* [PATCH/RFC v5 01/57] leds: Add brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-13 14:15 ` Andrew Lunn
2015-08-11 9:37 ` [PATCH/RFC v5 02/57] leds: Add LED_BRIGHTNESS_BLOCKING flag Jacek Anaszewski
` (55 subsequent siblings)
56 siblings, 1 reply; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski,
Andrew Lunn, Sakari Ailus, Pavel Machek
This patch adds a new brightness_set_nonblocking op to the LED subsystem.
The op is intended for drivers that set brightness in a non-blocking way,
i.e. they neither sleep nor use delays while setting brightness.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Bryan Wu <cooloney@gmail.com>
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: Sakari Ailus <sakari.ailus@linux.intel.com>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Stas Sergeev <stsp@users.sourceforge.net>
---
include/linux/leds.h | 3 +++
1 file changed, 3 insertions(+)
diff --git a/include/linux/leds.h b/include/linux/leds.h
index b122eea..c32f1b8 100644
--- a/include/linux/leds.h
+++ b/include/linux/leds.h
@@ -53,6 +53,9 @@ struct led_classdev {
/* Must not sleep, use a workqueue if needed */
void (*brightness_set)(struct led_classdev *led_cdev,
enum led_brightness brightness);
+ /* Intended for drivers that set brightness in a non-blocking way */
+ void (*brightness_set_nonblocking)(struct led_classdev *led_cdev,
+ enum led_brightness brightness);
/*
* Set LED brightness level immediately - it can block the caller for
* the time required for accessing a LED device register.
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* Re: [PATCH/RFC v5 01/57] leds: Add brightness_set_nonblocking op
2015-08-11 9:37 ` [PATCH/RFC v5 01/57] leds: Add brightness_set_nonblocking op Jacek Anaszewski
@ 2015-08-13 14:15 ` Andrew Lunn
2015-08-13 14:23 ` Pavel Machek
2015-08-14 7:57 ` Jacek Anaszewski
0 siblings, 2 replies; 72+ messages in thread
From: Andrew Lunn @ 2015-08-13 14:15 UTC (permalink / raw)
To: Jacek Anaszewski
Cc: linux-leds, linux-kernel, cooloney, rpurdie, stsp, Sakari Ailus,
Pavel Machek
On Tue, Aug 11, 2015 at 11:37:14AM +0200, Jacek Anaszewski wrote:
> This patch adds a new brightness_set_nonblocking op to the LED subsystem.
> The op is intended for drivers that set brightness in a non-blocking way,
> i.e. they neither sleep nor use delays while setting brightness.
>
> Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
> Cc: Bryan Wu <cooloney@gmail.com>
> Cc: Andrew Lunn <andrew@lunn.ch>
> Cc: Sakari Ailus <sakari.ailus@linux.intel.com>
> Cc: Pavel Machek <pavel@ucw.cz>
> Cc: Stas Sergeev <stsp@users.sourceforge.net>
> ---
> include/linux/leds.h | 3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/include/linux/leds.h b/include/linux/leds.h
> index b122eea..c32f1b8 100644
> --- a/include/linux/leds.h
> +++ b/include/linux/leds.h
> @@ -53,6 +53,9 @@ struct led_classdev {
> /* Must not sleep, use a workqueue if needed */
> void (*brightness_set)(struct led_classdev *led_cdev,
> enum led_brightness brightness);
> + /* Intended for drivers that set brightness in a non-blocking way */
> + void (*brightness_set_nonblocking)(struct led_classdev *led_cdev,
> + enum led_brightness brightness);
Hi Jacek
>From an API design point of view, i'm not sure this is the best way to
go. You now have two calls which do the same thing, with the plan that
you want to invert the meaning of brightness_set, the old well known
API call, sometime later. This inverting the meaning is going to catch
people out and introduce bugs.
I would rather add a brightness_set_blocking op. Then as you go
thought the drivers stripping out the work queue, move the driver to
use this brightness_set_blocking.
Andrew
^ permalink raw reply [flat|nested] 72+ messages in thread* Re: [PATCH/RFC v5 01/57] leds: Add brightness_set_nonblocking op
2015-08-13 14:15 ` Andrew Lunn
@ 2015-08-13 14:23 ` Pavel Machek
2015-08-14 7:57 ` Jacek Anaszewski
1 sibling, 0 replies; 72+ messages in thread
From: Pavel Machek @ 2015-08-13 14:23 UTC (permalink / raw)
To: Andrew Lunn
Cc: Jacek Anaszewski, linux-leds, linux-kernel, cooloney, rpurdie,
stsp, Sakari Ailus
On Thu 2015-08-13 16:15:47, Andrew Lunn wrote:
> On Tue, Aug 11, 2015 at 11:37:14AM +0200, Jacek Anaszewski wrote:
> > This patch adds a new brightness_set_nonblocking op to the LED subsystem.
> > The op is intended for drivers that set brightness in a non-blocking way,
> > i.e. they neither sleep nor use delays while setting brightness.
> >
> > Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
> > Cc: Bryan Wu <cooloney@gmail.com>
> > Cc: Andrew Lunn <andrew@lunn.ch>
> > Cc: Sakari Ailus <sakari.ailus@linux.intel.com>
> > Cc: Pavel Machek <pavel@ucw.cz>
> > Cc: Stas Sergeev <stsp@users.sourceforge.net>
> > ---
> > include/linux/leds.h | 3 +++
> > 1 file changed, 3 insertions(+)
> >
> > diff --git a/include/linux/leds.h b/include/linux/leds.h
> > index b122eea..c32f1b8 100644
> > --- a/include/linux/leds.h
> > +++ b/include/linux/leds.h
> > @@ -53,6 +53,9 @@ struct led_classdev {
> > /* Must not sleep, use a workqueue if needed */
> > void (*brightness_set)(struct led_classdev *led_cdev,
> > enum led_brightness brightness);
> > + /* Intended for drivers that set brightness in a non-blocking way */
> > + void (*brightness_set_nonblocking)(struct led_classdev *led_cdev,
> > + enum led_brightness brightness);
>
> Hi Jacek
>
> >From an API design point of view, i'm not sure this is the best way to
> go. You now have two calls which do the same thing, with the plan that
> you want to invert the meaning of brightness_set, the old well known
> API call, sometime later. This inverting the meaning is going to catch
> people out and introduce bugs.
Yes please.
Pavel
--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
^ permalink raw reply [flat|nested] 72+ messages in thread* Re: [PATCH/RFC v5 01/57] leds: Add brightness_set_nonblocking op
2015-08-13 14:15 ` Andrew Lunn
2015-08-13 14:23 ` Pavel Machek
@ 2015-08-14 7:57 ` Jacek Anaszewski
2015-08-19 14:03 ` Jacek Anaszewski
1 sibling, 1 reply; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-14 7:57 UTC (permalink / raw)
To: Andrew Lunn
Cc: linux-leds, linux-kernel, cooloney, rpurdie, stsp, Sakari Ailus,
Pavel Machek
H Andrew,
On 08/13/2015 04:15 PM, Andrew Lunn wrote:
> On Tue, Aug 11, 2015 at 11:37:14AM +0200, Jacek Anaszewski wrote:
>> This patch adds a new brightness_set_nonblocking op to the LED subsystem.
>> The op is intended for drivers that set brightness in a non-blocking way,
>> i.e. they neither sleep nor use delays while setting brightness.
>>
>> Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
>> Cc: Bryan Wu <cooloney@gmail.com>
>> Cc: Andrew Lunn <andrew@lunn.ch>
>> Cc: Sakari Ailus <sakari.ailus@linux.intel.com>
>> Cc: Pavel Machek <pavel@ucw.cz>
>> Cc: Stas Sergeev <stsp@users.sourceforge.net>
>> ---
>> include/linux/leds.h | 3 +++
>> 1 file changed, 3 insertions(+)
>>
>> diff --git a/include/linux/leds.h b/include/linux/leds.h
>> index b122eea..c32f1b8 100644
>> --- a/include/linux/leds.h
>> +++ b/include/linux/leds.h
>> @@ -53,6 +53,9 @@ struct led_classdev {
>> /* Must not sleep, use a workqueue if needed */
>> void (*brightness_set)(struct led_classdev *led_cdev,
>> enum led_brightness brightness);
>> + /* Intended for drivers that set brightness in a non-blocking way */
>> + void (*brightness_set_nonblocking)(struct led_classdev *led_cdev,
>> + enum led_brightness brightness);
>
> Hi Jacek
>
>>From an API design point of view, i'm not sure this is the best way to
> go. You now have two calls which do the same thing, with the plan that
> you want to invert the meaning of brightness_set, the old well known
> API call, sometime later. This inverting the meaning is going to catch
> people out and introduce bugs.
>
> I would rather add a brightness_set_blocking op. Then as you go
> thought the drivers stripping out the work queue, move the driver to
> use this brightness_set_blocking.
There are around 60 drivers in the other kernel subsystems that register
LED class devices. If we chose the way you proposed then we would have
to adjust all of them to the LED core changes, which could complicate
the situation during merge window if there were other modifications in
the affected drivers.
With my approach all old-fashion drivers will be treated by the
LED core in the old-fashion way. This is why I introduced
LED_BRIGHTNESS_BLOCKING flag - only drivers aware of the LED core
changes will set it, which will allow for the LED core to find out
which drivers are aware of brightness_set semantics change and
implement the op accordingly.
This way the change of brightness_set semantics will not be painful.
After all drivers across all subsystems are adapted to the LED core
changes the flag will be made redundant and we will remove it.
--
Best Regards,
Jacek Anaszewski
^ permalink raw reply [flat|nested] 72+ messages in thread* Re: [PATCH/RFC v5 01/57] leds: Add brightness_set_nonblocking op
2015-08-14 7:57 ` Jacek Anaszewski
@ 2015-08-19 14:03 ` Jacek Anaszewski
2015-08-19 14:09 ` Andrew Lunn
0 siblings, 1 reply; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-19 14:03 UTC (permalink / raw)
To: Andrew Lunn
Cc: linux-leds, linux-kernel, cooloney, rpurdie, stsp, Sakari Ailus,
Pavel Machek
On 08/14/2015 09:57 AM, Jacek Anaszewski wrote:
> H Andrew,
>
> On 08/13/2015 04:15 PM, Andrew Lunn wrote:
>> On Tue, Aug 11, 2015 at 11:37:14AM +0200, Jacek Anaszewski wrote:
>>> This patch adds a new brightness_set_nonblocking op to the LED
>>> subsystem.
>>> The op is intended for drivers that set brightness in a non-blocking
>>> way,
>>> i.e. they neither sleep nor use delays while setting brightness.
>>>
>>> Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
>>> Cc: Bryan Wu <cooloney@gmail.com>
>>> Cc: Andrew Lunn <andrew@lunn.ch>
>>> Cc: Sakari Ailus <sakari.ailus@linux.intel.com>
>>> Cc: Pavel Machek <pavel@ucw.cz>
>>> Cc: Stas Sergeev <stsp@users.sourceforge.net>
>>> ---
>>> include/linux/leds.h | 3 +++
>>> 1 file changed, 3 insertions(+)
>>>
>>> diff --git a/include/linux/leds.h b/include/linux/leds.h
>>> index b122eea..c32f1b8 100644
>>> --- a/include/linux/leds.h
>>> +++ b/include/linux/leds.h
>>> @@ -53,6 +53,9 @@ struct led_classdev {
>>> /* Must not sleep, use a workqueue if needed */
>>> void (*brightness_set)(struct led_classdev *led_cdev,
>>> enum led_brightness brightness);
>>> + /* Intended for drivers that set brightness in a non-blocking
>>> way */
>>> + void (*brightness_set_nonblocking)(struct led_classdev *led_cdev,
>>> + enum led_brightness brightness);
>>
>> Hi Jacek
>>
>>> From an API design point of view, i'm not sure this is the best way to
>> go. You now have two calls which do the same thing, with the plan that
>> you want to invert the meaning of brightness_set, the old well known
>> API call, sometime later. This inverting the meaning is going to catch
>> people out and introduce bugs.
>>
>> I would rather add a brightness_set_blocking op. Then as you go
>> thought the drivers stripping out the work queue, move the driver to
>> use this brightness_set_blocking.
>
> There are around 60 drivers in the other kernel subsystems that register
> LED class devices. If we chose the way you proposed then we would have
> to adjust all of them to the LED core changes, which could complicate
> the situation during merge window if there were other modifications in
> the affected drivers.
>
> With my approach all old-fashion drivers will be treated by the
> LED core in the old-fashion way. This is why I introduced
> LED_BRIGHTNESS_BLOCKING flag - only drivers aware of the LED core
> changes will set it, which will allow for the LED core to find out
> which drivers are aware of brightness_set semantics change and
> implement the op accordingly.
>
> This way the change of brightness_set semantics will not be painful.
>
> After all drivers across all subsystems are adapted to the LED core
> changes the flag will be made redundant and we will remove it.
>
Hi Andrew, Pavel and others,
Do you have any comments to my explanation? It would be great if
we could merge this during upcoming merge window, but I'd like to
obtain some acks for the LED core part beforehand.
--
Best Regards,
Jacek Anaszewski
^ permalink raw reply [flat|nested] 72+ messages in thread* Re: [PATCH/RFC v5 01/57] leds: Add brightness_set_nonblocking op
2015-08-19 14:03 ` Jacek Anaszewski
@ 2015-08-19 14:09 ` Andrew Lunn
2015-08-19 19:31 ` Pavel Machek
0 siblings, 1 reply; 72+ messages in thread
From: Andrew Lunn @ 2015-08-19 14:09 UTC (permalink / raw)
To: Jacek Anaszewski
Cc: linux-leds, linux-kernel, cooloney, rpurdie, stsp, Sakari Ailus,
Pavel Machek
On Wed, Aug 19, 2015 at 04:03:19PM +0200, Jacek Anaszewski wrote:
> On 08/14/2015 09:57 AM, Jacek Anaszewski wrote:
> >H Andrew,
> >
> >On 08/13/2015 04:15 PM, Andrew Lunn wrote:
> >>On Tue, Aug 11, 2015 at 11:37:14AM +0200, Jacek Anaszewski wrote:
> >>>This patch adds a new brightness_set_nonblocking op to the LED
> >>>subsystem.
> >>>The op is intended for drivers that set brightness in a non-blocking
> >>>way,
> >>>i.e. they neither sleep nor use delays while setting brightness.
> >>>
> >>>Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
> >>>Cc: Bryan Wu <cooloney@gmail.com>
> >>>Cc: Andrew Lunn <andrew@lunn.ch>
> >>>Cc: Sakari Ailus <sakari.ailus@linux.intel.com>
> >>>Cc: Pavel Machek <pavel@ucw.cz>
> >>>Cc: Stas Sergeev <stsp@users.sourceforge.net>
> >>>---
> >>> include/linux/leds.h | 3 +++
> >>> 1 file changed, 3 insertions(+)
> >>>
> >>>diff --git a/include/linux/leds.h b/include/linux/leds.h
> >>>index b122eea..c32f1b8 100644
> >>>--- a/include/linux/leds.h
> >>>+++ b/include/linux/leds.h
> >>>@@ -53,6 +53,9 @@ struct led_classdev {
> >>> /* Must not sleep, use a workqueue if needed */
> >>> void (*brightness_set)(struct led_classdev *led_cdev,
> >>> enum led_brightness brightness);
> >>>+ /* Intended for drivers that set brightness in a non-blocking
> >>>way */
> >>>+ void (*brightness_set_nonblocking)(struct led_classdev *led_cdev,
> >>>+ enum led_brightness brightness);
> >>
> >>Hi Jacek
> >>
> >>>From an API design point of view, i'm not sure this is the best way to
> >>go. You now have two calls which do the same thing, with the plan that
> >>you want to invert the meaning of brightness_set, the old well known
> >>API call, sometime later. This inverting the meaning is going to catch
> >>people out and introduce bugs.
> >>
> >>I would rather add a brightness_set_blocking op. Then as you go
> >>thought the drivers stripping out the work queue, move the driver to
> >>use this brightness_set_blocking.
> >
> >There are around 60 drivers in the other kernel subsystems that register
> >LED class devices. If we chose the way you proposed then we would have
> >to adjust all of them to the LED core changes, which could complicate
> >the situation during merge window if there were other modifications in
> >the affected drivers.
You don't need to change anything, if the semantics of
brightness_set() does not change. All current drivers don't
sleep. They use a work queue if needed to ensure they don't
sleep. Hence they are correct.
By adding a new operation, brightness_set_blocking(), you can strip
out this work queue and move to the new op member one driver at a
time. And you can take as long as you want doing this. No flag day
when an API suddenly means something totally different.
Andrew
^ permalink raw reply [flat|nested] 72+ messages in thread* Re: [PATCH/RFC v5 01/57] leds: Add brightness_set_nonblocking op
2015-08-19 14:09 ` Andrew Lunn
@ 2015-08-19 19:31 ` Pavel Machek
0 siblings, 0 replies; 72+ messages in thread
From: Pavel Machek @ 2015-08-19 19:31 UTC (permalink / raw)
To: Andrew Lunn
Cc: Jacek Anaszewski, linux-leds, linux-kernel, cooloney, rpurdie,
stsp, Sakari Ailus
> > >There are around 60 drivers in the other kernel subsystems that register
> > >LED class devices. If we chose the way you proposed then we would have
> > >to adjust all of them to the LED core changes, which could complicate
> > >the situation during merge window if there were other modifications in
> > >the affected drivers.
>
> You don't need to change anything, if the semantics of
> brightness_set() does not change. All current drivers don't
> sleep. They use a work queue if needed to ensure they don't
> sleep. Hence they are correct.
Exactly. As you explain above, there are 60 reasons not to change
existing semantics.
> By adding a new operation, brightness_set_blocking(), you can strip
> out this work queue and move to the new op member one driver at a
> time. And you can take as long as you want doing this. No flag day
> when an API suddenly means something totally different.
Yes please.
Pavel
--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
^ permalink raw reply [flat|nested] 72+ messages in thread
* [PATCH/RFC v5 02/57] leds: Add LED_BRIGHTNESS_BLOCKING flag
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 01/57] leds: Add brightness_set_nonblocking op Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-13 14:05 ` Andrew Lunn
2015-08-11 9:37 ` [PATCH/RFC v5 03/57] leds: Add led_set_brightness_sync to the public LED subsystem API Jacek Anaszewski
` (54 subsequent siblings)
56 siblings, 1 reply; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski,
Andrew Lunn, Sakari Ailus, Pavel Machek
This patch adds LED_BRIGHTNESS_BLOCKING flag to the LED core to make it
capable of discerning between drivers that implement brightness_set op
in the old manner, i.e. the ones which use work queue internally, and
drivers that implement brightness_set op in the new manner, i.e. they
are aware that brightness_set op will be called from a work queue task
by the LED core.
In the new approach all drivers that implement brightness_set op should
also set LED_BRIGHTNESS_BLOCKING flag.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Bryan Wu <cooloney@gmail.com>
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: Sakari Ailus <sakari.ailus@linux.intel.com>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Stas Sergeev <stsp@users.sourceforge.net>
---
drivers/leds/led-class.c | 3 +++
include/linux/leds.h | 1 +
2 files changed, 4 insertions(+)
diff --git a/drivers/leds/led-class.c b/drivers/leds/led-class.c
index ca51d58..2875e6a 100644
--- a/drivers/leds/led-class.c
+++ b/drivers/leds/led-class.c
@@ -279,6 +279,9 @@ int led_classdev_register(struct device *parent, struct led_classdev *led_cdev)
if (!led_cdev->max_brightness)
led_cdev->max_brightness = LED_FULL;
+ WARN_ON((led_cdev->flags & LED_BRIGHTNESS_BLOCKING) &&
+ !led_cdev->brightness_set);
+
led_cdev->flags |= SET_BRIGHTNESS_ASYNC;
led_update_brightness(led_cdev);
diff --git a/include/linux/leds.h b/include/linux/leds.h
index c32f1b8..ac9efe1 100644
--- a/include/linux/leds.h
+++ b/include/linux/leds.h
@@ -48,6 +48,7 @@ struct led_classdev {
#define SET_BRIGHTNESS_ASYNC (1 << 21)
#define SET_BRIGHTNESS_SYNC (1 << 22)
#define LED_DEV_CAP_FLASH (1 << 23)
+#define LED_BRIGHTNESS_BLOCKING (1 << 24)
/* Set LED brightness level */
/* Must not sleep, use a workqueue if needed */
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* Re: [PATCH/RFC v5 02/57] leds: Add LED_BRIGHTNESS_BLOCKING flag
2015-08-11 9:37 ` [PATCH/RFC v5 02/57] leds: Add LED_BRIGHTNESS_BLOCKING flag Jacek Anaszewski
@ 2015-08-13 14:05 ` Andrew Lunn
0 siblings, 0 replies; 72+ messages in thread
From: Andrew Lunn @ 2015-08-13 14:05 UTC (permalink / raw)
To: Jacek Anaszewski
Cc: linux-leds, linux-kernel, cooloney, rpurdie, stsp, Sakari Ailus,
Pavel Machek
On Tue, Aug 11, 2015 at 11:37:15AM +0200, Jacek Anaszewski wrote:
> This patch adds LED_BRIGHTNESS_BLOCKING flag to the LED core to make it
> capable of discerning between drivers that implement brightness_set op
> in the old manner, i.e. the ones which use work queue internally, and
> drivers that implement brightness_set op in the new manner, i.e. they
> are aware that brightness_set op will be called from a work queue task
> by the LED core.
> In the new approach all drivers that implement brightness_set op should
> also set LED_BRIGHTNESS_BLOCKING flag.
Hi Jacek
Why have the drivers set this flag? The core code can look at the two
ops function pointers and figure it out for itself.
Andrew
^ permalink raw reply [flat|nested] 72+ messages in thread
* [PATCH/RFC v5 03/57] leds: Add led_set_brightness_sync to the public LED subsystem API
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 01/57] leds: Add brightness_set_nonblocking op Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 02/57] leds: Add LED_BRIGHTNESS_BLOCKING flag Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 04/57] leds: Improve asynchronous path of setting brightness Jacek Anaszewski
` (53 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski,
Andrew Lunn, Sakari Ailus, Pavel Machek
led_set_brightness_sync function was visible only internally to the
LED subsystem. It is now being made publicly available since it has
become apparent that this is a caller who should decide whether
brightness is to be set in a synchronous or an asynchronous way.
The function is modified to use brightness_set_nonblocking or
brightness_set op if brightness_set_sync is not implemented.
Eventually all LED subsystem drivers will be modfified to set brightness
only in a synchronous way with use of brightness_set_nonblocking or
brightness_set op (brightness_set_sync op will be removed). LED core
will take care of delegating brightness_set op to the work queue task,
if needed.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Bryan Wu <cooloney@gmail.com>
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: Sakari Ailus <sakari.ailus@linux.intel.com>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Stas Sergeev <stsp@users.sourceforge.net>
---
drivers/leds/led-core.c | 25 +++++++++++++++++++++++++
drivers/leds/leds.h | 13 -------------
include/linux/leds.h | 15 +++++++++++++++
3 files changed, 40 insertions(+), 13 deletions(-)
diff --git a/drivers/leds/led-core.c b/drivers/leds/led-core.c
index 549de7e..192071d 100644
--- a/drivers/leds/led-core.c
+++ b/drivers/leds/led-core.c
@@ -141,6 +141,31 @@ void led_set_brightness(struct led_classdev *led_cdev,
}
EXPORT_SYMBOL(led_set_brightness);
+int led_set_brightness_sync(struct led_classdev *led_cdev,
+ enum led_brightness value)
+{
+ int ret = 0;
+
+ led_cdev->brightness = min(value, led_cdev->max_brightness);
+
+ if (led_cdev->flags & LED_SUSPENDED)
+ return 0;
+
+ if (led_cdev->brightness_set_sync)
+ ret = led_cdev->brightness_set_sync(led_cdev,
+ led_cdev->brightness);
+ else if (led_cdev->brightness_set_nonblocking)
+ led_cdev->brightness_set_nonblocking(led_cdev,
+ led_cdev->brightness);
+ else if (led_cdev->brightness_set)
+ led_cdev->brightness_set(led_cdev, led_cdev->brightness);
+ else
+ ret = -EINVAL;
+
+ return ret;
+}
+EXPORT_SYMBOL(led_set_brightness_sync);
+
int led_update_brightness(struct led_classdev *led_cdev)
{
int ret = 0;
diff --git a/drivers/leds/leds.h b/drivers/leds/leds.h
index bc89d7a..1c026c9 100644
--- a/drivers/leds/leds.h
+++ b/drivers/leds/leds.h
@@ -26,19 +26,6 @@ static inline void led_set_brightness_async(struct led_classdev *led_cdev,
led_cdev->brightness_set(led_cdev, value);
}
-static inline int led_set_brightness_sync(struct led_classdev *led_cdev,
- enum led_brightness value)
-{
- int ret = 0;
-
- led_cdev->brightness = min(value, led_cdev->max_brightness);
-
- if (!(led_cdev->flags & LED_SUSPENDED))
- ret = led_cdev->brightness_set_sync(led_cdev,
- led_cdev->brightness);
- return ret;
-}
-
static inline int led_get_brightness(struct led_classdev *led_cdev)
{
return led_cdev->brightness;
diff --git a/include/linux/leds.h b/include/linux/leds.h
index ac9efe1..0c3098ea 100644
--- a/include/linux/leds.h
+++ b/include/linux/leds.h
@@ -164,6 +164,21 @@ extern void led_blink_set_oneshot(struct led_classdev *led_cdev,
*/
extern void led_set_brightness(struct led_classdev *led_cdev,
enum led_brightness brightness);
+
+/**
+ * led_set_brightness_sync - set LED brightness synchronously
+ * @led_cdev: the LED to set
+ * @brightness: the brightness to set it to
+ *
+ * Set an LED's brightness immediately. This function will block
+ * the caller for the time required for accessing device register,
+ * and it can sleep.
+ *
+ * Returns: 0 on success or negative error value on failure
+ */
+extern int led_set_brightness_sync(struct led_classdev *led_cdev,
+ enum led_brightness value);
+
/**
* led_update_brightness - update LED brightness
* @led_cdev: the LED to query
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 04/57] leds: Improve asynchronous path of setting brightness
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (2 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 03/57] leds: Add led_set_brightness_sync to the public LED subsystem API Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 05/57] leds: Add an internal led_set_brightness_nosleep function Jacek Anaszewski
` (52 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski,
Andrew Lunn, Sakari Ailus, Pavel Machek
led_set_brightness_async function didn't set brightness in an
asynchronous way in all cases. It was mistakenly assuming that all
LED subsystem drivers used work queue in their brightness_set op,
whereas only half of them did that. Modify the function to assure
setting brightness asynchronously in all cases by using existing
set_brightness_delayed work queue. The work queue is now made using
led_set_brightness_sync for setting brightness.
All these modification change the initial purpose of the work queue,
which was used for setting brightness only if blink timer was active.
In order to keep this functionality LED_BLINK_DISABLE flag is being
introduced, as well as a 'new_brightness_value' field is being added
to the struct led_classdev. Aforementioned improvements entail changes
in the led_brightness_set function.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Bryan Wu <cooloney@gmail.com>
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: Sakari Ailus <sakari.ailus@linux.intel.com>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Stas Sergeev <stsp@users.sourceforge.net>
---
drivers/leds/led-class.c | 13 ++++++++-----
drivers/leds/led-core.c | 20 ++++++++++++++++----
drivers/leds/leds.h | 23 ++++++++++++++++++-----
include/linux/leds.h | 2 ++
4 files changed, 44 insertions(+), 14 deletions(-)
diff --git a/drivers/leds/led-class.c b/drivers/leds/led-class.c
index 2875e6a..853bd95 100644
--- a/drivers/leds/led-class.c
+++ b/drivers/leds/led-class.c
@@ -121,10 +121,10 @@ static void led_timer_function(unsigned long data)
brightness = led_get_brightness(led_cdev);
if (!brightness) {
/* Time to switch the LED on. */
- if (led_cdev->delayed_set_value) {
+ if (led_cdev->new_brightness_value) {
led_cdev->blink_brightness =
- led_cdev->delayed_set_value;
- led_cdev->delayed_set_value = 0;
+ led_cdev->new_brightness_value;
+ led_cdev->new_brightness_value = 0;
}
brightness = led_cdev->blink_brightness;
delay = led_cdev->blink_delay_on;
@@ -161,9 +161,12 @@ static void set_brightness_delayed(struct work_struct *ws)
struct led_classdev *led_cdev =
container_of(ws, struct led_classdev, set_brightness_work);
- led_stop_software_blink(led_cdev);
+ if (led_cdev->flags & LED_BLINK_DISABLE) {
+ led_stop_software_blink(led_cdev);
+ led_cdev->flags &= ~LED_BLINK_DISABLE;
+ }
- led_set_brightness_async(led_cdev, led_cdev->delayed_set_value);
+ led_set_brightness_sync(led_cdev, led_cdev->delayed_set_value);
}
/**
diff --git a/drivers/leds/led-core.c b/drivers/leds/led-core.c
index 192071d..3ea32ab 100644
--- a/drivers/leds/led-core.c
+++ b/drivers/leds/led-core.c
@@ -119,11 +119,23 @@ void led_set_brightness(struct led_classdev *led_cdev,
{
int ret = 0;
- /* delay brightness if soft-blink is active */
+ /*
+ * In case blinking is on delay brightness setting
+ * until the next timer tick.
+ */
if (led_cdev->blink_delay_on || led_cdev->blink_delay_off) {
- led_cdev->delayed_set_value = brightness;
- if (brightness == LED_OFF)
- schedule_work(&led_cdev->set_brightness_work);
+ led_cdev->new_brightness_value = brightness;
+
+ /* New brightness will be set on next timer tick. */
+ if (brightness != LED_OFF)
+ return;
+ /*
+ * If need to disable soft blinking delegate this to the
+ * work queue task to avoid problems in case we are
+ * called from hard irq context.
+ */
+ led_cdev->flags |= LED_BLINK_DISABLE;
+ led_set_brightness_async(led_cdev, brightness);
return;
}
diff --git a/drivers/leds/leds.h b/drivers/leds/leds.h
index 1c026c9..f700400 100644
--- a/drivers/leds/leds.h
+++ b/drivers/leds/leds.h
@@ -17,13 +17,26 @@
#include <linux/leds.h>
static inline void led_set_brightness_async(struct led_classdev *led_cdev,
- enum led_brightness value)
+ enum led_brightness value)
{
- value = min(value, led_cdev->max_brightness);
- led_cdev->brightness = value;
+ int ret;
+ /*
+ * Drivers that implement brightness_set op in the old manner also
+ * don't set LED_BRIGHTNESS_BLOCKING flag. They use work queue
+ * internally in case they set brightness in a blocking way, thus we
+ * avoid scheduling another work queue task by the LED core.
+ */
+ if (led_cdev->brightness_set &&
+ !(led_cdev->flags & LED_BRIGHTNESS_BLOCKING)) {
+ ret = led_set_brightness_sync(led_cdev, value);
+ if (ret < 0)
+ dev_err(led_cdev->dev,
+ "cannot set led brightness %d\n", ret);
+ return;
+ }
- if (!(led_cdev->flags & LED_SUSPENDED))
- led_cdev->brightness_set(led_cdev, value);
+ led_cdev->delayed_set_value = value;
+ schedule_work(&led_cdev->set_brightness_work);
}
static inline int led_get_brightness(struct led_classdev *led_cdev)
diff --git a/include/linux/leds.h b/include/linux/leds.h
index 0c3098ea..a004e84 100644
--- a/include/linux/leds.h
+++ b/include/linux/leds.h
@@ -49,6 +49,7 @@ struct led_classdev {
#define SET_BRIGHTNESS_SYNC (1 << 22)
#define LED_DEV_CAP_FLASH (1 << 23)
#define LED_BRIGHTNESS_BLOCKING (1 << 24)
+#define LED_BLINK_DISABLE (1 << 25)
/* Set LED brightness level */
/* Must not sleep, use a workqueue if needed */
@@ -91,6 +92,7 @@ struct led_classdev {
struct work_struct set_brightness_work;
int delayed_set_value;
+ int new_brightness_value;
#ifdef CONFIG_LEDS_TRIGGERS
/* Protects the trigger data below */
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 05/57] leds: Add an internal led_set_brightness_nosleep function
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (3 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 04/57] leds: Improve asynchronous path of setting brightness Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 06/57] leds: Improve setting brightness in a non sleeping way Jacek Anaszewski
` (51 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski,
Andrew Lunn, Sakari Ailus, Pavel Machek
This patch adds led_set_brightness_nosleep function. It guarantees setting
LED brightness in a non-blocking way.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Bryan Wu <cooloney@gmail.com>
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: Sakari Ailus <sakari.ailus@linux.intel.com>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Stas Sergeev <stsp@users.sourceforge.net>
---
drivers/leds/leds.h | 22 ++++++++++++++++++++++
include/linux/leds.h | 2 +-
2 files changed, 23 insertions(+), 1 deletion(-)
diff --git a/drivers/leds/leds.h b/drivers/leds/leds.h
index f700400..308f60f 100644
--- a/drivers/leds/leds.h
+++ b/drivers/leds/leds.h
@@ -2,8 +2,10 @@
* LED Core
*
* Copyright 2005 Openedhand Ltd.
+ * Copyright 2014, 2015 Samsung Electronics Co., Ltd.
*
* Author: Richard Purdie <rpurdie@openedhand.com>
+ * Author: Jacek Anaszewski <j.anaszewski@samsung.com>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2 as
@@ -39,6 +41,26 @@ static inline void led_set_brightness_async(struct led_classdev *led_cdev,
schedule_work(&led_cdev->set_brightness_work);
}
+static inline void led_set_brightness_nosleep(struct led_classdev *led_cdev,
+ enum led_brightness value)
+{
+ int ret;
+
+ /*
+ * Drivers that implement brightness_set_nonblocking op are guaranteed
+ * not to sleep while setting brightness.
+ */
+ if (led_cdev->brightness_set_nonblocking) {
+ ret = led_set_brightness_sync(led_cdev, value);
+ if (ret < 0)
+ dev_err(led_cdev->dev,
+ "cannot set led brightness %d\n", ret);
+ return;
+ }
+
+ led_set_brightness_async(led_cdev, value);
+}
+
static inline int led_get_brightness(struct led_classdev *led_cdev)
{
return led_cdev->brightness;
diff --git a/include/linux/leds.h b/include/linux/leds.h
index a004e84..9603a73 100644
--- a/include/linux/leds.h
+++ b/include/linux/leds.h
@@ -52,7 +52,7 @@ struct led_classdev {
#define LED_BLINK_DISABLE (1 << 25)
/* Set LED brightness level */
- /* Must not sleep, use a workqueue if needed */
+ /* Intended for drivers that may set brightness in a blocking way. */
void (*brightness_set)(struct led_classdev *led_cdev,
enum led_brightness brightness);
/* Intended for drivers that set brightness in a non-blocking way */
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 06/57] leds: Improve setting brightness in a non sleeping way
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (4 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 05/57] leds: Add an internal led_set_brightness_nosleep function Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 07/57] leds: Drivers shouldn't enforce SYNC/ASYNC brightness setting Jacek Anaszewski
` (50 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski,
Andrew Lunn, Sakari Ailus, Pavel Machek
This patch replaces led_set_brightness_async with
led_set_brightness_nosleep in all places where the most vital was setting
brightness in a non sleeping way but not necessarily asynchronously, which
is not needed for non-blocking drivers.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Bryan Wu <cooloney@gmail.com>
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: Sakari Ailus <sakari.ailus@linux.intel.com>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Stas Sergeev <stsp@users.sourceforge.net>
---
drivers/leds/led-class.c | 4 ++--
drivers/leds/led-core.c | 5 +++--
drivers/leds/trigger/ledtrig-backlight.c | 8 ++++----
drivers/leds/trigger/ledtrig-default-on.c | 2 +-
drivers/leds/trigger/ledtrig-gpio.c | 6 +++---
drivers/leds/trigger/ledtrig-heartbeat.c | 4 ++--
drivers/leds/trigger/ledtrig-oneshot.c | 4 ++--
drivers/leds/trigger/ledtrig-transient.c | 8 ++++----
8 files changed, 21 insertions(+), 20 deletions(-)
diff --git a/drivers/leds/led-class.c b/drivers/leds/led-class.c
index 853bd95..bd32919 100644
--- a/drivers/leds/led-class.c
+++ b/drivers/leds/led-class.c
@@ -109,7 +109,7 @@ static void led_timer_function(unsigned long data)
unsigned long delay;
if (!led_cdev->blink_delay_on || !led_cdev->blink_delay_off) {
- led_set_brightness_async(led_cdev, LED_OFF);
+ led_set_brightness_nosleep(led_cdev, LED_OFF);
return;
}
@@ -137,7 +137,7 @@ static void led_timer_function(unsigned long data)
delay = led_cdev->blink_delay_off;
}
- led_set_brightness_async(led_cdev, brightness);
+ led_set_brightness_nosleep(led_cdev, brightness);
/* Return in next iteration if led is in one-shot mode and we are in
* the final blink state so that the led is toggled each delay_on +
diff --git a/drivers/leds/led-core.c b/drivers/leds/led-core.c
index 3ea32ab..14377ea 100644
--- a/drivers/leds/led-core.c
+++ b/drivers/leds/led-core.c
@@ -42,13 +42,14 @@ static void led_set_software_blink(struct led_classdev *led_cdev,
/* never on - just set to off */
if (!delay_on) {
- led_set_brightness_async(led_cdev, LED_OFF);
+ led_set_brightness_nosleep(led_cdev, LED_OFF);
return;
}
/* never off - just set to brightness */
if (!delay_off) {
- led_set_brightness_async(led_cdev, led_cdev->blink_brightness);
+ led_set_brightness_nosleep(led_cdev,
+ led_cdev->blink_brightness);
return;
}
diff --git a/drivers/leds/trigger/ledtrig-backlight.c b/drivers/leds/trigger/ledtrig-backlight.c
index 59eca17..1ca1f16 100644
--- a/drivers/leds/trigger/ledtrig-backlight.c
+++ b/drivers/leds/trigger/ledtrig-backlight.c
@@ -51,9 +51,9 @@ static int fb_notifier_callback(struct notifier_block *p,
if ((n->old_status == UNBLANK) ^ n->invert) {
n->brightness = led->brightness;
- led_set_brightness_async(led, LED_OFF);
+ led_set_brightness_nosleep(led, LED_OFF);
} else {
- led_set_brightness_async(led, n->brightness);
+ led_set_brightness_nosleep(led, n->brightness);
}
n->old_status = new_status;
@@ -89,9 +89,9 @@ static ssize_t bl_trig_invert_store(struct device *dev,
/* After inverting, we need to update the LED. */
if ((n->old_status == BLANK) ^ n->invert)
- led_set_brightness_async(led, LED_OFF);
+ led_set_brightness_nosleep(led, LED_OFF);
else
- led_set_brightness_async(led, n->brightness);
+ led_set_brightness_nosleep(led, n->brightness);
return num;
}
diff --git a/drivers/leds/trigger/ledtrig-default-on.c b/drivers/leds/trigger/ledtrig-default-on.c
index 6f38f88..ff455cb 100644
--- a/drivers/leds/trigger/ledtrig-default-on.c
+++ b/drivers/leds/trigger/ledtrig-default-on.c
@@ -19,7 +19,7 @@
static void defon_trig_activate(struct led_classdev *led_cdev)
{
- led_set_brightness_async(led_cdev, led_cdev->max_brightness);
+ led_set_brightness_nosleep(led_cdev, led_cdev->max_brightness);
}
static struct led_trigger defon_led_trigger = {
diff --git a/drivers/leds/trigger/ledtrig-gpio.c b/drivers/leds/trigger/ledtrig-gpio.c
index 4cc7040..51288a4 100644
--- a/drivers/leds/trigger/ledtrig-gpio.c
+++ b/drivers/leds/trigger/ledtrig-gpio.c
@@ -54,12 +54,12 @@ static void gpio_trig_work(struct work_struct *work)
if (tmp) {
if (gpio_data->desired_brightness)
- led_set_brightness_async(gpio_data->led,
+ led_set_brightness_nosleep(gpio_data->led,
gpio_data->desired_brightness);
else
- led_set_brightness_async(gpio_data->led, LED_FULL);
+ led_set_brightness_nosleep(gpio_data->led, LED_FULL);
} else {
- led_set_brightness_async(gpio_data->led, LED_OFF);
+ led_set_brightness_nosleep(gpio_data->led, LED_OFF);
}
}
diff --git a/drivers/leds/trigger/ledtrig-heartbeat.c b/drivers/leds/trigger/ledtrig-heartbeat.c
index fea6871..3dc6f0c 100644
--- a/drivers/leds/trigger/ledtrig-heartbeat.c
+++ b/drivers/leds/trigger/ledtrig-heartbeat.c
@@ -37,7 +37,7 @@ static void led_heartbeat_function(unsigned long data)
unsigned long delay = 0;
if (unlikely(panic_heartbeats)) {
- led_set_brightness(led_cdev, LED_OFF);
+ led_set_brightness_nosleep(led_cdev, LED_OFF);
return;
}
@@ -74,7 +74,7 @@ static void led_heartbeat_function(unsigned long data)
break;
}
- led_set_brightness_async(led_cdev, brightness);
+ led_set_brightness_nosleep(led_cdev, brightness);
mod_timer(&heartbeat_data->timer, jiffies + delay);
}
diff --git a/drivers/leds/trigger/ledtrig-oneshot.c b/drivers/leds/trigger/ledtrig-oneshot.c
index fbd02cd..6729317 100644
--- a/drivers/leds/trigger/ledtrig-oneshot.c
+++ b/drivers/leds/trigger/ledtrig-oneshot.c
@@ -63,9 +63,9 @@ static ssize_t led_invert_store(struct device *dev,
oneshot_data->invert = !!state;
if (oneshot_data->invert)
- led_set_brightness_async(led_cdev, LED_FULL);
+ led_set_brightness_nosleep(led_cdev, LED_FULL);
else
- led_set_brightness_async(led_cdev, LED_OFF);
+ led_set_brightness_nosleep(led_cdev, LED_OFF);
return size;
}
diff --git a/drivers/leds/trigger/ledtrig-transient.c b/drivers/leds/trigger/ledtrig-transient.c
index 3c34de4..1dddd8f 100644
--- a/drivers/leds/trigger/ledtrig-transient.c
+++ b/drivers/leds/trigger/ledtrig-transient.c
@@ -41,7 +41,7 @@ static void transient_timer_function(unsigned long data)
struct transient_trig_data *transient_data = led_cdev->trigger_data;
transient_data->activate = 0;
- led_set_brightness_async(led_cdev, transient_data->restore_state);
+ led_set_brightness_nosleep(led_cdev, transient_data->restore_state);
}
static ssize_t transient_activate_show(struct device *dev,
@@ -72,7 +72,7 @@ static ssize_t transient_activate_store(struct device *dev,
if (state == 0 && transient_data->activate == 1) {
del_timer(&transient_data->timer);
transient_data->activate = state;
- led_set_brightness_async(led_cdev,
+ led_set_brightness_nosleep(led_cdev,
transient_data->restore_state);
return size;
}
@@ -81,7 +81,7 @@ static ssize_t transient_activate_store(struct device *dev,
if (state == 1 && transient_data->activate == 0 &&
transient_data->duration != 0) {
transient_data->activate = state;
- led_set_brightness_async(led_cdev, transient_data->state);
+ led_set_brightness_nosleep(led_cdev, transient_data->state);
transient_data->restore_state =
(transient_data->state == LED_FULL) ? LED_OFF : LED_FULL;
mod_timer(&transient_data->timer,
@@ -204,7 +204,7 @@ static void transient_trig_deactivate(struct led_classdev *led_cdev)
if (led_cdev->activated) {
del_timer_sync(&transient_data->timer);
- led_set_brightness_async(led_cdev,
+ led_set_brightness_nosleep(led_cdev,
transient_data->restore_state);
device_remove_file(led_cdev->dev, &dev_attr_activate);
device_remove_file(led_cdev->dev, &dev_attr_duration);
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 07/57] leds: Drivers shouldn't enforce SYNC/ASYNC brightness setting
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (5 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 06/57] leds: Improve setting brightness in a non sleeping way Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 08/57] leds: ktd2692: Remove work queue Jacek Anaszewski
` (49 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski,
Ingi Kim, Andrew Lunn, Sakari Ailus, Pavel Machek
This patch removes brightness_set_sync op as well as SET_BRIGHTNESS_ASYNC
and SET_BRIGHTNESS sync flags. led_set_brightness now calls
led_set_brightness_nosleep rather then choosing between sync and async op
basing on the flags defined by the driver. Now if a user wants to make
sure that brightness will be set synchronously it has to use
led_set_brightness_sync API.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Ingi Kim <ingi2.kim@samsung.com>
Cc: Bryan Wu <cooloney@gmail.com>
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: Sakari Ailus <sakari.ailus@linux.intel.com>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Stas Sergeev <stsp@users.sourceforge.net>
---
drivers/leds/led-class-flash.c | 7 -------
drivers/leds/led-class.c | 2 --
drivers/leds/led-core.c | 22 +++-------------------
drivers/leds/leds-aat1290.c | 12 ------------
drivers/leds/leds-ktd2692.c | 12 ------------
drivers/leds/leds-max77693.c | 12 ------------
include/linux/leds.h | 16 ++++------------
7 files changed, 7 insertions(+), 76 deletions(-)
diff --git a/drivers/leds/led-class-flash.c b/drivers/leds/led-class-flash.c
index 3b25734..68c3f36 100644
--- a/drivers/leds/led-class-flash.c
+++ b/drivers/leds/led-class-flash.c
@@ -298,9 +298,6 @@ int led_classdev_flash_register(struct device *parent,
led_cdev = &fled_cdev->led_cdev;
if (led_cdev->flags & LED_DEV_CAP_FLASH) {
- if (!led_cdev->brightness_set_sync)
- return -EINVAL;
-
ops = fled_cdev->ops;
if (!ops || !ops->strobe_set)
return -EINVAL;
@@ -316,10 +313,6 @@ int led_classdev_flash_register(struct device *parent,
if (ret < 0)
return ret;
- /* Setting a torch brightness needs to have immediate effect */
- led_cdev->flags &= ~SET_BRIGHTNESS_ASYNC;
- led_cdev->flags |= SET_BRIGHTNESS_SYNC;
-
return 0;
}
EXPORT_SYMBOL_GPL(led_classdev_flash_register);
diff --git a/drivers/leds/led-class.c b/drivers/leds/led-class.c
index bd32919..5932c4c 100644
--- a/drivers/leds/led-class.c
+++ b/drivers/leds/led-class.c
@@ -285,8 +285,6 @@ int led_classdev_register(struct device *parent, struct led_classdev *led_cdev)
WARN_ON((led_cdev->flags & LED_BRIGHTNESS_BLOCKING) &&
!led_cdev->brightness_set);
- led_cdev->flags |= SET_BRIGHTNESS_ASYNC;
-
led_update_brightness(led_cdev);
INIT_WORK(&led_cdev->set_brightness_work, set_brightness_delayed);
diff --git a/drivers/leds/led-core.c b/drivers/leds/led-core.c
index 14377ea..b36f50e 100644
--- a/drivers/leds/led-core.c
+++ b/drivers/leds/led-core.c
@@ -118,8 +118,6 @@ EXPORT_SYMBOL_GPL(led_stop_software_blink);
void led_set_brightness(struct led_classdev *led_cdev,
enum led_brightness brightness)
{
- int ret = 0;
-
/*
* In case blinking is on delay brightness setting
* until the next timer tick.
@@ -137,20 +135,9 @@ void led_set_brightness(struct led_classdev *led_cdev,
*/
led_cdev->flags |= LED_BLINK_DISABLE;
led_set_brightness_async(led_cdev, brightness);
- return;
+ } else {
+ led_set_brightness_nosleep(led_cdev, brightness);
}
-
- if (led_cdev->flags & SET_BRIGHTNESS_ASYNC) {
- led_set_brightness_async(led_cdev, brightness);
- return;
- } else if (led_cdev->flags & SET_BRIGHTNESS_SYNC)
- ret = led_set_brightness_sync(led_cdev, brightness);
- else
- ret = -EINVAL;
-
- if (ret < 0)
- dev_dbg(led_cdev->dev, "Setting LED brightness failed (%d)\n",
- ret);
}
EXPORT_SYMBOL(led_set_brightness);
@@ -164,10 +151,7 @@ int led_set_brightness_sync(struct led_classdev *led_cdev,
if (led_cdev->flags & LED_SUSPENDED)
return 0;
- if (led_cdev->brightness_set_sync)
- ret = led_cdev->brightness_set_sync(led_cdev,
- led_cdev->brightness);
- else if (led_cdev->brightness_set_nonblocking)
+ if (led_cdev->brightness_set_nonblocking)
led_cdev->brightness_set_nonblocking(led_cdev,
led_cdev->brightness);
else if (led_cdev->brightness_set)
diff --git a/drivers/leds/leds-aat1290.c b/drivers/leds/leds-aat1290.c
index fd7c25f..0ecaf76 100644
--- a/drivers/leds/leds-aat1290.c
+++ b/drivers/leds/leds-aat1290.c
@@ -180,17 +180,6 @@ static void aat1290_led_brightness_set(struct led_classdev *led_cdev,
schedule_work(&led->work_brightness_set);
}
-static int aat1290_led_brightness_set_sync(struct led_classdev *led_cdev,
- enum led_brightness brightness)
-{
- struct led_classdev_flash *fled_cdev = lcdev_to_flcdev(led_cdev);
- struct aat1290_led *led = fled_cdev_to_led(fled_cdev);
-
- aat1290_brightness_set(led, brightness);
-
- return 0;
-}
-
static int aat1290_led_flash_strobe_set(struct led_classdev_flash *fled_cdev,
bool state)
@@ -510,7 +499,6 @@ static int aat1290_led_probe(struct platform_device *pdev)
/* Initialize LED Flash class device */
led_cdev->brightness_set = aat1290_led_brightness_set;
- led_cdev->brightness_set_sync = aat1290_led_brightness_set_sync;
led_cdev->max_brightness = led_cfg.max_brightness;
led_cdev->flags |= LED_DEV_CAP_FLASH;
INIT_WORK(&led->work_brightness_set, aat1290_brightness_set_work);
diff --git a/drivers/leds/leds-ktd2692.c b/drivers/leds/leds-ktd2692.c
index 2ae8c4d..7f6758d 100644
--- a/drivers/leds/leds-ktd2692.c
+++ b/drivers/leds/leds-ktd2692.c
@@ -194,17 +194,6 @@ static void ktd2692_led_brightness_set(struct led_classdev *led_cdev,
schedule_work(&led->work_brightness_set);
}
-static int ktd2692_led_brightness_set_sync(struct led_classdev *led_cdev,
- enum led_brightness brightness)
-{
- struct led_classdev_flash *fled_cdev = lcdev_to_flcdev(led_cdev);
- struct ktd2692_context *led = fled_cdev_to_led(fled_cdev);
-
- ktd2692_brightness_set(led, brightness);
-
- return 0;
-}
-
static int ktd2692_led_flash_strobe_set(struct led_classdev_flash *fled_cdev,
bool state)
{
@@ -382,7 +371,6 @@ static int ktd2692_probe(struct platform_device *pdev)
led_cdev->max_brightness = led_cfg.max_brightness;
led_cdev->brightness_set = ktd2692_led_brightness_set;
- led_cdev->brightness_set_sync = ktd2692_led_brightness_set_sync;
led_cdev->flags |= LED_CORE_SUSPENDRESUME | LED_DEV_CAP_FLASH;
mutex_init(&led->lock);
diff --git a/drivers/leds/leds-max77693.c b/drivers/leds/leds-max77693.c
index df348a0..856d929 100644
--- a/drivers/leds/leds-max77693.c
+++ b/drivers/leds/leds-max77693.c
@@ -511,17 +511,6 @@ static void max77693_led_brightness_set_work(
/* LED subsystem callbacks */
-static int max77693_led_brightness_set_sync(
- struct led_classdev *led_cdev,
- enum led_brightness value)
-{
- struct led_classdev_flash *fled_cdev = lcdev_to_flcdev(led_cdev);
- struct max77693_sub_led *sub_led = flcdev_to_sub_led(fled_cdev);
- struct max77693_led_device *led = sub_led_to_led(sub_led);
-
- return __max77693_led_brightness_set(led, sub_led->fled_id, value);
-}
-
static void max77693_led_brightness_set(
struct led_classdev *led_cdev,
enum led_brightness value)
@@ -932,7 +921,6 @@ static void max77693_init_fled_cdev(struct max77693_sub_led *sub_led,
led_cdev->name = led_cfg->label[fled_id];
led_cdev->brightness_set = max77693_led_brightness_set;
- led_cdev->brightness_set_sync = max77693_led_brightness_set_sync;
led_cdev->max_brightness = (led->iout_joint ?
led_cfg->iout_torch_max[FLED1] +
led_cfg->iout_torch_max[FLED2] :
diff --git a/include/linux/leds.h b/include/linux/leds.h
index 9603a73..45cb93d 100644
--- a/include/linux/leds.h
+++ b/include/linux/leds.h
@@ -45,11 +45,9 @@ struct led_classdev {
#define LED_BLINK_ONESHOT_STOP (1 << 18)
#define LED_BLINK_INVERT (1 << 19)
#define LED_SYSFS_DISABLE (1 << 20)
-#define SET_BRIGHTNESS_ASYNC (1 << 21)
-#define SET_BRIGHTNESS_SYNC (1 << 22)
-#define LED_DEV_CAP_FLASH (1 << 23)
-#define LED_BRIGHTNESS_BLOCKING (1 << 24)
-#define LED_BLINK_DISABLE (1 << 25)
+#define LED_DEV_CAP_FLASH (1 << 21)
+#define LED_BRIGHTNESS_BLOCKING (1 << 22)
+#define LED_BLINK_DISABLE (1 << 23)
/* Set LED brightness level */
/* Intended for drivers that may set brightness in a blocking way. */
@@ -58,12 +56,6 @@ struct led_classdev {
/* Intended for drivers that set brightness in a non-blocking way */
void (*brightness_set_nonblocking)(struct led_classdev *led_cdev,
enum led_brightness brightness);
- /*
- * Set LED brightness level immediately - it can block the caller for
- * the time required for accessing a LED device register.
- */
- int (*brightness_set_sync)(struct led_classdev *led_cdev,
- enum led_brightness brightness);
/* Get LED brightness level */
enum led_brightness (*brightness_get)(struct led_classdev *led_cdev);
@@ -162,7 +154,7 @@ extern void led_blink_set_oneshot(struct led_classdev *led_cdev,
*
* Set an LED's brightness, and, if necessary, cancel the
* software blink timer that implements blinking when the
- * hardware doesn't.
+ * hardware doesn't. This function is guaranteed not to sleep.
*/
extern void led_set_brightness(struct led_classdev *led_cdev,
enum led_brightness brightness);
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 08/57] leds: ktd2692: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (6 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 07/57] leds: Drivers shouldn't enforce SYNC/ASYNC brightness setting Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 09/57] leds: aat1290: " Jacek Anaszewski
` (48 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski,
Sakari Ailus, Ingi Kim
Now the core implements the work queue, remove it from the drivers.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Sakari Ailus <sakari.ailus@linux.intel.com>
Cc: Bryan Wu <cooloney@gmail.com>
Cc: Ingi Kim <ingi2.kim@samsung.com>
---
drivers/leds/leds-ktd2692.c | 32 +++++++-------------------------
1 file changed, 7 insertions(+), 25 deletions(-)
diff --git a/drivers/leds/leds-ktd2692.c b/drivers/leds/leds-ktd2692.c
index 7f6758d..fc7f253 100644
--- a/drivers/leds/leds-ktd2692.c
+++ b/drivers/leds/leds-ktd2692.c
@@ -18,7 +18,6 @@
#include <linux/of.h>
#include <linux/platform_device.h>
#include <linux/regulator/consumer.h>
-#include <linux/workqueue.h>
/* Value related the movie mode */
#define KTD2692_MOVIE_MODE_CURRENT_LEVELS 16
@@ -82,7 +81,6 @@ struct ktd2692_context {
/* secures access to the device */
struct mutex lock;
struct regulator *regulator;
- struct work_struct work_brightness_set;
struct gpio_desc *aux_gpio;
struct gpio_desc *ctrl_gpio;
@@ -158,9 +156,12 @@ static void ktd2692_expresswire_write(struct ktd2692_context *led, u8 value)
ktd2692_expresswire_end(led);
}
-static void ktd2692_brightness_set(struct ktd2692_context *led,
- enum led_brightness brightness)
+static void ktd2692_led_brightness_set(struct led_classdev *led_cdev,
+ enum led_brightness brightness)
{
+ struct led_classdev_flash *fled_cdev = lcdev_to_flcdev(led_cdev);
+ struct ktd2692_context *led = fled_cdev_to_led(fled_cdev);
+
mutex_lock(&led->lock);
if (brightness == LED_OFF) {
@@ -176,24 +177,6 @@ static void ktd2692_brightness_set(struct ktd2692_context *led,
mutex_unlock(&led->lock);
}
-static void ktd2692_brightness_set_work(struct work_struct *work)
-{
- struct ktd2692_context *led =
- container_of(work, struct ktd2692_context, work_brightness_set);
-
- ktd2692_brightness_set(led, led->torch_brightness);
-}
-
-static void ktd2692_led_brightness_set(struct led_classdev *led_cdev,
- enum led_brightness brightness)
-{
- struct led_classdev_flash *fled_cdev = lcdev_to_flcdev(led_cdev);
- struct ktd2692_context *led = fled_cdev_to_led(fled_cdev);
-
- led->torch_brightness = brightness;
- schedule_work(&led->work_brightness_set);
-}
-
static int ktd2692_led_flash_strobe_set(struct led_classdev_flash *fled_cdev,
bool state)
{
@@ -371,10 +354,10 @@ static int ktd2692_probe(struct platform_device *pdev)
led_cdev->max_brightness = led_cfg.max_brightness;
led_cdev->brightness_set = ktd2692_led_brightness_set;
- led_cdev->flags |= LED_CORE_SUSPENDRESUME | LED_DEV_CAP_FLASH;
+ led_cdev->flags |= LED_CORE_SUSPENDRESUME | LED_DEV_CAP_FLASH |
+ LED_BRIGHTNESS_BLOCKING;
mutex_init(&led->lock);
- INIT_WORK(&led->work_brightness_set, ktd2692_brightness_set_work);
platform_set_drvdata(pdev, led);
@@ -396,7 +379,6 @@ static int ktd2692_remove(struct platform_device *pdev)
int ret;
led_classdev_flash_unregister(&led->fled_cdev);
- cancel_work_sync(&led->work_brightness_set);
if (led->regulator) {
ret = regulator_disable(led->regulator);
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 09/57] leds: aat1290: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (7 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 08/57] leds: ktd2692: Remove work queue Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 10/57] leds: max77693: " Jacek Anaszewski
` (47 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski, Sakari Ailus
Now the core implements the work queue, remove it from the drivers.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Sakari Ailus <sakari.ailus@linux.intel.com>
Cc: Bryan Wu <cooloney@gmail.com>
---
drivers/leds/leds-aat1290.c | 40 +++++++++++++---------------------------
1 file changed, 13 insertions(+), 27 deletions(-)
diff --git a/drivers/leds/leds-aat1290.c b/drivers/leds/leds-aat1290.c
index 0ecaf76..18af569 100644
--- a/drivers/leds/leds-aat1290.c
+++ b/drivers/leds/leds-aat1290.c
@@ -20,7 +20,6 @@
#include <linux/pinctrl/consumer.h>
#include <linux/platform_device.h>
#include <linux/slab.h>
-#include <linux/workqueue.h>
#include <media/v4l2-flash-led-class.h>
#define AAT1290_MOVIE_MODE_CURRENT_ADDR 17
@@ -82,8 +81,6 @@ struct aat1290_led {
/* brightness cache */
unsigned int torch_brightness;
- /* assures led-triggers compatibility */
- struct work_struct work_brightness_set;
};
static struct aat1290_led *fled_cdev_to_led(
@@ -92,6 +89,12 @@ static struct aat1290_led *fled_cdev_to_led(
return container_of(fled_cdev, struct aat1290_led, fled_cdev);
}
+static struct led_classdev_flash *led_cdev_to_fled_cdev(
+ struct led_classdev *led_cdev)
+{
+ return container_of(led_cdev, struct led_classdev_flash, led_cdev);
+}
+
static void aat1290_as2cwire_write(struct aat1290_led *led, int addr, int value)
{
int i;
@@ -134,9 +137,14 @@ static void aat1290_set_flash_safety_timer(struct aat1290_led *led,
flash_tm_reg);
}
-static void aat1290_brightness_set(struct aat1290_led *led,
+/* LED subsystem callbacks */
+
+static void aat1290_led_brightness_set(struct led_classdev *led_cdev,
enum led_brightness brightness)
{
+ struct led_classdev_flash *fled_cdev = led_cdev_to_fled_cdev(led_cdev);
+ struct aat1290_led *led = fled_cdev_to_led(fled_cdev);
+
mutex_lock(&led->lock);
if (brightness == 0) {
@@ -160,26 +168,6 @@ static void aat1290_brightness_set(struct aat1290_led *led,
mutex_unlock(&led->lock);
}
-/* LED subsystem callbacks */
-
-static void aat1290_brightness_set_work(struct work_struct *work)
-{
- struct aat1290_led *led =
- container_of(work, struct aat1290_led, work_brightness_set);
-
- aat1290_brightness_set(led, led->torch_brightness);
-}
-
-static void aat1290_led_brightness_set(struct led_classdev *led_cdev,
- enum led_brightness brightness)
-{
- struct led_classdev_flash *fled_cdev = lcdev_to_flcdev(led_cdev);
- struct aat1290_led *led = fled_cdev_to_led(fled_cdev);
-
- led->torch_brightness = brightness;
- schedule_work(&led->work_brightness_set);
-}
-
static int aat1290_led_flash_strobe_set(struct led_classdev_flash *fled_cdev,
bool state)
@@ -500,8 +488,7 @@ static int aat1290_led_probe(struct platform_device *pdev)
/* Initialize LED Flash class device */
led_cdev->brightness_set = aat1290_led_brightness_set;
led_cdev->max_brightness = led_cfg.max_brightness;
- led_cdev->flags |= LED_DEV_CAP_FLASH;
- INIT_WORK(&led->work_brightness_set, aat1290_brightness_set_work);
+ led_cdev->flags |= LED_DEV_CAP_FLASH | LED_BRIGHTNESS_BLOCKING;
aat1290_init_flash_timeout(led, &led_cfg);
@@ -536,7 +523,6 @@ static int aat1290_led_remove(struct platform_device *pdev)
v4l2_flash_release(led->v4l2_flash);
led_classdev_flash_unregister(&led->fled_cdev);
- cancel_work_sync(&led->work_brightness_set);
mutex_destroy(&led->lock);
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 10/57] leds: max77693: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (8 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 09/57] leds: aat1290: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 11/57] Documentation: leds: Add description of brightness_set* ops Jacek Anaszewski
` (46 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski, Sakari Ailus
Now the core implements the work queue, remove it from the drivers.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Sakari Ailus <sakari.ailus@linux.intel.com>
Cc: Bryan Wu <cooloney@gmail.com>
---
drivers/leds/leds-max77693.c | 45 ++++++++----------------------------------
1 file changed, 8 insertions(+), 37 deletions(-)
diff --git a/drivers/leds/leds-max77693.c b/drivers/leds/leds-max77693.c
index 856d929..cc4146a 100644
--- a/drivers/leds/leds-max77693.c
+++ b/drivers/leds/leds-max77693.c
@@ -20,7 +20,6 @@
#include <linux/platform_device.h>
#include <linux/regmap.h>
#include <linux/slab.h>
-#include <linux/workqueue.h>
#include <media/v4l2-flash-led-class.h>
#define MODE_OFF 0
@@ -62,8 +61,6 @@ struct max77693_sub_led {
int fled_id;
/* corresponding LED Flash class device */
struct led_classdev_flash fled_cdev;
- /* assures led-triggers compatibility */
- struct work_struct work_brightness_set;
/* V4L2 Flash device */
struct v4l2_flash *v4l2_flash;
@@ -463,10 +460,14 @@ static int max77693_setup(struct max77693_led_device *led,
return max77693_set_mode_reg(led, MODE_OFF);
}
-static int __max77693_led_brightness_set(struct max77693_led_device *led,
- int fled_id, enum led_brightness value)
+/* LED subsystem callbacks */
+static void max77693_led_brightness_set(struct led_classdev *led_cdev,
+ enum led_brightness value)
{
- int ret;
+ struct led_classdev_flash *fled_cdev = lcdev_to_flcdev(led_cdev);
+ struct max77693_sub_led *sub_led = flcdev_to_sub_led(fled_cdev);
+ struct max77693_led_device *led = sub_led_to_led(sub_led);
+ int fled_id = sub_led->fled_id, ret;
mutex_lock(&led->lock);
@@ -494,32 +495,6 @@ static int __max77693_led_brightness_set(struct max77693_led_device *led,
ret);
unlock:
mutex_unlock(&led->lock);
- return ret;
-}
-
-static void max77693_led_brightness_set_work(
- struct work_struct *work)
-{
- struct max77693_sub_led *sub_led =
- container_of(work, struct max77693_sub_led,
- work_brightness_set);
- struct max77693_led_device *led = sub_led_to_led(sub_led);
-
- __max77693_led_brightness_set(led, sub_led->fled_id,
- sub_led->torch_brightness);
-}
-
-/* LED subsystem callbacks */
-
-static void max77693_led_brightness_set(
- struct led_classdev *led_cdev,
- enum led_brightness value)
-{
- struct led_classdev_flash *fled_cdev = lcdev_to_flcdev(led_cdev);
- struct max77693_sub_led *sub_led = flcdev_to_sub_led(fled_cdev);
-
- sub_led->torch_brightness = value;
- schedule_work(&sub_led->work_brightness_set);
}
static int max77693_led_flash_brightness_set(
@@ -926,9 +901,7 @@ static void max77693_init_fled_cdev(struct max77693_sub_led *sub_led,
led_cfg->iout_torch_max[FLED2] :
led_cfg->iout_torch_max[fled_id]) /
TORCH_IOUT_STEP;
- led_cdev->flags |= LED_DEV_CAP_FLASH;
- INIT_WORK(&sub_led->work_brightness_set,
- max77693_led_brightness_set_work);
+ led_cdev->flags |= LED_DEV_CAP_FLASH | LED_BRIGHTNESS_BLOCKING;
max77693_init_flash_settings(sub_led, led_cfg);
@@ -1050,13 +1023,11 @@ static int max77693_led_remove(struct platform_device *pdev)
if (led->iout_joint || max77693_fled_used(led, FLED1)) {
v4l2_flash_release(sub_leds[FLED1].v4l2_flash);
led_classdev_flash_unregister(&sub_leds[FLED1].fled_cdev);
- cancel_work_sync(&sub_leds[FLED1].work_brightness_set);
}
if (!led->iout_joint && max77693_fled_used(led, FLED2)) {
v4l2_flash_release(sub_leds[FLED2].v4l2_flash);
led_classdev_flash_unregister(&sub_leds[FLED2].fled_cdev);
- cancel_work_sync(&sub_leds[FLED2].work_brightness_set);
}
mutex_destroy(&led->lock);
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 11/57] Documentation: leds: Add description of brightness_set* ops
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (9 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 10/57] leds: max77693: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 12/57] leds: tlc591xx: Remove work queue Jacek Anaszewski
` (45 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski,
Andrew Lunn, Sakari Ailus, Pavel Machek
This patch adds description of brightness_set and a recently introduced
brightness_set_nonblocking op.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Bryan Wu <cooloney@gmail.com>
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: Sakari Ailus <sakari.ailus@linux.intel.com>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Stas Sergeev <stsp@users.sourceforge.net>
---
Documentation/leds/leds-class.txt | 20 ++++++++++++++++++++
1 file changed, 20 insertions(+)
diff --git a/Documentation/leds/leds-class.txt b/Documentation/leds/leds-class.txt
index 62261c0..fde8fef 100644
--- a/Documentation/leds/leds-class.txt
+++ b/Documentation/leds/leds-class.txt
@@ -52,6 +52,26 @@ above leaves scope for further attributes should they be needed. If sections
of the name don't apply, just leave that section blank.
+Brightness setting callbacks
+============================
+
+LED subsystem core exposes two function pointers for setting brightness:
+
+ - brightness_set : Intended for drivers that may set brightness in
+ a blocking way. For a long time this was the only
+ op for setting brightness and it was drivers'
+ responsibility to assure that it will not block the
+ caller. With the new approach drivers are required to set
+ LED_BRIGHTNESS_BLOCKING flag, when using this op, to make
+ LED core aware that it should call it from a work queue
+ task, when brightness is set with led_set_brightness API.
+ In case synchronous operation is needed it should be
+ explicitly requested with led_set_brightness_sync API.
+
+ - brightness_set_nonblocking : Intended for drivers that set brightness in
+ a non-blocking way.
+
+
Hardware accelerated blink of LEDs
==================================
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 12/57] leds: tlc591xx: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (10 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 11/57] Documentation: leds: Add description of brightness_set* ops Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 13/57] leds: 88pm860x: " Jacek Anaszewski
` (44 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn, Jacek Anaszewski
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the driver.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-tlc591xx.c | 25 +++++--------------------
1 file changed, 5 insertions(+), 20 deletions(-)
diff --git a/drivers/leds/leds-tlc591xx.c b/drivers/leds/leds-tlc591xx.c
index b806eca..f7e6711 100644
--- a/drivers/leds/leds-tlc591xx.c
+++ b/drivers/leds/leds-tlc591xx.c
@@ -14,7 +14,6 @@
#include <linux/of_device.h>
#include <linux/regmap.h>
#include <linux/slab.h>
-#include <linux/workqueue.h>
#define TLC591XX_MAX_LEDS 16
@@ -42,13 +41,11 @@
#define LEDOUT_MASK 0x3
#define ldev_to_led(c) container_of(c, struct tlc591xx_led, ldev)
-#define work_to_led(work) container_of(work, struct tlc591xx_led, work)
struct tlc591xx_led {
bool active;
unsigned int led_no;
struct led_classdev ldev;
- struct work_struct work;
struct tlc591xx_priv *priv;
};
@@ -111,11 +108,11 @@ tlc591xx_set_pwm(struct tlc591xx_priv *priv, struct tlc591xx_led *led,
}
static void
-tlc591xx_led_work(struct work_struct *work)
+tlc591xx_brightness_set(struct led_classdev *led_cdev,
+ enum led_brightness brightness)
{
- struct tlc591xx_led *led = work_to_led(work);
+ struct tlc591xx_led *led = ldev_to_led(led_cdev);
struct tlc591xx_priv *priv = led->priv;
- enum led_brightness brightness = led->ldev.brightness;
int err;
switch (brightness) {
@@ -136,25 +133,13 @@ tlc591xx_led_work(struct work_struct *work)
}
static void
-tlc591xx_brightness_set(struct led_classdev *led_cdev,
- enum led_brightness brightness)
-{
- struct tlc591xx_led *led = ldev_to_led(led_cdev);
-
- led->ldev.brightness = brightness;
- schedule_work(&led->work);
-}
-
-static void
tlc591xx_destroy_devices(struct tlc591xx_priv *priv, unsigned int j)
{
int i = j;
while (--i >= 0) {
- if (priv->leds[i].active) {
+ if (priv->leds[i].active)
led_classdev_unregister(&priv->leds[i].ldev);
- cancel_work_sync(&priv->leds[i].work);
- }
}
}
@@ -177,7 +162,7 @@ tlc591xx_configure(struct device *dev,
led->led_no = i;
led->ldev.brightness_set = tlc591xx_brightness_set;
led->ldev.max_brightness = LED_FULL;
- INIT_WORK(&led->work, tlc591xx_led_work);
+ led->ldev.flags |= LED_BRIGHTNESS_BLOCKING;
err = led_classdev_register(dev, &led->ldev);
if (err < 0) {
dev_err(dev, "couldn't register LED %s\n",
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 13/57] leds: 88pm860x: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (11 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 12/57] leds: tlc591xx: Remove work queue Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 14/57] leds: adp5520: " Jacek Anaszewski
` (43 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn, Jacek Anaszewski
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the driver.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-88pm860x.c | 22 ++++++----------------
1 file changed, 6 insertions(+), 16 deletions(-)
diff --git a/drivers/leds/leds-88pm860x.c b/drivers/leds/leds-88pm860x.c
index 1497a09..a2f7b10 100644
--- a/drivers/leds/leds-88pm860x.c
+++ b/drivers/leds/leds-88pm860x.c
@@ -16,7 +16,6 @@
#include <linux/i2c.h>
#include <linux/leds.h>
#include <linux/slab.h>
-#include <linux/workqueue.h>
#include <linux/mfd/88pm860x.h>
#include <linux/module.h>
@@ -33,7 +32,6 @@
struct pm860x_led {
struct led_classdev cdev;
struct i2c_client *i2c;
- struct work_struct work;
struct pm860x_chip *chip;
struct mutex lock;
char name[MFD_NAME_SIZE];
@@ -69,17 +67,18 @@ static int led_power_set(struct pm860x_chip *chip, int port, int on)
return ret;
}
-static void pm860x_led_work(struct work_struct *work)
+static void pm860x_led_set(struct led_classdev *cdev,
+ enum led_brightness value)
{
-
- struct pm860x_led *led;
+ struct pm860x_led *led = container_of(cdev, struct pm860x_led, cdev);
struct pm860x_chip *chip;
unsigned char buf[3];
int ret;
- led = container_of(work, struct pm860x_led, work);
chip = led->chip;
mutex_lock(&led->lock);
+ led->brightness = value >> 3;
+
if ((led->current_brightness == 0) && led->brightness) {
led_power_set(chip, led->port, 1);
if (led->iset) {
@@ -114,15 +113,6 @@ static void pm860x_led_work(struct work_struct *work)
mutex_unlock(&led->lock);
}
-static void pm860x_led_set(struct led_classdev *cdev,
- enum led_brightness value)
-{
- struct pm860x_led *data = container_of(cdev, struct pm860x_led, cdev);
-
- data->brightness = value >> 3;
- schedule_work(&data->work);
-}
-
#ifdef CONFIG_OF
static int pm860x_led_dt_init(struct platform_device *pdev,
struct pm860x_led *data)
@@ -213,8 +203,8 @@ static int pm860x_led_probe(struct platform_device *pdev)
data->current_brightness = 0;
data->cdev.name = data->name;
data->cdev.brightness_set = pm860x_led_set;
+ data->cdev.flags = LED_BRIGHTNESS_BLOCKING;
mutex_init(&data->lock);
- INIT_WORK(&data->work, pm860x_led_work);
ret = led_classdev_register(chip->dev, &data->cdev);
if (ret < 0) {
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 14/57] leds: adp5520: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (12 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 13/57] leds: 88pm860x: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 15/57] leds: bd2802: " Jacek Anaszewski
` (42 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn,
Michael Hennerich, Jacek Anaszewski
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the driver.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Cc: Michael Hennerich <michael.hennerich@analog.com>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-adp5520.c | 23 ++++-------------------
1 file changed, 4 insertions(+), 19 deletions(-)
diff --git a/drivers/leds/leds-adp5520.c b/drivers/leds/leds-adp5520.c
index 07e66ca..d09f742 100644
--- a/drivers/leds/leds-adp5520.c
+++ b/drivers/leds/leds-adp5520.c
@@ -17,34 +17,24 @@
#include <linux/kernel.h>
#include <linux/platform_device.h>
#include <linux/leds.h>
-#include <linux/workqueue.h>
#include <linux/mfd/adp5520.h>
#include <linux/slab.h>
struct adp5520_led {
struct led_classdev cdev;
- struct work_struct work;
struct device *master;
- enum led_brightness new_brightness;
int id;
int flags;
};
-static void adp5520_led_work(struct work_struct *work)
-{
- struct adp5520_led *led = container_of(work, struct adp5520_led, work);
- adp5520_write(led->master, ADP5520_LED1_CURRENT + led->id - 1,
- led->new_brightness >> 2);
-}
-
static void adp5520_led_set(struct led_classdev *led_cdev,
enum led_brightness value)
{
struct adp5520_led *led;
led = container_of(led_cdev, struct adp5520_led, cdev);
- led->new_brightness = value;
- schedule_work(&led->work);
+ adp5520_write(led->master, ADP5520_LED1_CURRENT + led->id - 1,
+ value >> 2);
}
static int adp5520_led_setup(struct adp5520_led *led)
@@ -137,6 +127,7 @@ static int adp5520_led_probe(struct platform_device *pdev)
led_dat->cdev.default_trigger = cur_led->default_trigger;
led_dat->cdev.brightness_set = adp5520_led_set;
led_dat->cdev.brightness = LED_OFF;
+ led_dat->cdev.flags |= LED_BRIGHTNESS_BLOCKING;
if (cur_led->flags & ADP5520_FLAG_LED_MASK)
led_dat->flags = cur_led->flags;
@@ -146,9 +137,6 @@ static int adp5520_led_probe(struct platform_device *pdev)
led_dat->id = led_dat->flags & ADP5520_FLAG_LED_MASK;
led_dat->master = pdev->dev.parent;
- led_dat->new_brightness = LED_OFF;
-
- INIT_WORK(&led_dat->work, adp5520_led_work);
ret = led_classdev_register(led_dat->master, &led_dat->cdev);
if (ret) {
@@ -170,10 +158,8 @@ static int adp5520_led_probe(struct platform_device *pdev)
err:
if (i > 0) {
- for (i = i - 1; i >= 0; i--) {
+ for (i = i - 1; i >= 0; i--)
led_classdev_unregister(&led[i].cdev);
- cancel_work_sync(&led[i].work);
- }
}
return ret;
@@ -192,7 +178,6 @@ static int adp5520_led_remove(struct platform_device *pdev)
for (i = 0; i < pdata->num_leds; i++) {
led_classdev_unregister(&led[i].cdev);
- cancel_work_sync(&led[i].work);
}
return 0;
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 15/57] leds: bd2802: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (13 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 14/57] leds: adp5520: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 16/57] leds: blinkm: " Jacek Anaszewski
` (41 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn, Kim Kyuwon,
Jacek Anaszewski
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the driver.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Cc: Kim Kyuwon <q1.kim@samsung.com>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-bd2802.c | 32 +++++++++++++-------------------
1 file changed, 13 insertions(+), 19 deletions(-)
diff --git a/drivers/leds/leds-bd2802.c b/drivers/leds/leds-bd2802.c
index 6078c15d..9d1535a 100644
--- a/drivers/leds/leds-bd2802.c
+++ b/drivers/leds/leds-bd2802.c
@@ -72,7 +72,6 @@ struct bd2802_led {
struct bd2802_led_platform_data *pdata;
struct i2c_client *client;
struct rw_semaphore rwsem;
- struct work_struct work;
struct led_state led[2];
@@ -518,16 +517,6 @@ static struct device_attribute *bd2802_attributes[] = {
&bd2802_rgb_current_attr,
};
-static void bd2802_led_work(struct work_struct *work)
-{
- struct bd2802_led *led = container_of(work, struct bd2802_led, work);
-
- if (led->state)
- bd2802_turn_on(led, led->led_id, led->color, led->state);
- else
- bd2802_turn_off(led, led->led_id, led->color);
-}
-
#define BD2802_CONTROL_RGBS(name, id, clr) \
static void bd2802_set_##name##_brightness(struct led_classdev *led_cdev,\
enum led_brightness value) \
@@ -536,11 +525,13 @@ static void bd2802_set_##name##_brightness(struct led_classdev *led_cdev,\
container_of(led_cdev, struct bd2802_led, cdev_##name); \
led->led_id = id; \
led->color = clr; \
- if (value == LED_OFF) \
+ if (value == LED_OFF) { \
led->state = BD2802_OFF; \
- else \
+ bd2802_turn_off(led, led->led_id, led->color); \
+ } else { \
led->state = BD2802_ON; \
- schedule_work(&led->work); \
+ bd2802_turn_on(led, led->led_id, led->color, BD2802_ON);\
+ } \
} \
static int bd2802_set_##name##_blink(struct led_classdev *led_cdev, \
unsigned long *delay_on, unsigned long *delay_off) \
@@ -552,7 +543,7 @@ static int bd2802_set_##name##_blink(struct led_classdev *led_cdev, \
led->led_id = id; \
led->color = clr; \
led->state = BD2802_BLINK; \
- schedule_work(&led->work); \
+ bd2802_turn_on(led, led->led_id, led->color, BD2802_BLINK); \
return 0; \
}
@@ -567,12 +558,11 @@ static int bd2802_register_led_classdev(struct bd2802_led *led)
{
int ret;
- INIT_WORK(&led->work, bd2802_led_work);
-
led->cdev_led1r.name = "led1_R";
led->cdev_led1r.brightness = LED_OFF;
led->cdev_led1r.brightness_set = bd2802_set_led1r_brightness;
led->cdev_led1r.blink_set = bd2802_set_led1r_blink;
+ led->cdev_led1r.flags |= LED_BRIGHTNESS_BLOCKING;
ret = led_classdev_register(&led->client->dev, &led->cdev_led1r);
if (ret < 0) {
@@ -585,6 +575,7 @@ static int bd2802_register_led_classdev(struct bd2802_led *led)
led->cdev_led1g.brightness = LED_OFF;
led->cdev_led1g.brightness_set = bd2802_set_led1g_brightness;
led->cdev_led1g.blink_set = bd2802_set_led1g_blink;
+ led->cdev_led1g.flags |= LED_BRIGHTNESS_BLOCKING;
ret = led_classdev_register(&led->client->dev, &led->cdev_led1g);
if (ret < 0) {
@@ -597,6 +588,7 @@ static int bd2802_register_led_classdev(struct bd2802_led *led)
led->cdev_led1b.brightness = LED_OFF;
led->cdev_led1b.brightness_set = bd2802_set_led1b_brightness;
led->cdev_led1b.blink_set = bd2802_set_led1b_blink;
+ led->cdev_led1b.flags |= LED_BRIGHTNESS_BLOCKING;
ret = led_classdev_register(&led->client->dev, &led->cdev_led1b);
if (ret < 0) {
@@ -609,6 +601,7 @@ static int bd2802_register_led_classdev(struct bd2802_led *led)
led->cdev_led2r.brightness = LED_OFF;
led->cdev_led2r.brightness_set = bd2802_set_led2r_brightness;
led->cdev_led2r.blink_set = bd2802_set_led2r_blink;
+ led->cdev_led2r.flags |= LED_BRIGHTNESS_BLOCKING;
ret = led_classdev_register(&led->client->dev, &led->cdev_led2r);
if (ret < 0) {
@@ -621,6 +614,7 @@ static int bd2802_register_led_classdev(struct bd2802_led *led)
led->cdev_led2g.brightness = LED_OFF;
led->cdev_led2g.brightness_set = bd2802_set_led2g_brightness;
led->cdev_led2g.blink_set = bd2802_set_led2g_blink;
+ led->cdev_led2g.flags |= LED_BRIGHTNESS_BLOCKING;
ret = led_classdev_register(&led->client->dev, &led->cdev_led2g);
if (ret < 0) {
@@ -633,7 +627,8 @@ static int bd2802_register_led_classdev(struct bd2802_led *led)
led->cdev_led2b.brightness = LED_OFF;
led->cdev_led2b.brightness_set = bd2802_set_led2b_brightness;
led->cdev_led2b.blink_set = bd2802_set_led2b_blink;
- led->cdev_led2b.flags |= LED_CORE_SUSPENDRESUME;
+ led->cdev_led2b.flags |= LED_CORE_SUSPENDRESUME |
+ LED_BRIGHTNESS_BLOCKING;
ret = led_classdev_register(&led->client->dev, &led->cdev_led2b);
if (ret < 0) {
@@ -661,7 +656,6 @@ failed_unregister_led1_R:
static void bd2802_unregister_led_classdev(struct bd2802_led *led)
{
- cancel_work_sync(&led->work);
led_classdev_unregister(&led->cdev_led2b);
led_classdev_unregister(&led->cdev_led2g);
led_classdev_unregister(&led->cdev_led2r);
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 16/57] leds: blinkm: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (14 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 15/57] leds: bd2802: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 17/57] leds: lm3533: " Jacek Anaszewski
` (40 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn,
Jan-Simon Moeller, Jacek Anaszewski
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the driver.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Cc: Jan-Simon Moeller <dl9pf@gmx.de>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-blinkm.c | 69 +++++---------------------------------------
1 file changed, 8 insertions(+), 61 deletions(-)
diff --git a/drivers/leds/leds-blinkm.c b/drivers/leds/leds-blinkm.c
index d0452b0..f1c8668 100644
--- a/drivers/leds/leds-blinkm.c
+++ b/drivers/leds/leds-blinkm.c
@@ -39,16 +39,9 @@ struct blinkm_led {
struct i2c_client *i2c_client;
struct led_classdev led_cdev;
int id;
- atomic_t active;
-};
-
-struct blinkm_work {
- struct blinkm_led *blinkm_led;
- struct work_struct work;
};
#define cdev_to_blmled(c) container_of(c, struct blinkm_led, led_cdev)
-#define work_to_blmwork(c) container_of(c, struct blinkm_work, work)
struct blinkm_data {
struct i2c_client *i2c_client;
@@ -439,65 +432,30 @@ static int blinkm_transfer_hw(struct i2c_client *client, int cmd)
return 0;
}
-static void led_work(struct work_struct *work)
-{
- int ret;
- struct blinkm_led *led;
- struct blinkm_data *data;
- struct blinkm_work *blm_work = work_to_blmwork(work);
-
- led = blm_work->blinkm_led;
- data = i2c_get_clientdata(led->i2c_client);
- ret = blinkm_transfer_hw(led->i2c_client, BLM_GO_RGB);
- atomic_dec(&led->active);
- dev_dbg(&led->i2c_client->dev,
- "# DONE # next_red = %d, next_green = %d,"
- " next_blue = %d, active = %d\n",
- data->next_red, data->next_green,
- data->next_blue, atomic_read(&led->active));
- kfree(blm_work);
-}
-
static int blinkm_led_common_set(struct led_classdev *led_cdev,
enum led_brightness value, int color)
{
/* led_brightness is 0, 127 or 255 - we just use it here as-is */
struct blinkm_led *led = cdev_to_blmled(led_cdev);
struct blinkm_data *data = i2c_get_clientdata(led->i2c_client);
- struct blinkm_work *bl_work;
switch (color) {
case RED:
/* bail out if there's no change */
if (data->next_red == (u8) value)
return 0;
- /* we assume a quite fast sequence here ([off]->on->off)
- * think of network led trigger - we cannot blink that fast, so
- * in case we already have a off->on->off transition queued up,
- * we refuse to queue up more.
- * Revisit: fast-changing brightness. */
- if (atomic_read(&led->active) > 1)
- return 0;
data->next_red = (u8) value;
break;
case GREEN:
/* bail out if there's no change */
if (data->next_green == (u8) value)
return 0;
- /* we assume a quite fast sequence here ([off]->on->off)
- * Revisit: fast-changing brightness. */
- if (atomic_read(&led->active) > 1)
- return 0;
data->next_green = (u8) value;
break;
case BLUE:
/* bail out if there's no change */
if (data->next_blue == (u8) value)
return 0;
- /* we assume a quite fast sequence here ([off]->on->off)
- * Revisit: fast-changing brightness. */
- if (atomic_read(&led->active) > 1)
- return 0;
data->next_blue = (u8) value;
break;
@@ -506,23 +464,12 @@ static int blinkm_led_common_set(struct led_classdev *led_cdev,
return -EINVAL;
}
- bl_work = kzalloc(sizeof(*bl_work), GFP_ATOMIC);
- if (!bl_work)
- return -ENOMEM;
-
- atomic_inc(&led->active);
+ blinkm_transfer_hw(led->i2c_client, BLM_GO_RGB);
dev_dbg(&led->i2c_client->dev,
- "#TO_SCHED# next_red = %d, next_green = %d,"
- " next_blue = %d, active = %d\n",
+ "# DONE # next_red = %d, next_green = %d,"
+ " next_blue = %d\n",
data->next_red, data->next_green,
- data->next_blue, atomic_read(&led->active));
-
- /* a fresh work _item_ for each change */
- bl_work->blinkm_led = led;
- INIT_WORK(&bl_work->work, led_work);
- /* queue work in own queue for easy sync on exit*/
- schedule_work(&bl_work->work);
-
+ data->next_blue);
return 0;
}
@@ -669,7 +616,6 @@ static int blinkm_probe(struct i2c_client *client,
led[i]->id = i;
led[i]->led_cdev.max_brightness = 255;
led[i]->led_cdev.flags = LED_CORE_SUSPENDRESUME;
- atomic_set(&led[i]->active, 0);
switch (i) {
case RED:
snprintf(blinkm_led_name, sizeof(blinkm_led_name),
@@ -678,6 +624,7 @@ static int blinkm_probe(struct i2c_client *client,
client->addr);
led[i]->led_cdev.name = blinkm_led_name;
led[i]->led_cdev.brightness_set = blinkm_led_red_set;
+ led[i]->led_cdev.flags |= LED_BRIGHTNESS_BLOCKING;
err = led_classdev_register(&client->dev,
&led[i]->led_cdev);
if (err < 0) {
@@ -694,6 +641,7 @@ static int blinkm_probe(struct i2c_client *client,
client->addr);
led[i]->led_cdev.name = blinkm_led_name;
led[i]->led_cdev.brightness_set = blinkm_led_green_set;
+ led[i]->led_cdev.flags |= LED_BRIGHTNESS_BLOCKING;
err = led_classdev_register(&client->dev,
&led[i]->led_cdev);
if (err < 0) {
@@ -710,6 +658,7 @@ static int blinkm_probe(struct i2c_client *client,
client->addr);
led[i]->led_cdev.name = blinkm_led_name;
led[i]->led_cdev.brightness_set = blinkm_led_blue_set;
+ led[i]->led_cdev.flags |= LED_BRIGHTNESS_BLOCKING;
err = led_classdev_register(&client->dev,
&led[i]->led_cdev);
if (err < 0) {
@@ -746,10 +695,8 @@ static int blinkm_remove(struct i2c_client *client)
int i;
/* make sure no workqueue entries are pending */
- for (i = 0; i < 3; i++) {
- flush_scheduled_work();
+ for (i = 0; i < 3; i++)
led_classdev_unregister(&data->blinkm_leds[i].led_cdev);
- }
/* reset rgb */
data->next_red = 0x00;
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 17/57] leds: lm3533: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (15 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 16/57] leds: blinkm: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 18/57] leds: lm3642: " Jacek Anaszewski
` (39 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn,
Jacek Anaszewski, Johan Hovold
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the driver.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Johan Hovold <johan@kernel.org>
---
drivers/leds/leds-lm3533.c | 27 +++++----------------------
1 file changed, 5 insertions(+), 22 deletions(-)
diff --git a/drivers/leds/leds-lm3533.c b/drivers/leds/leds-lm3533.c
index 6e2e020..5a9ae74 100644
--- a/drivers/leds/leds-lm3533.c
+++ b/drivers/leds/leds-lm3533.c
@@ -17,7 +17,6 @@
#include <linux/mutex.h>
#include <linux/platform_device.h>
#include <linux/slab.h>
-#include <linux/workqueue.h>
#include <linux/mfd/lm3533.h>
@@ -53,9 +52,6 @@ struct lm3533_led {
struct mutex mutex;
unsigned long flags;
-
- struct work_struct work;
- u8 new_brightness;
};
@@ -123,18 +119,6 @@ out:
return ret;
}
-static void lm3533_led_work(struct work_struct *work)
-{
- struct lm3533_led *led = container_of(work, struct lm3533_led, work);
-
- dev_dbg(led->cdev.dev, "%s - %u\n", __func__, led->new_brightness);
-
- if (led->new_brightness == 0)
- lm3533_led_pattern_enable(led, 0); /* disable blink */
-
- lm3533_ctrlbank_set_brightness(&led->cb, led->new_brightness);
-}
-
static void lm3533_led_set(struct led_classdev *cdev,
enum led_brightness value)
{
@@ -142,8 +126,10 @@ static void lm3533_led_set(struct led_classdev *cdev,
dev_dbg(led->cdev.dev, "%s - %d\n", __func__, value);
- led->new_brightness = value;
- schedule_work(&led->work);
+ if (value == 0)
+ lm3533_led_pattern_enable(led, 0); /* disable blink */
+
+ lm3533_ctrlbank_set_brightness(&led->cb, value);
}
static enum led_brightness lm3533_led_get(struct led_classdev *cdev)
@@ -698,10 +684,10 @@ static int lm3533_led_probe(struct platform_device *pdev)
led->cdev.blink_set = lm3533_led_blink_set;
led->cdev.brightness = LED_OFF;
led->cdev.groups = lm3533_led_attribute_groups,
+ led->cdev.flags |= LED_BRIGHTNESS_BLOCKING;
led->id = pdev->id;
mutex_init(&led->mutex);
- INIT_WORK(&led->work, lm3533_led_work);
/* The class framework makes a callback to get brightness during
* registration so use parent device (for error reporting) until
@@ -733,7 +719,6 @@ static int lm3533_led_probe(struct platform_device *pdev)
err_unregister:
led_classdev_unregister(&led->cdev);
- flush_work(&led->work);
return ret;
}
@@ -746,7 +731,6 @@ static int lm3533_led_remove(struct platform_device *pdev)
lm3533_ctrlbank_disable(&led->cb);
led_classdev_unregister(&led->cdev);
- flush_work(&led->work);
return 0;
}
@@ -760,7 +744,6 @@ static void lm3533_led_shutdown(struct platform_device *pdev)
lm3533_ctrlbank_disable(&led->cb);
lm3533_led_set(&led->cdev, LED_OFF); /* disable blink */
- flush_work(&led->work);
}
static struct platform_driver lm3533_led_driver = {
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 18/57] leds: lm3642: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (16 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 17/57] leds: lm3533: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 19/57] leds: pca9532: Remove work queue for LEDs Jacek Anaszewski
` (38 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn, Daniel Jeong,
G.Shark Jeong, Jacek Anaszewski
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the driver.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Cc: Daniel Jeong <daniel.jeong@ti.com>
Cc: G.Shark Jeong <gshark.jeong@gmail.com>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-lm3642.c | 57 ++++++++++----------------------------------
1 file changed, 12 insertions(+), 45 deletions(-)
diff --git a/drivers/leds/leds-lm3642.c b/drivers/leds/leds-lm3642.c
index 02ebe34..38cbebb 100644
--- a/drivers/leds/leds-lm3642.c
+++ b/drivers/leds/leds-lm3642.c
@@ -15,7 +15,6 @@
#include <linux/platform_device.h>
#include <linux/fs.h>
#include <linux/regmap.h>
-#include <linux/workqueue.h>
#include <linux/platform_data/leds-lm3642.h>
#define REG_FILT_TIME (0x0)
@@ -73,10 +72,6 @@ struct lm3642_chip_data {
struct led_classdev cdev_torch;
struct led_classdev cdev_indicator;
- struct work_struct work_flash;
- struct work_struct work_torch;
- struct work_struct work_indicator;
-
u8 br_flash;
u8 br_torch;
u8 br_indicator;
@@ -209,24 +204,16 @@ out_strtoint:
static DEVICE_ATTR(torch_pin, S_IWUSR, NULL, lm3642_torch_pin_store);
-static void lm3642_deferred_torch_brightness_set(struct work_struct *work)
-{
- struct lm3642_chip_data *chip =
- container_of(work, struct lm3642_chip_data, work_torch);
-
- mutex_lock(&chip->lock);
- lm3642_control(chip, chip->br_torch, MODES_TORCH);
- mutex_unlock(&chip->lock);
-}
-
static void lm3642_torch_brightness_set(struct led_classdev *cdev,
enum led_brightness brightness)
{
struct lm3642_chip_data *chip =
container_of(cdev, struct lm3642_chip_data, cdev_torch);
+ mutex_lock(&chip->lock);
chip->br_torch = brightness;
- schedule_work(&chip->work_torch);
+ lm3642_control(chip, chip->br_torch, MODES_TORCH);
+ mutex_unlock(&chip->lock);
}
/* flash */
@@ -266,45 +253,29 @@ out_strtoint:
static DEVICE_ATTR(strobe_pin, S_IWUSR, NULL, lm3642_strobe_pin_store);
-static void lm3642_deferred_strobe_brightness_set(struct work_struct *work)
-{
- struct lm3642_chip_data *chip =
- container_of(work, struct lm3642_chip_data, work_flash);
-
- mutex_lock(&chip->lock);
- lm3642_control(chip, chip->br_flash, MODES_FLASH);
- mutex_unlock(&chip->lock);
-}
-
static void lm3642_strobe_brightness_set(struct led_classdev *cdev,
enum led_brightness brightness)
{
struct lm3642_chip_data *chip =
container_of(cdev, struct lm3642_chip_data, cdev_flash);
- chip->br_flash = brightness;
- schedule_work(&chip->work_flash);
-}
-
-/* indicator */
-static void lm3642_deferred_indicator_brightness_set(struct work_struct *work)
-{
- struct lm3642_chip_data *chip =
- container_of(work, struct lm3642_chip_data, work_indicator);
-
mutex_lock(&chip->lock);
- lm3642_control(chip, chip->br_indicator, MODES_INDIC);
+ chip->br_flash = brightness;
+ lm3642_control(chip, chip->br_flash, MODES_FLASH);
mutex_unlock(&chip->lock);
}
+/* indicator */
static void lm3642_indicator_brightness_set(struct led_classdev *cdev,
enum led_brightness brightness)
{
struct lm3642_chip_data *chip =
container_of(cdev, struct lm3642_chip_data, cdev_indicator);
+ mutex_lock(&chip->lock);
chip->br_indicator = brightness;
- schedule_work(&chip->work_indicator);
+ lm3642_control(chip, chip->br_indicator, MODES_INDIC);
+ mutex_unlock(&chip->lock);
}
static const struct regmap_config lm3642_regmap = {
@@ -371,12 +342,12 @@ static int lm3642_probe(struct i2c_client *client,
goto err_out;
/* flash */
- INIT_WORK(&chip->work_flash, lm3642_deferred_strobe_brightness_set);
chip->cdev_flash.name = "flash";
chip->cdev_flash.max_brightness = 16;
chip->cdev_flash.brightness_set = lm3642_strobe_brightness_set;
chip->cdev_flash.default_trigger = "flash";
chip->cdev_flash.groups = lm3642_flash_groups,
+ chip->cdev_flash.flags |= LED_BRIGHTNESS_BLOCKING;
err = led_classdev_register((struct device *)
&client->dev, &chip->cdev_flash);
if (err < 0) {
@@ -385,12 +356,12 @@ static int lm3642_probe(struct i2c_client *client,
}
/* torch */
- INIT_WORK(&chip->work_torch, lm3642_deferred_torch_brightness_set);
chip->cdev_torch.name = "torch";
chip->cdev_torch.max_brightness = 8;
chip->cdev_torch.brightness_set = lm3642_torch_brightness_set;
chip->cdev_torch.default_trigger = "torch";
chip->cdev_torch.groups = lm3642_torch_groups,
+ chip->cdev_torch.flags |= LED_BRIGHTNESS_BLOCKING;
err = led_classdev_register((struct device *)
&client->dev, &chip->cdev_torch);
if (err < 0) {
@@ -399,11 +370,10 @@ static int lm3642_probe(struct i2c_client *client,
}
/* indicator */
- INIT_WORK(&chip->work_indicator,
- lm3642_deferred_indicator_brightness_set);
chip->cdev_indicator.name = "indicator";
chip->cdev_indicator.max_brightness = 8;
chip->cdev_indicator.brightness_set = lm3642_indicator_brightness_set;
+ chip->cdev_indicator.flags |= LED_BRIGHTNESS_BLOCKING;
err = led_classdev_register((struct device *)
&client->dev, &chip->cdev_indicator);
if (err < 0) {
@@ -427,11 +397,8 @@ static int lm3642_remove(struct i2c_client *client)
struct lm3642_chip_data *chip = i2c_get_clientdata(client);
led_classdev_unregister(&chip->cdev_indicator);
- flush_work(&chip->work_indicator);
led_classdev_unregister(&chip->cdev_torch);
- flush_work(&chip->work_torch);
led_classdev_unregister(&chip->cdev_flash);
- flush_work(&chip->work_flash);
regmap_write(chip->regmap, REG_ENABLE, 0);
return 0;
}
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 19/57] leds: pca9532: Remove work queue for LEDs.
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (17 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 18/57] leds: lm3642: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 20/57] leds: lp3944: Remove work queue Jacek Anaszewski
` (37 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn, Jacek Anaszewski
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the driver.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-pca9532.c | 21 ++++++++-------------
1 file changed, 8 insertions(+), 13 deletions(-)
diff --git a/drivers/leds/leds-pca9532.c b/drivers/leds/leds-pca9532.c
index 5a6363d..18d4d61 100644
--- a/drivers/leds/leds-pca9532.c
+++ b/drivers/leds/leds-pca9532.c
@@ -174,7 +174,9 @@ static void pca9532_set_brightness(struct led_classdev *led_cdev,
if (err)
return; /* XXX: led api doesn't allow error code? */
}
- schedule_work(&led->work);
+ if (led->state == PCA9532_PWM0)
+ pca9532_setpwm(led->client, 0);
+ pca9532_setled(led);
}
static int pca9532_set_blink(struct led_classdev *led_cdev,
@@ -198,7 +200,10 @@ static int pca9532_set_blink(struct led_classdev *led_cdev,
err = pca9532_calcpwm(client, 0, psc, led_cdev->brightness);
if (err)
return err;
- schedule_work(&led->work);
+ if (led->state == PCA9532_PWM0)
+ pca9532_setpwm(led->client, 0);
+ pca9532_setled(led);
+
return 0;
}
@@ -233,15 +238,6 @@ static void pca9532_input_work(struct work_struct *work)
mutex_unlock(&data->update_lock);
}
-static void pca9532_led_work(struct work_struct *work)
-{
- struct pca9532_led *led;
- led = container_of(work, struct pca9532_led, work);
- if (led->state == PCA9532_PWM0)
- pca9532_setpwm(led->client, 0);
- pca9532_setled(led);
-}
-
#ifdef CONFIG_LEDS_PCA9532_GPIO
static int pca9532_gpio_request_pin(struct gpio_chip *gc, unsigned offset)
{
@@ -307,7 +303,6 @@ static int pca9532_destroy_devices(struct pca9532_data *data, int n_devs)
break;
case PCA9532_TYPE_LED:
led_classdev_unregister(&data->leds[i].ldev);
- cancel_work_sync(&data->leds[i].work);
break;
case PCA9532_TYPE_N2100_BEEP:
if (data->idev != NULL) {
@@ -361,7 +356,7 @@ static int pca9532_configure(struct i2c_client *client,
led->ldev.brightness = LED_OFF;
led->ldev.brightness_set = pca9532_set_brightness;
led->ldev.blink_set = pca9532_set_blink;
- INIT_WORK(&led->work, pca9532_led_work);
+ led->ldev.flags |= LED_BRIGHTNESS_BLOCKING;
err = led_classdev_register(&client->dev, &led->ldev);
if (err < 0) {
dev_err(&client->dev,
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 20/57] leds: lp3944: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (18 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 19/57] leds: pca9532: Remove work queue for LEDs Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 21/57] leds: lp55xx: " Jacek Anaszewski
` (36 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn,
Antonio Ospite, Jacek Anaszewski
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the driver.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Cc: Antonio Ospite <ospite@studenti.unina.it>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-lp3944.c | 30 +++++++-----------------------
1 file changed, 7 insertions(+), 23 deletions(-)
diff --git a/drivers/leds/leds-lp3944.c b/drivers/leds/leds-lp3944.c
index 53144fb..4b90663 100644
--- a/drivers/leds/leds-lp3944.c
+++ b/drivers/leds/leds-lp3944.c
@@ -31,7 +31,6 @@
#include <linux/slab.h>
#include <linux/leds.h>
#include <linux/mutex.h>
-#include <linux/workqueue.h>
#include <linux/leds-lp3944.h>
/* Read Only Registers */
@@ -68,10 +67,8 @@
struct lp3944_led_data {
u8 id;
enum lp3944_type type;
- enum lp3944_status status;
struct led_classdev ldev;
struct i2c_client *client;
- struct work_struct work;
};
struct lp3944_data {
@@ -275,8 +272,7 @@ static int lp3944_led_set_blink(struct led_classdev *led_cdev,
dev_dbg(&led->client->dev, "%s: OK hardware accelerated blink!\n",
__func__);
- led->status = LP3944_LED_STATUS_DIM0;
- schedule_work(&led->work);
+ lp3944_led_set(led, LP3944_LED_STATUS_DIM0);
return 0;
}
@@ -289,16 +285,7 @@ static void lp3944_led_set_brightness(struct led_classdev *led_cdev,
dev_dbg(&led->client->dev, "%s: %s, %d\n",
__func__, led_cdev->name, brightness);
- led->status = !!brightness;
- schedule_work(&led->work);
-}
-
-static void lp3944_led_work(struct work_struct *work)
-{
- struct lp3944_led_data *led;
-
- led = container_of(work, struct lp3944_led_data, work);
- lp3944_led_set(led, led->status);
+ lp3944_led_set(led, !!brightness);
}
static int lp3944_configure(struct i2c_client *client,
@@ -318,14 +305,13 @@ static int lp3944_configure(struct i2c_client *client,
case LP3944_LED_TYPE_LED:
case LP3944_LED_TYPE_LED_INVERTED:
led->type = pled->type;
- led->status = pled->status;
led->ldev.name = pled->name;
led->ldev.max_brightness = 1;
led->ldev.brightness_set = lp3944_led_set_brightness;
led->ldev.blink_set = lp3944_led_set_blink;
- led->ldev.flags = LED_CORE_SUSPENDRESUME;
+ led->ldev.flags = LED_CORE_SUSPENDRESUME |
+ LED_BRIGHTNESS_BLOCKING;
- INIT_WORK(&led->work, lp3944_led_work);
err = led_classdev_register(&client->dev, &led->ldev);
if (err < 0) {
dev_err(&client->dev,
@@ -336,14 +322,14 @@ static int lp3944_configure(struct i2c_client *client,
/* to expose the default value to userspace */
led->ldev.brightness =
- (enum led_brightness) led->status;
+ (enum led_brightness) pled->status;
/* Set the default led status */
- err = lp3944_led_set(led, led->status);
+ err = lp3944_led_set(led, pled->status);
if (err < 0) {
dev_err(&client->dev,
"%s couldn't set STATUS %d\n",
- led->ldev.name, led->status);
+ led->ldev.name, pled->status);
goto exit;
}
break;
@@ -364,7 +350,6 @@ exit:
case LP3944_LED_TYPE_LED:
case LP3944_LED_TYPE_LED_INVERTED:
led_classdev_unregister(&data->leds[i].ldev);
- cancel_work_sync(&data->leds[i].work);
break;
case LP3944_LED_TYPE_NONE:
@@ -424,7 +409,6 @@ static int lp3944_remove(struct i2c_client *client)
case LP3944_LED_TYPE_LED:
case LP3944_LED_TYPE_LED_INVERTED:
led_classdev_unregister(&data->leds[i].ldev);
- cancel_work_sync(&data->leds[i].work);
break;
case LP3944_LED_TYPE_NONE:
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 21/57] leds: lp55xx: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (19 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 20/57] leds: lp3944: Remove work queue Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 22/57] leds: lp8788: " Jacek Anaszewski
` (35 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn,
Jacek Anaszewski, Milo Kim
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the drivers.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Milo Kim <milo.kim@ti.com>
---
drivers/leds/leds-lp5521.c | 6 ++----
drivers/leds/leds-lp5523.c | 6 ++----
drivers/leds/leds-lp5562.c | 6 ++----
drivers/leds/leds-lp55xx-common.c | 9 ++++-----
drivers/leds/leds-lp55xx-common.h | 6 ++----
drivers/leds/leds-lp8501.c | 6 ++----
6 files changed, 14 insertions(+), 25 deletions(-)
diff --git a/drivers/leds/leds-lp5521.c b/drivers/leds/leds-lp5521.c
index 8ca197a..9893f7b 100644
--- a/drivers/leds/leds-lp5521.c
+++ b/drivers/leds/leds-lp5521.c
@@ -362,10 +362,8 @@ static int lp5521_run_selftest(struct lp55xx_chip *chip, char *buf)
return 0;
}
-static void lp5521_led_brightness_work(struct work_struct *work)
+static void lp5521_led_brightness(struct lp55xx_led *led)
{
- struct lp55xx_led *led = container_of(work, struct lp55xx_led,
- brightness_work);
struct lp55xx_chip *chip = led->chip;
mutex_lock(&chip->lock);
@@ -501,7 +499,7 @@ static struct lp55xx_device_config lp5521_cfg = {
},
.max_channel = LP5521_MAX_LEDS,
.post_init_device = lp5521_post_init_device,
- .brightness_work_fn = lp5521_led_brightness_work,
+ .brightness_fn = lp5521_led_brightness,
.set_led_current = lp5521_set_led_current,
.firmware_cb = lp5521_firmware_loaded,
.run_engine = lp5521_run_engine,
diff --git a/drivers/leds/leds-lp5523.c b/drivers/leds/leds-lp5523.c
index 584dbbc..35d58ac 100644
--- a/drivers/leds/leds-lp5523.c
+++ b/drivers/leds/leds-lp5523.c
@@ -802,10 +802,8 @@ leave:
return ret;
}
-static void lp5523_led_brightness_work(struct work_struct *work)
+static void lp5523_led_brightness(struct lp55xx_led *led)
{
- struct lp55xx_led *led = container_of(work, struct lp55xx_led,
- brightness_work);
struct lp55xx_chip *chip = led->chip;
mutex_lock(&chip->lock);
@@ -867,7 +865,7 @@ static struct lp55xx_device_config lp5523_cfg = {
},
.max_channel = LP5523_MAX_LEDS,
.post_init_device = lp5523_post_init_device,
- .brightness_work_fn = lp5523_led_brightness_work,
+ .brightness_fn = lp5523_led_brightness,
.set_led_current = lp5523_set_led_current,
.firmware_cb = lp5523_firmware_loaded,
.run_engine = lp5523_run_engine,
diff --git a/drivers/leds/leds-lp5562.c b/drivers/leds/leds-lp5562.c
index ca85724..0c0487b 100644
--- a/drivers/leds/leds-lp5562.c
+++ b/drivers/leds/leds-lp5562.c
@@ -311,10 +311,8 @@ static int lp5562_post_init_device(struct lp55xx_chip *chip)
return 0;
}
-static void lp5562_led_brightness_work(struct work_struct *work)
+static void lp5562_led_brightness(struct lp55xx_led *led)
{
- struct lp55xx_led *led = container_of(work, struct lp55xx_led,
- brightness_work);
struct lp55xx_chip *chip = led->chip;
u8 addr[] = {
LP5562_REG_R_PWM,
@@ -503,7 +501,7 @@ static struct lp55xx_device_config lp5562_cfg = {
},
.post_init_device = lp5562_post_init_device,
.set_led_current = lp5562_set_led_current,
- .brightness_work_fn = lp5562_led_brightness_work,
+ .brightness_fn = lp5562_led_brightness,
.run_engine = lp5562_run_engine,
.firmware_cb = lp5562_firmware_loaded,
.dev_attr_group = &lp5562_group,
diff --git a/drivers/leds/leds-lp55xx-common.c b/drivers/leds/leds-lp55xx-common.c
index 96d51e9..b3c834b 100644
--- a/drivers/leds/leds-lp55xx-common.c
+++ b/drivers/leds/leds-lp55xx-common.c
@@ -138,9 +138,10 @@ static void lp55xx_set_brightness(struct led_classdev *cdev,
enum led_brightness brightness)
{
struct lp55xx_led *led = cdev_to_lp55xx_led(cdev);
+ struct lp55xx_device_config *cfg = led->chip->cfg;
led->brightness = (u8)brightness;
- schedule_work(&led->brightness_work);
+ cfg->brightness_fn(led);
}
static int lp55xx_init_led(struct lp55xx_led *led,
@@ -174,6 +175,7 @@ static int lp55xx_init_led(struct lp55xx_led *led,
led->cdev.brightness_set = lp55xx_set_brightness;
led->cdev.groups = lp55xx_led_groups;
+ led->cdev.flags |= LED_BRIGHTNESS_BLOCKING;
if (pdata->led_config[chan].name) {
led->cdev.name = pdata->led_config[chan].name;
@@ -464,7 +466,7 @@ int lp55xx_register_leds(struct lp55xx_led *led, struct lp55xx_chip *chip)
int ret;
int i;
- if (!cfg->brightness_work_fn) {
+ if (!cfg->brightness_fn) {
dev_err(&chip->cl->dev, "empty brightness configuration\n");
return -EINVAL;
}
@@ -481,8 +483,6 @@ int lp55xx_register_leds(struct lp55xx_led *led, struct lp55xx_chip *chip)
if (ret)
goto err_init_led;
- INIT_WORK(&each->brightness_work, cfg->brightness_work_fn);
-
chip->num_leds++;
each->chip = chip;
@@ -507,7 +507,6 @@ void lp55xx_unregister_leds(struct lp55xx_led *led, struct lp55xx_chip *chip)
for (i = 0; i < chip->num_leds; i++) {
each = led + i;
led_classdev_unregister(&each->cdev);
- flush_work(&each->brightness_work);
}
}
EXPORT_SYMBOL_GPL(lp55xx_unregister_leds);
diff --git a/drivers/leds/leds-lp55xx-common.h b/drivers/leds/leds-lp55xx-common.h
index cceab48..0a045d0 100644
--- a/drivers/leds/leds-lp55xx-common.h
+++ b/drivers/leds/leds-lp55xx-common.h
@@ -95,7 +95,7 @@ struct lp55xx_reg {
* @enable : Chip specific enable command
* @max_channel : Maximum number of channels
* @post_init_device : Chip specific initialization code
- * @brightness_work_fn : Brightness work function
+ * @brightness_fn : Brightness function
* @set_led_current : LED current set function
* @firmware_cb : Call function when the firmware is loaded
* @run_engine : Run internal engine for pattern
@@ -110,7 +110,7 @@ struct lp55xx_device_config {
int (*post_init_device) (struct lp55xx_chip *chip);
/* access brightness register */
- void (*brightness_work_fn)(struct work_struct *work);
+ void (*brightness_fn)(struct lp55xx_led *led);
/* current setting function */
void (*set_led_current) (struct lp55xx_led *led, u8 led_current);
@@ -164,7 +164,6 @@ struct lp55xx_chip {
* @cdev : LED class device
* @led_current : Current setting at each led channel
* @max_current : Maximun current at each led channel
- * @brightness_work : Workqueue for brightness control
* @brightness : Brightness value
* @chip : The lp55xx chip data
*/
@@ -173,7 +172,6 @@ struct lp55xx_led {
struct led_classdev cdev;
u8 led_current;
u8 max_current;
- struct work_struct brightness_work;
u8 brightness;
struct lp55xx_chip *chip;
};
diff --git a/drivers/leds/leds-lp8501.c b/drivers/leds/leds-lp8501.c
index d3098e3..753ce15 100644
--- a/drivers/leds/leds-lp8501.c
+++ b/drivers/leds/leds-lp8501.c
@@ -272,10 +272,8 @@ static void lp8501_firmware_loaded(struct lp55xx_chip *chip)
lp8501_update_program_memory(chip, fw->data, fw->size);
}
-static void lp8501_led_brightness_work(struct work_struct *work)
+static void lp8501_led_brightness(struct lp55xx_led *led)
{
- struct lp55xx_led *led = container_of(work, struct lp55xx_led,
- brightness_work);
struct lp55xx_chip *chip = led->chip;
mutex_lock(&chip->lock);
@@ -296,7 +294,7 @@ static struct lp55xx_device_config lp8501_cfg = {
},
.max_channel = LP8501_MAX_LEDS,
.post_init_device = lp8501_post_init_device,
- .brightness_work_fn = lp8501_led_brightness_work,
+ .brightness_fn = lp8501_led_brightness,
.set_led_current = lp8501_set_led_current,
.firmware_cb = lp8501_firmware_loaded,
.run_engine = lp8501_run_engine,
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 22/57] leds: lp8788: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (20 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 21/57] leds: lp55xx: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 23/57] leds: lp8860: " Jacek Anaszewski
` (34 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn, Milo Kim,
Jacek Anaszewski
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the drivers.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Cc: Milo Kim <milo.kim@ti.com>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-lp8788.c | 23 ++++++-----------------
1 file changed, 6 insertions(+), 17 deletions(-)
diff --git a/drivers/leds/leds-lp8788.c b/drivers/leds/leds-lp8788.c
index 3409f03..b8adacb 100644
--- a/drivers/leds/leds-lp8788.c
+++ b/drivers/leds/leds-lp8788.c
@@ -26,10 +26,8 @@
struct lp8788_led {
struct lp8788 *lp;
struct mutex lock;
- struct work_struct work;
struct led_classdev led_dev;
enum lp8788_isink_number isink_num;
- enum led_brightness brightness;
int on;
};
@@ -88,12 +86,14 @@ static void lp8788_led_enable(struct lp8788_led *led,
led->on = on;
}
-static void lp8788_led_work(struct work_struct *work)
+static void lp8788_brightness_set(struct led_classdev *led_cdev,
+ enum led_brightness val)
{
- struct lp8788_led *led = container_of(work, struct lp8788_led, work);
+ struct lp8788_led *led =
+ container_of(led_cdev, struct lp8788_led, led_dev);
+
enum lp8788_isink_number num = led->isink_num;
int enable;
- u8 val = led->brightness;
mutex_lock(&led->lock);
@@ -115,16 +115,6 @@ static void lp8788_led_work(struct work_struct *work)
mutex_unlock(&led->lock);
}
-static void lp8788_brightness_set(struct led_classdev *led_cdev,
- enum led_brightness brt_val)
-{
- struct lp8788_led *led =
- container_of(led_cdev, struct lp8788_led, led_dev);
-
- led->brightness = brt_val;
- schedule_work(&led->work);
-}
-
static int lp8788_led_probe(struct platform_device *pdev)
{
struct lp8788 *lp = dev_get_drvdata(pdev->dev.parent);
@@ -140,6 +130,7 @@ static int lp8788_led_probe(struct platform_device *pdev)
led->lp = lp;
led->led_dev.max_brightness = MAX_BRIGHTNESS;
led->led_dev.brightness_set = lp8788_brightness_set;
+ led->led_dev.flags |= LED_BRIGHTNESS_BLOCKING;
led_pdata = lp->pdata ? lp->pdata->led_pdata : NULL;
@@ -149,7 +140,6 @@ static int lp8788_led_probe(struct platform_device *pdev)
led->led_dev.name = led_pdata->name;
mutex_init(&led->lock);
- INIT_WORK(&led->work, lp8788_led_work);
platform_set_drvdata(pdev, led);
@@ -173,7 +163,6 @@ static int lp8788_led_remove(struct platform_device *pdev)
struct lp8788_led *led = platform_get_drvdata(pdev);
led_classdev_unregister(&led->led_dev);
- flush_work(&led->work);
return 0;
}
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 23/57] leds: lp8860: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (21 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 22/57] leds: lp8788: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 24/57] leds: pca955x: " Jacek Anaszewski
` (33 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn, Dan Murphy,
Jacek Anaszewski
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the drivers.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Cc: Dan Murphy <dmurphy@ti.com>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-lp8860.c | 25 ++++++-------------------
1 file changed, 6 insertions(+), 19 deletions(-)
diff --git a/drivers/leds/leds-lp8860.c b/drivers/leds/leds-lp8860.c
index 79f0843..e0d7afc 100644
--- a/drivers/leds/leds-lp8860.c
+++ b/drivers/leds/leds-lp8860.c
@@ -91,26 +91,22 @@
/**
* struct lp8860_led -
* @lock - Lock for reading/writing the device
- * @work - Work item used to off load the brightness register writes
* @client - Pointer to the I2C client
* @led_dev - led class device pointer
* @regmap - Devices register map
* @eeprom_regmap - EEPROM register map
* @enable_gpio - VDDIO/EN gpio to enable communication interface
* @regulator - LED supply regulator pointer
- * @brightness - Current brightness value requested
* @label - LED label
**/
struct lp8860_led {
struct mutex lock;
- struct work_struct work;
struct i2c_client *client;
struct led_classdev led_dev;
struct regmap *regmap;
struct regmap *eeprom_regmap;
struct gpio_desc *enable_gpio;
struct regulator *regulator;
- enum led_brightness brightness;
const char *label;
};
@@ -212,11 +208,13 @@ out:
return ret;
}
-static void lp8860_led_brightness_work(struct work_struct *work)
+static void lp8860_brightness_set(struct led_classdev *led_cdev,
+ enum led_brightness brt_val)
{
- struct lp8860_led *led = container_of(work, struct lp8860_led, work);
+ struct lp8860_led *led =
+ container_of(led_cdev, struct lp8860_led, led_dev);
+ int disp_brightness = brt_val * 255;
int ret;
- int disp_brightness = led->brightness * 255;
mutex_lock(&led->lock);
@@ -243,16 +241,6 @@ out:
mutex_unlock(&led->lock);
}
-static void lp8860_brightness_set(struct led_classdev *led_cdev,
- enum led_brightness brt_val)
-{
- struct lp8860_led *led =
- container_of(led_cdev, struct lp8860_led, led_dev);
-
- led->brightness = brt_val;
- schedule_work(&led->work);
-}
-
static int lp8860_init(struct lp8860_led *led)
{
unsigned int read_buf;
@@ -407,9 +395,9 @@ static int lp8860_probe(struct i2c_client *client,
led->led_dev.name = led->label;
led->led_dev.max_brightness = LED_FULL;
led->led_dev.brightness_set = lp8860_brightness_set;
+ led->led_dev.flags |= LED_BRIGHTNESS_BLOCKING;
mutex_init(&led->lock);
- INIT_WORK(&led->work, lp8860_led_brightness_work);
i2c_set_clientdata(client, led);
@@ -448,7 +436,6 @@ static int lp8860_remove(struct i2c_client *client)
int ret;
led_classdev_unregister(&led->led_dev);
- cancel_work_sync(&led->work);
if (led->enable_gpio)
gpiod_direction_output(led->enable_gpio, 0);
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 24/57] leds: pca955x: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (22 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 23/57] leds: lp8860: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 25/57] leds: pca963x: " Jacek Anaszewski
` (32 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn, Nate Case,
Jacek Anaszewski
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the drivers.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Cc: Nate Case <ncase@xes-inc.com>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-pca955x.c | 38 ++++++++------------------------------
1 file changed, 8 insertions(+), 30 deletions(-)
diff --git a/drivers/leds/leds-pca955x.c b/drivers/leds/leds-pca955x.c
index b775e1e..32de9c6 100644
--- a/drivers/leds/leds-pca955x.c
+++ b/drivers/leds/leds-pca955x.c
@@ -47,7 +47,6 @@
#include <linux/leds.h>
#include <linux/err.h>
#include <linux/i2c.h>
-#include <linux/workqueue.h>
#include <linux/slab.h>
/* LED select registers determine the source that drives LED outputs */
@@ -110,8 +109,6 @@ struct pca955x {
struct pca955x_led {
struct pca955x *pca955x;
- struct work_struct work;
- enum led_brightness brightness;
struct led_classdev led_cdev;
int led_num; /* 0 .. 15 potentially */
char name[32];
@@ -193,7 +190,8 @@ static u8 pca955x_read_ls(struct i2c_client *client, int n)
pca95xx_num_input_regs(pca955x->chipdef->bits) + 4 + n);
}
-static void pca955x_led_work(struct work_struct *work)
+static void pca955x_led_set(struct led_classdev *led_cdev,
+ enum led_brightness value)
{
struct pca955x_led *pca955x_led;
struct pca955x *pca955x;
@@ -201,7 +199,7 @@ static void pca955x_led_work(struct work_struct *work)
int chip_ls; /* which LSx to use (0-3 potentially) */
int ls_led; /* which set of bits within LSx to use (0-3) */
- pca955x_led = container_of(work, struct pca955x_led, work);
+ pca955x_led = container_of(led_cdev, struct pca955x_led, led_cdev);
pca955x = pca955x_led->pca955x;
chip_ls = pca955x_led->led_num / 4;
@@ -211,7 +209,7 @@ static void pca955x_led_work(struct work_struct *work)
ls = pca955x_read_ls(pca955x->client, chip_ls);
- switch (pca955x_led->brightness) {
+ switch (value) {
case LED_FULL:
ls = pca955x_ledsel(ls, ls_led, PCA955X_LS_LED_ON);
break;
@@ -230,7 +228,7 @@ static void pca955x_led_work(struct work_struct *work)
* just turning off for all other values.
*/
pca955x_write_pwm(pca955x->client, 1,
- 255 - pca955x_led->brightness);
+ 255 - value);
ls = pca955x_ledsel(ls, ls_led, PCA955X_LS_BLINK1);
break;
}
@@ -240,21 +238,6 @@ static void pca955x_led_work(struct work_struct *work)
mutex_unlock(&pca955x->lock);
}
-static void pca955x_led_set(struct led_classdev *led_cdev, enum led_brightness value)
-{
- struct pca955x_led *pca955x;
-
- pca955x = container_of(led_cdev, struct pca955x_led, led_cdev);
-
- pca955x->brightness = value;
-
- /*
- * Must use workqueue for the actual I/O since I2C operations
- * can sleep.
- */
- schedule_work(&pca955x->work);
-}
-
static int pca955x_probe(struct i2c_client *client,
const struct i2c_device_id *id)
{
@@ -329,8 +312,7 @@ static int pca955x_probe(struct i2c_client *client,
pca955x_led->led_cdev.name = pca955x_led->name;
pca955x_led->led_cdev.brightness_set = pca955x_led_set;
-
- INIT_WORK(&pca955x_led->work, pca955x_led_work);
+ pca955x_led->led_cdev.flags |= LED_BRIGHTNESS_BLOCKING;
err = led_classdev_register(&client->dev,
&pca955x_led->led_cdev);
@@ -355,10 +337,8 @@ static int pca955x_probe(struct i2c_client *client,
return 0;
exit:
- while (i--) {
+ while (i--)
led_classdev_unregister(&pca955x->leds[i].led_cdev);
- cancel_work_sync(&pca955x->leds[i].work);
- }
return err;
}
@@ -368,10 +348,8 @@ static int pca955x_remove(struct i2c_client *client)
struct pca955x *pca955x = i2c_get_clientdata(client);
int i;
- for (i = 0; i < pca955x->chipdef->bits; i++) {
+ for (i = 0; i < pca955x->chipdef->bits; i++)
led_classdev_unregister(&pca955x->leds[i].led_cdev);
- cancel_work_sync(&pca955x->leds[i].work);
- }
return 0;
}
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 25/57] leds: pca963x: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (23 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 24/57] leds: pca955x: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:42 ` Peter Meerwald
2015-08-11 10:00 ` Ricardo Ribalda Delgado
2015-08-11 9:37 ` [PATCH/RFC v5 26/57] leds: wm831x: " Jacek Anaszewski
` (31 subsequent siblings)
56 siblings, 2 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn,
Jacek Anaszewski, Peter Meerwald, Ricardo Ribalda
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the drivers.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Peter Meerwald <p.meerwald@bct-electronic.com>
Cc: Ricardo Ribalda <ricardo.ribalda@gmail.com>
---
drivers/leds/leds-pca963x.c | 57 +++++++------------------------------------
1 file changed, 9 insertions(+), 48 deletions(-)
diff --git a/drivers/leds/leds-pca963x.c b/drivers/leds/leds-pca963x.c
index 3f63a1b..662d813 100644
--- a/drivers/leds/leds-pca963x.c
+++ b/drivers/leds/leds-pca963x.c
@@ -32,7 +32,6 @@
#include <linux/leds.h>
#include <linux/err.h>
#include <linux/i2c.h>
-#include <linux/workqueue.h>
#include <linux/slab.h>
#include <linux/of.h>
#include <linux/platform_data/leds-pca963x.h>
@@ -96,11 +95,6 @@ static const struct i2c_device_id pca963x_id[] = {
};
MODULE_DEVICE_TABLE(i2c, pca963x_id);
-enum pca963x_cmd {
- BRIGHTNESS_SET,
- BLINK_SET,
-};
-
struct pca963x_led;
struct pca963x {
@@ -112,17 +106,16 @@ struct pca963x {
struct pca963x_led {
struct pca963x *chip;
- struct work_struct work;
enum led_brightness brightness;
struct led_classdev led_cdev;
int led_num; /* 0 .. 15 potentially */
- enum pca963x_cmd cmd;
char name[32];
u8 gdc;
u8 gfrq;
};
-static void pca963x_brightness_work(struct pca963x_led *pca963x)
+static void pca963x_brightness(struct pca963x_led *pca963x,
+ enum led_brightness brightness)
{
u8 ledout_addr = pca963x->chip->chipdef->ledout_base
+ (pca963x->led_num / 4);
@@ -132,7 +125,7 @@ static void pca963x_brightness_work(struct pca963x_led *pca963x)
mutex_lock(&pca963x->chip->mutex);
ledout = i2c_smbus_read_byte_data(pca963x->chip->client, ledout_addr);
- switch (pca963x->brightness) {
+ switch (brightness) {
case LED_FULL:
i2c_smbus_write_byte_data(pca963x->chip->client, ledout_addr,
(ledout & ~mask) | (PCA963X_LED_ON << shift));
@@ -152,7 +145,7 @@ static void pca963x_brightness_work(struct pca963x_led *pca963x)
mutex_unlock(&pca963x->chip->mutex);
}
-static void pca963x_blink_work(struct pca963x_led *pca963x)
+static void pca963x_blink(struct pca963x_led *pca963x)
{
u8 ledout_addr = pca963x->chip->chipdef->ledout_base +
(pca963x->led_num / 4);
@@ -180,21 +173,6 @@ static void pca963x_blink_work(struct pca963x_led *pca963x)
mutex_unlock(&pca963x->chip->mutex);
}
-static void pca963x_work(struct work_struct *work)
-{
- struct pca963x_led *pca963x = container_of(work,
- struct pca963x_led, work);
-
- switch (pca963x->cmd) {
- case BRIGHTNESS_SET:
- pca963x_brightness_work(pca963x);
- break;
- case BLINK_SET:
- pca963x_blink_work(pca963x);
- break;
- }
-}
-
static void pca963x_led_set(struct led_classdev *led_cdev,
enum led_brightness value)
{
@@ -202,14 +180,7 @@ static void pca963x_led_set(struct led_classdev *led_cdev,
pca963x = container_of(led_cdev, struct pca963x_led, led_cdev);
- pca963x->cmd = BRIGHTNESS_SET;
- pca963x->brightness = value;
-
- /*
- * Must use workqueue for the actual I/O since I2C operations
- * can sleep.
- */
- schedule_work(&pca963x->work);
+ pca963x_brightness(pca963x, value);
}
static int pca963x_blink_set(struct led_classdev *led_cdev,
@@ -254,15 +225,10 @@ static int pca963x_blink_set(struct led_classdev *led_cdev,
*/
gfrq = (period * 24 / 1000) - 1;
- pca963x->cmd = BLINK_SET;
pca963x->gdc = gdc;
pca963x->gfrq = gfrq;
- /*
- * Must use workqueue for the actual I/O since I2C operations
- * can sleep.
- */
- schedule_work(&pca963x->work);
+ pca963x_blink(pca963x);
*delay_on = time_on;
*delay_off = time_off;
@@ -409,12 +375,11 @@ static int pca963x_probe(struct i2c_client *client,
pca963x[i].led_cdev.name = pca963x[i].name;
pca963x[i].led_cdev.brightness_set = pca963x_led_set;
+ pca963x[i].led_cdev.flags |= LED_BRIGHTNESS_BLOCKING;
if (pdata && pdata->blink_type == PCA963X_HW_BLINK)
pca963x[i].led_cdev.blink_set = pca963x_blink_set;
- INIT_WORK(&pca963x[i].work, pca963x_work);
-
err = led_classdev_register(&client->dev, &pca963x[i].led_cdev);
if (err < 0)
goto exit;
@@ -434,10 +399,8 @@ static int pca963x_probe(struct i2c_client *client,
return 0;
exit:
- while (i--) {
+ while (i--)
led_classdev_unregister(&pca963x[i].led_cdev);
- cancel_work_sync(&pca963x[i].work);
- }
return err;
}
@@ -447,10 +410,8 @@ static int pca963x_remove(struct i2c_client *client)
struct pca963x *pca963x = i2c_get_clientdata(client);
int i;
- for (i = 0; i < pca963x->chipdef->n_leds; i++) {
+ for (i = 0; i < pca963x->chipdef->n_leds; i++)
led_classdev_unregister(&pca963x->leds[i].led_cdev);
- cancel_work_sync(&pca963x->leds[i].work);
- }
return 0;
}
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* Re: [PATCH/RFC v5 25/57] leds: pca963x: Remove work queue
2015-08-11 9:37 ` [PATCH/RFC v5 25/57] leds: pca963x: " Jacek Anaszewski
@ 2015-08-11 9:42 ` Peter Meerwald
2015-08-11 10:00 ` Ricardo Ribalda Delgado
1 sibling, 0 replies; 72+ messages in thread
From: Peter Meerwald @ 2015-08-11 9:42 UTC (permalink / raw)
To: Jacek Anaszewski
Cc: linux-leds, linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn,
Ricardo Ribalda
> Now the core implements the work queue, remove it from the drivers.
nice!
> Signed-off-by: Andrew Lunn <andrew@lunn.ch>
> Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
> Cc: Peter Meerwald <p.meerwald@bct-electronic.com>
> Cc: Ricardo Ribalda <ricardo.ribalda@gmail.com>
Acked-by: Peter Meerwald <p.meerwald@bct-electronic.com>
> ---
> drivers/leds/leds-pca963x.c | 57 +++++++------------------------------------
> 1 file changed, 9 insertions(+), 48 deletions(-)
>
> diff --git a/drivers/leds/leds-pca963x.c b/drivers/leds/leds-pca963x.c
> index 3f63a1b..662d813 100644
> --- a/drivers/leds/leds-pca963x.c
> +++ b/drivers/leds/leds-pca963x.c
> @@ -32,7 +32,6 @@
> #include <linux/leds.h>
> #include <linux/err.h>
> #include <linux/i2c.h>
> -#include <linux/workqueue.h>
> #include <linux/slab.h>
> #include <linux/of.h>
> #include <linux/platform_data/leds-pca963x.h>
> @@ -96,11 +95,6 @@ static const struct i2c_device_id pca963x_id[] = {
> };
> MODULE_DEVICE_TABLE(i2c, pca963x_id);
>
> -enum pca963x_cmd {
> - BRIGHTNESS_SET,
> - BLINK_SET,
> -};
> -
> struct pca963x_led;
>
> struct pca963x {
> @@ -112,17 +106,16 @@ struct pca963x {
>
> struct pca963x_led {
> struct pca963x *chip;
> - struct work_struct work;
> enum led_brightness brightness;
> struct led_classdev led_cdev;
> int led_num; /* 0 .. 15 potentially */
> - enum pca963x_cmd cmd;
> char name[32];
> u8 gdc;
> u8 gfrq;
> };
>
> -static void pca963x_brightness_work(struct pca963x_led *pca963x)
> +static void pca963x_brightness(struct pca963x_led *pca963x,
> + enum led_brightness brightness)
> {
> u8 ledout_addr = pca963x->chip->chipdef->ledout_base
> + (pca963x->led_num / 4);
> @@ -132,7 +125,7 @@ static void pca963x_brightness_work(struct pca963x_led *pca963x)
>
> mutex_lock(&pca963x->chip->mutex);
> ledout = i2c_smbus_read_byte_data(pca963x->chip->client, ledout_addr);
> - switch (pca963x->brightness) {
> + switch (brightness) {
> case LED_FULL:
> i2c_smbus_write_byte_data(pca963x->chip->client, ledout_addr,
> (ledout & ~mask) | (PCA963X_LED_ON << shift));
> @@ -152,7 +145,7 @@ static void pca963x_brightness_work(struct pca963x_led *pca963x)
> mutex_unlock(&pca963x->chip->mutex);
> }
>
> -static void pca963x_blink_work(struct pca963x_led *pca963x)
> +static void pca963x_blink(struct pca963x_led *pca963x)
> {
> u8 ledout_addr = pca963x->chip->chipdef->ledout_base +
> (pca963x->led_num / 4);
> @@ -180,21 +173,6 @@ static void pca963x_blink_work(struct pca963x_led *pca963x)
> mutex_unlock(&pca963x->chip->mutex);
> }
>
> -static void pca963x_work(struct work_struct *work)
> -{
> - struct pca963x_led *pca963x = container_of(work,
> - struct pca963x_led, work);
> -
> - switch (pca963x->cmd) {
> - case BRIGHTNESS_SET:
> - pca963x_brightness_work(pca963x);
> - break;
> - case BLINK_SET:
> - pca963x_blink_work(pca963x);
> - break;
> - }
> -}
> -
> static void pca963x_led_set(struct led_classdev *led_cdev,
> enum led_brightness value)
> {
> @@ -202,14 +180,7 @@ static void pca963x_led_set(struct led_classdev *led_cdev,
>
> pca963x = container_of(led_cdev, struct pca963x_led, led_cdev);
>
> - pca963x->cmd = BRIGHTNESS_SET;
> - pca963x->brightness = value;
> -
> - /*
> - * Must use workqueue for the actual I/O since I2C operations
> - * can sleep.
> - */
> - schedule_work(&pca963x->work);
> + pca963x_brightness(pca963x, value);
> }
>
> static int pca963x_blink_set(struct led_classdev *led_cdev,
> @@ -254,15 +225,10 @@ static int pca963x_blink_set(struct led_classdev *led_cdev,
> */
> gfrq = (period * 24 / 1000) - 1;
>
> - pca963x->cmd = BLINK_SET;
> pca963x->gdc = gdc;
> pca963x->gfrq = gfrq;
>
> - /*
> - * Must use workqueue for the actual I/O since I2C operations
> - * can sleep.
> - */
> - schedule_work(&pca963x->work);
> + pca963x_blink(pca963x);
>
> *delay_on = time_on;
> *delay_off = time_off;
> @@ -409,12 +375,11 @@ static int pca963x_probe(struct i2c_client *client,
>
> pca963x[i].led_cdev.name = pca963x[i].name;
> pca963x[i].led_cdev.brightness_set = pca963x_led_set;
> + pca963x[i].led_cdev.flags |= LED_BRIGHTNESS_BLOCKING;
>
> if (pdata && pdata->blink_type == PCA963X_HW_BLINK)
> pca963x[i].led_cdev.blink_set = pca963x_blink_set;
>
> - INIT_WORK(&pca963x[i].work, pca963x_work);
> -
> err = led_classdev_register(&client->dev, &pca963x[i].led_cdev);
> if (err < 0)
> goto exit;
> @@ -434,10 +399,8 @@ static int pca963x_probe(struct i2c_client *client,
> return 0;
>
> exit:
> - while (i--) {
> + while (i--)
> led_classdev_unregister(&pca963x[i].led_cdev);
> - cancel_work_sync(&pca963x[i].work);
> - }
>
> return err;
> }
> @@ -447,10 +410,8 @@ static int pca963x_remove(struct i2c_client *client)
> struct pca963x *pca963x = i2c_get_clientdata(client);
> int i;
>
> - for (i = 0; i < pca963x->chipdef->n_leds; i++) {
> + for (i = 0; i < pca963x->chipdef->n_leds; i++)
> led_classdev_unregister(&pca963x->leds[i].led_cdev);
> - cancel_work_sync(&pca963x->leds[i].work);
> - }
>
> return 0;
> }
>
--
Peter Meerwald
+43-664-2444418 (mobile)
^ permalink raw reply [flat|nested] 72+ messages in thread* Re: [PATCH/RFC v5 25/57] leds: pca963x: Remove work queue
2015-08-11 9:37 ` [PATCH/RFC v5 25/57] leds: pca963x: " Jacek Anaszewski
2015-08-11 9:42 ` Peter Meerwald
@ 2015-08-11 10:00 ` Ricardo Ribalda Delgado
2015-08-11 10:35 ` Jacek Anaszewski
1 sibling, 1 reply; 72+ messages in thread
From: Ricardo Ribalda Delgado @ 2015-08-11 10:00 UTC (permalink / raw)
To: Jacek Anaszewski
Cc: Linux LED Subsystem, LKML, Bryan Wu, Richard Purdie, stsp,
Andrew Lunn, Peter Meerwald
Hello Jacek
Sorry for the late reply.
On Tue, Aug 11, 2015 at 11:37 AM, Jacek Anaszewski
<j.anaszewski@samsung.com> wrote:
> struct pca963x_led {
> struct pca963x *chip;
> - struct work_struct work;
> enum led_brightness brightness;
> struct led_classdev led_cdev;
> int led_num; /* 0 .. 15 potentially */
> - enum pca963x_cmd cmd;
> char name[32];
> u8 gdc;
> u8 gfrq;
> };
Maybe you want to remove also brightness, gdc and gfrq and pass them
as arguments to the functions.
> -static void pca963x_blink_work(struct pca963x_led *pca963x)
> +static void pca963x_blink(struct pca963x_led *pca963x)
something like: blink(pca963x, gdc, gfrq)
Big disclaimer: i havent tested it on real hardware. Maybe next week I
can get hold of a board with a pca chip.
BTW: great thing that the work queue is implemented in the core :)
Cheers!
--
Ricardo Ribalda
^ permalink raw reply [flat|nested] 72+ messages in thread* Re: [PATCH/RFC v5 25/57] leds: pca963x: Remove work queue
2015-08-11 10:00 ` Ricardo Ribalda Delgado
@ 2015-08-11 10:35 ` Jacek Anaszewski
0 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 10:35 UTC (permalink / raw)
To: Ricardo Ribalda Delgado
Cc: Linux LED Subsystem, LKML, Bryan Wu, Richard Purdie, stsp,
Andrew Lunn, Peter Meerwald
Hi Ricardo,
Thanks for the review.
On 08/11/2015 12:00 PM, Ricardo Ribalda Delgado wrote:
> Hello Jacek
>
> Sorry for the late reply.
>
> On Tue, Aug 11, 2015 at 11:37 AM, Jacek Anaszewski
> <j.anaszewski@samsung.com> wrote:
>
>
>> struct pca963x_led {
>> struct pca963x *chip;
>> - struct work_struct work;
>> enum led_brightness brightness;
>> struct led_classdev led_cdev;
>> int led_num; /* 0 .. 15 potentially */
>> - enum pca963x_cmd cmd;
>> char name[32];
>> u8 gdc;
>> u8 gfrq;
>> };
>
> Maybe you want to remove also brightness, gdc and gfrq and pass them
> as arguments to the functions.
>
>> -static void pca963x_blink_work(struct pca963x_led *pca963x)
>> +static void pca963x_blink(struct pca963x_led *pca963x)
>
> something like: blink(pca963x, gdc, gfrq)
OK, I'll check it.
>
> Big disclaimer: i havent tested it on real hardware. Maybe next week I
> can get hold of a board with a pca chip.
>
>
> BTW: great thing that the work queue is implemented in the core :)
So far it is only in this patch set, but hopefully we will manage
to merge it soon.
--
Best Regards,
Jacek Anaszewski
^ permalink raw reply [flat|nested] 72+ messages in thread
* [PATCH/RFC v5 26/57] leds: wm831x: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (24 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 25/57] leds: pca963x: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 27/57] leds: da903x: " Jacek Anaszewski
` (30 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn, Mark Brown,
Jacek Anaszewski
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the drivers.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Cc: Mark Brown <broonie@kernel.org>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-wm831x-status.c | 24 ++++++++----------------
1 file changed, 8 insertions(+), 16 deletions(-)
diff --git a/drivers/leds/leds-wm831x-status.c b/drivers/leds/leds-wm831x-status.c
index 56027ef..ed40dd7 100644
--- a/drivers/leds/leds-wm831x-status.c
+++ b/drivers/leds/leds-wm831x-status.c
@@ -23,7 +23,6 @@
struct wm831x_status {
struct led_classdev cdev;
struct wm831x *wm831x;
- struct work_struct work;
struct mutex mutex;
spinlock_t value_lock;
@@ -40,10 +39,8 @@ struct wm831x_status {
#define to_wm831x_status(led_cdev) \
container_of(led_cdev, struct wm831x_status, cdev)
-static void wm831x_status_work(struct work_struct *work)
+static void wm831x_status_set(struct wm831x_status *led)
{
- struct wm831x_status *led = container_of(work, struct wm831x_status,
- work);
unsigned long flags;
mutex_lock(&led->mutex);
@@ -70,8 +67,8 @@ static void wm831x_status_work(struct work_struct *work)
mutex_unlock(&led->mutex);
}
-static void wm831x_status_set(struct led_classdev *led_cdev,
- enum led_brightness value)
+static void wm831x_status_brightness_set(struct led_classdev *led_cdev,
+ enum led_brightness value)
{
struct wm831x_status *led = to_wm831x_status(led_cdev);
unsigned long flags;
@@ -80,8 +77,8 @@ static void wm831x_status_set(struct led_classdev *led_cdev,
led->brightness = value;
if (value == LED_OFF)
led->blink = 0;
- schedule_work(&led->work);
spin_unlock_irqrestore(&led->value_lock, flags);
+ wm831x_status_set(led);
}
static int wm831x_status_blink_set(struct led_classdev *led_cdev,
@@ -147,11 +144,8 @@ static int wm831x_status_blink_set(struct led_classdev *led_cdev,
else
led->blink = 0;
- /* Always update; if we fail turn off blinking since we expect
- * a software fallback. */
- schedule_work(&led->work);
-
spin_unlock_irqrestore(&led->value_lock, flags);
+ wm831x_status_set(led);
return ret;
}
@@ -206,11 +200,9 @@ static ssize_t wm831x_status_src_store(struct device *dev,
for (i = 0; i < ARRAY_SIZE(led_src_texts); i++) {
if (!strcmp(name, led_src_texts[i])) {
mutex_lock(&led->mutex);
-
led->src = i;
- schedule_work(&led->work);
-
mutex_unlock(&led->mutex);
+ wm831x_status_set(led);
}
}
@@ -262,7 +254,6 @@ static int wm831x_status_probe(struct platform_device *pdev)
pdata.name = dev_name(&pdev->dev);
mutex_init(&drvdata->mutex);
- INIT_WORK(&drvdata->work, wm831x_status_work);
spin_lock_init(&drvdata->value_lock);
/* We cache the configuration register and read startup values
@@ -287,9 +278,10 @@ static int wm831x_status_probe(struct platform_device *pdev)
drvdata->cdev.name = pdata.name;
drvdata->cdev.default_trigger = pdata.default_trigger;
- drvdata->cdev.brightness_set = wm831x_status_set;
+ drvdata->cdev.brightness_set = wm831x_status_brightness_set;
drvdata->cdev.blink_set = wm831x_status_blink_set;
drvdata->cdev.groups = wm831x_status_groups;
+ drvdata->cdev.flags |= LED_BRIGHTNESS_BLOCKING;
ret = led_classdev_register(wm831x->dev, &drvdata->cdev);
if (ret < 0) {
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 27/57] leds: da903x: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (25 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 26/57] leds: wm831x: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 28/57] leds: da9052: " Jacek Anaszewski
` (29 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn, Jacek Anaszewski
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the drivers.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-da903x.c | 33 ++++++++++-----------------------
1 file changed, 10 insertions(+), 23 deletions(-)
diff --git a/drivers/leds/leds-da903x.c b/drivers/leds/leds-da903x.c
index 952ba96e..bbb7935 100644
--- a/drivers/leds/leds-da903x.c
+++ b/drivers/leds/leds-da903x.c
@@ -16,7 +16,6 @@
#include <linux/kernel.h>
#include <linux/platform_device.h>
#include <linux/leds.h>
-#include <linux/workqueue.h>
#include <linux/mfd/da903x.h>
#include <linux/slab.h>
@@ -33,9 +32,7 @@
struct da903x_led {
struct led_classdev cdev;
- struct work_struct work;
struct device *master;
- enum led_brightness new_brightness;
int id;
int flags;
};
@@ -43,9 +40,11 @@ struct da903x_led {
#define DA9030_LED_OFFSET(id) ((id) - DA9030_ID_LED_1)
#define DA9034_LED_OFFSET(id) ((id) - DA9034_ID_LED_1)
-static void da903x_led_work(struct work_struct *work)
+static void da903x_led_set(struct led_classdev *led_cdev,
+ enum led_brightness value)
{
- struct da903x_led *led = container_of(work, struct da903x_led, work);
+ struct da903x_led *led =
+ container_of(led_cdev, struct da903x_led, cdev);
uint8_t val;
int offset;
@@ -57,39 +56,29 @@ static void da903x_led_work(struct work_struct *work)
case DA9030_ID_LED_PC:
offset = DA9030_LED_OFFSET(led->id);
val = led->flags & ~0x87;
- val |= (led->new_brightness) ? 0x80 : 0; /* EN bit */
- val |= (0x7 - (led->new_brightness >> 5)) & 0x7; /* PWM<2:0> */
+ val |= value ? 0x80 : 0; /* EN bit */
+ val |= (0x7 - (value >> 5)) & 0x7; /* PWM<2:0> */
da903x_write(led->master, DA9030_LED1_CONTROL + offset, val);
break;
case DA9030_ID_VIBRA:
val = led->flags & ~0x80;
- val |= (led->new_brightness) ? 0x80 : 0; /* EN bit */
+ val |= value ? 0x80 : 0; /* EN bit */
da903x_write(led->master, DA9030_MISC_CONTROL_A, val);
break;
case DA9034_ID_LED_1:
case DA9034_ID_LED_2:
offset = DA9034_LED_OFFSET(led->id);
- val = (led->new_brightness * 0x5f / LED_FULL) & 0x7f;
+ val = (value * 0x5f / LED_FULL) & 0x7f;
val |= (led->flags & DA9034_LED_RAMP) ? 0x80 : 0;
da903x_write(led->master, DA9034_LED1_CONTROL + offset, val);
break;
case DA9034_ID_VIBRA:
- val = led->new_brightness & 0xfe;
+ val = value & 0xfe;
da903x_write(led->master, DA9034_VIBRA, val);
break;
}
}
-static void da903x_led_set(struct led_classdev *led_cdev,
- enum led_brightness value)
-{
- struct da903x_led *led;
-
- led = container_of(led_cdev, struct da903x_led, cdev);
- led->new_brightness = value;
- schedule_work(&led->work);
-}
-
static int da903x_led_probe(struct platform_device *pdev)
{
struct led_info *pdata = dev_get_platdata(&pdev->dev);
@@ -115,13 +104,11 @@ static int da903x_led_probe(struct platform_device *pdev)
led->cdev.default_trigger = pdata->default_trigger;
led->cdev.brightness_set = da903x_led_set;
led->cdev.brightness = LED_OFF;
+ led->cdev.flags |= LED_BRIGHTNESS_BLOCKING;
led->id = id;
led->flags = pdata->flags;
led->master = pdev->dev.parent;
- led->new_brightness = LED_OFF;
-
- INIT_WORK(&led->work, da903x_led_work);
ret = led_classdev_register(led->master, &led->cdev);
if (ret) {
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 28/57] leds: da9052: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (26 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 27/57] leds: da903x: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 29/57] leds: dac124d085: " Jacek Anaszewski
` (28 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn, Jacek Anaszewski
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the drivers.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-da9052.c | 36 +++++++++++-------------------------
1 file changed, 11 insertions(+), 25 deletions(-)
diff --git a/drivers/leds/leds-da9052.c b/drivers/leds/leds-da9052.c
index 28291b6..8e5229e 100644
--- a/drivers/leds/leds-da9052.c
+++ b/drivers/leds/leds-da9052.c
@@ -16,7 +16,6 @@
#include <linux/kernel.h>
#include <linux/platform_device.h>
#include <linux/leds.h>
-#include <linux/workqueue.h>
#include <linux/slab.h>
#include <linux/mfd/da9052/reg.h>
@@ -32,11 +31,9 @@
struct da9052_led {
struct led_classdev cdev;
- struct work_struct work;
struct da9052 *da9052;
unsigned char led_index;
unsigned char id;
- int brightness;
};
static unsigned char led_reg[] = {
@@ -44,12 +41,13 @@ static unsigned char led_reg[] = {
DA9052_LED_CONT_5_REG,
};
-static int da9052_set_led_brightness(struct da9052_led *led)
+static int da9052_set_led_brightness(struct da9052_led *led,
+ enum led_brightness brightness)
{
u8 val;
int error;
- val = (led->brightness & 0x7f) | DA9052_LED_CONT_DIM;
+ val = (brightness & 0x7f) | DA9052_LED_CONT_DIM;
error = da9052_reg_write(led->da9052, led_reg[led->led_index], val);
if (error < 0)
@@ -58,21 +56,13 @@ static int da9052_set_led_brightness(struct da9052_led *led)
return error;
}
-static void da9052_led_work(struct work_struct *work)
-{
- struct da9052_led *led = container_of(work, struct da9052_led, work);
-
- da9052_set_led_brightness(led);
-}
-
static void da9052_led_set(struct led_classdev *led_cdev,
enum led_brightness value)
{
- struct da9052_led *led;
+ struct da9052_led *led =
+ container_of(led_cdev, struct da9052_led, cdev);
- led = container_of(led_cdev, struct da9052_led, cdev);
- led->brightness = value;
- schedule_work(&led->work);
+ da9052_set_led_brightness(led, value);
}
static int da9052_configure_leds(struct da9052 *da9052)
@@ -136,10 +126,9 @@ static int da9052_led_probe(struct platform_device *pdev)
led[i].cdev.brightness_set = da9052_led_set;
led[i].cdev.brightness = LED_OFF;
led[i].cdev.max_brightness = DA9052_MAX_BRIGHTNESS;
- led[i].brightness = LED_OFF;
+ led[i].cdev.flags |= LED_BRIGHTNESS_BLOCKING;
led[i].led_index = pled->leds[i].flags;
led[i].da9052 = dev_get_drvdata(pdev->dev.parent);
- INIT_WORK(&led[i].work, da9052_led_work);
error = led_classdev_register(pdev->dev.parent, &led[i].cdev);
if (error) {
@@ -148,7 +137,8 @@ static int da9052_led_probe(struct platform_device *pdev)
goto err_register;
}
- error = da9052_set_led_brightness(&led[i]);
+ error = da9052_set_led_brightness(&led[i],
+ led[i].cdev.brightness);
if (error) {
dev_err(&pdev->dev, "Unable to init led %d\n",
led[i].led_index);
@@ -166,10 +156,8 @@ static int da9052_led_probe(struct platform_device *pdev)
return 0;
err_register:
- for (i = i - 1; i >= 0; i--) {
+ for (i = i - 1; i >= 0; i--)
led_classdev_unregister(&led[i].cdev);
- cancel_work_sync(&led[i].work);
- }
err:
return error;
}
@@ -187,10 +175,8 @@ static int da9052_led_remove(struct platform_device *pdev)
pled = pdata->pled;
for (i = 0; i < pled->num_leds; i++) {
- led[i].brightness = 0;
- da9052_set_led_brightness(&led[i]);
+ da9052_set_led_brightness(&led[i], LED_OFF);
led_classdev_unregister(&led[i].cdev);
- cancel_work_sync(&led[i].work);
}
return 0;
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 29/57] leds: dac124d085: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (27 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 28/57] leds: da9052: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 30/57] leds: lt3593: " Jacek Anaszewski
` (27 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn,
Guennadi Liakhovetski, Jacek Anaszewski
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the drivers.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-dac124s085.c | 34 +++++++---------------------------
1 file changed, 7 insertions(+), 27 deletions(-)
diff --git a/drivers/leds/leds-dac124s085.c b/drivers/leds/leds-dac124s085.c
index db3ba8b..62039a1 100644
--- a/drivers/leds/leds-dac124s085.c
+++ b/drivers/leds/leds-dac124s085.c
@@ -13,20 +13,15 @@
#include <linux/module.h>
#include <linux/mutex.h>
#include <linux/slab.h>
-#include <linux/spinlock.h>
-#include <linux/workqueue.h>
#include <linux/spi/spi.h>
struct dac124s085_led {
struct led_classdev ldev;
struct spi_device *spi;
int id;
- int brightness;
char name[sizeof("dac124s085-3")];
struct mutex mutex;
- struct work_struct work;
- spinlock_t lock;
};
struct dac124s085 {
@@ -38,31 +33,20 @@ struct dac124s085 {
#define ALL_WRITE_UPDATE (2 << 12)
#define POWER_DOWN_OUTPUT (3 << 12)
-static void dac124s085_led_work(struct work_struct *work)
+static void dac124s085_set_brightness(struct led_classdev *ldev,
+ enum led_brightness brightness)
{
- struct dac124s085_led *led = container_of(work, struct dac124s085_led,
- work);
+ struct dac124s085_led *led = container_of(ldev, struct dac124s085_led,
+ ldev);
u16 word;
mutex_lock(&led->mutex);
word = cpu_to_le16(((led->id) << 14) | REG_WRITE_UPDATE |
- (led->brightness & 0xfff));
+ (brightness & 0xfff));
spi_write(led->spi, (const u8 *)&word, sizeof(word));
mutex_unlock(&led->mutex);
}
-static void dac124s085_set_brightness(struct led_classdev *ldev,
- enum led_brightness brightness)
-{
- struct dac124s085_led *led = container_of(ldev, struct dac124s085_led,
- ldev);
-
- spin_lock(&led->lock);
- led->brightness = brightness;
- schedule_work(&led->work);
- spin_unlock(&led->lock);
-}
-
static int dac124s085_probe(struct spi_device *spi)
{
struct dac124s085 *dac;
@@ -78,16 +62,14 @@ static int dac124s085_probe(struct spi_device *spi)
for (i = 0; i < ARRAY_SIZE(dac->leds); i++) {
led = dac->leds + i;
led->id = i;
- led->brightness = LED_OFF;
led->spi = spi;
snprintf(led->name, sizeof(led->name), "dac124s085-%d", i);
- spin_lock_init(&led->lock);
- INIT_WORK(&led->work, dac124s085_led_work);
mutex_init(&led->mutex);
led->ldev.name = led->name;
led->ldev.brightness = LED_OFF;
led->ldev.max_brightness = 0xfff;
led->ldev.brightness_set = dac124s085_set_brightness;
+ led->ldev.flags |= LED_BRIGHTNESS_BLOCKING;
ret = led_classdev_register(&spi->dev, &led->ldev);
if (ret < 0)
goto eledcr;
@@ -109,10 +91,8 @@ static int dac124s085_remove(struct spi_device *spi)
struct dac124s085 *dac = spi_get_drvdata(spi);
int i;
- for (i = 0; i < ARRAY_SIZE(dac->leds); i++) {
+ for (i = 0; i < ARRAY_SIZE(dac->leds); i++)
led_classdev_unregister(&dac->leds[i].ldev);
- cancel_work_sync(&dac->leds[i].work);
- }
return 0;
}
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 30/57] leds: lt3593: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (28 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 29/57] leds: dac124d085: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 31/57] leds: max8997: Remove unneeded workqueue include Jacek Anaszewski
` (26 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn, Daniel Mack,
Jacek Anaszewski
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the drivers.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Cc: Daniel Mack <daniel@caiaq.de>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-lt3593.c | 29 ++++++++---------------------
1 file changed, 8 insertions(+), 21 deletions(-)
diff --git a/drivers/leds/leds-lt3593.c b/drivers/leds/leds-lt3593.c
index 9f41124..ccd6589 100644
--- a/drivers/leds/leds-lt3593.c
+++ b/drivers/leds/leds-lt3593.c
@@ -19,7 +19,6 @@
#include <linux/kernel.h>
#include <linux/platform_device.h>
#include <linux/leds.h>
-#include <linux/workqueue.h>
#include <linux/delay.h>
#include <linux/gpio.h>
#include <linux/slab.h>
@@ -28,15 +27,14 @@
struct lt3593_led_data {
struct led_classdev cdev;
unsigned gpio;
- struct work_struct work;
- u8 new_level;
};
-static void lt3593_led_work(struct work_struct *work)
+static void lt3593_led_set(struct led_classdev *led_cdev,
+ enum led_brightness value)
{
- int pulses;
struct lt3593_led_data *led_dat =
- container_of(work, struct lt3593_led_data, work);
+ container_of(led_cdev, struct lt3593_led_data, cdev);
+ int pulses;
/*
* The LT3593 resets its internal current level register to the maximum
@@ -47,12 +45,12 @@ static void lt3593_led_work(struct work_struct *work)
* applied is to the output driver.
*/
- if (led_dat->new_level == 0) {
+ if (value == 0) {
gpio_set_value_cansleep(led_dat->gpio, 0);
return;
}
- pulses = 32 - (led_dat->new_level * 32) / 255;
+ pulses = 32 - (value * 32) / 255;
if (pulses == 0) {
gpio_set_value_cansleep(led_dat->gpio, 0);
@@ -71,16 +69,6 @@ static void lt3593_led_work(struct work_struct *work)
}
}
-static void lt3593_led_set(struct led_classdev *led_cdev,
- enum led_brightness value)
-{
- struct lt3593_led_data *led_dat =
- container_of(led_cdev, struct lt3593_led_data, cdev);
-
- led_dat->new_level = value;
- schedule_work(&led_dat->work);
-}
-
static int create_lt3593_led(const struct gpio_led *template,
struct lt3593_led_data *led_dat, struct device *parent)
{
@@ -105,14 +93,14 @@ static int create_lt3593_led(const struct gpio_led *template,
if (!template->retain_state_suspended)
led_dat->cdev.flags |= LED_CORE_SUSPENDRESUME;
+ led_dat->cdev.flags |= LED_BRIGHTNESS_BLOCKING;
+
ret = devm_gpio_request_one(parent, template->gpio, state ?
GPIOF_OUT_INIT_HIGH : GPIOF_OUT_INIT_LOW,
template->name);
if (ret < 0)
return ret;
- INIT_WORK(&led_dat->work, lt3593_led_work);
-
ret = led_classdev_register(parent, &led_dat->cdev);
if (ret < 0)
return ret;
@@ -129,7 +117,6 @@ static void delete_lt3593_led(struct lt3593_led_data *led)
return;
led_classdev_unregister(&led->cdev);
- cancel_work_sync(&led->work);
}
static int lt3593_led_probe(struct platform_device *pdev)
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 31/57] leds: max8997: Remove unneeded workqueue include
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (29 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 30/57] leds: lt3593: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 32/57] leds: mc13783: Remove work queue Jacek Anaszewski
` (25 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds; +Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn
From: Andrew Lunn <andrew@lunn.ch>
Work queues are not used in this driver, so remove the include.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
---
drivers/leds/leds-max8997.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/drivers/leds/leds-max8997.c b/drivers/leds/leds-max8997.c
index c592aa5..01b45906 100644
--- a/drivers/leds/leds-max8997.c
+++ b/drivers/leds/leds-max8997.c
@@ -13,7 +13,6 @@
#include <linux/module.h>
#include <linux/err.h>
#include <linux/slab.h>
-#include <linux/workqueue.h>
#include <linux/leds.h>
#include <linux/mfd/max8997.h>
#include <linux/mfd/max8997-private.h>
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 32/57] leds: mc13783: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (30 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 31/57] leds: max8997: Remove unneeded workqueue include Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 33/57] leds: regulator: " Jacek Anaszewski
` (24 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn, Jacek Anaszewski
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the drivers.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-mc13783.c | 32 ++++++++------------------------
1 file changed, 8 insertions(+), 24 deletions(-)
diff --git a/drivers/leds/leds-mc13783.c b/drivers/leds/leds-mc13783.c
index e2b847f..17909dd 100644
--- a/drivers/leds/leds-mc13783.c
+++ b/drivers/leds/leds-mc13783.c
@@ -20,7 +20,6 @@
#include <linux/platform_device.h>
#include <linux/leds.h>
#include <linux/of.h>
-#include <linux/workqueue.h>
#include <linux/mfd/mc13xxx.h>
struct mc13xxx_led_devtype {
@@ -32,8 +31,6 @@ struct mc13xxx_led_devtype {
struct mc13xxx_led {
struct led_classdev cdev;
- struct work_struct work;
- enum led_brightness new_brightness;
int id;
struct mc13xxx_leds *leds;
};
@@ -55,9 +52,11 @@ static unsigned int mc13xxx_max_brightness(int id)
return 0x3f;
}
-static void mc13xxx_led_work(struct work_struct *work)
+static void mc13xxx_led_set(struct led_classdev *led_cdev,
+ enum led_brightness value)
{
- struct mc13xxx_led *led = container_of(work, struct mc13xxx_led, work);
+ struct mc13xxx_led *led =
+ container_of(led_cdev, struct mc13xxx_led, cdev);
struct mc13xxx_leds *leds = led->leds;
unsigned int reg, bank, off, shift;
@@ -107,17 +106,7 @@ static void mc13xxx_led_work(struct work_struct *work)
mc13xxx_reg_rmw(leds->master, leds->devtype->ledctrl_base + reg,
mc13xxx_max_brightness(led->id) << shift,
- led->new_brightness << shift);
-}
-
-static void mc13xxx_led_set(struct led_classdev *led_cdev,
- enum led_brightness value)
-{
- struct mc13xxx_led *led =
- container_of(led_cdev, struct mc13xxx_led, cdev);
-
- led->new_brightness = value;
- schedule_work(&led->work);
+ value << shift);
}
#ifdef CONFIG_OF
@@ -259,8 +248,7 @@ static int __init mc13xxx_led_probe(struct platform_device *pdev)
leds->led[i].cdev.flags = LED_CORE_SUSPENDRESUME;
leds->led[i].cdev.brightness_set = mc13xxx_led_set;
leds->led[i].cdev.max_brightness = mc13xxx_max_brightness(id);
-
- INIT_WORK(&leds->led[i].work, mc13xxx_led_work);
+ leds->led[i].cdev.flags |= LED_BRIGHTNESS_BLOCKING;
ret = led_classdev_register(dev->parent, &leds->led[i].cdev);
if (ret) {
@@ -270,10 +258,8 @@ static int __init mc13xxx_led_probe(struct platform_device *pdev)
}
if (ret)
- while (--i >= 0) {
+ while (--i >= 0)
led_classdev_unregister(&leds->led[i].cdev);
- cancel_work_sync(&leds->led[i].work);
- }
return ret;
}
@@ -283,10 +269,8 @@ static int mc13xxx_led_remove(struct platform_device *pdev)
struct mc13xxx_leds *leds = platform_get_drvdata(pdev);
int i;
- for (i = 0; i < leds->num_leds; i++) {
+ for (i = 0; i < leds->num_leds; i++)
led_classdev_unregister(&leds->led[i].cdev);
- cancel_work_sync(&leds->led[i].work);
- }
return 0;
}
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 33/57] leds: regulator: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (31 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 32/57] leds: mc13783: Remove work queue Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 34/57] leds: wm8350: " Jacek Anaszewski
` (23 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn,
Antonio Ospite, Jacek Anaszewski
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the drivers.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Cc: Antonio Ospite <ospite@studenti.unina.it>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-regulator.c | 43 ++++++++++-------------------------------
1 file changed, 10 insertions(+), 33 deletions(-)
diff --git a/drivers/leds/leds-regulator.c b/drivers/leds/leds-regulator.c
index ffc2139..73af223 100644
--- a/drivers/leds/leds-regulator.c
+++ b/drivers/leds/leds-regulator.c
@@ -14,7 +14,6 @@
#include <linux/module.h>
#include <linux/err.h>
#include <linux/slab.h>
-#include <linux/workqueue.h>
#include <linux/leds.h>
#include <linux/leds-regulator.h>
#include <linux/platform_device.h>
@@ -25,10 +24,8 @@
struct regulator_led {
struct led_classdev cdev;
- enum led_brightness value;
int enabled;
struct mutex mutex;
- struct work_struct work;
struct regulator *vcc;
};
@@ -94,22 +91,24 @@ static void regulator_led_disable(struct regulator_led *led)
led->enabled = 0;
}
-static void regulator_led_set_value(struct regulator_led *led)
+static void regulator_led_brightness_set(struct led_classdev *led_cdev,
+ enum led_brightness value)
{
+ struct regulator_led *led = to_regulator_led(led_cdev);
int voltage;
int ret;
mutex_lock(&led->mutex);
- if (led->value == LED_OFF) {
+ if (value == LED_OFF) {
regulator_led_disable(led);
goto out;
}
if (led->cdev.max_brightness > 1) {
- voltage = led_regulator_get_voltage(led->vcc, led->value);
+ voltage = led_regulator_get_voltage(led->vcc, value);
dev_dbg(led->cdev.dev, "brightness: %d voltage: %d\n",
- led->value, voltage);
+ value, voltage);
ret = regulator_set_voltage(led->vcc, voltage, voltage);
if (ret != 0)
@@ -123,23 +122,6 @@ out:
mutex_unlock(&led->mutex);
}
-static void led_work(struct work_struct *work)
-{
- struct regulator_led *led;
-
- led = container_of(work, struct regulator_led, work);
- regulator_led_set_value(led);
-}
-
-static void regulator_led_brightness_set(struct led_classdev *led_cdev,
- enum led_brightness value)
-{
- struct regulator_led *led = to_regulator_led(led_cdev);
-
- led->value = value;
- schedule_work(&led->work);
-}
-
static int regulator_led_probe(struct platform_device *pdev)
{
struct led_regulator_platform_data *pdata =
@@ -169,11 +151,10 @@ static int regulator_led_probe(struct platform_device *pdev)
pdata->brightness);
return -EINVAL;
}
- led->value = pdata->brightness;
led->cdev.brightness_set = regulator_led_brightness_set;
led->cdev.name = pdata->name;
- led->cdev.flags |= LED_CORE_SUSPENDRESUME;
+ led->cdev.flags |= LED_CORE_SUSPENDRESUME | LED_BRIGHTNESS_BLOCKING;
led->vcc = vcc;
/* to handle correctly an already enabled regulator */
@@ -181,21 +162,18 @@ static int regulator_led_probe(struct platform_device *pdev)
led->enabled = 1;
mutex_init(&led->mutex);
- INIT_WORK(&led->work, led_work);
platform_set_drvdata(pdev, led);
ret = led_classdev_register(&pdev->dev, &led->cdev);
- if (ret < 0) {
- cancel_work_sync(&led->work);
+ if (ret < 0)
return ret;
- }
/* to expose the default value to userspace */
- led->cdev.brightness = led->value;
+ led->cdev.brightness = pdata->brightness;
/* Set the default led status */
- regulator_led_set_value(led);
+ regulator_led_brightness_set(&led->cdev, led->cdev.brightness);
return 0;
}
@@ -205,7 +183,6 @@ static int regulator_led_remove(struct platform_device *pdev)
struct regulator_led *led = platform_get_drvdata(pdev);
led_classdev_unregister(&led->cdev);
- cancel_work_sync(&led->work);
regulator_led_disable(led);
return 0;
}
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 34/57] leds: wm8350: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (32 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 33/57] leds: regulator: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 35/57] leds: gpio: " Jacek Anaszewski
` (22 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Andrew Lunn,
Jacek Anaszewski, Mark Brown
From: Andrew Lunn <andrew@lunn.ch>
Now the core implements the work queue, remove it from the drivers.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Mark Brown <broonie@kernel.org>
---
drivers/leds/leds-wm8350.c | 33 +++++++--------------------------
include/linux/mfd/wm8350/pmic.h | 1 -
2 files changed, 7 insertions(+), 27 deletions(-)
diff --git a/drivers/leds/leds-wm8350.c b/drivers/leds/leds-wm8350.c
index 0d12183..c01b180 100644
--- a/drivers/leds/leds-wm8350.c
+++ b/drivers/leds/leds-wm8350.c
@@ -138,21 +138,22 @@ static void wm8350_led_disable(struct wm8350_led *led)
led->enabled = 0;
}
-static void led_work(struct work_struct *work)
+static void wm8350_led_set(struct led_classdev *led_cdev,
+ enum led_brightness value)
{
- struct wm8350_led *led = container_of(work, struct wm8350_led, work);
+ struct wm8350_led *led = to_wm8350_led(led_cdev);
+ unsigned long flags;
int ret;
int uA;
- unsigned long flags;
- mutex_lock(&led->mutex);
+ led->value = value;
spin_lock_irqsave(&led->value_lock, flags);
if (led->value == LED_OFF) {
spin_unlock_irqrestore(&led->value_lock, flags);
wm8350_led_disable(led);
- goto out;
+ return;
}
/* This scales linearly into the index of valid current
@@ -171,31 +172,14 @@ static void led_work(struct work_struct *work)
isink_cur[uA], ret);
wm8350_led_enable(led);
-
-out:
- mutex_unlock(&led->mutex);
-}
-
-static void wm8350_led_set(struct led_classdev *led_cdev,
- enum led_brightness value)
-{
- struct wm8350_led *led = to_wm8350_led(led_cdev);
- unsigned long flags;
-
- spin_lock_irqsave(&led->value_lock, flags);
- led->value = value;
- schedule_work(&led->work);
- spin_unlock_irqrestore(&led->value_lock, flags);
}
static void wm8350_led_shutdown(struct platform_device *pdev)
{
struct wm8350_led *led = platform_get_drvdata(pdev);
- mutex_lock(&led->mutex);
led->value = LED_OFF;
wm8350_led_disable(led);
- mutex_unlock(&led->mutex);
}
static int wm8350_led_probe(struct platform_device *pdev)
@@ -235,7 +219,7 @@ static int wm8350_led_probe(struct platform_device *pdev)
led->cdev.brightness_set = wm8350_led_set;
led->cdev.default_trigger = pdata->default_trigger;
led->cdev.name = pdata->name;
- led->cdev.flags |= LED_CORE_SUSPENDRESUME;
+ led->cdev.flags |= LED_CORE_SUSPENDRESUME | LED_BRIGHTNESS_BLOCKING;
led->enabled = regulator_is_enabled(isink);
led->isink = isink;
led->dcdc = dcdc;
@@ -251,8 +235,6 @@ static int wm8350_led_probe(struct platform_device *pdev)
pdata->max_uA);
spin_lock_init(&led->value_lock);
- mutex_init(&led->mutex);
- INIT_WORK(&led->work, led_work);
led->value = LED_OFF;
platform_set_drvdata(pdev, led);
@@ -264,7 +246,6 @@ static int wm8350_led_remove(struct platform_device *pdev)
struct wm8350_led *led = platform_get_drvdata(pdev);
led_classdev_unregister(&led->cdev);
- flush_work(&led->work);
wm8350_led_disable(led);
return 0;
}
diff --git a/include/linux/mfd/wm8350/pmic.h b/include/linux/mfd/wm8350/pmic.h
index 579b50c..7a09e7f 100644
--- a/include/linux/mfd/wm8350/pmic.h
+++ b/include/linux/mfd/wm8350/pmic.h
@@ -715,7 +715,6 @@ struct wm8350_led_platform_data {
struct wm8350_led {
struct platform_device *pdev;
- struct mutex mutex;
struct work_struct work;
spinlock_t value_lock;
enum led_brightness value;
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 35/57] leds: gpio: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (33 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 34/57] leds: wm8350: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 36/57] leds: pwm: remove " Jacek Anaszewski
` (21 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski, Raphael Assenat
Now the core implements the work queue, remove it from the drivers.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Raphael Assenat <raph@8d.com>
---
drivers/leds/leds-gpio.c | 59 +++++++++++++---------------------------------
1 file changed, 16 insertions(+), 43 deletions(-)
diff --git a/drivers/leds/leds-gpio.c b/drivers/leds/leds-gpio.c
index af1876a..2a8e6a0 100644
--- a/drivers/leds/leds-gpio.c
+++ b/drivers/leds/leds-gpio.c
@@ -20,32 +20,16 @@
#include <linux/platform_device.h>
#include <linux/property.h>
#include <linux/slab.h>
-#include <linux/workqueue.h>
struct gpio_led_data {
struct led_classdev cdev;
struct gpio_desc *gpiod;
- struct work_struct work;
- u8 new_level;
u8 can_sleep;
u8 blinking;
int (*platform_gpio_blink_set)(struct gpio_desc *desc, int state,
unsigned long *delay_on, unsigned long *delay_off);
};
-static void gpio_led_work(struct work_struct *work)
-{
- struct gpio_led_data *led_dat =
- container_of(work, struct gpio_led_data, work);
-
- if (led_dat->blinking) {
- led_dat->platform_gpio_blink_set(led_dat->gpiod,
- led_dat->new_level, NULL, NULL);
- led_dat->blinking = 0;
- } else
- gpiod_set_value_cansleep(led_dat->gpiod, led_dat->new_level);
-}
-
static void gpio_led_set(struct led_classdev *led_cdev,
enum led_brightness value)
{
@@ -58,21 +42,12 @@ static void gpio_led_set(struct led_classdev *led_cdev,
else
level = 1;
- /* Setting GPIOs with I2C/etc requires a task context, and we don't
- * seem to have a reliable way to know if we're already in one; so
- * let's just assume the worst.
- */
- if (led_dat->can_sleep) {
- led_dat->new_level = level;
- schedule_work(&led_dat->work);
- } else {
- if (led_dat->blinking) {
- led_dat->platform_gpio_blink_set(led_dat->gpiod, level,
- NULL, NULL);
- led_dat->blinking = 0;
- } else
- gpiod_set_value(led_dat->gpiod, level);
- }
+ if (led_dat->blinking) {
+ led_dat->platform_gpio_blink_set(led_dat->gpiod, level,
+ NULL, NULL);
+ led_dat->blinking = 0;
+ } else
+ gpiod_set_value(led_dat->gpiod, level);
}
static int gpio_blink_set(struct led_classdev *led_cdev,
@@ -125,12 +100,17 @@ static int create_gpio_led(const struct gpio_led *template,
led_dat->cdev.name = template->name;
led_dat->cdev.default_trigger = template->default_trigger;
led_dat->can_sleep = gpiod_cansleep(led_dat->gpiod);
+ if (!led_dat->can_sleep) {
+ led_dat->cdev.brightness_set_nonblocking = gpio_led_set;
+ } else {
+ led_dat->cdev.flags |= LED_BRIGHTNESS_BLOCKING;
+ led_dat->cdev.brightness_set = gpio_led_set;
+ }
led_dat->blinking = 0;
if (blink_set) {
led_dat->platform_gpio_blink_set = blink_set;
led_dat->cdev.blink_set = gpio_blink_set;
}
- led_dat->cdev.brightness_set = gpio_led_set;
if (template->default_state == LEDS_GPIO_DEFSTATE_KEEP)
state = !!gpiod_get_value_cansleep(led_dat->gpiod);
else
@@ -143,17 +123,9 @@ static int create_gpio_led(const struct gpio_led *template,
if (ret < 0)
return ret;
- INIT_WORK(&led_dat->work, gpio_led_work);
-
return led_classdev_register(parent, &led_dat->cdev);
}
-static void delete_gpio_led(struct gpio_led_data *led)
-{
- led_classdev_unregister(&led->cdev);
- cancel_work_sync(&led->work);
-}
-
struct gpio_leds_priv {
int num_leds;
struct gpio_led_data leds[];
@@ -233,7 +205,7 @@ static struct gpio_leds_priv *gpio_leds_create(struct platform_device *pdev)
err:
for (count = priv->num_leds - 1; count >= 0; count--)
- delete_gpio_led(&priv->leds[count]);
+ led_classdev_unregister(&priv->leds[count].cdev);
return ERR_PTR(ret);
}
@@ -265,7 +237,8 @@ static int gpio_led_probe(struct platform_device *pdev)
if (ret < 0) {
/* On failure: unwind the led creations */
for (i = i - 1; i >= 0; i--)
- delete_gpio_led(&priv->leds[i]);
+ led_classdev_unregister(
+ &priv->leds[i].cdev);
return ret;
}
}
@@ -286,7 +259,7 @@ static int gpio_led_remove(struct platform_device *pdev)
int i;
for (i = 0; i < priv->num_leds; i++)
- delete_gpio_led(&priv->leds[i]);
+ led_classdev_unregister(&priv->leds[i].cdev);
return 0;
}
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 36/57] leds: pwm: remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (34 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 35/57] leds: gpio: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 37/57] leds: lm355x: Remove " Jacek Anaszewski
` (20 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski, Raphael Assenat
Now the core implements the work queue, remove it from the drivers.
If brightness setting is guaranteed not to sleep then implement
brightness_set_nonblocking op instead of brightness_set.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Raphael Assenat <raph@8d.com>
---
drivers/leds/leds-pwm.c | 29 ++++++++---------------------
1 file changed, 8 insertions(+), 21 deletions(-)
diff --git a/drivers/leds/leds-pwm.c b/drivers/leds/leds-pwm.c
index 1d07e3e..01e5838 100644
--- a/drivers/leds/leds-pwm.c
+++ b/drivers/leds/leds-pwm.c
@@ -22,12 +22,10 @@
#include <linux/pwm.h>
#include <linux/leds_pwm.h>
#include <linux/slab.h>
-#include <linux/workqueue.h>
struct led_pwm_data {
struct led_classdev cdev;
struct pwm_device *pwm;
- struct work_struct work;
unsigned int active_low;
unsigned int period;
int duty;
@@ -51,14 +49,6 @@ static void __led_pwm_set(struct led_pwm_data *led_dat)
pwm_enable(led_dat->pwm);
}
-static void led_pwm_work(struct work_struct *work)
-{
- struct led_pwm_data *led_dat =
- container_of(work, struct led_pwm_data, work);
-
- __led_pwm_set(led_dat);
-}
-
static void led_pwm_set(struct led_classdev *led_cdev,
enum led_brightness brightness)
{
@@ -75,10 +65,7 @@ static void led_pwm_set(struct led_classdev *led_cdev,
led_dat->duty = duty;
- if (led_dat->can_sleep)
- schedule_work(&led_dat->work);
- else
- __led_pwm_set(led_dat);
+ __led_pwm_set(led_dat);
}
static inline size_t sizeof_pwm_leds_priv(int num_leds)
@@ -89,11 +76,8 @@ static inline size_t sizeof_pwm_leds_priv(int num_leds)
static void led_pwm_cleanup(struct led_pwm_priv *priv)
{
- while (priv->num_leds--) {
+ while (priv->num_leds--)
led_classdev_unregister(&priv->leds[priv->num_leds].cdev);
- if (priv->leds[priv->num_leds].can_sleep)
- cancel_work_sync(&priv->leds[priv->num_leds].work);
- }
}
static int led_pwm_add(struct device *dev, struct led_pwm_priv *priv,
@@ -105,7 +89,6 @@ static int led_pwm_add(struct device *dev, struct led_pwm_priv *priv,
led_data->active_low = led->active_low;
led_data->cdev.name = led->name;
led_data->cdev.default_trigger = led->default_trigger;
- led_data->cdev.brightness_set = led_pwm_set;
led_data->cdev.brightness = LED_OFF;
led_data->cdev.max_brightness = led->max_brightness;
led_data->cdev.flags = LED_CORE_SUSPENDRESUME;
@@ -122,8 +105,12 @@ static int led_pwm_add(struct device *dev, struct led_pwm_priv *priv,
}
led_data->can_sleep = pwm_can_sleep(led_data->pwm);
- if (led_data->can_sleep)
- INIT_WORK(&led_data->work, led_pwm_work);
+ if (!led_data->can_sleep) {
+ led_data->cdev.brightness_set_nonblocking = led_pwm_set;
+ } else {
+ led_data->cdev.brightness_set = led_pwm_set;
+ led_data->cdev.flags |= LED_BRIGHTNESS_BLOCKING;
+ }
led_data->period = pwm_get_period(led_data->pwm);
if (!led_data->period && (led->pwm_period_ns > 0))
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 37/57] leds: lm355x: Remove work queue
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (35 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 36/57] leds: pwm: remove " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 38/57] leds: cobalt-raq: Use brightness_set_nonblocking op Jacek Anaszewski
` (19 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski,
Daniel Jeong, G.Shark Jeong
Now the core implements the work queue, remove it from the drivers.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Daniel Jeong <daniel.jeong@ti.com>
Cc: G.Shark Jeong <gshark.jeong@gmail.com>
---
drivers/leds/leds-lm355x.c | 61 +++++++++-----------------------------------
1 file changed, 12 insertions(+), 49 deletions(-)
diff --git a/drivers/leds/leds-lm355x.c b/drivers/leds/leds-lm355x.c
index 48872997..d2e74e9 100644
--- a/drivers/leds/leds-lm355x.c
+++ b/drivers/leds/leds-lm355x.c
@@ -16,7 +16,6 @@
#include <linux/platform_device.h>
#include <linux/fs.h>
#include <linux/regmap.h>
-#include <linux/workqueue.h>
#include <linux/platform_data/leds-lm355x.h>
enum lm355x_type {
@@ -59,14 +58,6 @@ struct lm355x_chip_data {
struct led_classdev cdev_torch;
struct led_classdev cdev_indicator;
- struct work_struct work_flash;
- struct work_struct work_torch;
- struct work_struct work_indicator;
-
- u8 br_flash;
- u8 br_torch;
- u8 br_indicator;
-
struct lm355x_platform_data *pdata;
struct regmap *regmap;
struct mutex lock;
@@ -316,15 +307,6 @@ out:
}
/* torch */
-static void lm355x_deferred_torch_brightness_set(struct work_struct *work)
-{
- struct lm355x_chip_data *chip =
- container_of(work, struct lm355x_chip_data, work_torch);
-
- mutex_lock(&chip->lock);
- lm355x_control(chip, chip->br_torch, MODE_TORCH);
- mutex_unlock(&chip->lock);
-}
static void lm355x_torch_brightness_set(struct led_classdev *cdev,
enum led_brightness brightness)
@@ -332,50 +314,35 @@ static void lm355x_torch_brightness_set(struct led_classdev *cdev,
struct lm355x_chip_data *chip =
container_of(cdev, struct lm355x_chip_data, cdev_torch);
- chip->br_torch = brightness;
- schedule_work(&chip->work_torch);
-}
-
-/* flash */
-static void lm355x_deferred_strobe_brightness_set(struct work_struct *work)
-{
- struct lm355x_chip_data *chip =
- container_of(work, struct lm355x_chip_data, work_flash);
-
mutex_lock(&chip->lock);
- lm355x_control(chip, chip->br_flash, MODE_FLASH);
+ lm355x_control(chip, brightness, MODE_TORCH);
mutex_unlock(&chip->lock);
}
+/* flash */
+
static void lm355x_strobe_brightness_set(struct led_classdev *cdev,
enum led_brightness brightness)
{
struct lm355x_chip_data *chip =
container_of(cdev, struct lm355x_chip_data, cdev_flash);
- chip->br_flash = brightness;
- schedule_work(&chip->work_flash);
-}
-
-/* indicator */
-static void lm355x_deferred_indicator_brightness_set(struct work_struct *work)
-{
- struct lm355x_chip_data *chip =
- container_of(work, struct lm355x_chip_data, work_indicator);
-
mutex_lock(&chip->lock);
- lm355x_control(chip, chip->br_indicator, MODE_INDIC);
+ lm355x_control(chip, brightness, MODE_FLASH);
mutex_unlock(&chip->lock);
}
+/* indicator */
+
static void lm355x_indicator_brightness_set(struct led_classdev *cdev,
enum led_brightness brightness)
{
struct lm355x_chip_data *chip =
container_of(cdev, struct lm355x_chip_data, cdev_indicator);
- chip->br_indicator = brightness;
- schedule_work(&chip->work_indicator);
+ mutex_lock(&chip->lock);
+ lm355x_control(chip, brightness, MODE_INDIC);
+ mutex_unlock(&chip->lock);
}
/* indicator pattern only for lm3556*/
@@ -479,28 +446,26 @@ static int lm355x_probe(struct i2c_client *client,
goto err_out;
/* flash */
- INIT_WORK(&chip->work_flash, lm355x_deferred_strobe_brightness_set);
chip->cdev_flash.name = "flash";
chip->cdev_flash.max_brightness = 16;
chip->cdev_flash.brightness_set = lm355x_strobe_brightness_set;
chip->cdev_flash.default_trigger = "flash";
+ chip->cdev_flash.flags |= LED_BRIGHTNESS_BLOCKING;
err = led_classdev_register((struct device *)
&client->dev, &chip->cdev_flash);
if (err < 0)
goto err_out;
/* torch */
- INIT_WORK(&chip->work_torch, lm355x_deferred_torch_brightness_set);
chip->cdev_torch.name = "torch";
chip->cdev_torch.max_brightness = 8;
chip->cdev_torch.brightness_set = lm355x_torch_brightness_set;
chip->cdev_torch.default_trigger = "torch";
+ chip->cdev_torch.flags |= LED_BRIGHTNESS_BLOCKING;
err = led_classdev_register((struct device *)
&client->dev, &chip->cdev_torch);
if (err < 0)
goto err_create_torch_file;
/* indicator */
- INIT_WORK(&chip->work_indicator,
- lm355x_deferred_indicator_brightness_set);
chip->cdev_indicator.name = "indicator";
if (id->driver_data == CHIP_LM3554)
chip->cdev_indicator.max_brightness = 4;
@@ -510,6 +475,7 @@ static int lm355x_probe(struct i2c_client *client,
/* indicator pattern control only for LM3556 */
if (id->driver_data == CHIP_LM3556)
chip->cdev_indicator.groups = lm355x_indicator_groups;
+ chip->cdev_indicator.flags |= LED_BRIGHTNESS_BLOCKING;
err = led_classdev_register((struct device *)
&client->dev, &chip->cdev_indicator);
if (err < 0)
@@ -534,11 +500,8 @@ static int lm355x_remove(struct i2c_client *client)
regmap_write(chip->regmap, preg[REG_OPMODE].regno, 0);
led_classdev_unregister(&chip->cdev_indicator);
- flush_work(&chip->work_indicator);
led_classdev_unregister(&chip->cdev_torch);
- flush_work(&chip->work_torch);
led_classdev_unregister(&chip->cdev_flash);
- flush_work(&chip->work_flash);
dev_info(&client->dev, "%s is removed\n", lm355x_name[chip->type]);
return 0;
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 38/57] leds: cobalt-raq: Use brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (36 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 37/57] leds: lm355x: Remove " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 39/57] leds: bcm6328: " Jacek Anaszewski
` (18 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds; +Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski
The driver neither sleeps nor uses delays in its brightness_set ops.
Use a new brightness_set_nonblocking op for setting brightness instead
of brightness_set then.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
---
drivers/leds/leds-cobalt-raq.c | 10 +++++-----
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/drivers/leds/leds-cobalt-raq.c b/drivers/leds/leds-cobalt-raq.c
index b316df4..ef81bb3 100644
--- a/drivers/leds/leds-cobalt-raq.c
+++ b/drivers/leds/leds-cobalt-raq.c
@@ -50,8 +50,8 @@ static void raq_web_led_set(struct led_classdev *led_cdev,
}
static struct led_classdev raq_web_led = {
- .name = "raq::web",
- .brightness_set = raq_web_led_set,
+ .name = "raq::web",
+ .brightness_set_nonblocking = raq_web_led_set,
};
static void raq_power_off_led_set(struct led_classdev *led_cdev,
@@ -71,9 +71,9 @@ static void raq_power_off_led_set(struct led_classdev *led_cdev,
}
static struct led_classdev raq_power_off_led = {
- .name = "raq::power-off",
- .brightness_set = raq_power_off_led_set,
- .default_trigger = "power-off",
+ .name = "raq::power-off",
+ .brightness_set_nonblocking = raq_power_off_led_set,
+ .default_trigger = "power-off",
};
static int cobalt_raq_led_probe(struct platform_device *pdev)
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 39/57] leds: bcm6328: Use brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (37 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 38/57] leds: cobalt-raq: Use brightness_set_nonblocking op Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 40/57] leds: bcm6358: " Jacek Anaszewski
` (17 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski,
Álvaro Fernández Rojas, Jonas Gorski
The driver neither sleeps nor uses delays in its brightness_set op.
Use a new brightness_set_nonblocking op instead of brightness_set for
setting brightness then.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Álvaro Fernández Rojas <noltari@gmail.com>
Cc: Jonas Gorski <jogo@openwrt.org>
---
drivers/leds/leds-bcm6328.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/leds/leds-bcm6328.c b/drivers/leds/leds-bcm6328.c
index 986fe1e..0b0a6b5 100644
--- a/drivers/leds/leds-bcm6328.c
+++ b/drivers/leds/leds-bcm6328.c
@@ -311,7 +311,7 @@ static int bcm6328_led(struct device *dev, struct device_node *nc, u32 reg,
spin_unlock_irqrestore(lock, flags);
}
- led->cdev.brightness_set = bcm6328_led_set;
+ led->cdev.brightness_set_nonblocking = bcm6328_led_set;
led->cdev.blink_set = bcm6328_blink_set;
rc = led_classdev_register(dev, &led->cdev);
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 40/57] leds: bcm6358: Use brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (38 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 39/57] leds: bcm6328: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 41/57] leds: syscon: " Jacek Anaszewski
` (16 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski,
Álvaro Fernández Rojas
The driver neither sleeps nor uses delays in its brightness_set op.
Use a new brightness_set_nonblocking op instead of brightness_set for
setting brightness then.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Álvaro Fernández Rojas <noltari@gmail.com>
---
drivers/leds/leds-bcm6358.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/leds/leds-bcm6358.c b/drivers/leds/leds-bcm6358.c
index 21f9693..a213160 100644
--- a/drivers/leds/leds-bcm6358.c
+++ b/drivers/leds/leds-bcm6358.c
@@ -144,7 +144,7 @@ static int bcm6358_led(struct device *dev, struct device_node *nc, u32 reg,
bcm6358_led_mode(led, led->cdev.brightness);
spin_unlock_irqrestore(lock, flags);
- led->cdev.brightness_set = bcm6358_led_set;
+ led->cdev.brightness_set_nonblocking = bcm6358_led_set;
rc = led_classdev_register(dev, &led->cdev);
if (rc < 0)
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 41/57] leds: syscon: Use brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (39 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 40/57] leds: bcm6358: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 12:54 ` Linus Walleij
2015-08-11 9:37 ` [PATCH/RFC v5 42/57] leds: ot200: " Jacek Anaszewski
` (15 subsequent siblings)
56 siblings, 1 reply; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski, Linus Walleij
The driver neither sleeps nor uses delays in its brightness_set op.
Use a new brightness_set_nonblocking op instead of brightness_set for
setting brightness then.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Linus Walleij <linus.walleij@linaro.org>
---
drivers/leds/leds-syscon.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/leds/leds-syscon.c b/drivers/leds/leds-syscon.c
index d1660b0..c3c0769 100644
--- a/drivers/leds/leds-syscon.c
+++ b/drivers/leds/leds-syscon.c
@@ -127,7 +127,7 @@ static int syscon_led_probe(struct platform_device *pdev)
return ret;
}
}
- sled->cdev.brightness_set = syscon_led_set;
+ sled->cdev.brightness_set_nonblocking = syscon_led_set;
ret = led_classdev_register(dev, &sled->cdev);
if (ret < 0)
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* Re: [PATCH/RFC v5 41/57] leds: syscon: Use brightness_set_nonblocking op
2015-08-11 9:37 ` [PATCH/RFC v5 41/57] leds: syscon: " Jacek Anaszewski
@ 2015-08-11 12:54 ` Linus Walleij
0 siblings, 0 replies; 72+ messages in thread
From: Linus Walleij @ 2015-08-11 12:54 UTC (permalink / raw)
To: Jacek Anaszewski
Cc: linux-leds, linux-kernel, Bryan Wu, Richard Purdie, Stas Sergeev
On Tue, Aug 11, 2015 at 11:37 AM, Jacek Anaszewski
<j.anaszewski@samsung.com> wrote:
> The driver neither sleeps nor uses delays in its brightness_set op.
> Use a new brightness_set_nonblocking op instead of brightness_set for
> setting brightness then.
>
> Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
> Cc: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Yours,
Linus Walleij
^ permalink raw reply [flat|nested] 72+ messages in thread
* [PATCH/RFC v5 42/57] leds: ot200: Use brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (40 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 41/57] leds: syscon: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 43/57] leds: s3c24xx: " Jacek Anaszewski
` (14 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski,
Sebastian A. Siewior
The driver neither sleeps nor uses delays in its brightness_set op.
Use a new brightness_set_nonblocking op instead of brightness_set for
setting brightness then.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Sebastian A. Siewior <bigeasy@linutronix.de>
---
drivers/leds/leds-ot200.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/leds/leds-ot200.c b/drivers/leds/leds-ot200.c
index 39870de..47de99e 100644
--- a/drivers/leds/leds-ot200.c
+++ b/drivers/leds/leds-ot200.c
@@ -122,7 +122,8 @@ static int ot200_led_probe(struct platform_device *pdev)
for (i = 0; i < ARRAY_SIZE(leds); i++) {
leds[i].cdev.name = leds[i].name;
- leds[i].cdev.brightness_set = ot200_led_brightness_set;
+ leds[i].cdev.brightness_set_nonblocking =
+ ot200_led_brightness_set;
ret = led_classdev_register(&pdev->dev, &leds[i].cdev);
if (ret < 0)
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 43/57] leds: s3c24xx: Use brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (41 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 42/57] leds: ot200: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 44/57] leds: ss4200: " Jacek Anaszewski
` (13 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski, Ben Dooks
The driver neither sleeps nor uses delays in its brightness_set op.
Use a new brightness_set_nonblocking op instead of brightness_set for
setting brightness then.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Ben Dooks <ben@simtec.co.uk>
---
drivers/leds/leds-s3c24xx.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/leds/leds-s3c24xx.c b/drivers/leds/leds-s3c24xx.c
index 83641a7..ae4b223 100644
--- a/drivers/leds/leds-s3c24xx.c
+++ b/drivers/leds/leds-s3c24xx.c
@@ -81,7 +81,7 @@ static int s3c24xx_led_probe(struct platform_device *dev)
platform_set_drvdata(dev, led);
- led->cdev.brightness_set = s3c24xx_led_set;
+ led->cdev.brightness_set_nonblocking = s3c24xx_led_set;
led->cdev.default_trigger = pdata->def_trigger;
led->cdev.name = pdata->name;
led->cdev.flags |= LED_CORE_SUSPENDRESUME;
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 44/57] leds: ss4200: Use brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (42 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 43/57] leds: s3c24xx: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:37 ` [PATCH/RFC v5 45/57] leds: versatile: " Jacek Anaszewski
` (12 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski, Dave Hansen
The driver neither sleeps nor uses delays in its brightness_set op.
Use a new brightness_set_nonblocking op instead of brightness_set for
setting brightness then.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Dave Hansen <dave@sr71.net>
---
drivers/leds/leds-ss4200.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/leds/leds-ss4200.c b/drivers/leds/leds-ss4200.c
index 046cb70..925305e 100644
--- a/drivers/leds/leds-ss4200.c
+++ b/drivers/leds/leds-ss4200.c
@@ -485,7 +485,7 @@ static int register_nasgpio_led(int led_nr)
led->brightness = LED_OFF;
if (nasgpio_led_get_attr(led, GP_LVL))
led->brightness = LED_FULL;
- led->brightness_set = nasgpio_led_set_brightness;
+ led->brightness_set_nonblocking = nasgpio_led_set_brightness;
led->blink_set = nasgpio_led_set_blink;
led->groups = nasgpio_led_groups;
ret = led_classdev_register(&nas_gpio_pci_dev->dev, led);
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 45/57] leds: versatile: Use brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (43 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 44/57] leds: ss4200: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 12:55 ` Linus Walleij
2015-08-11 9:37 ` [PATCH/RFC v5 46/57] leds: wrap: " Jacek Anaszewski
` (11 subsequent siblings)
56 siblings, 1 reply; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski, Linus Walleij
The driver neither sleeps nor uses delays in its brightness_set op.
Use a new brightness_set_nonblocking op instead of brightness_set for
setting brightness then.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Linus Walleij <linus.walleij@linaro.org>
---
drivers/leds/leds-versatile.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/leds/leds-versatile.c b/drivers/leds/leds-versatile.c
index 8055302..49b3339 100644
--- a/drivers/leds/leds-versatile.c
+++ b/drivers/leds/leds-versatile.c
@@ -82,7 +82,7 @@ static int versatile_leds_probe(struct platform_device *dev)
led->base = base;
led->cdev.name = versatile_leds[i].name;
- led->cdev.brightness_set = versatile_led_set;
+ led->cdev.brightness_set_nonblocking = versatile_led_set;
led->cdev.brightness_get = versatile_led_get;
led->cdev.default_trigger = versatile_leds[i].trigger;
led->mask = BIT(i);
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* Re: [PATCH/RFC v5 45/57] leds: versatile: Use brightness_set_nonblocking op
2015-08-11 9:37 ` [PATCH/RFC v5 45/57] leds: versatile: " Jacek Anaszewski
@ 2015-08-11 12:55 ` Linus Walleij
0 siblings, 0 replies; 72+ messages in thread
From: Linus Walleij @ 2015-08-11 12:55 UTC (permalink / raw)
To: Jacek Anaszewski
Cc: linux-leds, linux-kernel, Bryan Wu, Richard Purdie, Stas Sergeev
On Tue, Aug 11, 2015 at 11:37 AM, Jacek Anaszewski
<j.anaszewski@samsung.com> wrote:
> The driver neither sleeps nor uses delays in its brightness_set op.
> Use a new brightness_set_nonblocking op instead of brightness_set for
> setting brightness then.
>
> Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
> Cc: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Yours,
Linus Walleij
^ permalink raw reply [flat|nested] 72+ messages in thread
* [PATCH/RFC v5 46/57] leds: wrap: Use brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (44 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 45/57] leds: versatile: " Jacek Anaszewski
@ 2015-08-11 9:37 ` Jacek Anaszewski
2015-08-11 9:38 ` [PATCH/RFC v5 47/57] leds: net48xx: " Jacek Anaszewski
` (10 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:37 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski,
Kristian Kielhofner
The driver neither sleeps nor uses delays in its brightness_set op.
Use a new brightness_set_nonblocking op instead of brightness_set for
setting brightness then.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Kristian Kielhofner <kris@krisk.org>
---
drivers/leds/leds-wrap.c | 20 ++++++++++----------
1 file changed, 10 insertions(+), 10 deletions(-)
diff --git a/drivers/leds/leds-wrap.c b/drivers/leds/leds-wrap.c
index 1ba3def..aea72e2 100644
--- a/drivers/leds/leds-wrap.c
+++ b/drivers/leds/leds-wrap.c
@@ -54,22 +54,22 @@ static void wrap_extra_led_set(struct led_classdev *led_cdev,
}
static struct led_classdev wrap_power_led = {
- .name = "wrap::power",
- .brightness_set = wrap_power_led_set,
- .default_trigger = "default-on",
- .flags = LED_CORE_SUSPENDRESUME,
+ .name = "wrap::power",
+ .brightness_set_nonblocking = wrap_power_led_set,
+ .default_trigger = "default-on",
+ .flags = LED_CORE_SUSPENDRESUME,
};
static struct led_classdev wrap_error_led = {
- .name = "wrap::error",
- .brightness_set = wrap_error_led_set,
- .flags = LED_CORE_SUSPENDRESUME,
+ .name = "wrap::error",
+ .brightness_set_nonblocking = wrap_error_led_set,
+ .flags = LED_CORE_SUSPENDRESUME,
};
static struct led_classdev wrap_extra_led = {
- .name = "wrap::extra",
- .brightness_set = wrap_extra_led_set,
- .flags = LED_CORE_SUSPENDRESUME,
+ .name = "wrap::extra",
+ .brightness_set_nonblocking = wrap_extra_led_set,
+ .flags = LED_CORE_SUSPENDRESUME,
};
static int wrap_led_probe(struct platform_device *pdev)
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 47/57] leds: net48xx: Use brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (45 preceding siblings ...)
2015-08-11 9:37 ` [PATCH/RFC v5 46/57] leds: wrap: " Jacek Anaszewski
@ 2015-08-11 9:38 ` Jacek Anaszewski
2015-08-11 9:38 ` [PATCH/RFC v5 48/57] leds: asic3: " Jacek Anaszewski
` (9 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:38 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski, Chris Boot
The driver neither sleeps nor uses delays in its brightness_set op.
Use a new brightness_set_nonblocking op instead of brightness_set for
setting brightness then.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Chris Boot <bootc@bootc.net>
---
drivers/leds/leds-net48xx.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/leds/leds-net48xx.c b/drivers/leds/leds-net48xx.c
index ec3a2e8..b48fdeb 100644
--- a/drivers/leds/leds-net48xx.c
+++ b/drivers/leds/leds-net48xx.c
@@ -32,9 +32,9 @@ static void net48xx_error_led_set(struct led_classdev *led_cdev,
}
static struct led_classdev net48xx_error_led = {
- .name = "net48xx::error",
- .brightness_set = net48xx_error_led_set,
- .flags = LED_CORE_SUSPENDRESUME,
+ .name = "net48xx::error",
+ .brightness_set_nonblocking = net48xx_error_led_set,
+ .flags = LED_CORE_SUSPENDRESUME,
};
static int net48xx_led_probe(struct platform_device *pdev)
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 48/57] leds: asic3: Use brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (46 preceding siblings ...)
2015-08-11 9:38 ` [PATCH/RFC v5 47/57] leds: net48xx: " Jacek Anaszewski
@ 2015-08-11 9:38 ` Jacek Anaszewski
2015-08-11 9:38 ` [PATCH/RFC v5 49/57] leds: cobalt-qube: " Jacek Anaszewski
` (8 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:38 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski, Paul Parsons
The driver neither sleeps nor uses delays in its brightness_set op.
Use a new brightness_set_nonblocking op instead of brightness_set for
setting brightness then.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Paul Parsons <lost.distance@yahoo.com>
---
drivers/leds/leds-asic3.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/leds/leds-asic3.c b/drivers/leds/leds-asic3.c
index 1b71eac..c23f565 100644
--- a/drivers/leds/leds-asic3.c
+++ b/drivers/leds/leds-asic3.c
@@ -109,7 +109,7 @@ static int asic3_led_probe(struct platform_device *pdev)
led->cdev->name = led->name;
led->cdev->flags = LED_CORE_SUSPENDRESUME;
- led->cdev->brightness_set = brightness_set;
+ led->cdev->brightness_set_nonblocking = brightness_set;
led->cdev->blink_set = blink_set;
led->cdev->default_trigger = led->default_trigger;
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 49/57] leds: cobalt-qube: Use brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (47 preceding siblings ...)
2015-08-11 9:38 ` [PATCH/RFC v5 48/57] leds: asic3: " Jacek Anaszewski
@ 2015-08-11 9:38 ` Jacek Anaszewski
2015-08-11 9:38 ` [PATCH/RFC v5 50/57] leds: fsg: " Jacek Anaszewski
` (7 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:38 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski,
Florian Fainelli
The driver neither sleeps nor uses delays in its brightness_set op.
Use a new brightness_set_nonblocking op instead of brightness_set for
setting brightness then.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Florian Fainelli <florian@openwrt.org>
---
drivers/leds/leds-cobalt-qube.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/leds/leds-cobalt-qube.c b/drivers/leds/leds-cobalt-qube.c
index d975220..7cb986e 100644
--- a/drivers/leds/leds-cobalt-qube.c
+++ b/drivers/leds/leds-cobalt-qube.c
@@ -27,10 +27,10 @@ static void qube_front_led_set(struct led_classdev *led_cdev,
}
static struct led_classdev qube_front_led = {
- .name = "qube::front",
- .brightness = LED_FULL,
- .brightness_set = qube_front_led_set,
- .default_trigger = "default-on",
+ .name = "qube::front",
+ .brightness = LED_FULL,
+ .brightness_set_nonblocking = qube_front_led_set,
+ .default_trigger = "default-on",
};
static int cobalt_qube_led_probe(struct platform_device *pdev)
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 50/57] leds: fsg: Use brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (48 preceding siblings ...)
2015-08-11 9:38 ` [PATCH/RFC v5 49/57] leds: cobalt-qube: " Jacek Anaszewski
@ 2015-08-11 9:38 ` Jacek Anaszewski
2015-08-11 9:38 ` [PATCH/RFC v5 51/57] leds: hp6xx: " Jacek Anaszewski
` (6 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:38 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski, Rod Whitby
The driver neither sleeps nor uses delays in its brightness_set op.
Use a new brightness_set_nonblocking op instead of brightness_set for
setting brightness then.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Rod Whitby <rod@whitby.id.au>
---
drivers/leds/leds-fsg.c | 36 ++++++++++++++++++------------------
1 file changed, 18 insertions(+), 18 deletions(-)
diff --git a/drivers/leds/leds-fsg.c b/drivers/leds/leds-fsg.c
index 2b4dc73..605ef9f 100644
--- a/drivers/leds/leds-fsg.c
+++ b/drivers/leds/leds-fsg.c
@@ -107,39 +107,39 @@ static void fsg_led_ring_set(struct led_classdev *led_cdev,
static struct led_classdev fsg_wlan_led = {
- .name = "fsg:blue:wlan",
- .brightness_set = fsg_led_wlan_set,
- .flags = LED_CORE_SUSPENDRESUME,
+ .name = "fsg:blue:wlan",
+ .brightness_set_nonblocking = fsg_led_wlan_set,
+ .flags = LED_CORE_SUSPENDRESUME,
};
static struct led_classdev fsg_wan_led = {
- .name = "fsg:blue:wan",
- .brightness_set = fsg_led_wan_set,
- .flags = LED_CORE_SUSPENDRESUME,
+ .name = "fsg:blue:wan",
+ .brightness_set_nonblocking = fsg_led_wan_set,
+ .flags = LED_CORE_SUSPENDRESUME,
};
static struct led_classdev fsg_sata_led = {
- .name = "fsg:blue:sata",
- .brightness_set = fsg_led_sata_set,
- .flags = LED_CORE_SUSPENDRESUME,
+ .name = "fsg:blue:sata",
+ .brightness_set_nonblocking = fsg_led_sata_set,
+ .flags = LED_CORE_SUSPENDRESUME,
};
static struct led_classdev fsg_usb_led = {
- .name = "fsg:blue:usb",
- .brightness_set = fsg_led_usb_set,
- .flags = LED_CORE_SUSPENDRESUME,
+ .name = "fsg:blue:usb",
+ .brightness_set_nonblocking = fsg_led_usb_set,
+ .flags = LED_CORE_SUSPENDRESUME,
};
static struct led_classdev fsg_sync_led = {
- .name = "fsg:blue:sync",
- .brightness_set = fsg_led_sync_set,
- .flags = LED_CORE_SUSPENDRESUME,
+ .name = "fsg:blue:sync",
+ .brightness_set_nonblocking = fsg_led_sync_set,
+ .flags = LED_CORE_SUSPENDRESUME,
};
static struct led_classdev fsg_ring_led = {
- .name = "fsg:blue:ring",
- .brightness_set = fsg_led_ring_set,
- .flags = LED_CORE_SUSPENDRESUME,
+ .name = "fsg:blue:ring",
+ .brightness_set_nonblocking = fsg_led_ring_set,
+ .flags = LED_CORE_SUSPENDRESUME,
};
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 51/57] leds: hp6xx: Use brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (49 preceding siblings ...)
2015-08-11 9:38 ` [PATCH/RFC v5 50/57] leds: fsg: " Jacek Anaszewski
@ 2015-08-11 9:38 ` Jacek Anaszewski
2015-08-11 9:38 ` [PATCH/RFC v5 52/57] leds: locomo: " Jacek Anaszewski
` (5 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:38 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski,
Kristoffer Ericson
The driver neither sleeps nor uses delays in its brightness_set op.
Use a new brightness_set_nonblocking op instead of brightness_set for
setting brightness then.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Kristoffer Ericson <kristoffer.ericson@gmail.com>
---
drivers/leds/leds-hp6xx.c | 16 ++++++++--------
1 file changed, 8 insertions(+), 8 deletions(-)
diff --git a/drivers/leds/leds-hp6xx.c b/drivers/leds/leds-hp6xx.c
index 0b84c01..5848e2b 100644
--- a/drivers/leds/leds-hp6xx.c
+++ b/drivers/leds/leds-hp6xx.c
@@ -42,17 +42,17 @@ static void hp6xxled_red_set(struct led_classdev *led_cdev,
}
static struct led_classdev hp6xx_red_led = {
- .name = "hp6xx:red",
- .default_trigger = "hp6xx-charge",
- .brightness_set = hp6xxled_red_set,
- .flags = LED_CORE_SUSPENDRESUME,
+ .name = "hp6xx:red",
+ .default_trigger = "hp6xx-charge",
+ .brightness_set_nonblocking = hp6xxled_red_set,
+ .flags = LED_CORE_SUSPENDRESUME,
};
static struct led_classdev hp6xx_green_led = {
- .name = "hp6xx:green",
- .default_trigger = "ide-disk",
- .brightness_set = hp6xxled_green_set,
- .flags = LED_CORE_SUSPENDRESUME,
+ .name = "hp6xx:green",
+ .default_trigger = "ide-disk",
+ .brightness_set_nonblocking = hp6xxled_green_set,
+ .flags = LED_CORE_SUSPENDRESUME,
};
static int hp6xxled_probe(struct platform_device *pdev)
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 52/57] leds: locomo: Use brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (50 preceding siblings ...)
2015-08-11 9:38 ` [PATCH/RFC v5 51/57] leds: hp6xx: " Jacek Anaszewski
@ 2015-08-11 9:38 ` Jacek Anaszewski
2015-08-11 9:38 ` [PATCH/RFC v5 53/57] leds: dell: " Jacek Anaszewski
` (4 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:38 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski, John Lenz
The driver neither sleeps nor uses delays in its brightness_set op.
Use a new brightness_set_nonblocking op instead of brightness_set for
setting brightness then.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: John Lenz <lenz@cs.wisc.edu>
---
drivers/leds/leds-locomo.c | 12 ++++++------
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/drivers/leds/leds-locomo.c b/drivers/leds/leds-locomo.c
index 80ba048..2de7063 100644
--- a/drivers/leds/leds-locomo.c
+++ b/drivers/leds/leds-locomo.c
@@ -44,15 +44,15 @@ static void locomoled_brightness_set1(struct led_classdev *led_cdev,
}
static struct led_classdev locomo_led0 = {
- .name = "locomo:amber:charge",
- .default_trigger = "main-battery-charging",
- .brightness_set = locomoled_brightness_set0,
+ .name = "locomo:amber:charge",
+ .default_trigger = "main-battery-charging",
+ .brightness_set_nonblocking = locomoled_brightness_set0,
};
static struct led_classdev locomo_led1 = {
- .name = "locomo:green:mail",
- .default_trigger = "nand-disk",
- .brightness_set = locomoled_brightness_set1,
+ .name = "locomo:green:mail",
+ .default_trigger = "nand-disk",
+ .brightness_set_nonblocking = locomoled_brightness_set1,
};
static int locomoled_probe(struct locomo_dev *ldev)
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 53/57] leds: dell: Use brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (51 preceding siblings ...)
2015-08-11 9:38 ` [PATCH/RFC v5 52/57] leds: locomo: " Jacek Anaszewski
@ 2015-08-11 9:38 ` Jacek Anaszewski
2015-08-11 9:38 ` [PATCH/RFC v5 54/57] leds: rb532: " Jacek Anaszewski
` (3 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:38 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski,
Louis Davis, Jim Dailey
The driver neither sleeps nor uses delays in its brightness_set op.
Use a new brightness_set_nonblocking op instead of brightness_set for
setting brightness then.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Louis Davis <louis_davis@dell.com>
Cc: Jim Dailey <jim_dailey@dell.com>
---
drivers/leds/dell-led.c | 12 ++++++------
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/drivers/leds/dell-led.c b/drivers/leds/dell-led.c
index c36acaf..8a69849 100644
--- a/drivers/leds/dell-led.c
+++ b/drivers/leds/dell-led.c
@@ -315,12 +315,12 @@ static int dell_led_blink(struct led_classdev *led_cdev,
}
static struct led_classdev dell_led = {
- .name = "dell::lid",
- .brightness = LED_OFF,
- .max_brightness = 1,
- .brightness_set = dell_led_set,
- .blink_set = dell_led_blink,
- .flags = LED_CORE_SUSPENDRESUME,
+ .name = "dell::lid",
+ .brightness = LED_OFF,
+ .max_brightness = 1,
+ .brightness_set_nonblocking = dell_led_set,
+ .blink_set = dell_led_blink,
+ .flags = LED_CORE_SUSPENDRESUME,
};
static int __init dell_led_init(void)
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 54/57] leds: rb532: Use brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (52 preceding siblings ...)
2015-08-11 9:38 ` [PATCH/RFC v5 53/57] leds: dell: " Jacek Anaszewski
@ 2015-08-11 9:38 ` Jacek Anaszewski
2015-08-11 9:38 ` [PATCH/RFC v5 55/57] leds: sunfire: " Jacek Anaszewski
` (2 subsequent siblings)
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:38 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski, Phil Sutter
The driver neither sleeps nor uses delays in its brightness_set op.
Use a new brightness_set_nonblocking op instead of brightness_set for
setting brightness then.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Phil Sutter <n0-1@freewrt.org>
---
drivers/leds/leds-rb532.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/leds/leds-rb532.c b/drivers/leds/leds-rb532.c
index fcd1215b..67e220f 100644
--- a/drivers/leds/leds-rb532.c
+++ b/drivers/leds/leds-rb532.c
@@ -31,10 +31,10 @@ static enum led_brightness rb532_led_get(struct led_classdev *cdev)
}
static struct led_classdev rb532_uled = {
- .name = "uled",
- .brightness_set = rb532_led_set,
- .brightness_get = rb532_led_get,
- .default_trigger = "nand-disk",
+ .name = "uled",
+ .brightness_set_nonblocking = rb532_led_set,
+ .brightness_get = rb532_led_get,
+ .default_trigger = "nand-disk",
};
static int rb532_led_probe(struct platform_device *pdev)
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 55/57] leds: sunfire: Use brightness_set_nonblocking op
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (53 preceding siblings ...)
2015-08-11 9:38 ` [PATCH/RFC v5 54/57] leds: rb532: " Jacek Anaszewski
@ 2015-08-11 9:38 ` Jacek Anaszewski
2015-08-11 17:27 ` David Miller
2015-08-11 9:38 ` [PATCH/RFC v5 56/57] leds: ipaq-micro: Set LED_BRIGHTNESS_BLOCKING flag Jacek Anaszewski
2015-08-11 9:38 ` [PATCH/RFC v5 57/57] leds: clevo-mail: " Jacek Anaszewski
56 siblings, 1 reply; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:38 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski, David S. Miller
The driver neither sleeps nor uses delays in its brightness_set op.
Use a new brightness_set_nonblocking op instead of brightness_set for
setting brightness then.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: David S. Miller <davem@davemloft.net>
---
drivers/leds/leds-sunfire.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/leds/leds-sunfire.c b/drivers/leds/leds-sunfire.c
index c2553c5..0debc56 100644
--- a/drivers/leds/leds-sunfire.c
+++ b/drivers/leds/leds-sunfire.c
@@ -144,7 +144,7 @@ static int sunfire_led_generic_probe(struct platform_device *pdev,
p->leds[i].reg = (void __iomem *) pdev->resource[0].start;
lp->name = types[i].name;
lp->brightness = LED_FULL;
- lp->brightness_set = types[i].handler;
+ lp->brightness_set_nonblocking = types[i].handler;
lp->default_trigger = types[i].default_trigger;
err = led_classdev_register(&pdev->dev, lp);
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* Re: [PATCH/RFC v5 55/57] leds: sunfire: Use brightness_set_nonblocking op
2015-08-11 9:38 ` [PATCH/RFC v5 55/57] leds: sunfire: " Jacek Anaszewski
@ 2015-08-11 17:27 ` David Miller
0 siblings, 0 replies; 72+ messages in thread
From: David Miller @ 2015-08-11 17:27 UTC (permalink / raw)
To: j.anaszewski; +Cc: linux-leds, linux-kernel, cooloney, rpurdie, stsp
From: Jacek Anaszewski <j.anaszewski@samsung.com>
Date: Tue, 11 Aug 2015 11:38:08 +0200
> The driver neither sleeps nor uses delays in its brightness_set op.
> Use a new brightness_set_nonblocking op instead of brightness_set for
> setting brightness then.
>
> Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Acked-by: David S. Miller <davem@davemloft.net>
^ permalink raw reply [flat|nested] 72+ messages in thread
* [PATCH/RFC v5 56/57] leds: ipaq-micro: Set LED_BRIGHTNESS_BLOCKING flag
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (54 preceding siblings ...)
2015-08-11 9:38 ` [PATCH/RFC v5 55/57] leds: sunfire: " Jacek Anaszewski
@ 2015-08-11 9:38 ` Jacek Anaszewski
2015-08-11 12:55 ` Linus Walleij
2015-08-11 9:38 ` [PATCH/RFC v5 57/57] leds: clevo-mail: " Jacek Anaszewski
56 siblings, 1 reply; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:38 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski, Linus Walleij
Since implemented brightness_set op can sleep mark the LED blocking.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Linus Walleij <linus.walleij@linaro.org>
---
drivers/leds/leds-ipaq-micro.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/leds/leds-ipaq-micro.c b/drivers/leds/leds-ipaq-micro.c
index 3776f51..fb82bd3 100644
--- a/drivers/leds/leds-ipaq-micro.c
+++ b/drivers/leds/leds-ipaq-micro.c
@@ -104,7 +104,8 @@ static struct led_classdev micro_led = {
.name = "led-ipaq-micro",
.brightness_set = micro_leds_brightness_set,
.blink_set = micro_leds_blink_set,
- .flags = LED_CORE_SUSPENDRESUME,
+ .flags = LED_CORE_SUSPENDRESUME |
+ LED_BRIGHTNESS_BLOCKING,
};
static int micro_leds_probe(struct platform_device *pdev)
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread* [PATCH/RFC v5 57/57] leds: clevo-mail: Set LED_BRIGHTNESS_BLOCKING flag
2015-08-11 9:37 [PATCH/RFC v5 00/57] Remove work queues from LED class drivers Jacek Anaszewski
` (55 preceding siblings ...)
2015-08-11 9:38 ` [PATCH/RFC v5 56/57] leds: ipaq-micro: Set LED_BRIGHTNESS_BLOCKING flag Jacek Anaszewski
@ 2015-08-11 9:38 ` Jacek Anaszewski
56 siblings, 0 replies; 72+ messages in thread
From: Jacek Anaszewski @ 2015-08-11 9:38 UTC (permalink / raw)
To: linux-leds
Cc: linux-kernel, cooloney, rpurdie, stsp, Jacek Anaszewski,
Márton Németh
Since implemented brightness_set op uses delays mark the
LED blocking.
Signed-off-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Cc: Márton Németh <nm127@freemail.hu>
---
drivers/leds/leds-clevo-mail.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/leds/leds-clevo-mail.c b/drivers/leds/leds-clevo-mail.c
index 0f9ed1e..961da85 100644
--- a/drivers/leds/leds-clevo-mail.c
+++ b/drivers/leds/leds-clevo-mail.c
@@ -150,7 +150,8 @@ static struct led_classdev clevo_mail_led = {
.name = "clevo::mail",
.brightness_set = clevo_mail_led_set,
.blink_set = clevo_mail_led_blink,
- .flags = LED_CORE_SUSPENDRESUME,
+ .flags = LED_CORE_SUSPENDRESUME |
+ LED_BRIGHTNESS_BLOCKING,
};
static int __init clevo_mail_led_probe(struct platform_device *pdev)
--
1.7.9.5
^ permalink raw reply related [flat|nested] 72+ messages in thread