Re: [ACPI] Re: system slow since ~ 2.6.7

From: Guennadi Liakhovetski (g.liakhovetski_at_gmx.de)
Date: 11/20/04

  • Next message: Florian Schmidt: "Re: [patch] Real-Time Preemption, -RT-2.6.10-rc2-mm2-V0.7.29-0"
    Date:	Sat, 20 Nov 2004 21:35:46 +0100 (CET)
    To: Andrew Morton <akpm@osdl.org>
    
    
    

    On Wed, 17 Nov 2004, Andrew Morton wrote:

    > (added linux-acpi)

    Thanks for the reply and sorry for the delay - was away for 2 days.

    But now I've got some new data, that may give some hints.

    1) I played with BIOS setup, switched off, switched on the PC, booted into
    different kernels, maybe, re-configured and re-installed kernels, and
    then, at some point, the problem was gone.

    2) Today I added a coupld of points to the 2.6.9 config (attached), and
    the problem re-appeared.

    3) I noticed, the problem (blinking power LED) starts at
    /etc/init.d/lm_sensors.start. More precisely - after sensors -s is run. At
    the same time the system performance drops (you might know that this is
    related to sensors already - do you?)

    > readprofile -r
    > sleep 10
    > readprofile -n -v -m /boot/System.map | sort -n +2 | tail -40

    Not sure, if this is still needed, but here goes: before sensors:

    c0118b80 task_prio 1 0.0625
    c0119630 task_curr 1 0.0125
    c01197c0 add_wait_queue 1 0.0104
    c010aa80 do_gettimeofday 2 0.0114
    c01182d0 nr_running 2 0.1250
    c0119610 in_sched_functions 2 0.0625
    c0105143 syscall_exit 4 0.3636
    c0119d00 mmput 4 0.0250
    c0105138 syscall_call 6 0.5455
    c0119da0 get_task_mm 26 0.2321
    c0103d20 get_wchan 60 0.4167
    c010510c system_call 100 2.2727
    00000000 total 209 0.0019

    and after:

    c0105138 syscall_call 1 0.0909
    c0116710 pgd_ctor 1 0.0063
    c0118b80 task_prio 1 0.0625
    c0119ec0 copy_mm 1 0.0010
    c010aa80 do_gettimeofday 2 0.0114
    c0118b90 task_nice 2 0.1250
    c0119610 in_sched_functions 2 0.0625
    c0103db0 get_free_idx 3 0.0469
    c0105143 syscall_exit 3 0.2727
    c0119d00 mmput 3 0.0187
    c0116ab0 do_page_fault 4 0.0026
    c01197c0 add_wait_queue 5 0.0521
    c0119630 task_curr 8 0.1000
    c0119da0 get_task_mm 31 0.2768
    c0103d20 get_wchan 36 0.2500
    c010510c system_call 95 2.1591
    00000000 total 198 0.0018

    In my /etc/sysconfig/lm_sensors:

    MODULE_0=i2c_dev
    MODULE_1=i2c_isa
    MODULE_2=via686a

    Thanks
    Guennadi

    ---
    Guennadi Liakhovetski
    
    

    -
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    More majordomo info at http://vger.kernel.org/majordomo-info.html
    Please read the FAQ at http://www.tux.org/lkml/



  • Next message: Florian Schmidt: "Re: [patch] Real-Time Preemption, -RT-2.6.10-rc2-mm2-V0.7.29-0"

    Relevant Pages

    • Re: i2c_adapter i2c-0: Bus collision!
      ... > Are you able to reproduce the same behavior with kernel 2.4.24 and ... > lm-sensors 2.8.2? ... sometime has to reload the modules before "sensors" finds any sensor. ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)
    • Re: i2c: lost sensors with 2.6.10(-mm1)
      ... > This basically means that the i2c client was not registered. ... Disabling "Plug and Play ACPI support" freed the ioport range and sensors ... send the line "unsubscribe linux-kernel" in ... More majordomo info at http://vger.kernel.org/majordomo-info.html ...
      (Linux-Kernel)
    • Re: 2.6.5 Sensors & USB problems
      ... Seems that the w83781d driver no longer detects whatever the sensor chip on ... On Sat, 17 Apr 2004, Ian Morgan wrote: ... >> Your sensors problem will be resolved as soon as you switch to the just ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)
    • Q re /proc/bus/i2c
      ... I'm still trying to make sensors work when booted to a ... 2.6.x kernel. ... Copyright 2003 by Maurice Eugene Heskett, ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)