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 Not Applicable

Unit

Tests Not Applicable

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.