Announcement

Collapse
No announcement yet.

eth0 causing many interrupts

Collapse
This topic is closed.
X
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    eth0 causing many interrupts

    Info from powertop. I'm using a wireless network, so wlan0 should be the connection causing lots of interrupts, no? I don't even have eth0 plugged in right now. Any ideas why it's causing so many interrupts?

    Any thoughts are as always gratefully received.

    ---

    < Detailed C-state information is not P-states (frequencies)
    2.31 Ghz 0.0%
    2.21 Ghz 0.0%
    2.00 Ghz 0.0%
    1.80 Ghz 0.0%
    1000 Mhz 100.0%

    Wakeups-from-idle per second : 342.5 interval: 3.0s
    no ACPI power usage estimate available

    Top causes for wakeups:
    18.6% (113.0) amarok : schedule_hrtimeout_range (hrtimer_wakeup)
    16.6% (100.7) <interrupt> : eth0
    10.1% ( 61.0) <interrupt> : nvidia
    9.4% ( 57.0) <interrupt> : ehci_hcd:usb1
    9.1% ( 55.3) <kernel IPI> : Rescheduling interrupts
    7.8% ( 47.0) <interrupt> : ohci_hcd:usb2, HDA Intel
    6.8% ( 41.0) USB device 1-9 : 802.11 bg WLAN (Ralink)
    6.6% ( 40.0) firefox-3.5 : futex_wait (hrtimer_wakeup)
    3.6% ( 21.7) kwin : schedule_hrtimeout_range (hrtimer_wakeup)
    2.2% ( 13.3) USB device 1-10 : USB2.0-CRW (Generic)
    2.0% ( 12.0) usplash : scan_async (ehci_watchdog)
    1.5% ( 9.3) plasma : schedule_hrtimeout_range (hrtimer_wakeup)
    1.5% ( 9.0) <interrupt> : sata_nv
    1.4% ( 8.7) <kernel core> : hrtimer_start (tick_sched_timer)
    0.5% ( 3.0) knotify4 : schedule_hrtimeout_range (hrtimer_wakeup)
    0.4% ( 2.3) krunner : schedule_hrtimeout_range (hrtimer_wakeup)
    0.4% ( 2.3) <kernel core> : sk_reset_timer (tcp_delack_timer)

    Suggestion: increase the VM dirty writeback time from 5.00 to 15 seconds with:
    echo 1500 > /proc/sys/vm/dirty_writeback_centisecs
    This wakes the disk up less frequently for background VM activity

    #2
    Re: eth0 causing many interrupts

    Interesting.
    Were you listening to music when you ran Powertop? If you weren't I'd take Amarok out of the system tray and run it only when you are listening, and close it out of the tray when you are not. Saving a couple seconds of load time when you using minutes of interrupt time doesn't make sense.


    Increasing the VM write-back time as suggested is a good idea.

    Don't know why your eth0 is polling so often. Does it have a lan wakeup switch or reboot switch setting?
    "A nation that is afraid to let its people judge the truth and falsehood in an open market is a nation that is afraid of its people.”
    – John F. Kennedy, February 26, 1962.

    Comment


      #3
      Re: eth0 causing many interrupts

      The plot thickens...

      On fresh boot with eth0 running, it was causing 33% of the interrupts in my idle system!

      This was w/out music running. (When I posted the original information, Amarok was streaming music.)

      I went into ifconfig and disabled eth0. Now the interrupts are all gone... as below.

      I wonder why eth0 is sucking up so much of my processor's attention, since it's not even plugged in?

      My instinct is to just leave it off and call it a day. Is there a way to ensure it doesn't auto-on when I boot?

      Detailed C-state information is not P-states (frequencies)
      2.31 Ghz 0.0%
      2.21 Ghz 0.0%
      2.00 Ghz 0.0%
      1.80 Ghz 0.0%
      1000 Mhz 100.0%

      Wakeups-from-idle per second : 228.2 interval: 3.0s
      no ACPI power usage estimate available

      Top causes for wakeups:
      30.2% (128.3) <interrupt> : PS/2 keyboard/mouse/touchpad
      14.4% ( 61.0) <interrupt> : nvidia
      11.2% ( 47.7) <kernel IPI> : Rescheduling interrupts
      10.4% ( 44.0) <interrupt> : ehci_hcd:usb1
      6.3% ( 26.7) USB device 1-10 : USB2.0-CRW (Generic)
      6.1% ( 26.0) firefox-3.5 : futex_wait (hrtimer_wakeup)
      3.5% ( 15.0) kwin : schedule_hrtimeout_range (hrtimer_wakeup)
      3.5% ( 15.0) USB device 1-9 : 802.11 bg WLAN (Ralink)
      3.3% ( 14.0) <interrupt> : sata_nv
      2.5% ( 10.7) usplash : scan_async (ehci_watchdog)
      2.1% ( 9.0) konsole : schedule_hrtimeout_range (hrtimer_wakeup)
      2.0% ( 8.3) <kernel core> : hrtimer_start (tick_sched_timer)
      1.3% ( 5.3) plasma : schedule_hrtimeout_range (hrtimer_wakeup)
      0.8% ( 3.3) hald-addon-stor : schedule_hrtimeout_range (hrtimer_wakeup)
      0.7% ( 3.0) knotify4 : schedule_hrtimeout_range (hrtimer_wakeup)
      0.3% ( 1.3) hald-addon-stor : blk_plug_device (blk_unplug_timeout)
      0.2% ( 1.0) Xorg : nv_start_rc_timer (nv_kern_rc_timer)

      Suggestion: increase the VM dirty writeback time from 5.00 to 15 seconds with:
      echo 1500 > /proc/sys/vm/dirty_writeback_centisecs
      This wakes the disk up less frequently for background VM activity

      Comment


        #4
        Re: eth0 causing many interrupts

        Originally posted by fidesratioque
        My instinct is to just leave it off and call it a day. Is there a way to ensure it doesn't auto-on when I boot?
        You could disable it in the bios.

        Comment

        Working...
        X