linux-kernel - 2.6.24-rc8-mm1 - Tons of extra interrupts..

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <31145.1201198324@turing-police.cc.vt.edu>
Date:	Thu, 24 Jan 2008 13:12:04 -0500
From:	Valdis.Kletnieks@...edu
To:	linux-kernel@...r.kernel.org
Subject: 2.6.24-rc8-mm1 - Tons of extra interrupts..

Dell Latitude D820 laptop, T7200 Core2 Duo, x86_64 kernel built with
PREEMPT, NO_HZ, and HZ=1000. It's running basically idle with an X desktop session.

I'm seeing large masses of rescheduling interrupts that are basically killing
the C3 residency rate for no obvious reason/gain. I'm taking close to 500
interrupts/second, of which close to 350 are unexplained and apparently
gratuitous. Anybody have suggestions/clues on how to find out why I'm seeing
this? (Quite possibly, I've got something misconfigured because I misunderstood
the documentation - which would still indicate that the doc needs fixing at the
least...)

powertop reports:

     PowerTOP version 1.9       (C) 2007 Intel Corporation

Cn                Avg residency       P-states (frequencies)
C0 (cpu running)        ( 5.6%)         2.00 Ghz     1.2%
C1                0.0ms ( 0.0%)         1.67 Ghz     0.0%
C2                0.7ms ( 5.1%)         1333 Mhz     0.0%
C3                2.1ms (89.3%)         1000 Mhz    98.8%


Wakeups-from-idle per second : 494.9    interval: 10.0s
no ACPI power usage estimate available

Top causes for wakeups:
  43.9% (190.3)      <kernel IPI> : Rescheduling interrupts 
  29.4% (127.4)       <interrupt> : extra timer interrupt 
   5.9% ( 25.6)                 X : do_setitimer (it_real_fn) 
   5.0% ( 21.7)           gkrellm : schedule_timeout (process_timeout) 
   3.9% ( 16.7)     group_balance : schedule_timeout (process_timeout) 
   3.0% ( 13.2)       <interrupt> : iwl3945 
   2.3% ( 10.0)     <kernel core> : ehci_work (ehci_watchdog) 

/proc/interrupts (after about 10 hours of uptime, most of which was mostly
idle and "fetchmail check the mail server every 5 minutes"):

           CPU0       CPU1       
  0:   14507246          0   IO-APIC-edge      timer
  1:      10748          0   IO-APIC-edge      i8042
  8:          0          0   IO-APIC-edge      rtc
  9:          3          0   IO-APIC-fasteoi   acpi
 12:         86          0   IO-APIC-edge      i8042
 14:     147067          0   IO-APIC-edge      ata_piix
 15:         45          0   IO-APIC-edge      ata_piix
 16:      91584          0   IO-APIC-fasteoi   nvidia
 19:          9          0   IO-APIC-fasteoi   ohci1394, yenta
 20:      51139          0   IO-APIC-fasteoi   ehci_hcd:usb1, uhci_hcd:usb2
 21:      34708          0   IO-APIC-fasteoi   uhci_hcd:usb3, HDA Intel
 22:          0          0   IO-APIC-fasteoi   uhci_hcd:usb4
 23:          0          0   IO-APIC-fasteoi   uhci_hcd:usb5
312:     164320          0   PCI-MSI-edge      eth0
313:     385582          0   PCI-MSI-edge      iwl3945
314:       1343          0      none-edge    
NMI:          0          0   Non-maskable interrupts
LOC:    1263420    5737965   Local timer interrupts
RES:    2285220    2305246   Rescheduling interrupts
CAL:       4068       6172   function call interrupts
TLB:      41441      44272   TLB shootdowns
TRM:          0          0   Thermal event interrupts
THR:          0          0   Threshold APIC interrupts
SPU:          0          0   Spurious interrupts
ERR:          0

Any ideas/hints?

Content of type "application/pgp-signature" skipped