Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.kernel > #1702235 > unrolled thread
| Started by | "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> |
|---|---|
| First post | 2017-08-02 18:10 +0200 |
| Last post | 2017-08-02 19:50 +0200 |
| Articles | 3 — 2 participants |
Back to article view | Back to linux.kernel
This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by
below is the oldest one visible, not the original post.
Re: [PATCH tip/sched/core] swait: Remove the lockless swait_active() check in swake_up*() "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2017-08-02 18:10 +0200
Re: [PATCH tip/sched/core] swait: Remove the lockless swait_active() check in swake_up*() Peter Zijlstra <peterz@infradead.org> - 2017-08-02 19:20 +0200
Re: [PATCH tip/sched/core] swait: Remove the lockless swait_active() check in swake_up*() "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2017-08-02 19:50 +0200
| From | "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> |
|---|---|
| Date | 2017-08-02 18:10 +0200 |
| Subject | Re: [PATCH tip/sched/core] swait: Remove the lockless swait_active() check in swake_up*() |
| Message-ID | <ua4nH-3Eo-63@gated-at.bofh.it> |
On Sun, Jul 30, 2017 at 09:47:35PM +0800, Boqun Feng wrote:
> Steven Rostedt reported a potential race in RCU core because of
> swake_up():
>
> CPU0 CPU1
> ---- ----
> __call_rcu_core() {
>
> spin_lock(rnp_root)
> need_wake = __rcu_start_gp() {
> rcu_start_gp_advanced() {
> gp_flags = FLAG_INIT
> }
> }
>
> rcu_gp_kthread() {
> swait_event_interruptible(wq,
> gp_flags & FLAG_INIT) {
> spin_lock(q->lock)
>
> *fetch wq->task_list here! *
>
> list_add(wq->task_list, q->task_list)
> spin_unlock(q->lock);
>
> *fetch old value of gp_flags here *
>
> spin_unlock(rnp_root)
>
> rcu_gp_kthread_wake() {
> swake_up(wq) {
> swait_active(wq) {
> list_empty(wq->task_list)
>
> } * return false *
>
> if (condition) * false *
> schedule();
>
> In this case, a wakeup is missed, which could cause the rcu_gp_kthread
> waits for a long time.
>
> The reason of this is that we do a lockless swait_active() check in
> swake_up(). To fix this, we can either 1) add a smp_mb() in swake_up()
> before swait_active() to provide the proper order or 2) simply remove
> the swait_active() in swake_up().
>
> The solution 2 not only fixes this problem but also keeps the swait and
> wait API as close as possible, as wake_up() doesn't provide a full
> barrier and doesn't do a lockless check of the wait queue either.
> Moreover, there are users already using swait_active() to do their quick
> checks for the wait queues, so it make less sense that swake_up() and
> swake_up_all() do this on their own.
>
> This patch then removes the lockless swait_active() check in swake_up()
> and swake_up_all().
>
> Reported-by: Steven Rostedt <rostedt@goodmis.org>
> Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
> Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
> Tested-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Hearing no objections but not hearing anything else, either, I have
queued this for v4.14. If someone else would rather queue it, please
let me know.
Thanx, Paul
> ---
> kernel/sched/swait.c | 6 ------
> 1 file changed, 6 deletions(-)
>
> diff --git a/kernel/sched/swait.c b/kernel/sched/swait.c
> index 3d5610dcce11..2227e183e202 100644
> --- a/kernel/sched/swait.c
> +++ b/kernel/sched/swait.c
> @@ -33,9 +33,6 @@ void swake_up(struct swait_queue_head *q)
> {
> unsigned long flags;
>
> - if (!swait_active(q))
> - return;
> -
> raw_spin_lock_irqsave(&q->lock, flags);
> swake_up_locked(q);
> raw_spin_unlock_irqrestore(&q->lock, flags);
> @@ -51,9 +48,6 @@ void swake_up_all(struct swait_queue_head *q)
> struct swait_queue *curr;
> LIST_HEAD(tmp);
>
> - if (!swait_active(q))
> - return;
> -
> raw_spin_lock_irq(&q->lock);
> list_splice_init(&q->task_list, &tmp);
> while (!list_empty(&tmp)) {
> --
> 2.13.3
>
[toc] | [next] | [standalone]
| From | Peter Zijlstra <peterz@infradead.org> |
|---|---|
| Date | 2017-08-02 19:20 +0200 |
| Message-ID | <ua5tq-4lS-55@gated-at.bofh.it> |
| In reply to | #1702235 |
On Wed, Aug 02, 2017 at 09:07:13AM -0700, Paul E. McKenney wrote:
> On Sun, Jul 30, 2017 at 09:47:35PM +0800, Boqun Feng wrote:
> > Steven Rostedt reported a potential race in RCU core because of
> > swake_up():
> >
> > CPU0 CPU1
> > ---- ----
> > __call_rcu_core() {
> >
> > spin_lock(rnp_root)
> > need_wake = __rcu_start_gp() {
> > rcu_start_gp_advanced() {
> > gp_flags = FLAG_INIT
> > }
> > }
> >
> > rcu_gp_kthread() {
> > swait_event_interruptible(wq,
> > gp_flags & FLAG_INIT) {
> > spin_lock(q->lock)
> >
> > *fetch wq->task_list here! *
> >
> > list_add(wq->task_list, q->task_list)
> > spin_unlock(q->lock);
> >
> > *fetch old value of gp_flags here *
> >
> > spin_unlock(rnp_root)
> >
> > rcu_gp_kthread_wake() {
> > swake_up(wq) {
> > swait_active(wq) {
> > list_empty(wq->task_list)
> >
> > } * return false *
> >
> > if (condition) * false *
> > schedule();
> >
> > In this case, a wakeup is missed, which could cause the rcu_gp_kthread
> > waits for a long time.
> >
> > The reason of this is that we do a lockless swait_active() check in
> > swake_up(). To fix this, we can either 1) add a smp_mb() in swake_up()
> > before swait_active() to provide the proper order or 2) simply remove
> > the swait_active() in swake_up().
> >
> > The solution 2 not only fixes this problem but also keeps the swait and
> > wait API as close as possible, as wake_up() doesn't provide a full
> > barrier and doesn't do a lockless check of the wait queue either.
> > Moreover, there are users already using swait_active() to do their quick
> > checks for the wait queues, so it make less sense that swake_up() and
> > swake_up_all() do this on their own.
> >
> > This patch then removes the lockless swait_active() check in swake_up()
> > and swake_up_all().
> >
> > Reported-by: Steven Rostedt <rostedt@goodmis.org>
> > Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
> > Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
> > Tested-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
>
> Hearing no objections but not hearing anything else, either, I have
> queued this for v4.14. If someone else would rather queue it, please
> let me know.
I have it too. Lets see who can get it into -tip first :-)
[toc] | [prev] | [next] | [standalone]
| From | "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> |
|---|---|
| Date | 2017-08-02 19:50 +0200 |
| Message-ID | <ua5Wp-4AY-15@gated-at.bofh.it> |
| In reply to | #1702305 |
On Wed, Aug 02, 2017 at 07:12:23PM +0200, Peter Zijlstra wrote:
> On Wed, Aug 02, 2017 at 09:07:13AM -0700, Paul E. McKenney wrote:
> > On Sun, Jul 30, 2017 at 09:47:35PM +0800, Boqun Feng wrote:
> > > Steven Rostedt reported a potential race in RCU core because of
> > > swake_up():
> > >
> > > CPU0 CPU1
> > > ---- ----
> > > __call_rcu_core() {
> > >
> > > spin_lock(rnp_root)
> > > need_wake = __rcu_start_gp() {
> > > rcu_start_gp_advanced() {
> > > gp_flags = FLAG_INIT
> > > }
> > > }
> > >
> > > rcu_gp_kthread() {
> > > swait_event_interruptible(wq,
> > > gp_flags & FLAG_INIT) {
> > > spin_lock(q->lock)
> > >
> > > *fetch wq->task_list here! *
> > >
> > > list_add(wq->task_list, q->task_list)
> > > spin_unlock(q->lock);
> > >
> > > *fetch old value of gp_flags here *
> > >
> > > spin_unlock(rnp_root)
> > >
> > > rcu_gp_kthread_wake() {
> > > swake_up(wq) {
> > > swait_active(wq) {
> > > list_empty(wq->task_list)
> > >
> > > } * return false *
> > >
> > > if (condition) * false *
> > > schedule();
> > >
> > > In this case, a wakeup is missed, which could cause the rcu_gp_kthread
> > > waits for a long time.
> > >
> > > The reason of this is that we do a lockless swait_active() check in
> > > swake_up(). To fix this, we can either 1) add a smp_mb() in swake_up()
> > > before swait_active() to provide the proper order or 2) simply remove
> > > the swait_active() in swake_up().
> > >
> > > The solution 2 not only fixes this problem but also keeps the swait and
> > > wait API as close as possible, as wake_up() doesn't provide a full
> > > barrier and doesn't do a lockless check of the wait queue either.
> > > Moreover, there are users already using swait_active() to do their quick
> > > checks for the wait queues, so it make less sense that swake_up() and
> > > swake_up_all() do this on their own.
> > >
> > > This patch then removes the lockless swait_active() check in swake_up()
> > > and swake_up_all().
> > >
> > > Reported-by: Steven Rostedt <rostedt@goodmis.org>
> > > Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
> > > Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
> > > Tested-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
> >
> > Hearing no objections but not hearing anything else, either, I have
> > queued this for v4.14. If someone else would rather queue it, please
> > let me know.
>
> I have it too. Lets see who can get it into -tip first :-)
Heh! We could use it as a test of Ingo's handling of multiple identical
patches. ;-)
Thanx, Paul
[toc] | [prev] | [standalone]
Back to top | Article view | linux.kernel
csiph-web