rpmh_rsc_cpu_pm_callback(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… rpmh_rsc_cpu_pm_callback(9) πŸ’…

✨ NAME ✨

okay so this is the Kernel API Manual page for rpmh_rsc_cpu_pm_callback β€” Check if any of the AMCs are busy. That's the whole assignment, bestie. πŸ’…

🌸 SYNOPSIS 🌸

int rpmh_rsc_cpu_pm_callback (struct notifier_block *nfb , unsigned long
       action , void *v );

πŸ’… ARGUMENTS πŸ’…

  • nfb β€” Pointer to the notifier block in struct rsc_drv. 🫢
  • action β€” CPU_PM_ENTER, CPU_PM_ENTER_FAILED, or CPU_PM_EXIT. ✨
  • v β€” Unused. πŸ™ˆ

πŸ¦‹ DESCRIPTION πŸ¦‹

This function is literally given to cpu_pm_register_notifier so we can be informed about when CPUs go down. When all CPUs go down, we know no more active transfers will be started, so we write sleep/wake sets. no cap, that's the vibe.

It gets called from cpuidle code paths and also at system suspend time. 🌸

If it's the last CPU going down and the AMCs are not busy, then it writes cached sleep and wake messages to TCSes. The firmware then takes care of triggering them when entering deepest low power modes. that's the main character behavior πŸ”₯

πŸŽ€ RETURN πŸŽ€

See cpu_pm_register_notifier. That's the whole return tea β€” iykyk. πŸ’œ

πŸ”₯ SEE ALSO πŸ”₯

Kernel file ./drivers/soc/qcom/rpmh-rsc.c is where all this lore lives, and the whole friend group is:


okay bestie, that's the whole page πŸ’– this one comes from ./drivers/soc/qcom, dated October 2026 β€” go off and be iconic with your notifier πŸŽ€

πŸ“¦ where this page comes from