| Message ID | 1443301358-2131-6-git-send-email-simon.guinot@sequanux.org (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Hi Simon, Does your device support reading the brightness currently set? If so, it would be good to implement brightness_get op, because AFAIR you mentioned that the firmware you are working with sets always maximum brightness value. Having the op implemented would allow to find this out. On 09/26/2015 11:02 PM, Simon Guinot wrote: > This patch sets the led_classdev max_brightness to the maximum level > value supported by hardware. This allows to get rid of the brightness > conversion operation (from software [0:LED_FULL] to hardware ranges) in > brightness_set(). > > Signed-off-by: Simon Guinot <simon.guinot@sequanux.org> > --- > drivers/leds/leds-netxbig.c | 12 ++++-------- > 1 file changed, 4 insertions(+), 8 deletions(-) > > diff --git a/drivers/leds/leds-netxbig.c b/drivers/leds/leds-netxbig.c > index b166fd9f4186..4b88b93244be 100644 > --- a/drivers/leds/leds-netxbig.c > +++ b/drivers/leds/leds-netxbig.c > @@ -116,7 +116,6 @@ struct netxbig_led_data { > int mode_addr; > int *mode_val; > int bright_addr; > - int bright_max; > struct netxbig_led_timer *timer; > int num_timer; > enum netxbig_led_mode mode; > @@ -178,7 +177,7 @@ static void netxbig_led_set(struct led_classdev *led_cdev, > struct netxbig_led_data *led_dat = > container_of(led_cdev, struct netxbig_led_data, cdev); > enum netxbig_led_mode mode; > - int mode_val, bright_val; > + int mode_val; > int set_brightness = 1; > unsigned long flags; > > @@ -204,12 +203,9 @@ static void netxbig_led_set(struct led_classdev *led_cdev, > * SATA LEDs. So, change the brightness setting for a single > * SATA LED will affect all the others. > */ > - if (set_brightness) { > - bright_val = DIV_ROUND_UP(value * led_dat->bright_max, > - LED_FULL); > + if (set_brightness) > gpio_ext_set_value(led_dat->gpio_ext, > - led_dat->bright_addr, bright_val); > - } > + led_dat->bright_addr, value); > > spin_unlock_irqrestore(&led_dat->lock, flags); > } > @@ -306,11 +302,11 @@ static int create_netxbig_led(struct platform_device *pdev, > */ > led_dat->sata = 0; > led_dat->cdev.brightness = LED_OFF; > + led_dat->cdev.max_brightness = template->bright_max; > led_dat->cdev.flags |= LED_CORE_SUSPENDRESUME; > led_dat->mode_addr = template->mode_addr; > led_dat->mode_val = template->mode_val; > led_dat->bright_addr = template->bright_addr; > - led_dat->bright_max = template->bright_max; > led_dat->timer = pdata->timer; > led_dat->num_timer = pdata->num_timer; > /* >
On Mon, Sep 28, 2015 at 10:02:35AM +0200, Jacek Anaszewski wrote: > Hi Simon, Hi Jacek, > > Does your device support reading the brightness currently set? No it don't. > If so, it would be good to implement brightness_get op, because > AFAIR you mentioned that the firmware you are working with sets > always maximum brightness value. Having the op implemented would > allow to find this out. I don't understand how this can help. I mean, the only issue is that at startup the initial LED state is unknown. And the software brightness value could be false. But once the LED is configured, the brightness values for software and hardware are synchronized. The brightness value is stored/cached in led_classdev and it can be retrieved by the user via sysfs... For my own knowledge, is there some interest in having brightness_get(), aside of guessing the LED initial state ? What does the embedded firmware is writing 255 or 0 into the brightness sysfs attribute. The max_brightness value is ignored. After this patch, writing 255 and 0 still allows to configure the LED in the same way: maximum brightness or off. Thus, I believe there is no compatibility issue. But of course, I will update the firmware as well :) Thanks, Simon > > On 09/26/2015 11:02 PM, Simon Guinot wrote: > >This patch sets the led_classdev max_brightness to the maximum level > >value supported by hardware. This allows to get rid of the brightness > >conversion operation (from software [0:LED_FULL] to hardware ranges) in > >brightness_set(). > > > >Signed-off-by: Simon Guinot <simon.guinot@sequanux.org> > >--- > > drivers/leds/leds-netxbig.c | 12 ++++-------- > > 1 file changed, 4 insertions(+), 8 deletions(-) > > > >diff --git a/drivers/leds/leds-netxbig.c b/drivers/leds/leds-netxbig.c > >index b166fd9f4186..4b88b93244be 100644 > >--- a/drivers/leds/leds-netxbig.c > >+++ b/drivers/leds/leds-netxbig.c > >@@ -116,7 +116,6 @@ struct netxbig_led_data { > > int mode_addr; > > int *mode_val; > > int bright_addr; > >- int bright_max; > > struct netxbig_led_timer *timer; > > int num_timer; > > enum netxbig_led_mode mode; > >@@ -178,7 +177,7 @@ static void netxbig_led_set(struct led_classdev *led_cdev, > > struct netxbig_led_data *led_dat = > > container_of(led_cdev, struct netxbig_led_data, cdev); > > enum netxbig_led_mode mode; > >- int mode_val, bright_val; > >+ int mode_val; > > int set_brightness = 1; > > unsigned long flags; > > > >@@ -204,12 +203,9 @@ static void netxbig_led_set(struct led_classdev *led_cdev, > > * SATA LEDs. So, change the brightness setting for a single > > * SATA LED will affect all the others. > > */ > >- if (set_brightness) { > >- bright_val = DIV_ROUND_UP(value * led_dat->bright_max, > >- LED_FULL); > >+ if (set_brightness) > > gpio_ext_set_value(led_dat->gpio_ext, > >- led_dat->bright_addr, bright_val); > >- } > >+ led_dat->bright_addr, value); > > > > spin_unlock_irqrestore(&led_dat->lock, flags); > > } > >@@ -306,11 +302,11 @@ static int create_netxbig_led(struct platform_device *pdev, > > */ > > led_dat->sata = 0; > > led_dat->cdev.brightness = LED_OFF; > >+ led_dat->cdev.max_brightness = template->bright_max; > > led_dat->cdev.flags |= LED_CORE_SUSPENDRESUME; > > led_dat->mode_addr = template->mode_addr; > > led_dat->mode_val = template->mode_val; > > led_dat->bright_addr = template->bright_addr; > >- led_dat->bright_max = template->bright_max; > > led_dat->timer = pdata->timer; > > led_dat->num_timer = pdata->num_timer; > > /* > > > > > -- > Best Regards, > Jacek Anaszewski > -- > To unsubscribe from this list: send the line "unsubscribe linux-leds" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html
On 09/28/2015 11:19 AM, Simon Guinot wrote: > On Mon, Sep 28, 2015 at 10:02:35AM +0200, Jacek Anaszewski wrote: >> Hi Simon, > > Hi Jacek, > >> >> Does your device support reading the brightness currently set? > > No it don't. > >> If so, it would be good to implement brightness_get op, because >> AFAIR you mentioned that the firmware you are working with sets >> always maximum brightness value. Having the op implemented would >> allow to find this out. > > I don't understand how this can help. I mean, the only issue is that at > startup the initial LED state is unknown. And the software brightness > value could be false. But once the LED is configured, the brightness > values for software and hardware are synchronized. The brightness value > is stored/cached in led_classdev and it can be retrieved by the user via > sysfs... > > For my own knowledge, is there some interest in having brightness_get(), > aside of guessing the LED initial state ? Some LED controllers can adjust brightness in case battery voltage level falls below some threshold. > What does the embedded firmware is writing 255 or 0 into the brightness > sysfs attribute. The max_brightness value is ignored. After this patch, > writing 255 and 0 still allows to configure the LED in the same way: > maximum brightness or off. Thus, I believe there is no compatibility > issue. LED core always assures that brightness value passed to brightness_set op does not exceed max_brightness value. So, now after executing "echo 255 > brightness", LED core will adjust it to max_brightness (e.g. 7) before passing to brightness_set. In the message [1], you mentioned that "LEDs are only enabled at their maximum level", so IIUC following is possible: #echo 3 > "brightness" firmware sets brightness to max_brightness from DT (e.g. 7), but #cat brightness #3 Is it true? > But of course, I will update the firmware as well :) [1] http://www.spinics.net/lists/devicetree/msg94358.html
On Mon, Sep 28, 2015 at 12:15:22PM +0200, Jacek Anaszewski wrote: > On 09/28/2015 11:19 AM, Simon Guinot wrote: > >On Mon, Sep 28, 2015 at 10:02:35AM +0200, Jacek Anaszewski wrote: > >>Hi Simon, > > > >Hi Jacek, > > > >> > >>Does your device support reading the brightness currently set? > > > >No it don't. > > > >>If so, it would be good to implement brightness_get op, because > >>AFAIR you mentioned that the firmware you are working with sets > >>always maximum brightness value. Having the op implemented would > >>allow to find this out. > > > >I don't understand how this can help. I mean, the only issue is that at > >startup the initial LED state is unknown. And the software brightness > >value could be false. But once the LED is configured, the brightness > >values for software and hardware are synchronized. The brightness value > >is stored/cached in led_classdev and it can be retrieved by the user via > >sysfs... > > > >For my own knowledge, is there some interest in having brightness_get(), > >aside of guessing the LED initial state ? > > Some LED controllers can adjust brightness in case battery voltage level > falls below some threshold. OK, thanks for the explanation. > > >What does the embedded firmware is writing 255 or 0 into the brightness > >sysfs attribute. The max_brightness value is ignored. After this patch, > >writing 255 and 0 still allows to configure the LED in the same way: > >maximum brightness or off. Thus, I believe there is no compatibility > >issue. > > LED core always assures that brightness value passed to brightness_set > op does not exceed max_brightness value. So, now after executing > "echo 255 > brightness", LED core will adjust it to max_brightness > (e.g. 7) before passing to brightness_set. > > In the message [1], you mentioned that "LEDs are only enabled at their > maximum level", so IIUC following is possible: > > #echo 3 > "brightness" > > firmware sets brightness to max_brightness from DT (e.g. 7), but > > #cat brightness > #3 > > Is it true? Oh no sorry, it is a misunderstanding. By "LEDs are only enabled at their maximum level", I was meaning "LEDs are only enabled at their maximum level by the LaCie stock firmware". The firmware don't make use of the different hardware brightness levels available. But the feature works perfectly. If you write 3 into sysfs "brightness", then you get the third brightness level. OK, I understand you remark now. Sorry for not being very clear in a first place. Simon
On 09/28/2015 01:50 PM, Simon Guinot wrote: > On Mon, Sep 28, 2015 at 12:15:22PM +0200, Jacek Anaszewski wrote: >> On 09/28/2015 11:19 AM, Simon Guinot wrote: >>> On Mon, Sep 28, 2015 at 10:02:35AM +0200, Jacek Anaszewski wrote: >>>> Hi Simon, >>> >>> Hi Jacek, >>> >>>> >>>> Does your device support reading the brightness currently set? >>> >>> No it don't. >>> >>>> If so, it would be good to implement brightness_get op, because >>>> AFAIR you mentioned that the firmware you are working with sets >>>> always maximum brightness value. Having the op implemented would >>>> allow to find this out. >>> >>> I don't understand how this can help. I mean, the only issue is that at >>> startup the initial LED state is unknown. And the software brightness >>> value could be false. But once the LED is configured, the brightness >>> values for software and hardware are synchronized. The brightness value >>> is stored/cached in led_classdev and it can be retrieved by the user via >>> sysfs... >>> >>> For my own knowledge, is there some interest in having brightness_get(), >>> aside of guessing the LED initial state ? >> >> Some LED controllers can adjust brightness in case battery voltage level >> falls below some threshold. > > OK, thanks for the explanation. > >> >>> What does the embedded firmware is writing 255 or 0 into the brightness >>> sysfs attribute. The max_brightness value is ignored. After this patch, >>> writing 255 and 0 still allows to configure the LED in the same way: >>> maximum brightness or off. Thus, I believe there is no compatibility >>> issue. >> >> LED core always assures that brightness value passed to brightness_set >> op does not exceed max_brightness value. So, now after executing >> "echo 255 > brightness", LED core will adjust it to max_brightness >> (e.g. 7) before passing to brightness_set. >> >> In the message [1], you mentioned that "LEDs are only enabled at their >> maximum level", so IIUC following is possible: >> >> #echo 3 > "brightness" >> >> firmware sets brightness to max_brightness from DT (e.g. 7), but >> >> #cat brightness >> #3 >> >> Is it true? > > Oh no sorry, it is a misunderstanding. By "LEDs are only enabled at > their maximum level", I was meaning "LEDs are only enabled at their > maximum level by the LaCie stock firmware". The firmware don't make > use of the different hardware brightness levels available. But the > feature works perfectly. If you write 3 into sysfs "brightness", then > you get the third brightness level. I thought that driver talks to firmware, which controls the LEDs. From your explanation I infer that this driver replaces LaCie stock firmware, am I right? There must be however some circuit that controls LED brightness then. > OK, I understand you remark now. Sorry for not being very clear in a > first place.
On Mon, Sep 28, 2015 at 02:24:40PM +0200, Jacek Anaszewski wrote: > On 09/28/2015 01:50 PM, Simon Guinot wrote: > >On Mon, Sep 28, 2015 at 12:15:22PM +0200, Jacek Anaszewski wrote: > >>On 09/28/2015 11:19 AM, Simon Guinot wrote: > >>>On Mon, Sep 28, 2015 at 10:02:35AM +0200, Jacek Anaszewski wrote: > >>>>Hi Simon, > >>> > >>>Hi Jacek, > >>> > >>>> > >>>>Does your device support reading the brightness currently set? > >>> > >>>No it don't. > >>> > >>>>If so, it would be good to implement brightness_get op, because > >>>>AFAIR you mentioned that the firmware you are working with sets > >>>>always maximum brightness value. Having the op implemented would > >>>>allow to find this out. > >>> > >>>I don't understand how this can help. I mean, the only issue is that at > >>>startup the initial LED state is unknown. And the software brightness > >>>value could be false. But once the LED is configured, the brightness > >>>values for software and hardware are synchronized. The brightness value > >>>is stored/cached in led_classdev and it can be retrieved by the user via > >>>sysfs... > >>> > >>>For my own knowledge, is there some interest in having brightness_get(), > >>>aside of guessing the LED initial state ? > >> > >>Some LED controllers can adjust brightness in case battery voltage level > >>falls below some threshold. > > > >OK, thanks for the explanation. > > > >> > >>>What does the embedded firmware is writing 255 or 0 into the brightness > >>>sysfs attribute. The max_brightness value is ignored. After this patch, > >>>writing 255 and 0 still allows to configure the LED in the same way: > >>>maximum brightness or off. Thus, I believe there is no compatibility > >>>issue. > >> > >>LED core always assures that brightness value passed to brightness_set > >>op does not exceed max_brightness value. So, now after executing > >>"echo 255 > brightness", LED core will adjust it to max_brightness > >>(e.g. 7) before passing to brightness_set. > >> > >>In the message [1], you mentioned that "LEDs are only enabled at their > >>maximum level", so IIUC following is possible: > >> > >>#echo 3 > "brightness" > >> > >>firmware sets brightness to max_brightness from DT (e.g. 7), but > >> > >>#cat brightness > >>#3 > >> > >>Is it true? > > > >Oh no sorry, it is a misunderstanding. By "LEDs are only enabled at > >their maximum level", I was meaning "LEDs are only enabled at their > >maximum level by the LaCie stock firmware". The firmware don't make > >use of the different hardware brightness levels available. But the > >feature works perfectly. If you write 3 into sysfs "brightness", then > >you get the third brightness level. > > I thought that driver talks to firmware, which controls the LEDs. > From your explanation I infer that this driver replaces LaCie stock > firmware, am I right? There must be however some circuit that controls > LED brightness then. OK, I think I may have managed to confuse you completely. The LEDs are controlled by a CPLD. The leds-netxbig driver talks to the CPLD via the gpio-ext "kind of" bus: leds-netxbig -> gpio-ext bus -> CPLD -> LEDs "LaCie stock firmware" don't refer to the firmware embedded in the CPLD but to the stock LaCie Linux system running *on* the board. "LEDs are only enabled at their maximum level by the LaCie stock firmware" means: "In the Linux LaCie system running on the board, userland only configures the LED brightness to 0 or 255". It don't mean that the CPLD does some weird stuff implying that the feature is somewhat broken. No, the feature works. Simply userland (in the stock LaCie system aka firmware) don't use it. Remember that your original question was: "Doesn't specification of your device say what current value given brightness level reflects?" Let me rephrase my answer without using "LaCie stock firmware": No this information is not available in the board specification. It is probably because this feature is not used by the LaCie Linux userland. The LED brightness is configured to off or full. Other levels are not used. That's probably why no one took care of measuring the LED current consumption. Let me know if it is still unclear for you. Simon
On 09/28/2015 03:25 PM, Simon Guinot wrote: > On Mon, Sep 28, 2015 at 02:24:40PM +0200, Jacek Anaszewski wrote: >> On 09/28/2015 01:50 PM, Simon Guinot wrote: >>> On Mon, Sep 28, 2015 at 12:15:22PM +0200, Jacek Anaszewski wrote: >>>> On 09/28/2015 11:19 AM, Simon Guinot wrote: >>>>> On Mon, Sep 28, 2015 at 10:02:35AM +0200, Jacek Anaszewski wrote: >>>>>> Hi Simon, >>>>> >>>>> Hi Jacek, >>>>> >>>>>> >>>>>> Does your device support reading the brightness currently set? >>>>> >>>>> No it don't. >>>>> >>>>>> If so, it would be good to implement brightness_get op, because >>>>>> AFAIR you mentioned that the firmware you are working with sets >>>>>> always maximum brightness value. Having the op implemented would >>>>>> allow to find this out. >>>>> >>>>> I don't understand how this can help. I mean, the only issue is that at >>>>> startup the initial LED state is unknown. And the software brightness >>>>> value could be false. But once the LED is configured, the brightness >>>>> values for software and hardware are synchronized. The brightness value >>>>> is stored/cached in led_classdev and it can be retrieved by the user via >>>>> sysfs... >>>>> >>>>> For my own knowledge, is there some interest in having brightness_get(), >>>>> aside of guessing the LED initial state ? >>>> >>>> Some LED controllers can adjust brightness in case battery voltage level >>>> falls below some threshold. >>> >>> OK, thanks for the explanation. >>> >>>> >>>>> What does the embedded firmware is writing 255 or 0 into the brightness >>>>> sysfs attribute. The max_brightness value is ignored. After this patch, >>>>> writing 255 and 0 still allows to configure the LED in the same way: >>>>> maximum brightness or off. Thus, I believe there is no compatibility >>>>> issue. >>>> >>>> LED core always assures that brightness value passed to brightness_set >>>> op does not exceed max_brightness value. So, now after executing >>>> "echo 255 > brightness", LED core will adjust it to max_brightness >>>> (e.g. 7) before passing to brightness_set. >>>> >>>> In the message [1], you mentioned that "LEDs are only enabled at their >>>> maximum level", so IIUC following is possible: >>>> >>>> #echo 3 > "brightness" >>>> >>>> firmware sets brightness to max_brightness from DT (e.g. 7), but >>>> >>>> #cat brightness >>>> #3 >>>> >>>> Is it true? >>> >>> Oh no sorry, it is a misunderstanding. By "LEDs are only enabled at >>> their maximum level", I was meaning "LEDs are only enabled at their >>> maximum level by the LaCie stock firmware". The firmware don't make >>> use of the different hardware brightness levels available. But the >>> feature works perfectly. If you write 3 into sysfs "brightness", then >>> you get the third brightness level. >> >> I thought that driver talks to firmware, which controls the LEDs. >> From your explanation I infer that this driver replaces LaCie stock >> firmware, am I right? There must be however some circuit that controls >> LED brightness then. > > OK, I think I may have managed to confuse you completely. > > The LEDs are controlled by a CPLD. The leds-netxbig driver talks to the > CPLD via the gpio-ext "kind of" bus: > > leds-netxbig -> gpio-ext bus -> CPLD -> LEDs > > "LaCie stock firmware" don't refer to the firmware embedded in the CPLD > but to the stock LaCie Linux system running *on* the board. > > "LEDs are only enabled at their maximum level by the LaCie stock > firmware" means: "In the Linux LaCie system running on the board, > userland only configures the LED brightness to 0 or 255". > > It don't mean that the CPLD does some weird stuff implying that the > feature is somewhat broken. No, the feature works. Simply userland > (in the stock LaCie system aka firmware) don't use it. > > Remember that your original question was: > > "Doesn't specification of your device say what current value given > brightness level reflects?" > > Let me rephrase my answer without using "LaCie stock firmware": > > No this information is not available in the board specification. It is > probably because this feature is not used by the LaCie Linux userland. > The LED brightness is configured to off or full. Other levels are not > used. That's probably why no one took care of measuring the LED current > consumption. > > Let me know if it is still unclear for you. Thanks for this explanation. Now everything is clear.
diff --git a/drivers/leds/leds-netxbig.c b/drivers/leds/leds-netxbig.c index b166fd9f4186..4b88b93244be 100644 --- a/drivers/leds/leds-netxbig.c +++ b/drivers/leds/leds-netxbig.c @@ -116,7 +116,6 @@ struct netxbig_led_data { int mode_addr; int *mode_val; int bright_addr; - int bright_max; struct netxbig_led_timer *timer; int num_timer; enum netxbig_led_mode mode; @@ -178,7 +177,7 @@ static void netxbig_led_set(struct led_classdev *led_cdev, struct netxbig_led_data *led_dat = container_of(led_cdev, struct netxbig_led_data, cdev); enum netxbig_led_mode mode; - int mode_val, bright_val; + int mode_val; int set_brightness = 1; unsigned long flags; @@ -204,12 +203,9 @@ static void netxbig_led_set(struct led_classdev *led_cdev, * SATA LEDs. So, change the brightness setting for a single * SATA LED will affect all the others. */ - if (set_brightness) { - bright_val = DIV_ROUND_UP(value * led_dat->bright_max, - LED_FULL); + if (set_brightness) gpio_ext_set_value(led_dat->gpio_ext, - led_dat->bright_addr, bright_val); - } + led_dat->bright_addr, value); spin_unlock_irqrestore(&led_dat->lock, flags); } @@ -306,11 +302,11 @@ static int create_netxbig_led(struct platform_device *pdev, */ led_dat->sata = 0; led_dat->cdev.brightness = LED_OFF; + led_dat->cdev.max_brightness = template->bright_max; led_dat->cdev.flags |= LED_CORE_SUSPENDRESUME; led_dat->mode_addr = template->mode_addr; led_dat->mode_val = template->mode_val; led_dat->bright_addr = template->bright_addr; - led_dat->bright_max = template->bright_max; led_dat->timer = pdata->timer; led_dat->num_timer = pdata->num_timer; /*
This patch sets the led_classdev max_brightness to the maximum level value supported by hardware. This allows to get rid of the brightness conversion operation (from software [0:LED_FULL] to hardware ranges) in brightness_set(). Signed-off-by: Simon Guinot <simon.guinot@sequanux.org> --- drivers/leds/leds-netxbig.c | 12 ++++-------- 1 file changed, 4 insertions(+), 8 deletions(-)