[Bug 217247] [acpi] r265474 makes 11.0R unusable with Atom 330

[bugzilla-noreply at freebsd.org]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=217247

--- Comment #4 from Stéphane Lesimple <stephane_freebsd at lesimple.fr> ---
(In reply to Adrian Chadd from comment #2)

Compiled HEAD this morning, namely base r314326 and booted it.
It does behave differently: at boot I couldn't observe the slowdown.

However after some digging, I found out that HEAD's /etc/default/rc.conf is
different and now contains:

> performance_cx_lowest="NONE"

Where 11.0R contained

> performance_cx_lowest="Cmax"

(introduced by base r265474)

On my system, NONE translates to C1, as shown by sysctl -a, so that's why it
behaves properly: it never tries to use C2 or C3.

I tried to manually set dev.cpu.X.cx_lowest to C2, and see if the slowdown was
still there, but it wasn't. However sysctl indicates that my setting is
completely ignored and the system always uses C1, as seen by cx_usage:

dev.cpu.0.cx_method: C1/hlt C2/io C3/io
dev.cpu.0.cx_usage_counters: 4617 0 0
dev.cpu.0.cx_usage: 100.00% 0.00% 0.00% last 49467us
dev.cpu.0.cx_lowest: C2
dev.cpu.0.cx_supported: C1/1/0 C2/2/1 C3/3/85
dev.cpu.0.%parent: acpi0
dev.cpu.0.%pnpinfo: _HID=none _UID=0
dev.cpu.0.%location: handle=\_PR_.P001
dev.cpu.0.%driver: cpu
dev.cpu.0.%desc: ACPI CPU

So this *is* indeed different from 11.0R. For some reason the kernel now
chooses to keep C1 at all times, even if I tell it that C2 is okay. If you want
me to try anything, just ask, I'm keeping HEAD's memstick.img around just in
case.

-- 
You are receiving this mail because:
You are the assignee for the bug.