lkml.org 
[lkml]   [2011]   [Mar]   [18]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: [PATCH] ksm: add vm_stat and meminfo entry to reflect pte mapping to ksm pages
Date

>On Wednesday 02 March 2011, at 07:41:00, <Andrew Morton <akpm@linux-foundation.org>> wrote
> On Sat, 26 Feb 2011 22:56:31 +0800
> Nai Xia <nai.xia@gmail.com> wrote:
>
> > ksm_pages_sharing is updated by ksmd periodically. In some cases, it cannot
> > reflect the actual savings and makes the benchmarks on volatile VMAs very
> > inaccurate.
> >
> > This patch add a vm_stat entry and let the /proc/meminfo show information
> > about how much virutal address pte is being mapped to ksm pages. With default
> > ksm paramters (pages_to_scan==100 && sleep_millisecs==20), this can result in
> > 50% more accurate averaged savings result for the following test program.
> > Bigger sleep_millisecs values will increase this deviation.
>
> So I think you're saying that the existing ksm_pages_sharing sysfs file
> is no good.
>
> You added a new entry to /proc/meminfo and left ksm_pages_sharing
> as-is. Why not leave /proc/meminfo alone, and fix up the existing
> ksm_pages_sharing?


The ksm_pages_sharing is really a count for how many "rmap_item"s is currently
linked in stable_nodes. ksmd updates ksm_pages_sharing whenever it's waken up.
However, just during the time ksmd is sleeping, many shared KSM pages can be
broken because of page writes. So ksm_pages_sharing means much more an internal
state for ksm than a real count for how much pages is being shared at some time
point. Since the state of the internal data structures of ksm is only updated
by ksmd privately. I think it's hard to make it correctly reflect the real time
memory saving. So I just added another count and let ksm_pages_sharing still
focus on it original role.

>
> Also, the patch accumulates the NR_KSM_PAGES_SHARING counts on a
> per-zone basis as well as on a global basis, but only provides the
> global count to userspace. The per-zone counts are potentially
> interesting? If not, maintaining the per-zone counters is wasted
> overhead.

Yes, I will make it to the zoneinfo, soon.

>
> >
> > --- test.c-----
> >
>
> The "^---" token conventionally means "end of changelog". Please avoid
> inserting it into the middle of the changelog.

OK, I understand now. :)

>
> > +++ b/mm/ksm.c
> > @@ -897,6 +897,7 @@ static int try_to_merge_one_page(struct vm_area_struct
> > *vma,
>
> Your email client wordwraps the patches.

Oops, sorry for the crappy client it is also responsible for my late reply.
I will fix it soon.

Nai

>


\
 
 \ /
  Last update: 2011-03-18 08:19    [W:0.068 / U:5.584 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site