Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: [PATCH] counter: stm32-timer-cnt: add power management support
To:     Jonathan Cameron <jic23@kernel.org>
CC:     <vilhelm.gray@gmail.com>, <alexandre.torgue@st.com>,
        <mcoquelin.stm32@gmail.com>, <benjamin.gaignard@st.com>,
        <linux-iio@vger.kernel.org>,
        <linux-stm32@st-md-mailman.stormreply.com>,
        <linux-arm-kernel@lists.infradead.org>,
        <linux-kernel@vger.kernel.org>
References: <1580899678-26393-1-git-send-email-fabrice.gasnier@st.com>
 <20200208153541.448a0df3@archlinux>
From:   Fabrice Gasnier <fabrice.gasnier@st.com>
Message-ID: <dac6eb51-f13c-5369-d43e-77f21ed81172@st.com>
Date:   Mon, 10 Feb 2020 18:13:50 +0100
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101
 Thunderbird/60.9.0
MIME-Version: 1.0
In-Reply-To: <20200208153541.448a0df3@archlinux>
Content-Type: text/plain; charset="utf-8"
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 2/8/20 4:35 PM, Jonathan Cameron wrote:
> On Wed, 5 Feb 2020 11:47:58 +0100
> Fabrice Gasnier <fabrice.gasnier@st.com> wrote:
> 
>> Add suspend/resume PM sleep ops. When going to low power, enforce the
>> counter isn't active.
>>
>> Signed-off-by: Fabrice Gasnier <fabrice.gasnier@st.com>
>> ---
>>  drivers/counter/stm32-timer-cnt.c | 25 +++++++++++++++++++++++++
>>  1 file changed, 25 insertions(+)
>>
>> diff --git a/drivers/counter/stm32-timer-cnt.c b/drivers/counter/stm32-timer-cnt.c
>> index 3eafcce..3b84503 100644
>> --- a/drivers/counter/stm32-timer-cnt.c
>> +++ b/drivers/counter/stm32-timer-cnt.c
>> @@ -12,6 +12,7 @@
>>  #include <linux/iio/types.h>
>>  #include <linux/mfd/stm32-timers.h>
>>  #include <linux/module.h>
>> +#include <linux/pinctrl/consumer.h>
>>  #include <linux/platform_device.h>
>>  
>>  #define TIM_CCMR_CCXS	(BIT(8) | BIT(0))
>> @@ -358,10 +359,33 @@ static int stm32_timer_cnt_probe(struct platform_device *pdev)
>>  	priv->counter.num_signals = ARRAY_SIZE(stm32_signals);
>>  	priv->counter.priv = priv;
>>  
>> +	platform_set_drvdata(pdev, priv);
>> +
>>  	/* Register Counter device */
>>  	return devm_counter_register(dev, &priv->counter);
>>  }
>>  
>> +static int __maybe_unused stm32_timer_cnt_suspend(struct device *dev)
>> +{
>> +	struct stm32_timer_cnt *priv = dev_get_drvdata(dev);
>> +	u32 cr1;
>> +
>> +	/* Check for active counter */
>> +	regmap_read(priv->regmap, TIM_CR1, &cr1);
>> +	if (cr1 & TIM_CR1_CEN)
>> +		return -EBUSY;
> 
> Runtime suspend is allowed to refuse to happen (and that is
> explicitly documented)
> Is that allowed for normal suspend?  I didn't think it was.
> This will stop a suspend, but it's doing it by signalling a failure.

Hi Jonathan,

I'll rework this patch, to avoid stopping the suspend. Another approach
is to properly save/backup the registers here, and to disable the
counter in low power mode.
I'll update this in V2.

Thanks for reviewing,
Fabrice

> I'm not sure for example that the system will resume everything that
> was suspended before this call.
> 
> Normally you are in the tough luck category if a system decides
> to suspend underneath you.   There are ways to prevent it
> from userspace (wakelocks etc).
> 
>> +
>> +	return pinctrl_pm_select_sleep_state(dev);
>> +}
>> +
>> +static int __maybe_unused stm32_timer_cnt_resume(struct device *dev)
>> +{
>> +	return pinctrl_pm_select_default_state(dev);
>> +}
>> +
>> +static SIMPLE_DEV_PM_OPS(stm32_timer_cnt_pm_ops, stm32_timer_cnt_suspend,
>> +			 stm32_timer_cnt_resume);
>> +
>>  static const struct of_device_id stm32_timer_cnt_of_match[] = {
>>  	{ .compatible = "st,stm32-timer-counter", },
>>  	{},
>> @@ -373,6 +397,7 @@ static struct platform_driver stm32_timer_cnt_driver = {
>>  	.driver = {
>>  		.name = "stm32-timer-counter",
>>  		.of_match_table = stm32_timer_cnt_of_match,
>> +		.pm = &stm32_timer_cnt_pm_ops,
>>  	},
>>  };
>>  module_platform_driver(stm32_timer_cnt_driver);
>