[PATCHv4 2/2] ARM: imx: Add wake functionality to GPIO

[Nguyen Dinh-R00091]

Hi Uwe,

>-----Original Message-----
>From: Uwe Kleine-König [mailto:u.kleine-koenig at pengutronix.de]
>Sent: Thursday, October 21, 2010 2:15 PM
>To: Nguyen Dinh-R00091
>Cc: linux-kernel at vger.kernel.org; linux-arm-kernel at lists.infradead.org; linux at arm.linux.org.uk;
>s.hauer at pengutronix.de; valentin.longchamp at epfl.ch; daniel at caiaq.de; grant.likely at secretlab.ca;
>bryan.wu at canonical.com; amit.kucheria at canonical.com; Zhang Lily-R58066
>Subject: Re: [PATCHv4 2/2] ARM: imx: Add wake functionality to GPIO
>
>On Thu, Oct 21, 2010 at 01:55:24PM -0500, Dinh.Nguyen at freescale.com wrote:
>> From: Dinh Nguyen <Dinh.Nguyen at freescale.com>
>>
>> Add function definition for irq_chip.set_wake to enable GPIO to
>> wake-up the system.
>>
>> This patch is based on Sascha Hauer's imx/linux-2.6 for-rmk tree.
>>
>> This patch has been tested on a MX51 Babbage system that had suspend
>> code implemented. The set_wake implementation is necessary for a
>> GPIO to wake up a system from suspend.
>>
>> Signed-off-by: Dinh Nguyen <Dinh.Nguyen at freescale.com>
>> ---
>>  arch/arm/plat-mxc/gpio.c |   32 ++++++++++++++++++++++++++++++++
>>  1 files changed, 32 insertions(+), 0 deletions(-)
>>
>> diff --git a/arch/arm/plat-mxc/gpio.c b/arch/arm/plat-mxc/gpio.c
>> index 9d38da0..9c3e362 100644
>> --- a/arch/arm/plat-mxc/gpio.c
>> +++ b/arch/arm/plat-mxc/gpio.c
>> @@ -20,6 +20,7 @@
>>   */
>>
>>  #include <linux/init.h>
>> +#include <linux/interrupt.h>
>>  #include <linux/io.h>
>>  #include <linux/irq.h>
>>  #include <linux/gpio.h>
>> @@ -201,11 +202,42 @@ static void mx2_gpio_irq_handler(u32 irq, struct irq_desc *desc)
>>  	}
>>  }
>>
>> +/*
>> + * Set interrupt number "irq" in the GPIO as a wake-up source.
>> + * While system is running, all registered GPIO interrupts need to have
>> + * wake-up enabled. When system is suspended, only selected GPIO interrupts
>> + * need to have wake-up enabled.
>> + * @param  irq          interrupt source number
>> + * @param  enable       enable as wake-up if equal to non-zero
>> + * @return       This function returns 0 on success.
>> + */
>> +static int gpio_set_wake_irq(u32 irq, u32 enable)
>> +{
>> +	u32 gpio = irq_to_gpio(irq);
>> +	u32 gpio_idx = gpio & 0x1F;
>> +	struct mxc_gpio_port *port = &mxc_gpio_ports[gpio / 32];
>> +
>> +	if (enable) {
>> +		if (port->irq_high && (gpio_idx >= 16))
>> +			enable_irq_wake(port->irq_high);
>> +		else
>> +			enable_irq_wake(port->irq);
>> +	} else {
>> +		if (port->irq_high && (gpio_idx >= 16))
>> +			disable_irq_wake(port->irq_high);
>> +		else
>> +			disable_irq_wake(port->irq);
>> +	}
>> +
>> +	return 0;
>> +}
>> +
>Is this really enough to let the CPU wake up?  Is this implementation
>correct for all i.MX SoCs?

I tested the code to wakeup from suspend on my MX51 Babbage board, without this code the power button on the Babbage board will not wakeup the system. It appears that in the gpio-keys.c, gpio_keys_suspend() will call the enable_irq_wake, which then will call the GPIO's set_wake function on the GPIO pin. This implementation should be correct for all i.MX SoCs.

>
>Best regards
>Uwe
>
>>  static struct irq_chip gpio_irq_chip = {
>>  	.ack = gpio_ack_irq,
>>  	.mask = gpio_mask_irq,
>>  	.unmask = gpio_unmask_irq,
>>  	.set_type = gpio_set_irq_type,
>> +	.set_wake = gpio_set_wake_irq,
>>  };
>>
>>  static void _set_gpio_direction(struct gpio_chip *chip, unsigned offset,
>> --
>> 1.6.0.4
>>
>>
>>
>
>--
>Pengutronix e.K.                           | Uwe Kleine-König            |
>Industrial Linux Solutions                 | http://www.pengutronix.de/  |

Thanks,
Dinh