[PATCH 1/2 v2] cgroup: Remove unnecessary task_lock before fetching css_set on migration

[PATCH 1/2 v2] cgroup: Remove unnecessary task_lock before fetching css_set on migration

[Frederic Weisbecker]

When we fetch the css_set of the tasks on cgroup migration, we don't need
anymore to synchronize against cgroup_exit() that could swap the old one
with init_css_set. Now that we are using threadgroup_lock() during
the migrations, we don't need to worry about it anymore.

Signed-off-by: Frederic Weisbecker <fweisbec-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Reviewed-by: Mandeep Singh Baines <msb-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org>
Reviewed-by: Li Zefan <lizf-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org>
Cc: Tejun Heo <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
Cc: Containers <containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org>
Cc: Cgroups <cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu-+CUm20s59erQFUHtdCDX3A@public.gmane.org>
Cc: Oleg Nesterov <oleg-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
Cc: Andrew Morton <akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org>
Cc: Paul Menage <paul-inf54ven1CmVyaH7bEyXVA@public.gmane.org>
---
 kernel/cgroup.c |   20 ++++++++++----------
 1 files changed, 10 insertions(+), 10 deletions(-)

diff --git a/kernel/cgroup.c b/kernel/cgroup.c
index bc3caff..24f6d6f 100644
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -1850,14 +1850,14 @@ static int cgroup_task_migrate(struct cgroup *cgrp, struct cgroup *oldcgrp,
 	struct css_set *newcg;
 
 	/*
-	 * get old css_set. we need to take task_lock and refcount it, because
-	 * an exiting task can change its css_set to init_css_set and drop its
-	 * old one without taking cgroup_mutex.
+	 * get old css_set. We are synchronized through threadgroup_lock()
+	 * against PF_EXITING setting such that we can't race against
+	 * cgroup_exit() changing the css_set to init_css_set and dropping the
+	 * old one.
 	 */
-	task_lock(tsk);
+	WARN_ON_ONCE(tsk->flags & PF_EXITING);
 	oldcg = tsk->cgroups;
 	get_css_set(oldcg);
-	task_unlock(tsk);
 
 	/* locate or allocate a new css_set for this task. */
 	if (guarantee) {
@@ -1879,9 +1879,7 @@ static int cgroup_task_migrate(struct cgroup *cgrp, struct cgroup *oldcgrp,
 	}
 	put_css_set(oldcg);
 
-	/* @tsk can't exit as its threadgroup is locked */
 	task_lock(tsk);
-	WARN_ON_ONCE(tsk->flags & PF_EXITING);
 	rcu_assign_pointer(tsk->cgroups, newcg);
 	task_unlock(tsk);
 
@@ -2182,11 +2180,13 @@ int cgroup_attach_proc(struct cgroup *cgrp, struct task_struct *leader)
 		/* nothing to do if this task is already in the cgroup */
 		if (tc->cgrp == cgrp)
 			continue;
-		/* get old css_set pointer */
-		task_lock(tc->task);
+		/*
+		 * get old css_set pointer. threadgroup is locked so this is
+		 * safe against concurrent cgroup_exit() changing this to
+		 * init_css_set.
+		 */
 		oldcg = tc->task->cgroups;
 		get_css_set(oldcg);
-		task_unlock(tc->task);
 		/* see if the new one for us is already in the list? */
 		if (css_set_check_fetched(cgrp, tc->task, oldcg, &newcg_list)) {
 			/* was already there, nothing to do. */
-- 
1.7.5.4

[PATCH 2/2 v2] cgroup: Drop task_lock(parent) on cgroup_fork()

[Frederic Weisbecker]

We don't need to hold the parent task_lock() on the
parent in cgroup_fork() because we are already synchronized
against the two places that may change the parent css_set
concurrently:

- cgroup_exit(), but the parent obviously can't exit concurrently
- cgroup migration: we are synchronized against threadgroup_lock()

So we can safely remove the task_lock() there.

Signed-off-by: Frederic Weisbecker <fweisbec-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Cc: Tejun Heo <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
Cc: Li Zefan <lizf-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org>
Cc: Containers <containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org>
Cc: Cgroups <cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu-+CUm20s59erQFUHtdCDX3A@public.gmane.org>
Cc: Oleg Nesterov <oleg-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
Cc: Andrew Morton <akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org>
Cc: Paul Menage <paul-inf54ven1CmVyaH7bEyXVA@public.gmane.org>
Cc: Mandeep Singh Baines <msb-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org>
---
 kernel/cgroup.c |   10 +++++++---
 1 files changed, 7 insertions(+), 3 deletions(-)

diff --git a/kernel/cgroup.c b/kernel/cgroup.c
index 24f6d6f..1999f60 100644
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -4556,7 +4556,7 @@ static const struct file_operations proc_cgroupstats_operations = {
  *
  * A pointer to the shared css_set was automatically copied in
  * fork.c by dup_task_struct().  However, we ignore that copy, since
- * it was not made under the protection of RCU or cgroup_mutex, so
+ * it was not made under the protection of threadgroup_change_begin(), so
  * might no longer be a valid cgroup pointer.  cgroup_attach_task() might
  * have already changed current->cgroups, allowing the previously
  * referenced cgroup group to be removed and freed.
@@ -4566,10 +4566,14 @@ static const struct file_operations proc_cgroupstats_operations = {
  */
 void cgroup_fork(struct task_struct *child)
 {
-	task_lock(current);
+	/*
+	 * We don't need to task_lock() current because current->cgroups
+	 * can't be changed concurrently here. The parent obviously hasn't
+	 * exited and called cgroup_exit(), and we are synchronized against
+	 * cgroup migration through threadgroup_change_begin().
+	 */
 	child->cgroups = current->cgroups;
 	get_css_set(child->cgroups);
-	task_unlock(current);
 	INIT_LIST_HEAD(&child->cg_list);
 }
 
-- 
1.7.5.4

Re: [PATCH 2/2 v2] cgroup: Drop task_lock(parent) on cgroup_fork()

[Li Zefan]

Frederic Weisbecker wrote:
> We don't need to hold the parent task_lock() on the
> parent in cgroup_fork() because we are already synchronized
> against the two places that may change the parent css_set
> concurrently:
> 
> - cgroup_exit(), but the parent obviously can't exit concurrently
> - cgroup migration: we are synchronized against threadgroup_lock()
> 
> So we can safely remove the task_lock() there.
> 
> Signed-off-by: Frederic Weisbecker <fweisbec-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
> Cc: Tejun Heo <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
> Cc: Li Zefan <lizf-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org>
> Cc: Containers <containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org>
> Cc: Cgroups <cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org>
> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu-+CUm20s59erQFUHtdCDX3A@public.gmane.org>
> Cc: Oleg Nesterov <oleg-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
> Cc: Andrew Morton <akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org>
> Cc: Paul Menage <paul-inf54ven1CmVyaH7bEyXVA@public.gmane.org>
> Cc: Mandeep Singh Baines <msb-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org>
> ---
>  kernel/cgroup.c |   10 +++++++---
>  1 files changed, 7 insertions(+), 3 deletions(-)
> 
> diff --git a/kernel/cgroup.c b/kernel/cgroup.c
> index 24f6d6f..1999f60 100644
> --- a/kernel/cgroup.c
> +++ b/kernel/cgroup.c
> @@ -4556,7 +4556,7 @@ static const struct file_operations proc_cgroupstats_operations = {
>   *
>   * A pointer to the shared css_set was automatically copied in
>   * fork.c by dup_task_struct().  However, we ignore that copy, since
> - * it was not made under the protection of RCU or cgroup_mutex, so
> + * it was not made under the protection of threadgroup_change_begin(), so

I think the original comment still stands, but now threadgroup_change_begin()
can also protect the cgroup pointer from becoming invalid.

>   * might no longer be a valid cgroup pointer.  cgroup_attach_task() might
>   * have already changed current->cgroups, allowing the previously
>   * referenced cgroup group to be removed and freed.
> @@ -4566,10 +4566,14 @@ static const struct file_operations proc_cgroupstats_operations = {
>   */
>  void cgroup_fork(struct task_struct *child)
>  {
> -	task_lock(current);
> +	/*
> +	 * We don't need to task_lock() current because current->cgroups
> +	 * can't be changed concurrently here. The parent obviously hasn't
> +	 * exited and called cgroup_exit(), and we are synchronized against
> +	 * cgroup migration through threadgroup_change_begin().
> +	 */
>  	child->cgroups = current->cgroups;
>  	get_css_set(child->cgroups);
> -	task_unlock(current);
>  	INIT_LIST_HEAD(&child->cg_list);
>  }
>  

Re: [PATCH 2/2 v2] cgroup: Drop task_lock(parent) on cgroup_fork()

[Frederic Weisbecker]

On Wed, Dec 21, 2011 at 10:16:22AM +0800, Li Zefan wrote:
> Frederic Weisbecker wrote:
> > We don't need to hold the parent task_lock() on the
> > parent in cgroup_fork() because we are already synchronized
> > against the two places that may change the parent css_set
> > concurrently:
> > 
> > - cgroup_exit(), but the parent obviously can't exit concurrently
> > - cgroup migration: we are synchronized against threadgroup_lock()
> > 
> > So we can safely remove the task_lock() there.
> > 
> > Signed-off-by: Frederic Weisbecker <fweisbec-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
> > Cc: Tejun Heo <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
> > Cc: Li Zefan <lizf-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org>
> > Cc: Containers <containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org>
> > Cc: Cgroups <cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org>
> > Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu-+CUm20s59erQFUHtdCDX3A@public.gmane.org>
> > Cc: Oleg Nesterov <oleg-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
> > Cc: Andrew Morton <akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org>
> > Cc: Paul Menage <paul-inf54ven1CmVyaH7bEyXVA@public.gmane.org>
> > Cc: Mandeep Singh Baines <msb-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org>
> > ---
> >  kernel/cgroup.c |   10 +++++++---
> >  1 files changed, 7 insertions(+), 3 deletions(-)
> > 
> > diff --git a/kernel/cgroup.c b/kernel/cgroup.c
> > index 24f6d6f..1999f60 100644
> > --- a/kernel/cgroup.c
> > +++ b/kernel/cgroup.c
> > @@ -4556,7 +4556,7 @@ static const struct file_operations proc_cgroupstats_operations = {
> >   *
> >   * A pointer to the shared css_set was automatically copied in
> >   * fork.c by dup_task_struct().  However, we ignore that copy, since
> > - * it was not made under the protection of RCU or cgroup_mutex, so
> > + * it was not made under the protection of threadgroup_change_begin(), so
> 
> I think the original comment still stands, but now threadgroup_change_begin()
> can also protect the cgroup pointer from becoming invalid.

Right but I'm not sure it's worth quoting RCU and cgroup_mutex. The reason
why we use threadgroup_change_begin() is not only to ensure the pointer
validity but also to synchronize the whole cgroup proc logic. This way
when we attach a whole proc with cgroup_attach_proc(), we are sure that
no thread forked too soon or too late such that it wouldn't be migrated with
the rest.

RCU or cgroup_mutex on dup_task_struct() (+ a get_css_set()) would have
protected the pointer validity but not the whole above described machinery.
So I don't think it's even worth quoting those solutions. But if you prefer
I can keep the old comment.

OTOH what I think is missing in the comment is that explanation on the synchronization
against entire proc migration. I can edit that.

> 
> >   * might no longer be a valid cgroup pointer.  cgroup_attach_task() might
> >   * have already changed current->cgroups, allowing the previously
> >   * referenced cgroup group to be removed and freed.
> > @@ -4566,10 +4566,14 @@ static const struct file_operations proc_cgroupstats_operations = {
> >   */
> >  void cgroup_fork(struct task_struct *child)
> >  {
> > -	task_lock(current);
> > +	/*
> > +	 * We don't need to task_lock() current because current->cgroups
> > +	 * can't be changed concurrently here. The parent obviously hasn't
> > +	 * exited and called cgroup_exit(), and we are synchronized against
> > +	 * cgroup migration through threadgroup_change_begin().
> > +	 */
> >  	child->cgroups = current->cgroups;
> >  	get_css_set(child->cgroups);
> > -	task_unlock(current);
> >  	INIT_LIST_HEAD(&child->cg_list);
> >  }
> >  

Re: [PATCH 2/2 v2] cgroup: Drop task_lock(parent) on cgroup_fork()

[Li Zefan]

>>> @@ -4556,7 +4556,7 @@ static const struct file_operations proc_cgroupstats_operations = {
>>>   *
>>>   * A pointer to the shared css_set was automatically copied in
>>>   * fork.c by dup_task_struct().  However, we ignore that copy, since
>>> - * it was not made under the protection of RCU or cgroup_mutex, so
>>> + * it was not made under the protection of threadgroup_change_begin(), so
>>
>> I think the original comment still stands, but now threadgroup_change_begin()
>> can also protect the cgroup pointer from becoming invalid.
> 
> Right but I'm not sure it's worth quoting RCU and cgroup_mutex. The reason
> why we use threadgroup_change_begin() is not only to ensure the pointer
> validity but also to synchronize the whole cgroup proc logic. This way
> when we attach a whole proc with cgroup_attach_proc(), we are sure that
> no thread forked too soon or too late such that it wouldn't be migrated with
> the rest.
> 
> RCU or cgroup_mutex on dup_task_struct() (+ a get_css_set()) would have
> protected the pointer validity but not the whole above described machinery.
> So I don't think it's even worth quoting those solutions. But if you prefer
> I can keep the old comment.
> 

No, I don't have strong opinion here.

So I'll ack this patch.

> OTOH what I think is missing in the comment is that explanation on the synchronization
> against entire proc migration. I can edit that.
> 

I would appreciate this. :)

>>
>>>   * might no longer be a valid cgroup pointer.  cgroup_attach_task() might
>>>   * have already changed current->cgroups, allowing the previously
>>>   * referenced cgroup group to be removed and freed.
>>> @@ -4566,10 +4566,14 @@ static const struct file_operations proc_cgroupstats_operations = {
>>>   */
>>>  void cgroup_fork(struct task_struct *child)
>>>  {
>>> -	task_lock(current);
>>> +	/*
>>> +	 * We don't need to task_lock() current because current->cgroups
>>> +	 * can't be changed concurrently here. The parent obviously hasn't
>>> +	 * exited and called cgroup_exit(), and we are synchronized against
>>> +	 * cgroup migration through threadgroup_change_begin().
>>> +	 */
>>>  	child->cgroups = current->cgroups;
>>>  	get_css_set(child->cgroups);
>>> -	task_unlock(current);
>>>  	INIT_LIST_HEAD(&child->cg_list);
>>>  }
>>>  
> 

Re: [PATCH 2/2 v2] cgroup: Drop task_lock(parent) on cgroup_fork()

[Tejun Heo]

Hello,

On Wed, Dec 21, 2011 at 11:16:11AM +0800, Li Zefan wrote:
> So I'll ack this patch.

Do I have your ack on the first one too?  If so, I'll go ahead and
apply these two.

Thanks.

-- 
tejun

Re: [PATCH 2/2 v2] cgroup: Drop task_lock(parent) on cgroup_fork()

[Tejun Heo]

On Wed, Dec 21, 2011 at 09:50:24AM -0800, Tejun Heo wrote:
> Hello,
> 
> On Wed, Dec 21, 2011 at 11:16:11AM +0800, Li Zefan wrote:
> > So I'll ack this patch.
> 
> Do I have your ack on the first one too?  If so, I'll go ahead and
> apply these two.

Ooh, you already did in the previous posting.  Applying to
cgroup/for-3.3.

Thanks.

-- 
tejun

Re: [PATCH 2/2 v2] cgroup: Drop task_lock(parent) on cgroup_fork()

[Frederic Weisbecker]

On Wed, Dec 21, 2011 at 09:50:24AM -0800, Tejun Heo wrote:
> Hello,
> 
> On Wed, Dec 21, 2011 at 11:16:11AM +0800, Li Zefan wrote:
> > So I'll ack this patch.
> 
> Do I have your ack on the first one too?  If so, I'll go ahead and
> apply these two.
> 
> Thanks.

Lemme just update the comment on this one before, I'll resend a new version
quickly, thanks!