I read somewhere that the kernel devs really don't see this as an issue. They just expect that laptop owners should know to try these boot parameters. So, definitely not by 3.1.
from https://bugzilla.redhat.com/show_bug.cgi?id=727579:
Dave Jones 2011-08-03 15:32:21 EDT
enabling it by default in 3.0 caused regressions for some people, so it was
disabled. Hopefully Intel figures it out, and we can switch it back on by
default in 3.1 / 2.6.41 (until then, you'll have to set it by hand).
Interesting stuff, also recommend this thread on phoronix:
http://phoronix.com/forums/showthread.p … Regression
Can confirm this helps lower power consumption on the Thinkpad T420. I am wondering though does this boot parameter sacrifice any performance? Also when are we going to see a fix for this power regression issue? 3.1?
I read somewhere that the kernel devs really don't see this as an issue. They just expect that laptop owners should know to try these boot parameters. So, definitely not by 3.1.
]]>I don't use laptop mode, nor any specific laptop optimization as i'm never in battery mode.
]]>Core temperature with wifi and light use stays around 47 °C, whilst the temperature in the room is 28°C.
]]>My power usage is still higher than it was with 2.6.39 even with all these boot parameters. I'm really upset at all the regression I've experienced and read about in the kernel since around 2.6.35.
]]>pcie_aspm=force i915.i915_enable_rc6=1
to the kernel boot parameters, the power consumption when idle dropped to 10 W!
Thank you very much
]]>pcie_aspm=force i915.i915_enable_rc6=1
to the kernel line in your bootloader config file.
]]>I have recently bought the new lenovo thinkpad X1 equipped with a Core I5 processors and 4GB of RAM.
I am quite satisfied with this machine except for the fact that the fan is extremely loud (but perhaps a bios upgrade will fix problem).
Now my problem:
When I first installed Arch (first days of August) I ran powertop and the power consumption was around 10 Watts when idle with wifi card turned off.
Now, after some upgrades, the power consumption raises to 18W in idle and consequently the battery lasts less than 2h.
Here my current configuration:
Linux think-x1 3.0-ARCH #1 SMP PREEMPT Tue Aug 30 08:53:25 CEST 2011 x86_64 Intel(R) Core(TM) i5-2520M CPU @ 2.50GHz GenuineIntel GNU/Linux
and here the powertop output obtained with the cpu frequency governor set to "On demand":
Cn Avg residency P-states (frequencies)
C0 (cpu running) ( 5.3%) Turbo Mode 0.1%
polling 0.0ms ( 0.0%) 2.50 Ghz 0.0%
C1 mwait 0.2ms ( 0.1%) 2.21 Ghz 0.0%
C2 mwait 0.3ms ( 0.0%) 1200 Mhz 0.1%
C3 mwait 2.4ms ( 0.0%) 800 Mhz 99.8%
C4 mwait 13.1ms (94.6%)
Wakeups-from-idle per second : 76.1 interval: 3.0s
Power usage (ACPI estimate): 15.7W (2.8 hours)
Top causes for wakeups:
46.5% (135.7) kworker/0:0
17.7% ( 51.7) PS/2 keyboard/mouse/touchpad interrupt
11.9% ( 34.7) [kernel scheduler] Load balancing tick
8.6% ( 25.0) [iwlagn] <interrupt>
2.9% ( 8.3) [i915] <interrupt>
2.6% ( 7.7) [acpi] <interrupt>
1.8% ( 5.3) chromium
1.6% ( 4.7) [kernel core] hrtimer_start (tick_sched_timer)
1.5% ( 4.3) kworker/0:1
0.7% ( 2.0) minilogd
0.7% ( 2.0) [kernel core] iwl_bg_watchdog (iwl_bg_watchdog)
0.6% ( 1.7) X
0.6% ( 1.7) [kernel core] intel_gpu_idle_timer (intel_gpu_idle_timer)
0.3% ( 1.0) Terminal
0.3% ( 1.0) [kernel core] tpt_trig_timer (tpt_trig_timer)
0.2% ( 0.7) [mmc0, mei, ehci_hcd:usb3] <interrupt>
0.2% ( 0.7) upowerd
0.1% ( 0.3) init
0.1% ( 0.3) gpg-agent
0.1% ( 0.3) [kernel core] ieee80211_sta_reset_conn_monitor (ieee80211_sta_conn_mon_timer)
0.1% ( 0.3) wicd-client
0.1% ( 0.3) kworker/u:3
0.1% ( 0.3) watchdog/0
I have also tried to disable most of the running daemons and unload some modules, but I have never got less than 14/15W when idle, which in my opinion is definitely too much!
As far as I'm concerned, the kworker process is responsible most of the wakeups.
Any ideas?
]]>