Received-SPF: pass (google.com: domain of linux-kernel+bounces-129731-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:40f1:3f00::1 as permitted sender) client-ip=2604:1380:40f1:3f00::1;
Message-ID: <1d956aab-2892-4a2b-a4b3-0a93504668eb@gmail.com>
Date: Wed, 3 Apr 2024 15:47:12 +0300
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: [RFC PATCH 5/6] watchdog: ROHM BD96801 PMIC WDG driver
Content-Language: en-US, en-GB
To: Guenter Roeck <linux@roeck-us.net>
Cc: Matti Vaittinen <matti.vaittinen@fi.rohmeurope.com>,
 Lee Jones <lee@kernel.org>, Wim Van Sebroeck <wim@linux-watchdog.org>,
 devicetree@vger.kernel.org, linux-kernel@vger.kernel.org,
 linux-watchdog@vger.kernel.org
References: <cover.1712058690.git.mazziesaccount@gmail.com>
 <f8e743a6c49607de0dd7a27778383477e051b130.1712058690.git.mazziesaccount@gmail.com>
 <4fa3a64b-60fb-4e5e-8785-0f14da37eea2@roeck-us.net>
 <279336b3-f28d-48ee-a10f-47abba7b0b89@gmail.com>
 <d2ab33e6-4d3e-472a-b4d7-b703955989ba@roeck-us.net>
From: Matti Vaittinen <mazziesaccount@gmail.com>
In-Reply-To: <d2ab33e6-4d3e-472a-b4d7-b703955989ba@roeck-us.net>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit

On 4/3/24 15:41, Guenter Roeck wrote:
> On Wed, Apr 03, 2024 at 09:34:35AM +0300, Matti Vaittinen wrote:
>> Hi Guenter,
>>
>> First of all, thanks for the review. It was quick! Especially when we speak
>> of a RFC series. Very much appreciated.
>>
>> On 4/2/24 20:11, Guenter Roeck wrote:
>>> On Tue, Apr 02, 2024 at 04:11:41PM +0300, Matti Vaittinen wrote >> +static int init_wdg_hw(struct wdtbd96801 *w)
>>>> +{
>>>> +	u32 hw_margin[2];
>>>> +	int count, ret;
>>>> +	u32 hw_margin_max = BD96801_WDT_DEFAULT_MARGIN, hw_margin_min = 0;
>>>> +
>>>> +	count = device_property_count_u32(w->dev->parent, "rohm,hw-timeout-ms");
>>>> +	if (count < 0 && count != -EINVAL)
>>>> +		return count;
>>>> +
>>>> +	if (count > 0) {
>>>> +		if (count > ARRAY_SIZE(hw_margin))
>>>> +			return -EINVAL;
>>>> +
>>>> +		ret = device_property_read_u32_array(w->dev->parent,
>>>> +						     "rohm,hw-timeout-ms",
>>>> +						     &hw_margin[0], count);
>>>> +		if (ret < 0)
>>>> +			return ret;
>>>> +
>>>> +		if (count == 1)
>>>> +			hw_margin_max = hw_margin[0];
>>>> +
>>>> +		if (count == 2) {
>>>> +			hw_margin_max = hw_margin[1];
>>>> +			hw_margin_min = hw_margin[0];
>>>> +		}
>>>> +	}
>>>> +
>>>> +	ret = bd96801_set_wdt_mode(w, hw_margin_max, hw_margin_min);
>>>> +	if (ret)
>>>> +		return ret;
>>>> +
>>>> +	ret = device_property_match_string(w->dev->parent, "rohm,wdg-action",
>>>> +					   "prstb");
>>>> +	if (ret >= 0) {
>>>> +		ret = regmap_update_bits(w->regmap, BD96801_REG_WD_CONF,
>>>> +				 BD96801_WD_ASSERT_MASK,
>>>> +				 BD96801_WD_ASSERT_RST);
>>>> +		return ret;
>>>> +	}
>>>> +
>>>> +	ret = device_property_match_string(w->dev->parent, "rohm,wdg-action",
>>>> +					   "intb-only");
>>>> +	if (ret >= 0) {
>>>> +		ret = regmap_update_bits(w->regmap, BD96801_REG_WD_CONF,
>>>> +				 BD96801_WD_ASSERT_MASK,
>>>> +				 BD96801_WD_ASSERT_IRQ);
>>>> +		return ret;
>>>> +	}
>>>
>>> I don't see the devicetree bindings documented in the series.
>>
>> Seems like I have missed this WDG binding. But after reading your comment
>> below, I am wondering if I should just drop the binding and default to
>> "prstb" (shutdown should the feeding be skipped) - and leave the "intb-only"
>> case for one who actually needs such.
>>
>>> I am also a bit surprised that the interrupt isn't handled in the driver.
>>> Please explain.
>>
>> Basically, I just had no idea what the IRQ should do in the generic case. If
>> we get an interrupt, it means the WDG feeding has failed. My thinking is
>> that, what should happen is forced reset. I don't see how that can be done
>> in reliably manner from an IRQ handler.
>>
>> When the "prstb WDG action" is set (please, see the above DT binding
>> handling), the PMIC shall shut down power outputs. This should get the
>> watchdog's job done.
>>
>> With the "intb-only"-option, PMIC will not turn off the power. I'd expect
>> there to be some external HW connection which handles the reset by HW.
>>
>> After all this being said, I wonder if I should just unconditionally
>> configure the PMIC to always turn off the power (prstb option) should the
>> feeding fail? Or do someone have some suggestion what the IRQ handler should
>> do (except maybe print an error msg)?
>>
> 
> Other watchdog drivers call emergency_restart() if the watchdog times out
> and triggers an interrupt. Are you saying this won't work for this system ?
> If so, please explain.
> 

Thanks Guenter. If it works with systems using other devices, then it 
should work (to the same extent) with systems using this PMIC. Thanks.

I'll add the IRQ handling to next version - but it may take a while as 
I'm currently having some problems with the IRQs in general, and because 
I'll wait for feedback from Mark to the regulator part.

Yours,
	-- Matti

-- 
Matti Vaittinen
Linux kernel developer at ROHM Semiconductors
Oulu Finland

~~ When things go utterly wrong vim users can always type :help! ~~