Paul Menzel wrote:
> Dear list,
> 
> 
> thanks to the Debian packagers I updated to Powertop 1.11 today. (I will
> send in tests for the other thing (units for calculations) some other
> day.)
> 
> Here is the new dump for the EeePC 4G and the USB device. Sorry that I
> did not change to English language.
> 
> The dump was taken using an ssh-session and only GDM was running and
> nothing else done after boot.
> 
> 
> PowerTOP 1.11   (C) 2007, 2008 Intel Corporation 
> 
> Sammle Daten für 15 Sekunden 
> 
> 
> Cn                 Verweildauer
> C0 (Prozessor läuft)    ( 0,9%)
> zyklisches AbfraC1                0,0ms ( 0,0%)
> C1                0,0ms ( 0,0%)
> C2                0,0ms ( 0,0%)
> C3               69,8ms (99,1%)
> P-States (Frequenzen)
> Aufwachen pro Sekunde : 14,3    Intervall: 15,0s
> Keine ACPI Stromverbrauch-Schätzung verfügbar
> Häufigste Ursachen für das Aufwachen:
>   64,1% ( 13,3)       <interrupt> : uhci_hcd:usb4, HDA Intel, 
> i...@pci:0000:00:02.0 
>    8,0% (  1,7)          gdmlogin : schedule_timeout (process_timeout) 
>    4,8% (  1,0)              ntpd : do_setitimer (it_real_fn) 
>    4,8% (  1,0)            dhcdbd : schedule_timeout (process_timeout) 
>    4,8% (  1,0)              Xorg : schedule_timeout (process_timeout) 
>    2,6% (  0,5)    <Kernel Modul> : neigh_table_init_no_netlink 
> (neigh_periodic_timer) 
>    2,6% (  0,5)     <Kernel Kern> : schedule_delayed_work_on 
> (delayed_work_timer_fn) 
>    1,6% (  0,3)       <interrupt> : ata_piix 
>    1,3% (  0,3)     <Kernel Kern> : neigh_table_init_no_netlink 
> (neigh_periodic_timer) 
>    1,0% (  0,2)              init : schedule_timeout (process_timeout) 
>    1,0% (  0,2)          ifconfig : atl2_open (atl2_watchdog) 
>    1,0% (  0,2)          events/0 : __netdev_watchdog_up (dev_watchdog) 
>    0,3% (  0,1)            screen : do_setitimer (it_real_fn) 
>    0,3% (  0,1)              sshd : sk_reset_timer (tcp_write_timer) 
>    0,3% (  0,1)     <Kernel Kern> : neigh_add_timer (neigh_timer_handler) 
>    0,3% (  0,1)             touch : start_this_handle (commit_timeout) 
>    0,3% (  0,1)              hald : schedule_timeout (process_timeout) 
>    0,3% (  0,1)          rsyslogd : schedule_timeout (process_timeout) 
>    0,3% (  0,1)     <Kernel Kern> : page_writeback_init (wb_timer_fn) 
>    0,3% (  0,1)           pdflush : blk_plug_device (blk_unplug_timeout) 
> 
> aktuelle "USB suspend" Statistiken
> Name des aktiven Gerätes
>   0,0%  USB Gerät usb5 : EHCI Host Controller (Linux 2.6.26-1-686 ehci_hcd)
>   0,0%  USB Gerät usb4 : UHCI Host Controller (Linux 2.6.26-1-686 uhci_hcd)
>   0,0%  USB Gerät usb3 : UHCI Host Controller (Linux 2.6.26-1-686 uhci_hcd)
>   0,0%  USB Gerät usb2 : UHCI Host Controller (Linux 2.6.26-1-686 uhci_hcd)
>   0,0%  USB Gerät usb1 : UHCI Host Controller (Linux 2.6.26-1-686 uhci_hcd)
> 
> 
> Am Sonntag, den 23.11.2008, 09:11 -0700 schrieb Nathan Coulson:
> 
> […]
> 
>> I've been researching eeepc power reduction tweaks back when I had
>> some spare time,  and I came up with the following suggestions.
> 
> […]
> 
>>     * CONFIG_SND_HDA_POWER_SAVE=y &
>> CONFIG_SND_HDA_POWER_SAVE_DEFAULT=10: These are kernel configuration
>> options that will allow your soundcard to suspend itself when it has
>> not been used for 10 seconds. The above kernel configuration kernel
>> already has this enabled.
> 
> So this seems to be the culprit. It is strange for me or new to me that
> these sound devices communicate using USB. I will test with those
> setting is some days on to see if it goes away.

they might be sharing an irq together with the USB subsystem, which is quite
common, and confuses powertop.

Auke


_______________________________________________
Power mailing list
[email protected]
http://www.bughost.org/mailman/listinfo/power

Reply via email to