lkml.org 
[lkml]   [2006]   [Mar]   [21]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    Date
    From
    SubjectRe: [RFC] - Move call to calc_load()

    On Tue, Mar 21, 2006 at 09:36:47PM +0100, Ingo Molnar wrote:
    > * Jack Steiner <steiner@sgi.com> wrote:
    >
    > > Here is the patch that I am proposing. This patch is incomplete
    > > because it addresses only the IA64 architecture. If this approach is
    > > acceptible, I'll update the patch to cover all architectures.
    > >
    > > Signed-off-by: Jack Steiner <steiner@sgi.com>
    >
    > i agree with your analysis - there is no reason calc_load() should be
    > under xtime_lock. I guess no-one noticed this so far because calc_load()
    > iterating over hundreds of CPUs isnt too common.

    I tested this patch and it works well for eliminating latencies due to
    contention for the xtime_lock. Without the patch the latencies are
    quite substantial at higher cpu counts.
    -
    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/

    \
     
     \ /
      Last update: 2006-03-21 21:51    [W:0.038 / U:0.936 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site