Re: [PATCH] at91: use gpiolib calls for USB vbus pin on at91sam9g45

[Nicolas Ferre <nicolas.ferre@atmel.com>]

Jean-Christophe PLAGNIOL-VILLARD :
> On 10:13 Thu 17 Sep     , Nicolas Ferre wrote:
>> Jean-Christophe PLAGNIOL-VILLARD :
>>> On 19:29 Wed 16 Sep     , Nicolas Ferre wrote:
>>>> Change pin configuration for USB vbus on at91sam9g45: use the generic gpiolib
>>>> call instead of the at91 specific one.
>>>> Use gpio_request() function with same identifier for OHCI and EHCI hosts as
>>>> they are sharing the same pin.
>>>>
>>>> Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
>>>> ---
>>>> This patch is on top of at91sam9g45 USB integration one:
>>>> "[PATCH 2/2] at91/USB: at91sam9g45 series USB host integration"
>>>> http://lkml.org/lkml/2009/6/9/221
>>>>
>>>>  arch/arm/mach-at91/at91sam9g45_devices.c |   12 ++++++++----
>>>>  1 files changed, 8 insertions(+), 4 deletions(-)
>>>>
>>>> diff --git a/arch/arm/mach-at91/at91sam9g45_devices.c b/arch/arm/mach-at91/at91sam9g45_devices.c
>>>> index 5be8cf2..7d939c0 100644
>>>> --- a/arch/arm/mach-at91/at91sam9g45_devices.c
>>>> +++ b/arch/arm/mach-at91/at91sam9g45_devices.c
>>>> @@ -118,8 +118,10 @@ void __init at91_add_device_usbh_ohci(struct at91_usbh_data *data)
>>>>  
>>>>  	/* Enable VBus control for UHP ports */
>>>>  	for (i = 0; i < data->ports; i++) {
>>>> -		if (data->vbus_pin[i])
>>>> -			at91_set_gpio_output(data->vbus_pin[i], 0);
>>>> +		if (data->vbus_pin[i]) {
>>>> +			gpio_request(data->vbus_pin[i], "usb host vbus");
>>>> +			gpio_direction_output(data->vbus_pin[i], 0);
>>>> +		}
>>>>  	}
>>>>  
>>>>  	usbh_ohci_data = *data;
>>>> @@ -173,8 +175,10 @@ void __init at91_add_device_usbh_ehci(struct at91_usbh_data *data)
>>>>  
>>>>  	/* Enable VBus control for UHP ports */
>>>>  	for (i = 0; i < data->ports; i++) {
>>>> -		if (data->vbus_pin[i])
>>>> -			at91_set_gpio_output(data->vbus_pin[i], 0);
>>>> +		if (data->vbus_pin[i]) {
>>>> +			gpio_request(data->vbus_pin[i], "usb host vbus");
>>>> +			gpio_direction_output(data->vbus_pin[i], 0);
>>>> +		}
>>>>  	}
>>> as you do the same think for ehci & ohci why not factorize it?
>> Just because you may choose only one or the other controller.
>> For instance, if you only want ohci, you can disable ehci selection and
>> always have its configuration done.
> I agree but both configuration function share some code which could be
> factorize, is it not?

Well it is true that it is duplication *but* we are talking about only a
very few lines with very linear configuration instructions.
I do not think it is even worth doing.

Bye,
-- 
Nicolas Ferre