Re: [PATCH v2 1/3] leds: leds-pwm: Convert to use devm_get_pwm

From: Bryan Wu
Date: Mon Dec 03 2012 - 13:32:12 EST


On Mon, Dec 3, 2012 at 6:13 AM, Peter Ujfalusi <peter.ujfalusi@xxxxxx> wrote:
> Hi Bryan,
>
> On 11/14/2012 02:14 AM, Bryan Wu wrote:
>> On Mon, Nov 12, 2012 at 6:41 AM, Peter Ujfalusi <peter.ujfalusi@xxxxxx> wrote:
>>> Update the driver to use the new API for requesting pwm so we can take
>>> advantage of the pwm_lookup table to find the correct pwm to be used for the
>>> LED functionality.
>>> If the devm_get_pwm fails we fall back to legacy mode to try to get the pwm.
>>>
>>> Signed-off-by: Peter Ujfalusi <peter.ujfalusi@xxxxxx>
>>> ---
>>> drivers/leds/leds-pwm.c | 19 ++++++-------------
>>> include/linux/leds_pwm.h | 2 +-
>>> 2 files changed, 7 insertions(+), 14 deletions(-)
>>>
>>> diff --git a/drivers/leds/leds-pwm.c b/drivers/leds/leds-pwm.c
>>> index f2e44c7..c953c75 100644
>>> --- a/drivers/leds/leds-pwm.c
>>> +++ b/drivers/leds/leds-pwm.c
>>> @@ -67,12 +67,11 @@ static int led_pwm_probe(struct platform_device *pdev)
>>> cur_led = &pdata->leds[i];
>>> led_dat = &leds_data[i];
>>>
>>> - led_dat->pwm = pwm_request(cur_led->pwm_id,
>>> - cur_led->name);
>>> + led_dat->pwm = devm_pwm_get(&pdev->dev, cur_led->name);
>>> if (IS_ERR(led_dat->pwm)) {
>>> ret = PTR_ERR(led_dat->pwm);
>>> - dev_err(&pdev->dev, "unable to request PWM %d\n",
>>> - cur_led->pwm_id);
>>> + dev_err(&pdev->dev, "unable to request PWM for %s\n",
>>> + cur_led->name);
>>> goto err;
>>> }
>>>
>>> @@ -86,10 +85,8 @@ static int led_pwm_probe(struct platform_device *pdev)
>>> led_dat->cdev.flags |= LED_CORE_SUSPENDRESUME;
>>>
>>> ret = led_classdev_register(&pdev->dev, &led_dat->cdev);
>>> - if (ret < 0) {
>>> - pwm_free(led_dat->pwm);
>>> + if (ret < 0)
>>> goto err;
>>> - }
>>> }
>>>
>>> platform_set_drvdata(pdev, leds_data);
>>> @@ -98,10 +95,8 @@ static int led_pwm_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(&leds_data[i].cdev);
>>> - pwm_free(leds_data[i].pwm);
>>> - }
>>> }
>>>
>>> return ret;
>>> @@ -115,10 +110,8 @@ static int __devexit led_pwm_remove(struct platform_device *pdev)
>>>
>>> leds_data = platform_get_drvdata(pdev);
>>>
>>> - for (i = 0; i < pdata->num_leds; i++) {
>>> + for (i = 0; i < pdata->num_leds; i++)
>>> led_classdev_unregister(&leds_data[i].cdev);
>>> - pwm_free(leds_data[i].pwm);
>>> - }
>>>
>>> return 0;
>>> }
>>> diff --git a/include/linux/leds_pwm.h b/include/linux/leds_pwm.h
>>> index 33a0711..a65e964 100644
>>> --- a/include/linux/leds_pwm.h
>>> +++ b/include/linux/leds_pwm.h
>>> @@ -7,7 +7,7 @@
>>> struct led_pwm {
>>> const char *name;
>>> const char *default_trigger;
>>> - unsigned pwm_id;
>>> + unsigned pwm_id __deprecated;
>>
>> I suggest we remove this later, we can provide patches from this from
>> platform data of board file. And I think this patch is good for me to
>> merge, will do it soon.
>
> I have marked the pwm_id as deprecated for now to allow one kernel cycle for
> external (?) drivers to adopt. We can remove it for 3.9 I think.
>
> I just checked linux-next today and this series is still not there. Do you
> want me to resend it for you? I can rebase the patches on top of linux-next or
> if you have preference on which tree should I use please let me know.
>

Actually, I'm waiting for some feedback from DT maintainers about this
new binding. But it looks find to me.
I plan to put this series into my -devel branch and target for 3.9, is
that OK for you? Right now, it's -rc7. It's better to put this new
thing for next cycle.

-Bryan

>>
>> Thanks,
>> -Bryan
>>
>>> u8 active_low;
>>> unsigned max_brightness;
>>> unsigned pwm_period_ns;
>>> --
>>> 1.8.0
>>>
>
>
> Thanks,
> Péter
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/