lkml.org 
[lkml]   [2015]   [Nov]   [16]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    Date
    From
    SubjectRe: [PATCH] mm: vmpressure: fix scan window after SWAP_CLUSTER_MAX increase
    Dear Hugh,

    [ sorry, I just noticed this email now ]

    On Wed, Oct 21, 2015 at 01:05:53PM -0700, Hugh Dickins wrote:
    > On Wed, 21 Oct 2015, Johannes Weiner wrote:
    > > On Mon, Oct 19, 2015 at 02:13:01PM -0400, Johannes Weiner wrote:
    > > > mm-increase-swap_cluster_max-to-batch-tlb-flushes.patch changed
    > > > SWAP_CLUSTER_MAX from 32 pages to 256 pages, inadvertantly switching
    > > > the scan window for vmpressure detection from 2MB to 16MB. Revert.
    > > >
    > > > Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
    > > > ---
    > > > mm/vmpressure.c | 2 +-
    > > > 1 file changed, 1 insertion(+), 1 deletion(-)
    > > >
    > > > diff --git a/mm/vmpressure.c b/mm/vmpressure.c
    > > > index c5afd57..74f206b 100644
    > > > --- a/mm/vmpressure.c
    > > > +++ b/mm/vmpressure.c
    > > > @@ -38,7 +38,7 @@
    > > > * TODO: Make the window size depend on machine size, as we do for vmstat
    > > > * thresholds. Currently we set it to 512 pages (2MB for 4KB pages).
    > > > */
    > > > -static const unsigned long vmpressure_win = SWAP_CLUSTER_MAX * 16;
    > > > +static const unsigned long vmpressure_win = SWAP_CLUSTER_MAX;
    > >
    > > Argh, Mel's patch sets SWAP_CLUSTER_MAX to 256, so this should be
    > > SWAP_CLUSTER_MAX * 2 to retain the 512 pages scan window.
    > >
    > > Andrew could you please update this fix in-place? Otherwise I'll
    > > resend a corrected version.
    > >
    > > Thanks, and sorry about that.
    >
    > I don't understand why "SWAP_CLUSTER_MAX * 2" is thought better than "512".
    > Retaining a level of obscurity, that just bit us twice, is a good thing?

    I'm not sure it is. But it doesn't seem entirely wrong to link it to
    the reclaim scan window, either--at least be a multiple of it so that
    the vmpressure reporting happens cleanly at the end of a scan cycle?

    I don't mind changing it to 512, but it doesn't feel like an obvious
    improvement, either.


    \
     
     \ /
      Last update: 2015-11-16 21:41    [W:2.509 / U:0.012 seconds]
    ©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site