lkml.org 
[lkml]   [2012]   [Jul]   [31]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    Date
    From
    SubjectRe: [PATCH 18/19] sched, numa: Per task memory placement for big processes
    On 07/31/2012 03:12 PM, Peter Zijlstra wrote:
    > Probability says that the task faulting on a page after we protect it,
    > is most likely to be the task that uses that page most.
    >
    > To decrease the likelyhood of acting on a false relation, we only
    > migrate a page when two consecutive samples are from the same task.
    >
    > I'm still not entirely convinced this scheme is sound, esp. for things
    > like virtualization and n:m threading solutions in general the
    > compute<->task relation is fundamentally untrue.

    Again, we may need some additional code on top in the future,
    eg. something like Andrea's policy that tries grouping related
    tasks/threads together, but this looks like a very good way
    to start.

    We can introduce complexity if it is needed. Simplicity is good.

    Acked-by: Rik van Riel

    --
    All rights reversed


    \
     
     \ /
      Last update: 2012-08-01 00:21    [W:4.023 / U:23.592 seconds]
    ©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site