linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
To: Hugh Dickins <hugh@veritas.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
	Nick Piggin <nickpiggin@yahoo.com.au>,
	Rik van Riel <riel@redhat.com>,
	linux-mm@kvack.org
Subject: Re: [PATCH 6/8] mm: remove try_to_munlock from vmscan
Date: Mon, 24 Nov 2008 12:34:52 -0500	[thread overview]
Message-ID: <1227548092.6937.23.camel@lts-notebook> (raw)
In-Reply-To: <Pine.LNX.4.64.0811232202040.4142@blonde.site>

On Sun, 2008-11-23 at 22:03 +0000, Hugh Dickins wrote:
> An unfortunate feature of the Unevictable LRU work was that reclaiming an
> anonymous page involved an extra scan through the anon_vma: to check that
> the page is evictable before allocating swap, because the swap could not
> be freed reliably soon afterwards.
> 
> Now try_to_free_swap() has replaced remove_exclusive_swap_page(), that's
> not an issue any more: remove try_to_munlock() call from shrink_page_list(),
> leaving it to try_to_munmap() to discover if the page is one to be culled
> to the unevictable list - in which case then try_to_free_swap().
> 
> Update unevictable-lru.txt to remove comments on the try_to_munlock()
> in shrink_page_list(), and shorten some lines over 80 columns.


Hugh:  Thanks for doing this.   Another item on my to-do list, as noted
in the document.

> 
> Signed-off-by: Hugh Dickins <hugh@veritas.com>
> ---
> I've not tested this against whatever test showed the need for that
> try_to_munlock() in shrink_page_list() in the first place.  Rik or Lee,
> please, would you have the time to run that test on the next -mm that has
> this patch in, to check that I've not messed things up?  Alternatively,
> please point me to such a test - but I think you've been targeting
> larger machines than I have access to - thanks.

I will rerun my test workload when this shows up in mmotm.  

I added the extra try_to_munlock() [TODO:  maybe "page_mlocked()" is
better name?] to prevent using swap space for pages that were destined
for the unevictable list.  This is more likely, I think, now that we've
removed the lru_drain_all() calls from the mlock[all]() handlers.  Back
when I added this, I wasn't sure that we could reliably remove swap from
a page with an arbitrary number of mappers.  Rik had warned against
making that assumption.

Lee

<snip>

--
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>

  parent reply	other threads:[~2008-11-24 17:34 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-11-23 21:53 [PATCH 0/8] mm: from gup to vmscan Hugh Dickins
2008-11-23 21:55 ` [PATCH 1/8] mm: gup persist for write permission Hugh Dickins
2008-11-23 21:56 ` [PATCH 2/8] mm: wp lock page before deciding cow Hugh Dickins
2008-11-23 21:58 ` [PATCH 3/8] mm: reuse_swap_page replaces can_share_swap_page Hugh Dickins
2008-11-23 22:11   ` [PATCH] memcg: memswap controller core swapcache fixes Hugh Dickins
2008-11-24  5:43     ` KAMEZAWA Hiroyuki
2008-11-24  6:15       ` KAMEZAWA Hiroyuki
2008-11-24 12:29         ` Hugh Dickins
2008-11-24 12:57           ` KAMEZAWA Hiroyuki
2008-11-23 22:43   ` [PATCH 3/8] mm: reuse_swap_page replaces can_share_swap_page Rik van Riel
2008-11-23 22:00 ` [PATCH 4/8] mm: try_to_free_swap replaces remove_exclusive_swap_page Hugh Dickins
2008-11-23 22:01 ` [PATCH 5/8] mm: try_to_unuse check removing right swap Hugh Dickins
2008-11-23 22:03 ` [PATCH 6/8] mm: remove try_to_munlock from vmscan Hugh Dickins
2008-11-23 22:53   ` Rik van Riel
2008-11-24 17:34   ` Lee Schermerhorn [this message]
2008-11-24 19:29     ` Hugh Dickins
2008-12-01 20:16       ` Lee Schermerhorn
2008-12-02  0:51         ` Hugh Dickins
2008-11-23 22:05 ` [PATCH 7/8] mm: remove gfp_mask from add_to_swap Hugh Dickins
2008-11-23 22:07 ` [PATCH 8/8] mm: add add_to_swap stub Hugh Dickins
2008-11-23 22:55   ` Rik van Riel
2008-11-24 13:49     ` Hugh Dickins
2008-11-24 13:53       ` [PATCH 9/8] mm: optimize get_scan_ratio for no swap Hugh Dickins
2008-11-24 14:11         ` Rik van Riel

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=1227548092.6937.23.camel@lts-notebook \
    --to=lee.schermerhorn@hp.com \
    --cc=akpm@linux-foundation.org \
    --cc=hugh@veritas.com \
    --cc=linux-mm@kvack.org \
    --cc=nickpiggin@yahoo.com.au \
    --cc=riel@redhat.com \
    /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