linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Minchan Kim <minchan@kernel.org>
To: Michal Hocko <mhocko@kernel.org>
Cc: akpm@linux-foundation.org, ebru.akagunduz@gmail.com,
	aarcange@redhat.com, aneesh.kumar@linux.vnet.ibm.com,
	boaz@plexistor.com, gorcunov@openvz.org, hannes@cmpxchg.org,
	hughd@google.com, iamjoonsoo.kim@lge.com,
	kirill.shutemov@linux.intel.com, mgorman@suse.de,
	n-horiguchi@ah.jp.nec.com, riel@redhat.com, rientjes@google.com,
	vbabka@suse.cz, mm-commits@vger.kernel.org, linux-mm@kvack.org
Subject: Re: + mm-thp-avoid-unnecessary-swapin-in-khugepaged.patch added to -mm tree
Date: Fri, 20 May 2016 17:26:19 +0900	[thread overview]
Message-ID: <20160520082619.GB14049@bbox> (raw)
In-Reply-To: <20160520080217.GG19172@dhcp22.suse.cz>

On Fri, May 20, 2016 at 10:02:18AM +0200, Michal Hocko wrote:
> On Fri 20-05-16 16:44:50, Minchan Kim wrote:
> > > > > That being said khugepaged_max_ptes_none = HPAGE_PMD_NR/2 sounds like a
> > > > 
> > > > max_ptes_none?
> > > 
> > > Not sure I understand what you mean here.
> > 
> > We are talking about max_ptes_swap and max_active_pages(i.e., pte_young)
> > but suddenly you are saying max_ptes_none so I was curious it was just
> > typo.
> 
> Because the default for pte_none resp. zero pages collapsing into THP is
> khugepaged_max_ptes_none and the current default means that a single
> present page is sufficient. That is way too optimistic. So I consider
> this to be a good start. I am not so sure about minimum young pages
> because that would probably require yet another tunable and we have more
> than enough of them. Anyway I guess we are getting off-topic here...

Optimistic swapin collapsing

1. it could be too optimisitic to lose the gain due to evicting workingset
2. let's detect memory pressure
3. current allocstall magic is not a good idea.
4. let's change the design from optimistic to conservative
5. how we can be conservative
6. two things - detect hot pages and threshold of swap pte
7. threhsold of swap pte is already done so remained thing is detect hot page
8. how to detect hot page - let's use young bit
9. Now, we are conservatie so we will swap in when it's worth
10. let's remove allocstall magic

I think it's not off-topic.
Anyway, it's just my thought and don't have any real workload and objection.
Feel free to ignore.

> -- 
> Michal Hocko
> SUSE Labs
> 
> --
> To unsubscribe, send a message with 'unsubscribe linux-mm' in
> the body to majordomo@kvack.org.  For more info on Linux MM,
> see: http://www.linux-mm.org/ .
> Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

      reply	other threads:[~2016-05-20  8:26 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <57212c60.fUSE244UFwhXE+az%akpm@linux-foundation.org>
2016-04-28 15:19 ` + mm-thp-avoid-unnecessary-swapin-in-khugepaged.patch added to -mm tree Michal Hocko
2016-05-17  7:58   ` Michal Hocko
2016-05-17  9:02     ` Michal Hocko
2016-05-17 11:31       ` Kirill A. Shutemov
2016-05-17 12:25         ` Michal Hocko
2016-05-19  5:00       ` Minchan Kim
2016-05-19  7:03         ` Michal Hocko
2016-05-19  7:27           ` Minchan Kim
2016-05-19  7:39             ` Michal Hocko
2016-05-20  0:21               ` Minchan Kim
2016-05-20  6:39                 ` Michal Hocko
2016-05-20  7:26                   ` Minchan Kim
2016-05-20  7:34                     ` Michal Hocko
2016-05-20  7:44                       ` Minchan Kim
2016-05-20  8:02                         ` Michal Hocko
2016-05-20  8:26                           ` Minchan Kim [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20160520082619.GB14049@bbox \
    --to=minchan@kernel.org \
    --cc=aarcange@redhat.com \
    --cc=akpm@linux-foundation.org \
    --cc=aneesh.kumar@linux.vnet.ibm.com \
    --cc=boaz@plexistor.com \
    --cc=ebru.akagunduz@gmail.com \
    --cc=gorcunov@openvz.org \
    --cc=hannes@cmpxchg.org \
    --cc=hughd@google.com \
    --cc=iamjoonsoo.kim@lge.com \
    --cc=kirill.shutemov@linux.intel.com \
    --cc=linux-mm@kvack.org \
    --cc=mgorman@suse.de \
    --cc=mhocko@kernel.org \
    --cc=mm-commits@vger.kernel.org \
    --cc=n-horiguchi@ah.jp.nec.com \
    --cc=riel@redhat.com \
    --cc=rientjes@google.com \
    --cc=vbabka@suse.cz \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).