On 25/02/2022 17:46, Benjamin Li wrote:
On Fri, Feb 25, 2022 at 6:02 AM Daniel Lezcano
<daniel.lezcano@xxxxxxxxxx> wrote:
Some drivers that support thermal zone disabling implement a set_mode
operation and simply disable the sensor or the relevant IRQ(s), so they
actually don't log anything when zones are disabled. These drivers are
imx_thermal.c, intel_quark_dts_thermal.c, and int3400_thermal.c.
For tsens.c, implementing a change_mode would require migrating the driver
from devm_thermal_zone_of_sensor_register to thermal_zone_device_register
(or updating thermal_of.c to add a change_mode operation in thermal_zone_
of_device_ops).
stm_thermal.c seems to use this patch's model of not disabling IRQs when
the zone is disabled (they still perform the thermal_zone_device_update
upon IRQ, but return -EAGAIN from their get_temp).
What is the concern by changing the core code to have a correct handling
of the disabled / enabled state in this driver ? (and by this way give
the opportunity to other drivers to fix their code)'
It seems fine, is that the preference? Updating thermal_of.c to add a
change_mode
operation in thermal_zone_of_device_ops?
I'm not a big fan of this duplicated ops structure but preferably it
would be better to put it there (except if you see a better way to do it)
--
<http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs
Follow Linaro: <http://www.facebook.com/pages/Linaro> Facebook |
<http://twitter.com/#!/linaroorg> Twitter |
<http://www.linaro.org/linaro-blog/> Blog
