[PATCH v3 1/2] driver core: bus.h: document bus notifiers better

[PATCH v3 1/2] driver core: bus.h: document bus notifiers better

[Greg Kroah-Hartman]

The bus notifier values are not documented all that well, so clean this
up and make a real enumerated type for them and document them much
better.  When doing this, remove the hex values and just rely on the
enumerated type instead as that is all that is needed.

Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Cc: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
v3: - fix up the documentation to be proper kernel-doc format thanks to
      Randy.
    - Remove the values entirely now that this is an enumerated type,
      they are not needed.

v2: - move the values to decimal from hex as pointed out by Rafael.

 include/linux/device/bus.h | 42 +++++++++++++++++++++++++-------------
 1 file changed, 28 insertions(+), 14 deletions(-)

diff --git a/include/linux/device/bus.h b/include/linux/device/bus.h
index d529f644e92b..7b4a48b5159b 100644
--- a/include/linux/device/bus.h
+++ b/include/linux/device/bus.h
@@ -257,21 +257,35 @@ extern int bus_register_notifier(struct bus_type *bus,
 extern int bus_unregister_notifier(struct bus_type *bus,
 				   struct notifier_block *nb);
 
-/* All 4 notifers below get called with the target struct device *
- * as an argument. Note that those functions are likely to be called
- * with the device lock held in the core, so be careful.
+/**
+ * enum bus_notifier_event - Bus Notifier events that have happened
+ * @BUS_NOTIFY_ADD_DEVICE: device is added to this bus
+ * @BUS_NOTIFY_DEL_DEVICE: device is about to be removed from this bus
+ * @BUS_NOTIFY_REMOVED_DEVICE: device is successfully removed from this bus
+ * @BUS_NOTIFY_BIND_DRIVER: a driver is about to be bound to this device on this bus
+ * @BUS_NOTIFY_BOUND_DRIVER: a driver is successfully bound to this device on this bus
+ * @BUS_NOTIFY_UNBIND_DRIVER: a driver is about to be unbound from this device on this bus
+ * @BUS_NOTIFY_UNBOUND_DRIVER: a driver is successfully unbound from this device on this bus
+ * @BUS_NOTIFY_DRIVER_NOT_BOUND: a driver failed to be bound to this device on this bus
+ *
+ * These are the value passed to a bus notifier when a specific event happens.
+ *
+ * Note that bus notifiers are likely to be called with the device lock already
+ * held by the driver core, so be careful in any notifier callback as to what
+ * you do with the device structure.
+ *
+ * All bus notifiers are called with the target struct device * as an argument.
  */
-#define BUS_NOTIFY_ADD_DEVICE		0x00000001 /* device added */
-#define BUS_NOTIFY_DEL_DEVICE		0x00000002 /* device to be removed */
-#define BUS_NOTIFY_REMOVED_DEVICE	0x00000003 /* device removed */
-#define BUS_NOTIFY_BIND_DRIVER		0x00000004 /* driver about to be
-						      bound */
-#define BUS_NOTIFY_BOUND_DRIVER		0x00000005 /* driver bound to device */
-#define BUS_NOTIFY_UNBIND_DRIVER	0x00000006 /* driver about to be
-						      unbound */
-#define BUS_NOTIFY_UNBOUND_DRIVER	0x00000007 /* driver is unbound
-						      from the device */
-#define BUS_NOTIFY_DRIVER_NOT_BOUND	0x00000008 /* driver fails to be bound */
+enum bus_notifier_event {
+	BUS_NOTIFY_ADD_DEVICE,
+	BUS_NOTIFY_DEL_DEVICE,
+	BUS_NOTIFY_REMOVED_DEVICE,
+	BUS_NOTIFY_BIND_DRIVER,
+	BUS_NOTIFY_BOUND_DRIVER,
+	BUS_NOTIFY_UNBIND_DRIVER,
+	BUS_NOTIFY_UNBOUND_DRIVER,
+	BUS_NOTIFY_DRIVER_NOT_BOUND,
+};
 
 extern struct kset *bus_get_kset(struct bus_type *bus);
 
-- 
2.39.0

[PATCH v3 2/2] driver core: bus: move bus notifier logic into bus.c

[Greg Kroah-Hartman]

The logic to touch the bus notifier was open-coded in numberous places
in the driver core.  Clean that up by creating a local bus_notify()
function and have everyone call this function instead, making the
reading of the caller code simpler and easier to maintain over time.

Reviewed-by: Rafael J. Wysocki <rafael@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
v3: - no change
v2: - Added Rafael's reviewed-by

 drivers/base/base.h |  1 +
 drivers/base/bus.c  |  8 ++++++++
 drivers/base/core.c | 13 +++----------
 drivers/base/dd.c   | 28 +++++++---------------------
 4 files changed, 19 insertions(+), 31 deletions(-)

diff --git a/drivers/base/base.h b/drivers/base/base.h
index 7d4803c03d3e..2e08258ce82e 100644
--- a/drivers/base/base.h
+++ b/drivers/base/base.h
@@ -130,6 +130,7 @@ struct kobject *virtual_device_parent(struct device *dev);
 extern int bus_add_device(struct device *dev);
 extern void bus_probe_device(struct device *dev);
 extern void bus_remove_device(struct device *dev);
+void bus_notify(struct device *dev, enum bus_notifier_event value);
 
 extern int bus_add_driver(struct device_driver *drv);
 extern void bus_remove_driver(struct device_driver *drv);
diff --git a/drivers/base/bus.c b/drivers/base/bus.c
index 428c26c6b615..cf1b8f00b4c0 100644
--- a/drivers/base/bus.c
+++ b/drivers/base/bus.c
@@ -850,6 +850,14 @@ int bus_unregister_notifier(struct bus_type *bus, struct notifier_block *nb)
 }
 EXPORT_SYMBOL_GPL(bus_unregister_notifier);
 
+void bus_notify(struct device *dev, enum bus_notifier_event value)
+{
+	struct bus_type *bus = dev->bus;
+
+	if (bus)
+		blocking_notifier_call_chain(&bus->p->bus_notifier, value, dev);
+}
+
 struct kset *bus_get_kset(struct bus_type *bus)
 {
 	return &bus->p->subsys;
diff --git a/drivers/base/core.c b/drivers/base/core.c
index a3e14143ec0c..af6a2761b31d 100644
--- a/drivers/base/core.c
+++ b/drivers/base/core.c
@@ -3453,10 +3453,7 @@ int device_add(struct device *dev)
 	/* Notify clients of device addition.  This call must come
 	 * after dpm_sysfs_add() and before kobject_uevent().
 	 */
-	if (dev->bus)
-		blocking_notifier_call_chain(&dev->bus->p->bus_notifier,
-					     BUS_NOTIFY_ADD_DEVICE, dev);
-
+	bus_notify(dev, BUS_NOTIFY_ADD_DEVICE);
 	kobject_uevent(&dev->kobj, KOBJ_ADD);
 
 	/*
@@ -3636,9 +3633,7 @@ void device_del(struct device *dev)
 	 * before dpm_sysfs_remove().
 	 */
 	noio_flag = memalloc_noio_save();
-	if (dev->bus)
-		blocking_notifier_call_chain(&dev->bus->p->bus_notifier,
-					     BUS_NOTIFY_DEL_DEVICE, dev);
+	bus_notify(dev, BUS_NOTIFY_DEL_DEVICE);
 
 	dpm_sysfs_remove(dev);
 	if (parent)
@@ -3669,9 +3664,7 @@ void device_del(struct device *dev)
 	device_platform_notify_remove(dev);
 	device_links_purge(dev);
 
-	if (dev->bus)
-		blocking_notifier_call_chain(&dev->bus->p->bus_notifier,
-					     BUS_NOTIFY_REMOVED_DEVICE, dev);
+	bus_notify(dev, BUS_NOTIFY_REMOVED_DEVICE);
 	kobject_uevent(&dev->kobj, KOBJ_REMOVE);
 	glue_dir = get_glue_dir(dev);
 	kobject_del(&dev->kobj);
diff --git a/drivers/base/dd.c b/drivers/base/dd.c
index e9b2f9c25efe..a519eaf1990c 100644
--- a/drivers/base/dd.c
+++ b/drivers/base/dd.c
@@ -413,10 +413,7 @@ static void driver_bound(struct device *dev)
 	driver_deferred_probe_del(dev);
 	driver_deferred_probe_trigger();
 
-	if (dev->bus)
-		blocking_notifier_call_chain(&dev->bus->p->bus_notifier,
-					     BUS_NOTIFY_BOUND_DRIVER, dev);
-
+	bus_notify(dev, BUS_NOTIFY_BOUND_DRIVER);
 	kobject_uevent(&dev->kobj, KOBJ_BIND);
 }
 
@@ -435,9 +432,7 @@ static int driver_sysfs_add(struct device *dev)
 {
 	int ret;
 
-	if (dev->bus)
-		blocking_notifier_call_chain(&dev->bus->p->bus_notifier,
-					     BUS_NOTIFY_BIND_DRIVER, dev);
+	bus_notify(dev, BUS_NOTIFY_BIND_DRIVER);
 
 	ret = sysfs_create_link(&dev->driver->p->kobj, &dev->kobj,
 				kobject_name(&dev->kobj));
@@ -502,9 +497,8 @@ int device_bind_driver(struct device *dev)
 		device_links_force_bind(dev);
 		driver_bound(dev);
 	}
-	else if (dev->bus)
-		blocking_notifier_call_chain(&dev->bus->p->bus_notifier,
-					     BUS_NOTIFY_DRIVER_NOT_BOUND, dev);
+	else
+		bus_notify(dev, BUS_NOTIFY_DRIVER_NOT_BOUND);
 	return ret;
 }
 EXPORT_SYMBOL_GPL(device_bind_driver);
@@ -695,9 +689,7 @@ static int really_probe(struct device *dev, struct device_driver *drv)
 probe_failed:
 	driver_sysfs_remove(dev);
 sysfs_failed:
-	if (dev->bus)
-		blocking_notifier_call_chain(&dev->bus->p->bus_notifier,
-					     BUS_NOTIFY_DRIVER_NOT_BOUND, dev);
+	bus_notify(dev, BUS_NOTIFY_DRIVER_NOT_BOUND);
 	if (dev->bus && dev->bus->dma_cleanup)
 		dev->bus->dma_cleanup(dev);
 pinctrl_bind_failed:
@@ -1243,10 +1235,7 @@ static void __device_release_driver(struct device *dev, struct device *parent)
 
 		driver_sysfs_remove(dev);
 
-		if (dev->bus)
-			blocking_notifier_call_chain(&dev->bus->p->bus_notifier,
-						     BUS_NOTIFY_UNBIND_DRIVER,
-						     dev);
+		bus_notify(dev, BUS_NOTIFY_UNBIND_DRIVER);
 
 		pm_runtime_put_sync(dev);
 
@@ -1260,11 +1249,8 @@ static void __device_release_driver(struct device *dev, struct device *parent)
 
 		klist_remove(&dev->p->knode_driver);
 		device_pm_check_callbacks(dev);
-		if (dev->bus)
-			blocking_notifier_call_chain(&dev->bus->p->bus_notifier,
-						     BUS_NOTIFY_UNBOUND_DRIVER,
-						     dev);
 
+		bus_notify(dev, BUS_NOTIFY_UNBOUND_DRIVER);
 		kobject_uevent(&dev->kobj, KOBJ_UNBIND);
 	}
 }
-- 
2.39.0

Re: [PATCH v3 1/2] driver core: bus.h: document bus notifiers better

[Rafael J. Wysocki]

On Wed, Jan 11, 2023 at 10:23 AM Greg Kroah-Hartman
<gregkh@linuxfoundation.org> wrote:
>
> The bus notifier values are not documented all that well, so clean this
> up and make a real enumerated type for them and document them much
> better.  When doing this, remove the hex values and just rely on the
> enumerated type instead as that is all that is needed.
>
> Cc: "Rafael J. Wysocki" <rafael@kernel.org>
> Cc: Randy Dunlap <rdunlap@infradead.org>

I would add a Co-developed-by for Randy along with his sign-off (which
was given IIRC).

Anyway

Reviewed-by: Rafael J. Wysocki <rafael@kernel.org>

> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> ---
> v3: - fix up the documentation to be proper kernel-doc format thanks to
>       Randy.
>     - Remove the values entirely now that this is an enumerated type,
>       they are not needed.
>
> v2: - move the values to decimal from hex as pointed out by Rafael.
>
>  include/linux/device/bus.h | 42 +++++++++++++++++++++++++-------------
>  1 file changed, 28 insertions(+), 14 deletions(-)
>
> diff --git a/include/linux/device/bus.h b/include/linux/device/bus.h
> index d529f644e92b..7b4a48b5159b 100644
> --- a/include/linux/device/bus.h
> +++ b/include/linux/device/bus.h
> @@ -257,21 +257,35 @@ extern int bus_register_notifier(struct bus_type *bus,
>  extern int bus_unregister_notifier(struct bus_type *bus,
>                                    struct notifier_block *nb);
>
> -/* All 4 notifers below get called with the target struct device *
> - * as an argument. Note that those functions are likely to be called
> - * with the device lock held in the core, so be careful.
> +/**
> + * enum bus_notifier_event - Bus Notifier events that have happened
> + * @BUS_NOTIFY_ADD_DEVICE: device is added to this bus
> + * @BUS_NOTIFY_DEL_DEVICE: device is about to be removed from this bus
> + * @BUS_NOTIFY_REMOVED_DEVICE: device is successfully removed from this bus
> + * @BUS_NOTIFY_BIND_DRIVER: a driver is about to be bound to this device on this bus
> + * @BUS_NOTIFY_BOUND_DRIVER: a driver is successfully bound to this device on this bus
> + * @BUS_NOTIFY_UNBIND_DRIVER: a driver is about to be unbound from this device on this bus
> + * @BUS_NOTIFY_UNBOUND_DRIVER: a driver is successfully unbound from this device on this bus
> + * @BUS_NOTIFY_DRIVER_NOT_BOUND: a driver failed to be bound to this device on this bus
> + *
> + * These are the value passed to a bus notifier when a specific event happens.
> + *
> + * Note that bus notifiers are likely to be called with the device lock already
> + * held by the driver core, so be careful in any notifier callback as to what
> + * you do with the device structure.
> + *
> + * All bus notifiers are called with the target struct device * as an argument.
>   */
> -#define BUS_NOTIFY_ADD_DEVICE          0x00000001 /* device added */
> -#define BUS_NOTIFY_DEL_DEVICE          0x00000002 /* device to be removed */
> -#define BUS_NOTIFY_REMOVED_DEVICE      0x00000003 /* device removed */
> -#define BUS_NOTIFY_BIND_DRIVER         0x00000004 /* driver about to be
> -                                                     bound */
> -#define BUS_NOTIFY_BOUND_DRIVER                0x00000005 /* driver bound to device */
> -#define BUS_NOTIFY_UNBIND_DRIVER       0x00000006 /* driver about to be
> -                                                     unbound */
> -#define BUS_NOTIFY_UNBOUND_DRIVER      0x00000007 /* driver is unbound
> -                                                     from the device */
> -#define BUS_NOTIFY_DRIVER_NOT_BOUND    0x00000008 /* driver fails to be bound */
> +enum bus_notifier_event {
> +       BUS_NOTIFY_ADD_DEVICE,
> +       BUS_NOTIFY_DEL_DEVICE,
> +       BUS_NOTIFY_REMOVED_DEVICE,
> +       BUS_NOTIFY_BIND_DRIVER,
> +       BUS_NOTIFY_BOUND_DRIVER,
> +       BUS_NOTIFY_UNBIND_DRIVER,
> +       BUS_NOTIFY_UNBOUND_DRIVER,
> +       BUS_NOTIFY_DRIVER_NOT_BOUND,
> +};
>
>  extern struct kset *bus_get_kset(struct bus_type *bus);
>
> --
> 2.39.0
>

Re: [PATCH v3 1/2] driver core: bus.h: document bus notifiers better

[Randy Dunlap]

On 1/11/23 01:54, Rafael J. Wysocki wrote:
> On Wed, Jan 11, 2023 at 10:23 AM Greg Kroah-Hartman
> <gregkh@linuxfoundation.org> wrote:
>>
>> The bus notifier values are not documented all that well, so clean this
>> up and make a real enumerated type for them and document them much
>> better.  When doing this, remove the hex values and just rely on the
>> enumerated type instead as that is all that is needed.
>>
>> Cc: "Rafael J. Wysocki" <rafael@kernel.org>
>> Cc: Randy Dunlap <rdunlap@infradead.org>
> 
> I would add a Co-developed-by for Randy along with his sign-off (which
> was given IIRC).

or
Reviewed-by: Randy Dunlap <rdunlap@infradead.org>
Either way; doesn't matter.

> Anyway
> 
> Reviewed-by: Rafael J. Wysocki <rafael@kernel.org>
> 
>> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
>> ---
>> v3: - fix up the documentation to be proper kernel-doc format thanks to
>>       Randy.
>>     - Remove the values entirely now that this is an enumerated type,
>>       they are not needed.
>>
>> v2: - move the values to decimal from hex as pointed out by Rafael.
>>
>>  include/linux/device/bus.h | 42 +++++++++++++++++++++++++-------------
>>  1 file changed, 28 insertions(+), 14 deletions(-)
>>
>> diff --git a/include/linux/device/bus.h b/include/linux/device/bus.h
>> index d529f644e92b..7b4a48b5159b 100644
>> --- a/include/linux/device/bus.h
>> +++ b/include/linux/device/bus.h
>> @@ -257,21 +257,35 @@ extern int bus_register_notifier(struct bus_type *bus,
>>  extern int bus_unregister_notifier(struct bus_type *bus,
>>                                    struct notifier_block *nb);
>>
>> -/* All 4 notifers below get called with the target struct device *
>> - * as an argument. Note that those functions are likely to be called
>> - * with the device lock held in the core, so be careful.
>> +/**
>> + * enum bus_notifier_event - Bus Notifier events that have happened
>> + * @BUS_NOTIFY_ADD_DEVICE: device is added to this bus
>> + * @BUS_NOTIFY_DEL_DEVICE: device is about to be removed from this bus
>> + * @BUS_NOTIFY_REMOVED_DEVICE: device is successfully removed from this bus
>> + * @BUS_NOTIFY_BIND_DRIVER: a driver is about to be bound to this device on this bus
>> + * @BUS_NOTIFY_BOUND_DRIVER: a driver is successfully bound to this device on this bus
>> + * @BUS_NOTIFY_UNBIND_DRIVER: a driver is about to be unbound from this device on this bus
>> + * @BUS_NOTIFY_UNBOUND_DRIVER: a driver is successfully unbound from this device on this bus
>> + * @BUS_NOTIFY_DRIVER_NOT_BOUND: a driver failed to be bound to this device on this bus
>> + *
>> + * These are the value passed to a bus notifier when a specific event happens.
>> + *
>> + * Note that bus notifiers are likely to be called with the device lock already
>> + * held by the driver core, so be careful in any notifier callback as to what
>> + * you do with the device structure.
>> + *
>> + * All bus notifiers are called with the target struct device * as an argument.
>>   */
>> -#define BUS_NOTIFY_ADD_DEVICE          0x00000001 /* device added */
>> -#define BUS_NOTIFY_DEL_DEVICE          0x00000002 /* device to be removed */
>> -#define BUS_NOTIFY_REMOVED_DEVICE      0x00000003 /* device removed */
>> -#define BUS_NOTIFY_BIND_DRIVER         0x00000004 /* driver about to be
>> -                                                     bound */
>> -#define BUS_NOTIFY_BOUND_DRIVER                0x00000005 /* driver bound to device */
>> -#define BUS_NOTIFY_UNBIND_DRIVER       0x00000006 /* driver about to be
>> -                                                     unbound */
>> -#define BUS_NOTIFY_UNBOUND_DRIVER      0x00000007 /* driver is unbound
>> -                                                     from the device */
>> -#define BUS_NOTIFY_DRIVER_NOT_BOUND    0x00000008 /* driver fails to be bound */
>> +enum bus_notifier_event {
>> +       BUS_NOTIFY_ADD_DEVICE,
>> +       BUS_NOTIFY_DEL_DEVICE,
>> +       BUS_NOTIFY_REMOVED_DEVICE,
>> +       BUS_NOTIFY_BIND_DRIVER,
>> +       BUS_NOTIFY_BOUND_DRIVER,
>> +       BUS_NOTIFY_UNBIND_DRIVER,
>> +       BUS_NOTIFY_UNBOUND_DRIVER,
>> +       BUS_NOTIFY_DRIVER_NOT_BOUND,
>> +};
>>
>>  extern struct kset *bus_get_kset(struct bus_type *bus);
>>
>> --
>> 2.39.0
>>

-- 
~Randy

Re: [PATCH v3 1/2] driver core: bus.h: document bus notifiers better

[Greg Kroah-Hartman]

On Wed, Jan 11, 2023 at 11:29:12AM -0800, Randy Dunlap wrote:
> 
> 
> On 1/11/23 01:54, Rafael J. Wysocki wrote:
> > On Wed, Jan 11, 2023 at 10:23 AM Greg Kroah-Hartman
> > <gregkh@linuxfoundation.org> wrote:
> >>
> >> The bus notifier values are not documented all that well, so clean this
> >> up and make a real enumerated type for them and document them much
> >> better.  When doing this, remove the hex values and just rely on the
> >> enumerated type instead as that is all that is needed.
> >>
> >> Cc: "Rafael J. Wysocki" <rafael@kernel.org>
> >> Cc: Randy Dunlap <rdunlap@infradead.org>
> > 
> > I would add a Co-developed-by for Randy along with his sign-off (which
> > was given IIRC).
> 
> or
> Reviewed-by: Randy Dunlap <rdunlap@infradead.org>
> Either way; doesn't matter.

Thanks, I'll take it this way I don't have to respin it again :)

It's always the "simple" patches that take the most revisions...

greg k-h