[PATCH 1/2] PM / wakeup: Remove timer from wakeup_source_remove()

[PATCH 1/2] PM / wakeup: Remove timer from wakeup_source_remove()

[Viresh Kumar]

wakeup_source_remove() is the counterpart of wakeup_source_add() helper
and must undo the initializations done by wakeup_source_add(). Currently
the timer is initialized by wakeup_source_add() but removed from
wakeup_source_drop(), which doesn't look logically correct. Also it
should be okay to call wakeup_source_add() right after calling
wakeup_source_remove(), and in that case we may end up calling
timer_setup() for a potentially scheduled timer which is surely
incorrect.

Move the timer removal part to wakeup_source_remove() instead.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
---
 drivers/base/power/wakeup.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/base/power/wakeup.c b/drivers/base/power/wakeup.c
index f1fee72ed970..18333962e3da 100644
--- a/drivers/base/power/wakeup.c
+++ b/drivers/base/power/wakeup.c
@@ -118,7 +118,6 @@ void wakeup_source_drop(struct wakeup_source *ws)
 	if (!ws)
 		return;
 
-	del_timer_sync(&ws->timer);
 	__pm_relax(ws);
 }
 EXPORT_SYMBOL_GPL(wakeup_source_drop);
@@ -205,6 +204,8 @@ void wakeup_source_remove(struct wakeup_source *ws)
 	list_del_rcu(&ws->entry);
 	raw_spin_unlock_irqrestore(&events_lock, flags);
 	synchronize_srcu(&wakeup_srcu);
+
+	del_timer_sync(&ws->timer);
 }
 EXPORT_SYMBOL_GPL(wakeup_source_remove);
 
-- 
2.21.0.rc0.269.g1a574e7a288b

[PATCH 2/2] PM / wakeup: Clear timer.function in wakeup_source_remove()

[Viresh Kumar]

wakeup_source_activate() performs a check to see if the wakeup source is
registered or not. It works fine for a newly added wakeup source which
may not have been registered but fails to catch the case where a wakeup
source is unregistered as the timer.function is still valid.

Fix it by setting the timer.function to NULL from
wakeup_source_remove().

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
---
 drivers/base/power/wakeup.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/drivers/base/power/wakeup.c b/drivers/base/power/wakeup.c
index 18333962e3da..3699faca24af 100644
--- a/drivers/base/power/wakeup.c
+++ b/drivers/base/power/wakeup.c
@@ -206,6 +206,13 @@ void wakeup_source_remove(struct wakeup_source *ws)
 	synchronize_srcu(&wakeup_srcu);
 
 	del_timer_sync(&ws->timer);
+
+	/*
+	 * Clear timer.function so that wakeup_source_not_registered() can
+	 * detect an unregistered wakeup source.
+	 */
+	ws->timer.function = NULL;
+
 }
 EXPORT_SYMBOL_GPL(wakeup_source_remove);
 
-- 
2.21.0.rc0.269.g1a574e7a288b

Re: [PATCH 1/2] PM / wakeup: Remove timer from wakeup_source_remove()

[Rafael J. Wysocki]

On Friday, March 8, 2019 10:53:11 AM CET Viresh Kumar wrote:
> wakeup_source_remove() is the counterpart of wakeup_source_add() helper
> and must undo the initializations done by wakeup_source_add(). Currently
> the timer is initialized by wakeup_source_add() but removed from
> wakeup_source_drop(), which doesn't look logically correct. Also it
> should be okay to call wakeup_source_add() right after calling
> wakeup_source_remove(), and in that case we may end up calling
> timer_setup() for a potentially scheduled timer which is surely
> incorrect.
> 
> Move the timer removal part to wakeup_source_remove() instead.
> 
> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
> ---
>  drivers/base/power/wakeup.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/base/power/wakeup.c b/drivers/base/power/wakeup.c
> index f1fee72ed970..18333962e3da 100644
> --- a/drivers/base/power/wakeup.c
> +++ b/drivers/base/power/wakeup.c
> @@ -118,7 +118,6 @@ void wakeup_source_drop(struct wakeup_source *ws)
>  	if (!ws)
>  		return;
>  
> -	del_timer_sync(&ws->timer);
>  	__pm_relax(ws);
>  }
>  EXPORT_SYMBOL_GPL(wakeup_source_drop);
> @@ -205,6 +204,8 @@ void wakeup_source_remove(struct wakeup_source *ws)
>  	list_del_rcu(&ws->entry);
>  	raw_spin_unlock_irqrestore(&events_lock, flags);
>  	synchronize_srcu(&wakeup_srcu);
> +
> +	del_timer_sync(&ws->timer);
>  }
>  EXPORT_SYMBOL_GPL(wakeup_source_remove);
>  
> 

I've merged it with the [2/2], rewritten the subject and changelog and
queued the result as commit d856f39ac1cc ("PM / wakeup: Rework wakeup
source timer cancellation").

Re: [PATCH 1/2] PM / wakeup: Remove timer from wakeup_source_remove()

[Viresh Kumar]

On 11-03-19, 13:05, Rafael J. Wysocki wrote:
> On Friday, March 8, 2019 10:53:11 AM CET Viresh Kumar wrote:
> > wakeup_source_remove() is the counterpart of wakeup_source_add() helper
> > and must undo the initializations done by wakeup_source_add(). Currently
> > the timer is initialized by wakeup_source_add() but removed from
> > wakeup_source_drop(), which doesn't look logically correct. Also it
> > should be okay to call wakeup_source_add() right after calling
> > wakeup_source_remove(), and in that case we may end up calling
> > timer_setup() for a potentially scheduled timer which is surely
> > incorrect.
> > 
> > Move the timer removal part to wakeup_source_remove() instead.
> > 
> > Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
> > ---
> >  drivers/base/power/wakeup.c | 3 ++-
> >  1 file changed, 2 insertions(+), 1 deletion(-)
> > 
> > diff --git a/drivers/base/power/wakeup.c b/drivers/base/power/wakeup.c
> > index f1fee72ed970..18333962e3da 100644
> > --- a/drivers/base/power/wakeup.c
> > +++ b/drivers/base/power/wakeup.c
> > @@ -118,7 +118,6 @@ void wakeup_source_drop(struct wakeup_source *ws)
> >  	if (!ws)
> >  		return;
> >  
> > -	del_timer_sync(&ws->timer);
> >  	__pm_relax(ws);
> >  }
> >  EXPORT_SYMBOL_GPL(wakeup_source_drop);
> > @@ -205,6 +204,8 @@ void wakeup_source_remove(struct wakeup_source *ws)
> >  	list_del_rcu(&ws->entry);
> >  	raw_spin_unlock_irqrestore(&events_lock, flags);
> >  	synchronize_srcu(&wakeup_srcu);
> > +
> > +	del_timer_sync(&ws->timer);
> >  }
> >  EXPORT_SYMBOL_GPL(wakeup_source_remove);
> >  
> > 
> 
> I've merged it with the [2/2], rewritten the subject and changelog and
> queued the result as commit d856f39ac1cc ("PM / wakeup: Rework wakeup
> source timer cancellation").

Okay, thanks. We (Android guys) want this to be backported into 4.4+
kernels via the stable tree. Can we mark this for stable in the commit
itself ? Else I would be required to send this separately for all the
kernels. I should have marked it for stable initially though, sorry
about forgetting then.

-- 
viresh

Re: [PATCH 1/2] PM / wakeup: Remove timer from wakeup_source_remove()

[Rafael J. Wysocki]

On Tue, Mar 12, 2019 at 4:28 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
>
> On 11-03-19, 13:05, Rafael J. Wysocki wrote:
> > On Friday, March 8, 2019 10:53:11 AM CET Viresh Kumar wrote:
> > > wakeup_source_remove() is the counterpart of wakeup_source_add() helper
> > > and must undo the initializations done by wakeup_source_add(). Currently
> > > the timer is initialized by wakeup_source_add() but removed from
> > > wakeup_source_drop(), which doesn't look logically correct. Also it
> > > should be okay to call wakeup_source_add() right after calling
> > > wakeup_source_remove(), and in that case we may end up calling
> > > timer_setup() for a potentially scheduled timer which is surely
> > > incorrect.
> > >
> > > Move the timer removal part to wakeup_source_remove() instead.
> > >
> > > Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
> > > ---
> > >  drivers/base/power/wakeup.c | 3 ++-
> > >  1 file changed, 2 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/drivers/base/power/wakeup.c b/drivers/base/power/wakeup.c
> > > index f1fee72ed970..18333962e3da 100644
> > > --- a/drivers/base/power/wakeup.c
> > > +++ b/drivers/base/power/wakeup.c
> > > @@ -118,7 +118,6 @@ void wakeup_source_drop(struct wakeup_source *ws)
> > >     if (!ws)
> > >             return;
> > >
> > > -   del_timer_sync(&ws->timer);
> > >     __pm_relax(ws);
> > >  }
> > >  EXPORT_SYMBOL_GPL(wakeup_source_drop);
> > > @@ -205,6 +204,8 @@ void wakeup_source_remove(struct wakeup_source *ws)
> > >     list_del_rcu(&ws->entry);
> > >     raw_spin_unlock_irqrestore(&events_lock, flags);
> > >     synchronize_srcu(&wakeup_srcu);
> > > +
> > > +   del_timer_sync(&ws->timer);
> > >  }
> > >  EXPORT_SYMBOL_GPL(wakeup_source_remove);
> > >
> > >
> >
> > I've merged it with the [2/2], rewritten the subject and changelog and
> > queued the result as commit d856f39ac1cc ("PM / wakeup: Rework wakeup
> > source timer cancellation").
>
> Okay, thanks. We (Android guys) want this to be backported into 4.4+
> kernels via the stable tree. Can we mark this for stable in the commit
> itself ? Else I would be required to send this separately for all the
> kernels. I should have marked it for stable initially though, sorry
> about forgetting then.

Queued up with a CC-stable tag for 4.4 an later, thanks!

Re: [PATCH 1/2] PM / wakeup: Remove timer from wakeup_source_remove()

[Pavel Machek]

[-- Attachment #1: Type: text/plain, Size: 1637 bytes --]

On Tue 2019-03-12 08:58:02, Viresh Kumar wrote:
> On 11-03-19, 13:05, Rafael J. Wysocki wrote:
> > On Friday, March 8, 2019 10:53:11 AM CET Viresh Kumar wrote:
> > > wakeup_source_remove() is the counterpart of wakeup_source_add() helper
> > > and must undo the initializations done by wakeup_source_add(). Currently
> > > the timer is initialized by wakeup_source_add() but removed from
> > > wakeup_source_drop(), which doesn't look logically correct. Also it
> > > should be okay to call wakeup_source_add() right after calling
> > > wakeup_source_remove(), and in that case we may end up calling
> > > timer_setup() for a potentially scheduled timer which is surely
> > > incorrect.
> > > 
> > > Move the timer removal part to wakeup_source_remove() instead.

> > 
> > I've merged it with the [2/2], rewritten the subject and changelog and
> > queued the result as commit d856f39ac1cc ("PM / wakeup: Rework wakeup
> > source timer cancellation").
> 
> Okay, thanks. We (Android guys) want this to be backported into 4.4+
> kernels via the stable tree. Can we mark this for stable in the commit
> itself ? Else I would be required to send this separately for all the
> kernels. I should have marked it for stable initially though, sorry
> about forgetting then.

Greg is normally pretty agressive about backporting everything
remotely looking like a fix...

But better changelog would help. How is the bug actually affecting
users?

Best regards,
									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 181 bytes --]

Re: [PATCH 1/2] PM / wakeup: Remove timer from wakeup_source_remove()

[Viresh Kumar]

On 12-03-19, 10:04, Pavel Machek wrote:
> On Tue 2019-03-12 08:58:02, Viresh Kumar wrote:
> > On 11-03-19, 13:05, Rafael J. Wysocki wrote:
> > > On Friday, March 8, 2019 10:53:11 AM CET Viresh Kumar wrote:
> > > > wakeup_source_remove() is the counterpart of wakeup_source_add() helper
> > > > and must undo the initializations done by wakeup_source_add(). Currently
> > > > the timer is initialized by wakeup_source_add() but removed from
> > > > wakeup_source_drop(), which doesn't look logically correct. Also it
> > > > should be okay to call wakeup_source_add() right after calling
> > > > wakeup_source_remove(), and in that case we may end up calling
> > > > timer_setup() for a potentially scheduled timer which is surely
> > > > incorrect.
> > > > 
> > > > Move the timer removal part to wakeup_source_remove() instead.
> 
> > > 
> > > I've merged it with the [2/2], rewritten the subject and changelog and
> > > queued the result as commit d856f39ac1cc ("PM / wakeup: Rework wakeup
> > > source timer cancellation").
> > 
> > Okay, thanks. We (Android guys) want this to be backported into 4.4+
> > kernels via the stable tree. Can we mark this for stable in the commit
> > itself ? Else I would be required to send this separately for all the
> > kernels. I should have marked it for stable initially though, sorry
> > about forgetting then.
> 
> Greg is normally pretty agressive about backporting everything
> remotely looking like a fix...
> 
> But better changelog would help. How is the bug actually affecting
> users?

The Android wakelock infrastructure (not in mainline) is based on the
wakeup sources stuff and that is where we found the issue. A wakelock
was taken due to a bug in user driver and the board never attempts
going into suspend anymore.

If this patch doesn't go into the stable kernels, I would be required
to send this separately for Android kernels and the Android guys will
suggest getting it via stable tree.

And because this fixes a potential issue it was worth trying getting
it via the stable kernel :)

-- 
viresh