From: Li Zefan <lizefan@huawei.com>
To: Tejun Heo <tj@kernel.org>
Cc: LKML <linux-kernel@vger.kernel.org>,
Cgroups <cgroups@vger.kernel.org>,
Containers <containers@lists.linux-foundation.org>
Subject: [PATCH 08/11] cpuset: separate configured masks and efffective masks
Date: Wed, 21 Aug 2013 18:00:42 +0800 [thread overview]
Message-ID: <52148FCA.8010704@huawei.com> (raw)
In-Reply-To: <52148F52.0@huawei.com>
Now we've used effective cpumasks to enforce hierarchical manner,
we can use cs->{cpus,mems}_allowed as configured masks.
Configured masks can be changed by writing cpuset.cpus and cpuset.mems
only. The new behaviors are:
- They won't be changed by hotplug anymore.
- They won't be limited by its parent's masks.
This behavior change won't take effect unless mount with sane_behavior.
Signed-off-by: Li Zefan <lizefan@huawei.com>
---
kernel/cpuset.c | 33 ++++++++++++++++++++++-----------
1 file changed, 22 insertions(+), 11 deletions(-)
diff --git a/kernel/cpuset.c b/kernel/cpuset.c
index e7ad4a7..c3a02a9 100644
--- a/kernel/cpuset.c
+++ b/kernel/cpuset.c
@@ -457,9 +457,13 @@ static int validate_change(struct cpuset *cur, struct cpuset *trial)
par = parent_cs(cur);
- /* We must be a subset of our parent cpuset */
+ /*
+ * We must be a subset of our parent cpuset, unless sane_behavior
+ * flag is set.
+ */
ret = -EACCES;
- if (!is_cpuset_subset(trial, par))
+ if (!cgroup_sane_behavior(cur->css.cgroup) &&
+ !is_cpuset_subset(trial, par))
goto out;
/*
@@ -780,7 +784,7 @@ static void rebuild_sched_domains_locked(void)
* passing doms with offlined cpu to partition_sched_domains().
* Anyways, hotplug work item will rebuild sched domains.
*/
- if (!cpumask_equal(top_cpuset.cpus_allowed, cpu_active_mask))
+ if (!cpumask_equal(top_cpuset.real_cpus_allowed, cpu_active_mask))
goto out;
/* Generate domain masks and attrs */
@@ -2159,11 +2163,14 @@ retry:
goto retry;
}
- cpumask_andnot(&off_cpus, cs->cpus_allowed, top_cpuset.cpus_allowed);
- nodes_andnot(off_mems, cs->mems_allowed, top_cpuset.mems_allowed);
+ cpumask_andnot(&off_cpus, cs->real_cpus_allowed,
+ top_cpuset.real_cpus_allowed);
+ nodes_andnot(off_mems, cs->real_mems_allowed,
+ top_cpuset.real_mems_allowed);
mutex_lock(&callback_mutex);
- cpumask_andnot(cs->cpus_allowed, cs->cpus_allowed, &off_cpus);
+ if (!sane)
+ cpumask_andnot(cs->cpus_allowed, cs->cpus_allowed, &off_cpus);
cpumask_andnot(cs->real_cpus_allowed, cs->real_cpus_allowed,
&off_cpus);
@@ -2184,7 +2191,8 @@ retry:
update_tasks_cpumask(cs, NULL);
mutex_lock(&callback_mutex);
- nodes_andnot(cs->mems_allowed, cs->mems_allowed, off_mems);
+ if (!sane)
+ nodes_andnot(cs->mems_allowed, cs->mems_allowed, off_mems);
nodes_andnot(cs->real_mems_allowed, cs->real_mems_allowed, off_mems);
/* Inherite the effective mask of the parent, if it becomes empty */
@@ -2239,6 +2247,7 @@ static void cpuset_hotplug_workfn(struct work_struct *work)
static cpumask_t new_cpus;
static nodemask_t new_mems;
bool cpus_updated, mems_updated;
+ bool sane = cgroup_sane_behavior(top_cpuset.css.cgroup);
mutex_lock(&cpuset_mutex);
@@ -2246,13 +2255,14 @@ static void cpuset_hotplug_workfn(struct work_struct *work)
cpumask_copy(&new_cpus, cpu_active_mask);
new_mems = node_states[N_MEMORY];
- cpus_updated = !cpumask_equal(top_cpuset.cpus_allowed, &new_cpus);
- mems_updated = !nodes_equal(top_cpuset.mems_allowed, new_mems);
+ cpus_updated = !cpumask_equal(top_cpuset.real_cpus_allowed, &new_cpus);
+ mems_updated = !nodes_equal(top_cpuset.real_mems_allowed, new_mems);
/* synchronize cpus_allowed to cpu_active_mask */
if (cpus_updated) {
mutex_lock(&callback_mutex);
- cpumask_copy(top_cpuset.cpus_allowed, &new_cpus);
+ if (!sane)
+ cpumask_copy(top_cpuset.cpus_allowed, &new_cpus);
cpumask_copy(top_cpuset.real_cpus_allowed, &new_cpus);
mutex_unlock(&callback_mutex);
/* we don't mess with cpumasks of tasks in top_cpuset */
@@ -2261,7 +2271,8 @@ static void cpuset_hotplug_workfn(struct work_struct *work)
/* synchronize mems_allowed to N_MEMORY */
if (mems_updated) {
mutex_lock(&callback_mutex);
- top_cpuset.mems_allowed = new_mems;
+ if (!sane)
+ top_cpuset.mems_allowed = new_mems;
top_cpuset.real_mems_allowed = new_mems;
mutex_unlock(&callback_mutex);
update_tasks_nodemask(&top_cpuset, NULL);
--
1.8.0.2
next prev parent reply other threads:[~2013-08-21 10:00 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-08-21 9:58 [PATCH 00/11] cpuset: separate configured masks and effective masks Li Zefan
2013-08-21 9:58 ` [PATCH 01/11] cgroup: allow subsystems to create files for sane_behavior only Li Zefan
2013-08-21 9:59 ` [PATCH 03/11] cpuset: update cpuset->real_{cpus,mems}_allowed at hotplug Li Zefan
2013-08-21 9:59 ` [PATCH 06/11] cpuset: apply cs->real_{cpus,mems}_allowed Li Zefan
[not found] ` <52148F9C.2080600-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
2013-08-21 14:01 ` Tejun Heo
2013-08-21 10:00 ` Li Zefan [this message]
[not found] ` <52148FCA.8010704-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
2013-08-21 14:08 ` [PATCH 08/11] cpuset: separate configured masks and efffective masks Tejun Heo
[not found] ` <20130821140846.GH19286-9pTldWuhBndy/B6EtB590w@public.gmane.org>
2013-08-23 7:46 ` Li Zefan
[not found] ` <52171367.90005-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
2013-08-23 15:33 ` Tejun Heo
[not found] ` <52148F52.0-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
2013-08-21 9:59 ` [PATCH 02/11] cpuset: add cs->real_cpus_allowed and cs->real_mems_allowed Li Zefan
[not found] ` <52148F6F.4070507-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
2013-08-21 13:22 ` Tejun Heo
2013-08-21 9:59 ` [PATCH 04/11] cpuset: update cs->real_{cpus,mems}_allowed when config changes Li Zefan
[not found] ` <52148F84.9050309-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
2013-08-21 13:39 ` Tejun Heo
2013-08-21 9:59 ` [PATCH 05/11] cpuset: inherite ancestor's masks if real_{cpus, mems}_allowed become empty Li Zefan
[not found] ` <52148F90.7070809-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
2013-08-21 13:44 ` [PATCH 05/11] cpuset: inherite ancestor's masks if real_{cpus,mems}_allowed " Tejun Heo
2013-08-21 10:00 ` [PATCH 07/11] cpuset: use effective cpumask to build sched domains Li Zefan
[not found] ` <52148FA9.806-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
2013-08-21 14:04 ` Tejun Heo
2013-08-21 10:01 ` [PATCH 09/11] cpuset: enable onlined cpu/node in effective masks Li Zefan
[not found] ` <52148FE1.3080806-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
2013-08-21 14:11 ` Tejun Heo
2013-08-21 10:01 ` [PATCH 10/11] cpuset: allow writing offlined masks to cpuset.cpus/mems Li Zefan
[not found] ` <52148FF1.5060503-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
2013-08-21 14:18 ` Tejun Heo
[not found] ` <20130821141851.GJ19286-9pTldWuhBndy/B6EtB590w@public.gmane.org>
2013-08-23 7:37 ` Li Zefan
2013-08-21 10:01 ` [PATCH 11/11] cpuset: export effective masks to userspace Li Zefan
[not found] ` <52148FFC.4080701-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
2013-08-21 14:20 ` Tejun Heo
[not found] ` <20130821142001.GK19286-9pTldWuhBndy/B6EtB590w@public.gmane.org>
2013-08-23 7:53 ` Li Zefan
[not found] ` <52171501.8050401-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
2013-08-23 12:34 ` Tejun Heo
2013-08-21 14:21 ` [PATCH 00/11] cpuset: separate configured masks and effective masks Tejun Heo
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=52148FCA.8010704@huawei.com \
--to=lizefan@huawei.com \
--cc=cgroups@vger.kernel.org \
--cc=containers@lists.linux-foundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=tj@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox