Details

Reviewers

jkim
imp
emaste
jhb
markj
mckusick

Commits

rGe779891327b1: sys/power: Sleep type reporting by PM backends

Summary

Allow PM backends to report supported sleep types when registering through power_pm_register. Expose this information through kern.power.supported_stype sysctl, and set defaults for power_standby/suspend/hibernate_stype based on this.

Implement this in ACPI PM backend.

Diff Detail

Repository

rG FreeBSD src repository

Lint

Lint Skipped

Unit

Tests Skipped

Build Status

Buildable 66943
Build 63826: arc lint + arc unit

Event Timeline

obiwac created this revision.Aug 20 2025, 10:38 AM

Herald added a subscriber: olce. · View Herald TranscriptAug 20 2025, 10:38 AM

obiwac requested review of this revision.Aug 20 2025, 10:38 AM

Harbormaster completed remote builds in B66374: Diff 160655.Aug 20 2025, 10:38 AM

obiwac added a parent revision: D52043: acpi: Use sleep types defined in sys/power.h.Aug 20 2025, 10:38 AM

obiwac mentioned this in D48732: acpi: Abstract over ACPI Sx sleep states with `enum sleep_type`.

Check if stype supported in power_sysctl_stype

Harbormaster completed remote builds in B66375: Diff 160658.Aug 20 2025, 10:45 AM

obiwac added inline comments.Aug 20 2025, 10:47 AM

sys/kern/subr_power.c
106–107	okay to have defensive checks like this (i mean `new_stype < 0 \|\| new_stype >= POWER_STYPE_COUNT`)? or should this just be the `new_stype == POWER_STYPE_UNKNOWN` check?

obiwac mentioned this in D52036: sys/power: Generic sleep types.Aug 20 2025, 10:54 AM

markj added inline comments.Sep 4 2025, 2:30 PM

sys/kern/subr_power.c
106–107	Given that the return type of `power_name_to_stype()` is an enum, it's weird to me to check for values outside of the enum. I think it's ok to assume that `new_stype` is valid.
143
sys/sys/power.h
94–95	To enable some extra checking.