Re: [PATCH 2/3] input/keyboard: Add support for Dollar Cove TI power button

[Takashi Iwai]

On Thu, 31 Aug 2017 20:33:55 +0200,
Dmitry Torokhov wrote:
> 
> Hi Takashi,
> 
> On Tue, Aug 22, 2017 at 07:57:09AM +0200, Takashi Iwai wrote:
> > This provides a new input driver for supporting the power button on
> > Dollar Cove TI PMIC, found on Cherrytrail-based devices.
> > The patch is based on the original work by Intel, found at:
> >   https://github.com/01org/ProductionKernelQuilts
> > 
> > Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=193891
> > Signed-off-by: Takashi Iwai <tiwai@xxxxxxx>
> > ---
> >  drivers/input/keyboard/Kconfig        |  7 +++
> >  drivers/input/keyboard/Makefile       |  1 +
> >  drivers/input/keyboard/dc_ti_pwrbtn.c | 85 +++++++++++++++++++++++++++++++++++
> 
> Not sure if it ends up in drivers/input/keyboard, or drivers/input/misc/
> (where most power buttons live) or in platform drivers, still a few
> comments below.

Oh sorry, I forgot to put you to Cc after v2 patch.
Per Andy's suggestion, the driver was moved to drivers/platform/x86
now.  For v3 patchset, see the following thread.
  https://marc.info/?i=20170825134443.14843-1-tiwai%40suse.de

> > +config KEYBOARD_DC_TI_PWRBTN
> > +	tristate "Dollar Cove TI power button driver"
> > +	depends on INTEL_SOC_PMIC_DC_TI
> > +	help
> > +	  Say Y here fi you want to have a power button driver for
> > +	  Dollar Cove TI PMIC.
> 
> If keeping in input we customarily call out the module name (see a few
> lines above).

I changed the driver and config names to follow the platform driver.


> > +#define DC_TI_SIRQ_REG		0x3
> > +#define SIRQ_PWRBTN_REL		(1 << 0)
> 
> BIT()?

OK.

> > +#define DRIVER_NAME "dc_ti_pwrbtn"
> > +
> > +static irqreturn_t dc_ti_pwrbtn_interrupt(int irq, void *dev_id)
> > +{
> > +	struct input_dev *pwrbtn_input = dev_id;
> > +	struct device *dev = pwrbtn_input->dev.parent;
> > +	struct regmap *regmap = dev_get_drvdata(dev);
> > +	int state;
> > +
> > +	if (!regmap_read(regmap, DC_TI_SIRQ_REG, &state)) {
> > +		dev_dbg(dev, "SIRQ_REG=0x%x\n", state);
> > +		state &= SIRQ_PWRBTN_REL;
> > +		input_event(pwrbtn_input, EV_KEY, KEY_POWER, !state);
> 
> Why not
> 
> 		input_report_key(pwrbtn_input, KEY_POWER,
> 				 !(state & SIRQ_PWRBTN_REL));

Sounds better, indeed.


> > +		input_sync(pwrbtn_input);
> > +	}
> > +
> > +	return IRQ_HANDLED;
> > +}
> > +
> > +static int dc_ti_pwrbtn_probe(struct platform_device *pdev)
> > +{
> > +	struct device *dev = &pdev->dev;
> > +	struct intel_soc_pmic *pmic = dev_get_drvdata(dev->parent);
> > +	struct input_dev *pwrbtn_input;
> > +	int irq;
> > +	int ret;
> > +
> > +	irq = platform_get_irq(pdev, 0);
> > +	if (irq < 0)
> > +		return -EINVAL;
> 
> Why do you clobber return value? Simply return "irq".

OK.


> > +	pwrbtn_input = devm_input_allocate_device(dev);
> > +	if (!pwrbtn_input)
> > +		return -ENOMEM;
> > +	pwrbtn_input->name = pdev->name;
> > +	pwrbtn_input->phys = "dc-ti-power/input0";
> > +	pwrbtn_input->id.bustype = BUS_HOST;
> > +	pwrbtn_input->dev.parent = dev;
> 
> Not needed since devm_input_allocate_device() does it for us.

OK.

> > +	input_set_capability(pwrbtn_input, EV_KEY, KEY_POWER);
> > +	ret = input_register_device(pwrbtn_input);
> > +	if (ret)
> > +		return ret;
> 
> If staying in input, can we please call this variable err or error?

OK, I don't mind to change it.

> > +
> > +	dev_set_drvdata(dev, pmic->regmap);
> > +
> > +	ret = devm_request_threaded_irq(dev, irq, NULL, dc_ti_pwrbtn_interrupt,
> > +					0, KBUILD_MODNAME, pwrbtn_input);
> > +	if (ret)
> > +		return ret;
> > +
> > +	ret = enable_irq_wake(irq);
> > +	if (ret)
> > +		dev_warn(dev, "Can't enable IRQ as wake source: %d\n", ret);
> 
> We do not normally enable wake IRQs in probe, but instead do:
> 
> 	device_init_wakeup(&pdev->dev, true);

> in probe()

Right, this was already changed in the later version.
But...

> and then check it in suspend/resume:
> 
> 	if (device_may_wakeup(dev)) {
> 		err = enable_irq_wake(XXX->irq);
> 		if (!err)
> 			XXX->irq_wake_enabled = true;
> 	}
> 
> ...
> 
> 	if (XXX->irq_wake_enabled)
> 		disable_irq_wake(XXX->irq);
> 
> This allows userspace to inhibit wakeup, if needed.

... this wasn't.  Will put it.

Thank your for the review!


Takashi