From mboxrd@z Thu Jan  1 00:00:00 1970
From: Sourav Poddar <sourav.poddar@ti.com>
Subject: Re: [PATCH 3/6] driver: serial: omap: add prepare/complete callback
 for "no_console_suspend" case
Date: Thu, 18 Apr 2013 23:41:43 +0530
Message-ID: <5170375F.2020603@ti.com>
References: <1366198467-6757-1-git-send-email-sourav.poddar@ti.com>	<1366198467-6757-4-git-send-email-sourav.poddar@ti.com> <874nf367th.fsf@linaro.org>
Mime-Version: 1.0
Content-Type: text/plain; charset="ISO-8859-1"; format=flowed
Content-Transfer-Encoding: 7bit
Return-path: <linux-kernel-owner@vger.kernel.org>
In-Reply-To: <874nf367th.fsf@linaro.org>
Sender: linux-kernel-owner@vger.kernel.org
To: Kevin Hilman <khilman@linaro.org>
Cc: gregkh@linuxfoundation.org, tony@atomide.com, rmk+kernel@arm.linux.org.uk, linux-serial@vger.kernel.org, linux-omap@vger.kernel.org, linux-kernel@vger.kernel.org
List-Id: linux-serial@vger.kernel.org

Hi Kevin,
On Thursday 18 April 2013 11:26 PM, Kevin Hilman wrote:
> Sourav Poddar<sourav.poddar@ti.com>  writes:
>
>> The patch adapt the serial core/driver to take care of the case when "no_console_suspend"
>> is used in the bootargs. The patch will remove dependency to set od->flags to
>> "OMAP_DEVICE_NO_IDLE_ON_SUSPEND" in serial.c(non dt case) and omap_device.c(dt case).
>>
>> Prepare and complete callbacks will ensure that clocks remain active for the console
>> uart when "no_console_suspend" is used in the bootargs.
>>
>> Signed-off-by: Sourav Poddar<sourav.poddar@ti.com>
> This changelog needs a rework.  The driver itself was not aware of
> od->flags and omap_device stuff in general, so it's not really
> relevant.  The driver is also not directly managing clocks, int's only
> doing runtime PM callbacks.
>
> What you want to say in the changelog is that the driver manages
> "no_console_suspend" by preventing runtime PM during the suspend path,
> which forces the console UART to stay awake.
>
Yes, looks to the point. Will update the changelog in the next version.
>> ---
>>   drivers/tty/serial/omap-serial.c |   20 ++++++++++++++++++++
>>   1 files changed, 20 insertions(+), 0 deletions(-)
>>
>> diff --git a/drivers/tty/serial/omap-serial.c b/drivers/tty/serial/omap-serial.c
>> index 08332f3..9ef80cf 100644
>> --- a/drivers/tty/serial/omap-serial.c
>> +++ b/drivers/tty/serial/omap-serial.c
>> @@ -1278,6 +1278,24 @@ static struct uart_driver serial_omap_reg = {
>>   };
>>
>>   #ifdef CONFIG_PM_SLEEP
>> +static int serial_omap_prepare(struct device *dev)
>> +{
>> +	struct uart_omap_port *up = dev_get_drvdata(dev);
>> +
>> +	if (!console_suspend_enabled&&  uart_console(&up->port))
>> +		pm_runtime_disable(dev);
>> +
>> +	return 0;
>> +}
>> +
>> +static void serial_omap_complete(struct device *dev)
>> +{
>> +	struct uart_omap_port *up = dev_get_drvdata(dev);
>> +
>> +	if (!console_suspend_enabled&&  uart_console(&up->port))
>> +		pm_runtime_enable(dev);
>> +}
>> +
> For compilation with !CONFIG_PM_SLEEP, you'll also need:
>
> #else
> #define serial_omap_prepare NULL
> #define serial_omap_prepare NULL
> #endif /* CONFIG_PM_SLEEP */
>
Ok. Will change this.
Though, just a query/proposal on this, will it be correct if we try to 
create a macro[1]
in include/linux/pm.h for prepare/complete as it is done for 
suspend/resume. ?

[1]:
#ifdef CONFIG_PM_SLEEP
#define SET_SYSTEM_SLEEP_PM_PREP_COMP_OPS(prepare_fn, complete_fn) \
         .prepare = prepare_fn, \
         .complete = complete_fn, \
#else
#define SET_SYSTEM_SLEEP_PM_PREP_COMP_OPS(prepare_fn, complete_fn)
#endif

~Sourav
>>   static int serial_omap_suspend(struct device *dev)
>>   {
>>   	struct uart_omap_port *up = dev_get_drvdata(dev);
>> @@ -1632,6 +1650,8 @@ static const struct dev_pm_ops serial_omap_dev_pm_ops = {
>>   	SET_SYSTEM_SLEEP_PM_OPS(serial_omap_suspend, serial_omap_resume)
>>   	SET_RUNTIME_PM_OPS(serial_omap_runtime_suspend,
>>   				serial_omap_runtime_resume, NULL)
>> +	.prepare        = serial_omap_prepare,
>> +	.complete       = serial_omap_complete,
>>   };
>>
>>   #if defined(CONFIG_OF)
> Kevin