Using gpio_keys with regmapped gpio?

Using gpio_keys with regmapped gpio?

[Thor Thayer]

Hi,

I have a SPI gpio expander chip that is using the regmap framework. 
There are some pushbuttons that I'd like to use the gpio_keys framework 
for but only the first button press creates an input event.

It seems like my problem occurs because I'm only getting events on a 
button press.  The input_get_disposition() function toggles a local copy 
of the button state [__change_bit(code, dev->key);] in the input.c file. 
I think this function also needs to be called when the button is 
released to toggle the dev->key state and re-arm for the next button press.

I tried enabling autorepeat but that causes an infinite loop in the 
input_repeat_key() function.

In both cases, it seems like the button press condition must be cleared 
when the button is released. This seems like it should be common for all 
GPIOs so I'm probably missing something.

Which brings up my next question. Can the gpio_keys framework be used 
with a regmapped gpio? I haven't been able to find any examples of 
gpio_keys with an external gpio expander and maybe this isn't valid usage.

Thanks,

Thor


--
To unsubscribe from this list: send the line "unsubscribe linux-spi" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Using gpio_keys with regmapped gpio?

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 503 bytes --]

On Tue, Mar 31, 2015 at 02:57:35PM -0500, Thor Thayer wrote:

> Which brings up my next question. Can the gpio_keys framework be used with a
> regmapped gpio? I haven't been able to find any examples of gpio_keys with
> an external gpio expander and maybe this isn't valid usage.

Why would there be any problem doing that?  I suspect a lack of examples
is more to do with it being rare to build hardware like that (usually on
SoC GPIOs are fairly abundant and much more performant) than anything
else.

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 473 bytes --]

Re: Using gpio_keys with regmapped gpio?

[Dmitry Torokhov]

On Tue, Mar 31, 2015 at 02:57:35PM -0500, Thor Thayer wrote:
> Hi,
> 
> I have a SPI gpio expander chip that is using the regmap framework.
> There are some pushbuttons that I'd like to use the gpio_keys
> framework for but only the first button press creates an input
> event.
> 
> It seems like my problem occurs because I'm only getting events on a
> button press.  The input_get_disposition() function toggles a local
> copy of the button state [__change_bit(code, dev->key);] in the
> input.c file. I think this function also needs to be called when the
> button is released to toggle the dev->key state and re-arm for the
> next button press.

Yes.

> 
> I tried enabling autorepeat but that causes an infinite loop in the
> input_repeat_key() function.
> 
> In both cases, it seems like the button press condition must be
> cleared when the button is released. This seems like it should be
> common for all GPIOs so I'm probably missing something.

It sounds like your expander does not generate interrupt when the button
is released and so gpio-keys driver is not aware of it.

Thanks.

-- 
Dmitry
--
To unsubscribe from this list: send the line "unsubscribe linux-spi" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Using gpio_keys with regmapped gpio?

[Thor Thayer]

On 03/31/2015 03:17 PM, Dmitry Torokhov wrote:
> On Tue, Mar 31, 2015 at 02:57:35PM -0500, Thor Thayer wrote:
>> Hi,
>>
>> I have a SPI gpio expander chip that is using the regmap framework.
>> There are some pushbuttons that I'd like to use the gpio_keys
>> framework for but only the first button press creates an input
>> event.
>>
>> It seems like my problem occurs because I'm only getting events on a
>> button press.  The input_get_disposition() function toggles a local
>> copy of the button state [__change_bit(code, dev->key);] in the
>> input.c file. I think this function also needs to be called when the
>> button is released to toggle the dev->key state and re-arm for the
>> next button press.
>
> Yes.
>
>>
>> I tried enabling autorepeat but that causes an infinite loop in the
>> input_repeat_key() function.
>>
>> In both cases, it seems like the button press condition must be
>> cleared when the button is released. This seems like it should be
>> common for all GPIOs so I'm probably missing something.
>
> It sounds like your expander does not generate interrupt when the button
> is released and so gpio-keys driver is not aware of it.
>
> Thanks.
>

Thanks for the quick reply. Is giving an interrupt on both press and 
release a requirement for gpio-keys? or is there another setup that I'm 
not aware of?

How does the gpio-key framework work with register mapped GPIOs that are 
rising-edge or falling-edge triggered? I tend to think of these IRQ 
states as the general case for GPIO.

Thanks


--
To unsubscribe from this list: send the line "unsubscribe linux-spi" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Using gpio_keys with regmapped gpio?

[Dmitry Torokhov]

On Tue, Mar 31, 2015 at 2:06 PM, Thor Thayer
<tthayer-yzvPICuk2ABMcg4IHK0kFoH6Mc4MB0Vx@public.gmane.org> wrote:
>
>
> On 03/31/2015 03:17 PM, Dmitry Torokhov wrote:
>>
>> On Tue, Mar 31, 2015 at 02:57:35PM -0500, Thor Thayer wrote:
>>>
>>> Hi,
>>>
>>> I have a SPI gpio expander chip that is using the regmap framework.
>>> There are some pushbuttons that I'd like to use the gpio_keys
>>> framework for but only the first button press creates an input
>>> event.
>>>
>>> It seems like my problem occurs because I'm only getting events on a
>>> button press.  The input_get_disposition() function toggles a local
>>> copy of the button state [__change_bit(code, dev->key);] in the
>>> input.c file. I think this function also needs to be called when the
>>> button is released to toggle the dev->key state and re-arm for the
>>> next button press.
>>
>>
>> Yes.
>>
>>>
>>> I tried enabling autorepeat but that causes an infinite loop in the
>>> input_repeat_key() function.
>>>
>>> In both cases, it seems like the button press condition must be
>>> cleared when the button is released. This seems like it should be
>>> common for all GPIOs so I'm probably missing something.
>>
>>
>> It sounds like your expander does not generate interrupt when the button
>> is released and so gpio-keys driver is not aware of it.
>>
>> Thanks.
>>
>
> Thanks for the quick reply. Is giving an interrupt on both press and release
> a requirement for gpio-keys? or is there another setup that I'm not aware
> of?
>

Yes.

> How does the gpio-key framework work with register mapped GPIOs that are
> rising-edge or falling-edge triggered? I tend to think of these IRQ states
> as the general case for GPIO.

I am not sure why it matters if it is register mapped gpio or not. The
driver needs to know when the key is pressed and released, so it
expects to get an interrupt when gpio changes state.

Thanks.

-- 
Dmitry
--
To unsubscribe from this list: send the line "unsubscribe linux-spi" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Using gpio_keys with regmapped gpio?

[Dmitry Torokhov]

On Wed, Apr 1, 2015 at 8:59 AM, Dmitry Torokhov
<dmitry.torokhov-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
> On Tue, Mar 31, 2015 at 2:06 PM, Thor Thayer
> <tthayer-yzvPICuk2ABMcg4IHK0kFoH6Mc4MB0Vx@public.gmane.org> wrote:
>>
>>
>> On 03/31/2015 03:17 PM, Dmitry Torokhov wrote:
>>>
>>> On Tue, Mar 31, 2015 at 02:57:35PM -0500, Thor Thayer wrote:
>>>>
>>>> Hi,
>>>>
>>>> I have a SPI gpio expander chip that is using the regmap framework.
>>>> There are some pushbuttons that I'd like to use the gpio_keys
>>>> framework for but only the first button press creates an input
>>>> event.
>>>>
>>>> It seems like my problem occurs because I'm only getting events on a
>>>> button press.  The input_get_disposition() function toggles a local
>>>> copy of the button state [__change_bit(code, dev->key);] in the
>>>> input.c file. I think this function also needs to be called when the
>>>> button is released to toggle the dev->key state and re-arm for the
>>>> next button press.
>>>
>>>
>>> Yes.
>>>
>>>>
>>>> I tried enabling autorepeat but that causes an infinite loop in the
>>>> input_repeat_key() function.
>>>>
>>>> In both cases, it seems like the button press condition must be
>>>> cleared when the button is released. This seems like it should be
>>>> common for all GPIOs so I'm probably missing something.
>>>
>>>
>>> It sounds like your expander does not generate interrupt when the button
>>> is released and so gpio-keys driver is not aware of it.
>>>
>>> Thanks.
>>>
>>
>> Thanks for the quick reply. Is giving an interrupt on both press and release
>> a requirement for gpio-keys? or is there another setup that I'm not aware
>> of?
>>
>
> Yes.

And by that I meant "Yes, it is a requirement".

Thanks.

-- 
Dmitry
--
To unsubscribe from this list: send the line "unsubscribe linux-spi" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Using gpio_keys with regmapped gpio?

[Geert Uytterhoeven]

Hi Thor,

On Tue, Mar 31, 2015 at 9:57 PM, Thor Thayer
<tthayer-yzvPICuk2ABMcg4IHK0kFoH6Mc4MB0Vx@public.gmane.org> wrote:
> Which brings up my next question. Can the gpio_keys framework be used with a
> regmapped gpio? I haven't been able to find any examples of gpio_keys with
> an external gpio expander and maybe this isn't valid usage.

Kzm9g uses gpio-keys with a PCF8575 i2c GPIO expander, cfr.
arch/arm/boot/dts/sh73a0-kzm9g.dts. It doesn't use regmap, though.

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert-Td1EMuHUCqxL1ZNQvxDV9g@public.gmane.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds
--
To unsubscribe from this list: send the line "unsubscribe linux-spi" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html