Powerd on Atom machine does not seem to be working

GoldServe

I enabled powerd on my Lanner 7530 atom machine and it draws 18W from the AC side no mater if i have it enabled or not.

dev.cpu.0.%desc: ACPI CPU
dev.cpu.0.%driver: cpu
dev.cpu.0.%location: handle=\_PR_.P001
dev.cpu.0.%pnpinfo: _HID=none _UID=0
dev.cpu.0.%parent: acpi0
dev.cpu.0.freq: 1596
dev.cpu.0.freq_levels: 1596/-1 1396/-1 1197/-1 997/-1 798/-1 598/-1 399/-1 199/-1
dev.cpu.0.cx_supported: C1/0
dev.cpu.0.cx_lowest: C1
dev.cpu.0.cx_usage: 100.00% last 500us
dev.cpu.1.%desc: ACPI CPU
dev.cpu.1.%driver: cpu
dev.cpu.1.%location: handle=\_PR_.P002
dev.cpu.1.%pnpinfo: _HID=none _UID=0
dev.cpu.1.%parent: acpi0
dev.cpu.1.cx_supported: C1/0
dev.cpu.1.cx_lowest: C1
dev.cpu.1.cx_usage: 100.00% last 500us

Any ideas on how I can get some power savings and reduce the heat on the fanless system?

jimp

powerd hasn't seen a lot of testing that I know of. Are you sure that it is running when you have it enabled?

Can you look in the output of sysctl and see if you have any thermal zones that ACPI can understand on that board? (They'd be under hw.acpi.thermal )

GoldServe

When I enable it, I see it running in TOP.

Here are the hw.acpi sysctls without powered running (if that makes any difference):

# sysctl hw.acpi
hw.acpi.supported_sleep_state: S1 S4 S5
hw.acpi.power_button_state: S5
hw.acpi.sleep_button_state: S1
hw.acpi.lid_switch_state: NONE
hw.acpi.standby_state: S1
hw.acpi.suspend_state: NONE
hw.acpi.sleep_delay: 1
hw.acpi.s4bios: 0
hw.acpi.verbose: 0
hw.acpi.disable_on_reboot: 0
hw.acpi.handle_reboot: 0
hw.acpi.reset_video: 0
hw.acpi.cpu.cx_lowest: C1

jimp

Not sure

What kind of CPU usage are you seeing on that box?

powerd's adaptive mode, which pfsense uses, only kicks in if the system is under 50% cpu load, or so the man page implies. I haven't tried to use it on mine so I'm not sure how visible its effects may be. When you run commands it might be driving up the CPU usage and making it shift back up.

GoldServe

It is an atom N270 using SMP kernel.

Cry Havok

Atom processors do not have variable clock speeds - they run at full speed all the time. As such you can't reduce the power draw.

jimp

@Cry:

Atom processors do not have variable clock speeds - they run at full speed all the time. As such you can't reduce the power draw.

If that were the case, it shouldn't be reporting that it does:

dev.cpu.0.freq_levels: 1596/-1 1396/-1 1197/-1 997/-1 798/-1 598/-1 399/-1 199/-1

But it's possible that is the cause.

GoldServe

so I ran powerd with -v and it does seem like it is setting the frequencies correctly. I also did a test by setting the cpu.0.frequency manually but the power does not decrease. I think the CPU is throttling and not actually changing the voltage to the CPU. Now, if I can figure that part out…