lkml.org 
[lkml]   [2009]   [Jan]   [13]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    SubjectRe: ALLANC: Proposal to add progress records to process accounting
    From
    Date
    > Are you planning to update the log based upon wall time or cpu used ?

    I planned on using the exact same algorithm as is used currently. I believe
    the current->signal->pacct structure holds system and user CPU time which
    is copied into the log ac_stime and ac_utime entries respectively in
    kernel/acct.c/do_acct_proces() thus:
    ac.ac_utime =
    encode_comp_t(jiffies_to_AHZ(cputime_to_jiffies(pacct->ac_utime)));
    ac.ac_stime =
    encode_comp_t(jiffies_to_AHZ(cputime_to_jiffies(pacct->ac_stime)));

    The log ac_etime is elapsed time (which is what I assume you mean by wall
    time) and is calculated as current_uptime - process_start_time
    (paraphrased) in the same function.

    Cheers,
    Pax.
    ==========
    Two strings walk into a bar. The first one says: "Hello, I'd like some
    Vodka please.ytewsr@)W$(#*$&!^Y@)^&30@#!".
    "You'll have to excuse my friend," the second one says, "he's not
    null-terminated".
    ==========
    Allan Chandler (Pax), Senior IT Specialist, Australia Development
    Laboratory (ADL), Perth, IBM Corporation
    Internet: allachan@au1.ibm.com, Phone: +61 8 926 13019, Fax: +61 8 921
    89201
    Location code MP25, 77 St George's Terrace (Allendale Square), Level 25,
    Perth WA 6000
    _______________________________________________




    Alan Cox
    <alan@lxorguk.uku
    u.org.uk> To
    Sent by: Allan Chandler/Australia/IBM@IBMAU
    linux-kernel-owne cc
    r@vger.kernel.org linux-kernel@vger.kernel.org
    Subject
    Re: ALLANC: Proposal to add
    13/01/2009 06:17 progress records to process
    PM accounting










    > will be written for any process that hasn't written a record in a while
    > (for our purposes, 24 hours is fine but it may be better to make this
    > configurable).

    Agreed and this sounds sensible.

    > too much overhead to the context switch (that's putting it mildly). I
    > assume this would be somewhere in kernel/sched.c/context_switch() or
    > kernel/sched.c/prepare_task_switch() although I haven't yet looked in
    great
    > detail.

    Are you planning to update the log based upon wall time or cpu used ?

    Alan





    \
     
     \ /
      Last update: 2009-01-13 10:49    [W:0.060 / U:2.436 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site