lkml.org 
[lkml]   [2008]   [Mar]   [28]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    From
    SubjectRe: 2.6.25 regression: powertop says 120K wakeups/sec
    Date
    On Saturday 22 March 2008, Andrew Morton wrote:
    > On Sat, 22 Mar 2008 13:24:54 -0700 David Brownell <david-b@pacbell.net> wrote:
    >
    > > I noticed this with 2.6.25-rc2 (if not before), and the problem
    > > is still there with 2.6.25-rc6-git (as of this AM).

    A 2.6.24 kernel I still had stashed away didn't act odd; the
    problem joined us before 2.6.25-rc1 was tagged.


    > > System is an Athlon64 single CPU laptop, and instead of reading a
    > > few dozen wakeups per second, it says a many tens of thousands...
    > > clearly wrong. In previous kernels it gave more plausible counts;
    > > unfortunately high because of various un-evolved desktop tools in
    > > this Ubuntu system (Feisty).
    > >
    > > Possibly more truthful, it says that the system never enters
    > > C1 or C2, and spends all its time in C0. Though if I look at
    > > /sys/devices/system/cpu/cpu0/cpuidle/state[01]/usage, that
    > > seems to tell a different story ... it's C0 that's never used.
    > > In previous kernels it reported time in both C0 and C2. ISTR
    > > some patch to avoid C2, which would explain part of this.
    > >
    > > Comments or fixes, anyone?
    >
    > This is likely to be an acpi regression, isn't it?
    >
    > A git-bisect would be nice, please.

    The git-bisect says the 120K wakeups/second comes from a patch
    which unfortunately can't be directly reverted, so I didn't
    verify that reverting it resolves that problem. I've not tried
    to do anything with the other C0/C1/C2 stuff.


    $ git bisect good
    bc71bec91f9875ef825d12104acf3bf4ca215fa4 is first bad commit
    commit bc71bec91f9875ef825d12104acf3bf4ca215fa4
    Author: venkatesh.pallipadi@intel.com <venkatesh.pallipadi@intel.com>
    Date: Thu Jan 31 17:35:04 2008 -0800

    ACPI: enable MWAIT for C1 idle

    Add MWAIT idle for C1 state instead of halt, on platforms that support
    C1 state with MWAIT.

    Renames cx->space_id to something more appropriate.

    Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
    Signed-off-by: Len Brown <len.brown@intel.com>

    :040000 040000 88ebe48d024f7fb21237cd75dbc9c681c43252b1 8af87317facbd018b47e717ede6907d9a831f92c M drivers
    :040000 040000 ecd73d87c1b7b7004e06ffa3a2b3e7260c045543 934c38290353186cbaaaf27094d5c1712e548fcc M include
    $





    \
     
     \ /
      Last update: 2008-03-28 20:05    [W:0.022 / U:63.156 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site