* Re: [PATCH 1/2] acpi: video: add function to support unregister backlight
2013-04-22 12:39 ` [PATCH 1/2] acpi: video: add function to support unregister backlight Chun-Yi Lee
@ 2013-04-22 23:09 ` Rafael J. Wysocki
2013-04-23 4:12 ` Dmitry Torokhov
2013-04-26 5:24 ` Aaron Lu
2 siblings, 0 replies; 10+ messages in thread
From: Rafael J. Wysocki @ 2013-04-22 23:09 UTC (permalink / raw)
To: Chun-Yi Lee, Zhang Rui
Cc: mjg, platform-driver-x86, linux-acpi, linux-kernel, Lee, Chun-Yi,
Carlos Corbacho, Dmitry Torokhov, Corentin Chary, Aaron Lu,
Thomas Renninger
On Monday, April 22, 2013 08:39:15 PM Chun-Yi Lee wrote:
> From: "Lee, Chun-Yi" <jlee@suse.com>
>
> There have situation we unregister whole acpi/video driver by downstream driver
> just want to remove backlight control interface of acpi/video. It caues we lost
> other functions of acpi/video, e.g. transfer acpi event to input event.
>
> So, this patch add a new function, find_video_unregister_backlight, it provide
> the interface let downstream driver can tell acpi/video to unregister backlight
> interface of all acpi video devices. Then we can keep functions of acpi/video
> but only remove backlight support.
>
> Reference: bko#35622
> https://bugzilla.kernel.org/show_bug.cgi?id=35622
Rui, any comments?
Rafael
> Tested-by: Andrzej Krentosz <endrjux@gmail.com>
> Cc: Carlos Corbacho <carlos@strangeworlds.co.uk>
> Cc: Matthew Garrett <mjg@redhat.com>
> Cc: Dmitry Torokhov <dtor@mail.ru>
> Cc: Corentin Chary <corentincj@iksaif.net>
> Cc: Rafael J. Wysocki <rjw@sisk.pl>
> Cc: Aaron Lu <aaron.lu@intel.com>
> Cc: Thomas Renninger <trenn@suse.de>
> Signed-off-by: Lee, Chun-Yi <jlee@suse.com>
> ---
> drivers/acpi/video.c | 46 ++++++++++++++++++++++++++++++++++++++++++++++
> include/acpi/video.h | 2 ++
> 2 files changed, 48 insertions(+)
>
> diff --git a/drivers/acpi/video.c b/drivers/acpi/video.c
> index 313f959..acd2e7a 100644
> --- a/drivers/acpi/video.c
> +++ b/drivers/acpi/video.c
> @@ -1793,6 +1793,52 @@ static int __init intel_opregion_present(void)
> return opregion;
> }
>
> +static acpi_status
> +find_video_unregister_backlight(acpi_handle handle, u32 lvl, void *context,
> + void **rv)
> +{
> + struct acpi_device *acpi_dev;
> + struct acpi_video_bus *video = NULL;
> + struct acpi_video_device *dev, *next;
> +
> + if (acpi_bus_get_device(handle, &acpi_dev))
> + return AE_OK;
> +
> + if (!acpi_match_device_ids(acpi_dev, video_device_ids)) {
> + video = acpi_driver_data(acpi_dev);
> + acpi_video_bus_stop_devices(video);
> + mutex_lock(&video->device_list_lock);
> + list_for_each_entry_safe(dev, next, &video->video_device_list,
> + entry) {
> + if (dev->backlight) {
> + backlight_device_unregister(dev->backlight);
> + dev->backlight = NULL;
> + kfree(dev->brightness->levels);
> + kfree(dev->brightness);
> + }
> + }
> + mutex_unlock(&video->device_list_lock);
> + acpi_video_bus_start_devices(video);
> + }
> + return AE_OK;
> +}
> +
> +void acpi_video_backlight_unregister(void)
> +{
> + if (!register_count) {
> + /*
> + * If the acpi video bus is already unloaded, don't
> + * unregister backlight of devices and return directly.
> + */
> + return;
> + }
> + acpi_walk_namespace(ACPI_TYPE_DEVICE, ACPI_ROOT_OBJECT,
> + ACPI_UINT32_MAX, find_video_unregister_backlight,
> + NULL, NULL, NULL);
> + return;
> +}
> +EXPORT_SYMBOL(acpi_video_backlight_unregister);
> +
> int acpi_video_register(void)
> {
> int result = 0;
> diff --git a/include/acpi/video.h b/include/acpi/video.h
> index 61109f2..1e810a1 100644
> --- a/include/acpi/video.h
> +++ b/include/acpi/video.h
> @@ -19,11 +19,13 @@ struct acpi_device;
> #if (defined CONFIG_ACPI_VIDEO || defined CONFIG_ACPI_VIDEO_MODULE)
> extern int acpi_video_register(void);
> extern void acpi_video_unregister(void);
> +extern void acpi_video_backlight_unregister(void);
> extern int acpi_video_get_edid(struct acpi_device *device, int type,
> int device_id, void **edid);
> #else
> static inline int acpi_video_register(void) { return 0; }
> static inline void acpi_video_unregister(void) { return; }
> +static inline void acpi_video_backlight_unregister(void) { return; }
> static inline int acpi_video_get_edid(struct acpi_device *device, int type,
> int device_id, void **edid)
> {
>
--
I speak only for myself.
Rafael J. Wysocki, Intel Open Source Technology Center.
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH 1/2] acpi: video: add function to support unregister backlight
2013-04-22 12:39 ` [PATCH 1/2] acpi: video: add function to support unregister backlight Chun-Yi Lee
2013-04-22 23:09 ` Rafael J. Wysocki
@ 2013-04-23 4:12 ` Dmitry Torokhov
2013-04-29 9:19 ` joeyli
2013-04-26 5:24 ` Aaron Lu
2 siblings, 1 reply; 10+ messages in thread
From: Dmitry Torokhov @ 2013-04-23 4:12 UTC (permalink / raw)
To: Chun-Yi Lee
Cc: mjg, rjw, platform-driver-x86, linux-acpi, linux-kernel, Lee,
Chun-Yi, Carlos Corbacho, Corentin Chary, Aaron Lu,
Thomas Renninger
On Mon, Apr 22, 2013 at 08:39:15PM +0800, Chun-Yi Lee wrote:
> From: "Lee, Chun-Yi" <jlee@suse.com>
> +static acpi_status
> +find_video_unregister_backlight(acpi_handle handle, u32 lvl, void *context,
> + void **rv)
> +{
> + struct acpi_device *acpi_dev;
> + struct acpi_video_bus *video = NULL;
Gratuitous initialization of local variables prevents compiler from
warning when you using variable uninitialized.
> + struct acpi_video_device *dev, *next;
> +
> + if (acpi_bus_get_device(handle, &acpi_dev))
> + return AE_OK;
> +
> + if (!acpi_match_device_ids(acpi_dev, video_device_ids)) {
> + video = acpi_driver_data(acpi_dev);
> + acpi_video_bus_stop_devices(video);
> + mutex_lock(&video->device_list_lock);
> + list_for_each_entry_safe(dev, next, &video->video_device_list,
> + entry) {
> + if (dev->backlight) {
> + backlight_device_unregister(dev->backlight);
> + dev->backlight = NULL;
> + kfree(dev->brightness->levels);
> + kfree(dev->brightness);
> + }
> + }
> + mutex_unlock(&video->device_list_lock);
> + acpi_video_bus_start_devices(video);
> + }
> + return AE_OK;
> +}
> +
> +void acpi_video_backlight_unregister(void)
> +{
> + if (!register_count) {
Locking? It looks like the rest of the driver ignores locking too...
> + /*
> + * If the acpi video bus is already unloaded, don't
> + * unregister backlight of devices and return directly.
> + */
> + return;
> + }
> + acpi_walk_namespace(ACPI_TYPE_DEVICE, ACPI_ROOT_OBJECT,
> + ACPI_UINT32_MAX, find_video_unregister_backlight,
> + NULL, NULL, NULL);
> + return;
> +}
> +EXPORT_SYMBOL(acpi_video_backlight_unregister);
> +
> int acpi_video_register(void)
> {
> int result = 0;
> diff --git a/include/acpi/video.h b/include/acpi/video.h
> index 61109f2..1e810a1 100644
> --- a/include/acpi/video.h
> +++ b/include/acpi/video.h
> @@ -19,11 +19,13 @@ struct acpi_device;
> #if (defined CONFIG_ACPI_VIDEO || defined CONFIG_ACPI_VIDEO_MODULE)
> extern int acpi_video_register(void);
> extern void acpi_video_unregister(void);
> +extern void acpi_video_backlight_unregister(void);
> extern int acpi_video_get_edid(struct acpi_device *device, int type,
> int device_id, void **edid);
> #else
> static inline int acpi_video_register(void) { return 0; }
> static inline void acpi_video_unregister(void) { return; }
> +static inline void acpi_video_backlight_unregister(void) { return; }
> static inline int acpi_video_get_edid(struct acpi_device *device, int type,
> int device_id, void **edid)
> {
> --
> 1.8.1.4
>
>
Thanks.
--
Dmitry
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH 1/2] acpi: video: add function to support unregister backlight
2013-04-23 4:12 ` Dmitry Torokhov
@ 2013-04-29 9:19 ` joeyli
0 siblings, 0 replies; 10+ messages in thread
From: joeyli @ 2013-04-29 9:19 UTC (permalink / raw)
To: Dmitry Torokhov
Cc: mjg, rjw, platform-driver-x86, linux-acpi, linux-kernel,
Carlos Corbacho, Corentin Chary, Aaron Lu, Thomas Renninger
Hi Dmitry,
Thanks for your review and suggestions first!
於 一,2013-04-22 於 21:12 -0700,Dmitry Torokhov 提到:
> On Mon, Apr 22, 2013 at 08:39:15PM +0800, Chun-Yi Lee wrote:
> > From: "Lee, Chun-Yi" <jlee@suse.com>
> > +static acpi_status
> > +find_video_unregister_backlight(acpi_handle handle, u32 lvl, void *context,
> > + void **rv)
> > +{
> > + struct acpi_device *acpi_dev;
> > + struct acpi_video_bus *video = NULL;
>
> Gratuitous initialization of local variables prevents compiler from
> warning when you using variable uninitialized.
Yes, I agreed should not put gratuitous initialization, I will remove it
in v2 patch.
>
> > + struct acpi_video_device *dev, *next;
> > +
> > + if (acpi_bus_get_device(handle, &acpi_dev))
> > + return AE_OK;
> > +
> > + if (!acpi_match_device_ids(acpi_dev, video_device_ids)) {
> > + video = acpi_driver_data(acpi_dev);
> > + acpi_video_bus_stop_devices(video);
> > + mutex_lock(&video->device_list_lock);
> > + list_for_each_entry_safe(dev, next, &video->video_device_list,
> > + entry) {
> > + if (dev->backlight) {
> > + backlight_device_unregister(dev->backlight);
> > + dev->backlight = NULL;
> > + kfree(dev->brightness->levels);
> > + kfree(dev->brightness);
> > + }
> > + }
> > + mutex_unlock(&video->device_list_lock);
> > + acpi_video_bus_start_devices(video);
> > + }
> > + return AE_OK;
> > +}
> > +
> > +void acpi_video_backlight_unregister(void)
> > +{
> > + if (!register_count) {
>
> Locking? It looks like the rest of the driver ignores locking too...
Did you mean locking video_device_list?
The acpi/video locks video_device_list when add, remove and notify acpi
video bus driver. It always do the mutex lock before control
video_device_list, so I also add lock when unregister all backlight of
devices.
Thanks a lot!
Joey Lee
--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH 1/2] acpi: video: add function to support unregister backlight
@ 2013-04-29 9:19 ` joeyli
0 siblings, 0 replies; 10+ messages in thread
From: joeyli @ 2013-04-29 9:19 UTC (permalink / raw)
To: Dmitry Torokhov
Cc: mjg, rjw, platform-driver-x86, linux-acpi, linux-kernel,
Carlos Corbacho, Corentin Chary, Aaron Lu, Thomas Renninger
Hi Dmitry,
Thanks for your review and suggestions first!
於 一,2013-04-22 於 21:12 -0700,Dmitry Torokhov 提到:
> On Mon, Apr 22, 2013 at 08:39:15PM +0800, Chun-Yi Lee wrote:
> > From: "Lee, Chun-Yi" <jlee@suse.com>
> > +static acpi_status
> > +find_video_unregister_backlight(acpi_handle handle, u32 lvl, void *context,
> > + void **rv)
> > +{
> > + struct acpi_device *acpi_dev;
> > + struct acpi_video_bus *video = NULL;
>
> Gratuitous initialization of local variables prevents compiler from
> warning when you using variable uninitialized.
Yes, I agreed should not put gratuitous initialization, I will remove it
in v2 patch.
>
> > + struct acpi_video_device *dev, *next;
> > +
> > + if (acpi_bus_get_device(handle, &acpi_dev))
> > + return AE_OK;
> > +
> > + if (!acpi_match_device_ids(acpi_dev, video_device_ids)) {
> > + video = acpi_driver_data(acpi_dev);
> > + acpi_video_bus_stop_devices(video);
> > + mutex_lock(&video->device_list_lock);
> > + list_for_each_entry_safe(dev, next, &video->video_device_list,
> > + entry) {
> > + if (dev->backlight) {
> > + backlight_device_unregister(dev->backlight);
> > + dev->backlight = NULL;
> > + kfree(dev->brightness->levels);
> > + kfree(dev->brightness);
> > + }
> > + }
> > + mutex_unlock(&video->device_list_lock);
> > + acpi_video_bus_start_devices(video);
> > + }
> > + return AE_OK;
> > +}
> > +
> > +void acpi_video_backlight_unregister(void)
> > +{
> > + if (!register_count) {
>
> Locking? It looks like the rest of the driver ignores locking too...
Did you mean locking video_device_list?
The acpi/video locks video_device_list when add, remove and notify acpi
video bus driver. It always do the mutex lock before control
video_device_list, so I also add lock when unregister all backlight of
devices.
Thanks a lot!
Joey Lee
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH 1/2] acpi: video: add function to support unregister backlight
2013-04-22 12:39 ` [PATCH 1/2] acpi: video: add function to support unregister backlight Chun-Yi Lee
2013-04-22 23:09 ` Rafael J. Wysocki
2013-04-23 4:12 ` Dmitry Torokhov
@ 2013-04-26 5:24 ` Aaron Lu
2013-04-29 10:06 ` joeyli
2 siblings, 1 reply; 10+ messages in thread
From: Aaron Lu @ 2013-04-26 5:24 UTC (permalink / raw)
To: Chun-Yi Lee
Cc: mjg, rjw, platform-driver-x86, linux-acpi, linux-kernel, Lee,
Chun-Yi, Carlos Corbacho, Dmitry Torokhov, Corentin Chary,
Thomas Renninger
On 04/22/2013 08:39 PM, Chun-Yi Lee wrote:
> From: "Lee, Chun-Yi" <jlee@suse.com>
>
> There have situation we unregister whole acpi/video driver by downstream driver
> just want to remove backlight control interface of acpi/video. It caues we lost
> other functions of acpi/video, e.g. transfer acpi event to input event.
>
> So, this patch add a new function, find_video_unregister_backlight, it provide
> the interface let downstream driver can tell acpi/video to unregister backlight
> interface of all acpi video devices. Then we can keep functions of acpi/video
> but only remove backlight support.
>
> Reference: bko#35622
> https://bugzilla.kernel.org/show_bug.cgi?id=35622
>
> Tested-by: Andrzej Krentosz <endrjux@gmail.com>
> Cc: Carlos Corbacho <carlos@strangeworlds.co.uk>
> Cc: Matthew Garrett <mjg@redhat.com>
> Cc: Dmitry Torokhov <dtor@mail.ru>
> Cc: Corentin Chary <corentincj@iksaif.net>
> Cc: Rafael J. Wysocki <rjw@sisk.pl>
> Cc: Aaron Lu <aaron.lu@intel.com>
> Cc: Thomas Renninger <trenn@suse.de>
> Signed-off-by: Lee, Chun-Yi <jlee@suse.com>
> ---
> drivers/acpi/video.c | 46 ++++++++++++++++++++++++++++++++++++++++++++++
> include/acpi/video.h | 2 ++
> 2 files changed, 48 insertions(+)
>
> diff --git a/drivers/acpi/video.c b/drivers/acpi/video.c
> index 313f959..acd2e7a 100644
> --- a/drivers/acpi/video.c
> +++ b/drivers/acpi/video.c
> @@ -1793,6 +1793,52 @@ static int __init intel_opregion_present(void)
> return opregion;
> }
>
> +static acpi_status
> +find_video_unregister_backlight(acpi_handle handle, u32 lvl, void *context,
> + void **rv)
> +{
> + struct acpi_device *acpi_dev;
> + struct acpi_video_bus *video = NULL;
> + struct acpi_video_device *dev, *next;
> +
> + if (acpi_bus_get_device(handle, &acpi_dev))
> + return AE_OK;
> +
> + if (!acpi_match_device_ids(acpi_dev, video_device_ids)) {
> + video = acpi_driver_data(acpi_dev);
> + acpi_video_bus_stop_devices(video);
> + mutex_lock(&video->device_list_lock);
> + list_for_each_entry_safe(dev, next, &video->video_device_list,
> + entry) {
> + if (dev->backlight) {
> + backlight_device_unregister(dev->backlight);
> + dev->backlight = NULL;
> + kfree(dev->brightness->levels);
> + kfree(dev->brightness);
> + }
The cooling_dev should also be unregistered I think.
Thanks,
Aaron
> + }
> + mutex_unlock(&video->device_list_lock);
> + acpi_video_bus_start_devices(video);
> + }
> + return AE_OK;
> +}
> +
> +void acpi_video_backlight_unregister(void)
> +{
> + if (!register_count) {
> + /*
> + * If the acpi video bus is already unloaded, don't
> + * unregister backlight of devices and return directly.
> + */
> + return;
> + }
> + acpi_walk_namespace(ACPI_TYPE_DEVICE, ACPI_ROOT_OBJECT,
> + ACPI_UINT32_MAX, find_video_unregister_backlight,
> + NULL, NULL, NULL);
> + return;
> +}
> +EXPORT_SYMBOL(acpi_video_backlight_unregister);
> +
> int acpi_video_register(void)
> {
> int result = 0;
> diff --git a/include/acpi/video.h b/include/acpi/video.h
> index 61109f2..1e810a1 100644
> --- a/include/acpi/video.h
> +++ b/include/acpi/video.h
> @@ -19,11 +19,13 @@ struct acpi_device;
> #if (defined CONFIG_ACPI_VIDEO || defined CONFIG_ACPI_VIDEO_MODULE)
> extern int acpi_video_register(void);
> extern void acpi_video_unregister(void);
> +extern void acpi_video_backlight_unregister(void);
> extern int acpi_video_get_edid(struct acpi_device *device, int type,
> int device_id, void **edid);
> #else
> static inline int acpi_video_register(void) { return 0; }
> static inline void acpi_video_unregister(void) { return; }
> +static inline void acpi_video_backlight_unregister(void) { return; }
> static inline int acpi_video_get_edid(struct acpi_device *device, int type,
> int device_id, void **edid)
> {
>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH 1/2] acpi: video: add function to support unregister backlight
2013-04-26 5:24 ` Aaron Lu
@ 2013-04-29 10:06 ` joeyli
0 siblings, 0 replies; 10+ messages in thread
From: joeyli @ 2013-04-29 10:06 UTC (permalink / raw)
To: Aaron Lu
Cc: mjg, rjw, platform-driver-x86, linux-acpi, linux-kernel,
Carlos Corbacho, Dmitry Torokhov, Corentin Chary,
Thomas Renninger
於 五,2013-04-26 於 13:24 +0800,Aaron Lu 提到:
> On 04/22/2013 08:39 PM, Chun-Yi Lee wrote:
> > From: "Lee, Chun-Yi" <jlee@suse.com>
> >
> > There have situation we unregister whole acpi/video driver by downstream driver
> > just want to remove backlight control interface of acpi/video. It caues we lost
> > other functions of acpi/video, e.g. transfer acpi event to input event.
> >
> > So, this patch add a new function, find_video_unregister_backlight, it provide
> > the interface let downstream driver can tell acpi/video to unregister backlight
> > interface of all acpi video devices. Then we can keep functions of acpi/video
> > but only remove backlight support.
> >
> > Reference: bko#35622
> > https://bugzilla.kernel.org/show_bug.cgi?id=35622
> >
> > Tested-by: Andrzej Krentosz <endrjux@gmail.com>
> > Cc: Carlos Corbacho <carlos@strangeworlds.co.uk>
> > Cc: Matthew Garrett <mjg@redhat.com>
> > Cc: Dmitry Torokhov <dtor@mail.ru>
> > Cc: Corentin Chary <corentincj@iksaif.net>
> > Cc: Rafael J. Wysocki <rjw@sisk.pl>
> > Cc: Aaron Lu <aaron.lu@intel.com>
> > Cc: Thomas Renninger <trenn@suse.de>
> > Signed-off-by: Lee, Chun-Yi <jlee@suse.com>
> > ---
> > drivers/acpi/video.c | 46 ++++++++++++++++++++++++++++++++++++++++++++++
> > include/acpi/video.h | 2 ++
> > 2 files changed, 48 insertions(+)
> >
> > diff --git a/drivers/acpi/video.c b/drivers/acpi/video.c
> > index 313f959..acd2e7a 100644
> > --- a/drivers/acpi/video.c
> > +++ b/drivers/acpi/video.c
> > @@ -1793,6 +1793,52 @@ static int __init intel_opregion_present(void)
> > return opregion;
> > }
> >
> > +static acpi_status
> > +find_video_unregister_backlight(acpi_handle handle, u32 lvl, void *context,
> > + void **rv)
> > +{
> > + struct acpi_device *acpi_dev;
> > + struct acpi_video_bus *video = NULL;
> > + struct acpi_video_device *dev, *next;
> > +
> > + if (acpi_bus_get_device(handle, &acpi_dev))
> > + return AE_OK;
> > +
> > + if (!acpi_match_device_ids(acpi_dev, video_device_ids)) {
> > + video = acpi_driver_data(acpi_dev);
> > + acpi_video_bus_stop_devices(video);
> > + mutex_lock(&video->device_list_lock);
> > + list_for_each_entry_safe(dev, next, &video->video_device_list,
> > + entry) {
> > + if (dev->backlight) {
> > + backlight_device_unregister(dev->backlight);
> > + dev->backlight = NULL;
> > + kfree(dev->brightness->levels);
> > + kfree(dev->brightness);
> > + }
>
> The cooling_dev should also be unregistered I think.
>
> Thanks,
> Aaron
>
Yes, you are right!
Unregistering backlight interface of video devices means the backlight
control functions broken, so it can not used for throttling.
I will unregister cooling_dev in v2 patch.
Thanks a lot!
Joey Lee
^ permalink raw reply [flat|nested] 10+ messages in thread