mbox series

[v2,0/8] ACPI: thermal: Use trip point table to register thermal zones

Message ID 5710197.DvuYhMxLoT@kreacher
Headers show
Series ACPI: thermal: Use trip point table to register thermal zones | expand

Message

Rafael J. Wysocki July 21, 2023, 12:44 p.m. UTC 
Hi Everyone,

This is the second iteration of the $subject patch series and its original
description below is still applicable

On Tuesday, July 18, 2023 8:01:20 PM CEST Rafael J. Wysocki wrote:
> 
> This patch series makes the ACPI thermal driver register thermal zones
> with the help of thermal_zone_device_register_with_trips(), so it
> doesn't need to use the thermal zone callbacks related to trip points
> any more (and they are dropped in the last patch).
> 
> The approach presented here is quite radically different from the
> previous attempts, as it doesn't really rearrange the driver's
> internal data structures, but adds the trip table support on top of
> them.  For this purpose, it uses an additional field in struct thermal_trip
> introduced in the first patch.

In the meantime I have updated the patches and tested them on a system with
a couple of ACPI thermal zones.  No differences in functionality after applying
the patches have been observed.

The update is mostly related to adding extra locking around trip point
temperature updates and some hardening of the .get_trend() callback routine
against invalid trip point indices.

Please see patch changelogs for details.

Thanks!

Comments

Daniel Lezcano July 23, 2023, 10:19 a.m. UTC | #1 
Hi Rafael,

could you wait before applying I would like to review the series but I'm 
OoO ATM, coming back next week?



On 21/07/2023 14:44, Rafael J. Wysocki wrote:
> Hi Everyone,
> 
> This is the second iteration of the $subject patch series and its original
> description below is still applicable
> 
> On Tuesday, July 18, 2023 8:01:20 PM CEST Rafael J. Wysocki wrote:
>>
>> This patch series makes the ACPI thermal driver register thermal zones
>> with the help of thermal_zone_device_register_with_trips(), so it
>> doesn't need to use the thermal zone callbacks related to trip points
>> any more (and they are dropped in the last patch).
>>
>> The approach presented here is quite radically different from the
>> previous attempts, as it doesn't really rearrange the driver's
>> internal data structures, but adds the trip table support on top of
>> them.  For this purpose, it uses an additional field in struct thermal_trip
>> introduced in the first patch.
> 
> In the meantime I have updated the patches and tested them on a system with
> a couple of ACPI thermal zones.  No differences in functionality after applying
> the patches have been observed.
> 
> The update is mostly related to adding extra locking around trip point
> temperature updates and some hardening of the .get_trend() callback routine
> against invalid trip point indices.
> 
> Please see patch changelogs for details.
> 
> Thanks!
> 
> 
>
Rafael J. Wysocki July 24, 2023, 7:59 a.m. UTC | #2 
Hi Daniel,

On Sun, Jul 23, 2023 at 12:19 PM Daniel Lezcano
<daniel.lezcano@linaro.org> wrote:
>
>
> Hi Rafael,
>
> could you wait before applying I would like to review the series but I'm
> OoO ATM, coming back next week?

Yes, I can wait.

Thanks!