Re: [PATCH 2/2] arm: shmobile: lager: enable HS-USB

[Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>]

Hello.

(2014/10/07 12:25), Yoshihiro Shimoda wrote:
> Hello.
> 
> (2014/10/06 9:59), Yoshihiro Shimoda wrote:
>> Hello.
>>
>> (2014/10/04 4:50), Sergei Shtylyov wrote:
>>> On 10/02/2014 12:04 PM, Yoshihiro Shimoda wrote:
>>>
< snip >
>>>> +&hsusb {
>>>> +	status = "okay";
>>>> +	renesas,enable-gpio = <&gpio5 18 GPIO_ACTIVE_LOW>;
>>>
>>>     It's certainly active-high.
>>
>> Since the current code has the following, we have to set the active_low...
>> However, the code is unreadable, I think. So, I will modify the code.
>>
>> 	/* check GPIO determining if USB function should be enabled */
>> 	if (priv->dparam.enable_gpio) {
>> 		gpio_request_one(priv->dparam.enable_gpio, GPIOF_IN, NULL);
>> 		ret = !gpio_get_value(priv->dparam.enable_gpio);
>> 		gpio_free(priv->dparam.enable_gpio);
>> 		if (ret) {
>> 			dev_warn(&pdev->dev,
>> 				 "USB function not selected (GPIO %d)\n",
>> 				 priv->dparam.enable_gpio);
>> 			ret = -ENOTSUPP;
>> 			goto probe_end_mod_exit;
>> 		}
>> 	}
> 
> I am confusing about the gpio_get_value()...
> In case of ARM, gpio_get_value() will call gpiod_get_raw_value() finally.
> 
> gpio_get_value() in arch/arm/include/asm/gpio.h
>  --> __gpio_get_value() in include/asm-generic/gpio.h
>   --> gpiod_get_raw_value() in drivers/gpio/gpiolib.c
> 
> The gpiod_get_raw_value() doesn't care of the GPIO_ACTIVE_{HIGH,LOW}.
> So, should I add gpiod_is_active_low() or someting in the renesas_usbhs driver?
> Or, Do I misunderstand something?

I looked at the gpio-keys driver, and then I understand the renesas_usbhs driver should
have the gpio flags from of_get_named_gpio_flags().

So, I will try to modify the renesas_usbhs driver.

Best regards,
Yoshihiro Shimoda

> Best regards,
> Yoshihiro Shimoda
> 
>> Best regards,
>> Yoshihiro Shimoda
>>
>>> WBR, Sergei
>>>