diff mbox series

[2/2] media: v4l: async: Properly check for a notifier initialised or registered

Message ID 20240311121741.1249976-3-sakari.ailus@linux.intel.com (mailing list archive)
State Superseded
Delegated to: Sakari Ailus
Headers
Series V4L2 async fixes |

Commit Message

Sakari Ailus March 11, 2024, 12:17 p.m. UTC 
Properly check that a notifier was never initialised or register. This can
now be done by looking at the entry in the notifier list, not the V4L2
device or sub-device that are set in the initialiser now.

Fixes: b8ec754ae4c5 ("media: v4l: async: Set v4l2_device and subdev in async notifier init")
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
---
 drivers/media/v4l2-core/v4l2-async.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

Comments

Alexander Stein March 14, 2024, 2:04 p.m. UTC | #1 
Hi Sakari,

I suppose on of your intentions of this series is to replace my patch, no?

Am Montag, 11. März 2024, 13:17:41 CET schrieb Sakari Ailus:
> Properly check that a notifier was never initialised or register. This can
> now be done by looking at the entry in the notifier list, not the V4L2
> device or sub-device that are set in the initialiser now.
> 
> Fixes: b8ec754ae4c5 ("media: v4l: async: Set v4l2_device and subdev in async notifier init")
> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
> ---
>  drivers/media/v4l2-core/v4l2-async.c | 6 ++++--
>  1 file changed, 4 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/media/v4l2-core/v4l2-async.c b/drivers/media/v4l2-core/v4l2-async.c
> index 2ff35d5d60f2..3b43d6285dfe 100644
> --- a/drivers/media/v4l2-core/v4l2-async.c
> +++ b/drivers/media/v4l2-core/v4l2-async.c
> @@ -630,12 +630,14 @@ EXPORT_SYMBOL(v4l2_async_nf_register);
>  static void
>  __v4l2_async_nf_unregister(struct v4l2_async_notifier *notifier)
>  {
> -	if (!notifier || (!notifier->v4l2_dev && !notifier->sd))
> +	/* Return here if the notifier is never initialised or registered. */
> +	if (!notifier->notifier_entry.next ||

I don't like the idea checking the next pointer of a list.
Despite that it's not even necessary.

Best regards,
Alexander

> +	    list_empty(&notifier->notifier_entry))
>  		return;
>  
>  	v4l2_async_nf_unbind_all_subdevs(notifier);
>  
> -	list_del(&notifier->notifier_entry);
> +	list_del_init(&notifier->notifier_entry);
>  }
>  
>  void v4l2_async_nf_unregister(struct v4l2_async_notifier *notifier)
>
Sakari Ailus March 20, 2024, 7:07 a.m. UTC | #2 
Hi Alexander,

On Thu, Mar 14, 2024 at 03:04:44PM +0100, Alexander Stein wrote:
> Hi Sakari,
> 
> I suppose on of your intentions of this series is to replace my patch, no?

No, these are somewhat unrelated issues.

> 
> Am Montag, 11. März 2024, 13:17:41 CET schrieb Sakari Ailus:
> > Properly check that a notifier was never initialised or register. This can
> > now be done by looking at the entry in the notifier list, not the V4L2
> > device or sub-device that are set in the initialiser now.
> > 
> > Fixes: b8ec754ae4c5 ("media: v4l: async: Set v4l2_device and subdev in async notifier init")
> > Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
> > ---
> >  drivers/media/v4l2-core/v4l2-async.c | 6 ++++--
> >  1 file changed, 4 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/media/v4l2-core/v4l2-async.c b/drivers/media/v4l2-core/v4l2-async.c
> > index 2ff35d5d60f2..3b43d6285dfe 100644
> > --- a/drivers/media/v4l2-core/v4l2-async.c
> > +++ b/drivers/media/v4l2-core/v4l2-async.c
> > @@ -630,12 +630,14 @@ EXPORT_SYMBOL(v4l2_async_nf_register);
> >  static void
> >  __v4l2_async_nf_unregister(struct v4l2_async_notifier *notifier)
> >  {
> > -	if (!notifier || (!notifier->v4l2_dev && !notifier->sd))
> > +	/* Return here if the notifier is never initialised or registered. */
> > +	if (!notifier->notifier_entry.next ||
> 
> I don't like the idea checking the next pointer of a list.
> Despite that it's not even necessary.

Actually I think we can drop the above change. But the list_del_init()
below is still necessary.

> 
> Best regards,
> Alexander
> 
> > +	    list_empty(&notifier->notifier_entry))
> >  		return;
> >  
> >  	v4l2_async_nf_unbind_all_subdevs(notifier);
> >  
> > -	list_del(&notifier->notifier_entry);
> > +	list_del_init(&notifier->notifier_entry);
> >  }
> >  
> >  void v4l2_async_nf_unregister(struct v4l2_async_notifier *notifier)
> > 
> 
>
Alexander Stein March 21, 2024, 8:28 a.m. UTC | #3 
Hi Sakari,

Am Mittwoch, 20. März 2024, 08:07:51 CET schrieb Sakari Ailus:
> Hi Alexander,
> 
> On Thu, Mar 14, 2024 at 03:04:44PM +0100, Alexander Stein wrote:
> > Hi Sakari,
> > 
> > I suppose on of your intentions of this series is to replace my patch, no?
> 
> No, these are somewhat unrelated issues.

Okay, thanks for confirmation.

> > 
> > Am Montag, 11. März 2024, 13:17:41 CET schrieb Sakari Ailus:
> > > Properly check that a notifier was never initialised or register. This can
> > > now be done by looking at the entry in the notifier list, not the V4L2
> > > device or sub-device that are set in the initialiser now.
> > > 
> > > Fixes: b8ec754ae4c5 ("media: v4l: async: Set v4l2_device and subdev in async notifier init")
> > > Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
> > > ---
> > >  drivers/media/v4l2-core/v4l2-async.c | 6 ++++--
> > >  1 file changed, 4 insertions(+), 2 deletions(-)
> > > 
> > > diff --git a/drivers/media/v4l2-core/v4l2-async.c b/drivers/media/v4l2-core/v4l2-async.c
> > > index 2ff35d5d60f2..3b43d6285dfe 100644
> > > --- a/drivers/media/v4l2-core/v4l2-async.c
> > > +++ b/drivers/media/v4l2-core/v4l2-async.c
> > > @@ -630,12 +630,14 @@ EXPORT_SYMBOL(v4l2_async_nf_register);
> > >  static void
> > >  __v4l2_async_nf_unregister(struct v4l2_async_notifier *notifier)
> > >  {
> > > -	if (!notifier || (!notifier->v4l2_dev && !notifier->sd))
> > > +	/* Return here if the notifier is never initialised or registered. */
> > > +	if (!notifier->notifier_entry.next ||
> > 
> > I don't like the idea checking the next pointer of a list.
> > Despite that it's not even necessary.
> 
> Actually I think we can drop the above change. But the list_del_init()
> below is still necessary.

I agree about the list_del_init() change.

Best regards,
Alexander

> > 
> > Best regards,
> > Alexander
> > 
> > > +	    list_empty(&notifier->notifier_entry))
> > >  		return;
> > >  
> > >  	v4l2_async_nf_unbind_all_subdevs(notifier);
> > >  
> > > -	list_del(&notifier->notifier_entry);
> > > +	list_del_init(&notifier->notifier_entry);
> > >  }
> > >  
> > >  void v4l2_async_nf_unregister(struct v4l2_async_notifier *notifier)
> > > 
> > 
> > 
> 
>
diff mbox series

Patch

diff --git a/drivers/media/v4l2-core/v4l2-async.c b/drivers/media/v4l2-core/v4l2-async.c
index 2ff35d5d60f2..3b43d6285dfe 100644
--- a/drivers/media/v4l2-core/v4l2-async.c
+++ b/drivers/media/v4l2-core/v4l2-async.c
@@ -630,12 +630,14 @@  EXPORT_SYMBOL(v4l2_async_nf_register);
 static void
 __v4l2_async_nf_unregister(struct v4l2_async_notifier *notifier)
 {
-	if (!notifier || (!notifier->v4l2_dev && !notifier->sd))
+	/* Return here if the notifier is never initialised or registered. */
+	if (!notifier->notifier_entry.next ||
+	    list_empty(&notifier->notifier_entry))
 		return;
 
 	v4l2_async_nf_unbind_all_subdevs(notifier);
 
-	list_del(&notifier->notifier_entry);
+	list_del_init(&notifier->notifier_entry);
 }
 
 void v4l2_async_nf_unregister(struct v4l2_async_notifier *notifier)