Re: [PATCH 1/3] hwmon: Add Texas Instruments TMP108 temperature sensor driver.

[John Muir]

> On Nov 28, 2016, at 11:58 AM, Guenter Roeck <linux@xxxxxxxxxxxx> wrote:
> 
> On Mon, Nov 28, 2016 at 11:40:42AM -0800, John Muir wrote:
>>>> +static void tmp108_update_ready_time(struct tmp108 *tmp108)
>>>> +{
>>>> +	tmp108->ready_time = jiffies;
>>>> +	if ((tmp108->config & TMP108_CONF_MODE_MASK)
>>>> +	    == TMP108_MODE_CONTINUOUS) {
>>> 
>>> Don't you want a "!" here ? Presumably the delay is only needed
>>> if the original configuration was not for continuous mode.
>>> 
>>>> +		tmp108->ready_time +=
>>>> +			msecs_to_jiffies(TMP108_CONVERSION_TIME_MS);
>>>> +	}
>>>> +}
>> The delay is required for both really. When the device is set into continuous mode it starts converting and the first temperature is ready (just under) 30ms later.
>> 
> 
> The datasheet states, though, that the chip will start converting as soon
> as the device powers up. The kernel would have to start really fast after that
> to have to wait another 30 ms.
> 
> The current code ends up waiting if it doesn't have to (because it waits
> if the chip was originally configured for continuous mode), and not waiting
> if it has to (because the chip was not configured for continuous mode),
> which doesn't seem to be such a good idea.

You are referring to the statement under âConversion Rateâ in the data sheet?

"After power-up or a general-call reset, the TMP108 immediately starts a conversion, as shown in Figure 9. The first result is available after 27 ms (typical).â

The datasheet also states that in âshutdownâ mode, the device is only powering the serial interface (see "Mode Bitsâ). So, I assumed that the conversions werenât happening during in that state, and that there would be the delay after moving from the âshutdown' state (as is described by the one-shot mode as well where the device state returns to âshutdownâ when the conversion has taken place).

My intention was that the delay is enforced after PM resume, and for convenience I re-used the same function during probe to initialize the ready_time. I was assuming that the device could be in the shutdown state - for example if the module had previously been unloaded. I agree that this is not required at system startup time. Perhaps I should check to see if the previous configuration had the device in the âshutdownâ state, and in that instance apply the delay? In my opinion it doesnât really matter either way.

FYI, this same logic exists in the TMP102 device driver. The TMP102 device is very similar.

> 
>> For the (unsupported at this time) one-shot mode, there would likewise be a delay, but I was envisioning having the requesting task sleep while waiting for the data to be ready, rather than get an -EAGAIN until the data is ready. 
>> 
> 
> We had that discussion before; the thermal subsystem doesn't like to be kept
> waiting. You would end up adding a lot of complexity for very little gain.
> It might make more sense to consider implementing runtime idle support
> instead of one-shot mode if power consumption is a concern.

OK thanks. Iâll leave this as unsupported for now.

> 
>>>> +	hwmon_dev = hwmon_device_register_with_groups(dev, client->name,
>>>> +						      tmp108, tmp108_groups);
>>> 
>>> Please consider using the devm_ function here.
>>> 
>>>> +	if (IS_ERR(hwmon_dev)) {
>>>> +		dev_dbg(dev, "unable to register hwmon device\n");
>>>> +		return PTR_ERR(hwmon_dev);
>>>> +	}
>>>> +
>>>> +	tmp108->hwmon_dev = hwmon_dev;
>>>> +	tmp108->tz = thermal_zone_of_sensor_register(hwmon_dev, 0, hwmon_dev,
>>>> +						     &tmp108_of_thermal_ops);
>>>> +	if (IS_ERR(tmp108->tz))
>>>> +		return PTR_ERR(tmp108->tz);
>>> 
>>> hwmon device not unregistered here. That would be fixed by using the devm
>>> function above.
>>> 
>> For the above two registrations and .remove function, I was worried that there would be an order problem between the i2c->remove and the device-managed cleanup. Iâll get deeper into that code to determine if that is a problem.
>> 
> 
> The thermal sensor registration will be removed first (in the remove function).
> devm_ functions are all unregistered / removed after the remove function was
> called (in the order of the devm_ call). Otherwise you would have trouble
> with devm_kzalloc() as well.

Great.

> Note that my followup-patch had a problem with the minimum hysteresis
> temperature; it needs to be higher than the lower limit, not lower.

Iâll keep that in mind.

Cheers,

John.