
@ 2013-08-21 14:21   ` Tejun Heo
  0 siblings, 0 replies; 2+ messages in thread
From: Tejun Heo @ 2013-08-21 14:21 UTC (permalink / raw)
  To: Li Zefan; +Cc: Cgroups, Containers, LKML

Hello, Li.

On Wed, Aug 21, 2013 at 05:58:42PM +0800, Li Zefan wrote:
> This patcheset introduces behavior changes, but only if you mount cgroupfs
> with sane_behavior option:
> 
> - We introduce new interfaces cpuset.effective_cpus and cpuset.effective_mems,
>   while cpuset.cpus and cpuset.mems will be configured masks.
> 
> - The configured masks can be changed by writing cpuset.cpus/mems only. They
>   won't be changed when hotplug happens.
> 
> - Users can config cpus and mems without restrictions from the parent cpuset.
>   effective masks will enforce the hierarchical behavior.
> 
> - Users can also config cpus and mems to have already offlined CPU/nodes.
> 
> - When a CPU/node is onlined, it will be brought back to the effective masks
>   if it's in the configured masks.
> 
> - We build sched domains based on effective cpumask but not configured cpumask.

Overall, it looks great.  Thank you so much for doing this.  Most of
my reviews are concerned with patch descriptions and documentation.  I
wish they showed clearer what are changed how and why that doesn't
affect !sane behaviors.

Thanks!

-- 
tejun

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2013-08-21 14:21 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-08-21  9:58 [PATCH 00/11] cpuset: separate configured masks and effective masks Li Zefan
     [not found] <52148F52.0@huawei.com>
     [not found] ` <52148F52.0-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
2013-08-21 14:21   ` Tejun Heo

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).