From: Andrew Morton <akpm@zip.com.au>
To: Rik van Riel <riel@conectiva.com.br>
Cc: Daniel Phillips <phillips@bonn-fries.net>,
Alan Cox <alan@lxorguk.ukuu.org.uk>,
yodaiken@fsmlabs.com, Jeff Dike <jdike@karaya.com>,
Benjamin LaHaise <bcrl@redhat.com>,
"H. Peter Anvin" <hpa@zytor.com>,
linux-kernel@vger.kernel.org
Subject: Re: [RFC] Arch option to touch newly allocated pages
Date: Thu, 07 Mar 2002 12:56:44 -0800 [thread overview]
Message-ID: <3C87D40C.603DE513@zip.com.au> (raw)
In-Reply-To: <3C87BD22.BBBF4A86@zip.com.au> <Pine.LNX.4.44L.0203071709400.2181-100000@imladris.surriel.com>
Rik van Riel wrote:
>
> On Thu, 7 Mar 2002, Andrew Morton wrote:
> > Daniel Phillips wrote:
> > >
> > > a GFP flag that says 'fail if this looks hard to get'.
> >
> > Something like that would provide a solution to the
> > readahead thrashing problem.
>
> Nope. Readahead pages are clean and very easy to evict, so
> it's still trivial to evict all the pages from another readahead
> window because everybody's readahead window is too large.
>
I was thinking an explicit GFP_READAHEAD and PG_readahead.
Where a GFP_READAHEAD allocation would fail if it can't
find any non-readahead pages. And it would fail if it
had to perform I/O.
That's not nice - it'd result in large LRU walks. But it'd
be better than the 10x slowdown which readahead thrashing
causes.
Any clever ideas?
-
next prev parent reply other threads:[~2002-03-07 20:59 UTC|newest]
Thread overview: 96+ messages / expand[flat|nested] mbox.gz Atom feed top
2002-03-03 21:12 [RFC] Arch option to touch newly allocated pages Jeff Dike
2002-03-03 22:01 ` Alan Cox
2002-03-03 23:27 ` Jeff Dike
2002-03-03 23:48 ` Alan Cox
2002-03-04 3:16 ` Jeff Dike
2002-03-04 3:35 ` Alan Cox
2002-03-04 5:04 ` Jeff Dike
2002-03-04 15:09 ` Alan Cox
2002-03-04 17:42 ` Jeff Dike
2002-03-04 18:29 ` Alan Cox
2002-03-04 18:36 ` Jeff Dike
2002-03-04 18:49 ` Alan Cox
2002-03-04 20:46 ` Jeff Dike
2002-03-04 22:49 ` Alan Cox
2002-03-04 17:46 ` H. Peter Anvin
2002-03-04 18:34 ` Jeff Dike
2002-03-04 18:33 ` H. Peter Anvin
2002-03-04 20:36 ` Jeff Dike
2002-03-04 22:51 ` Alan Cox
2002-03-05 4:15 ` Jeff Dike
2002-03-05 4:28 ` Benjamin LaHaise
2002-03-05 4:40 ` Jeff Dike
2002-03-05 5:34 ` H. Peter Anvin
2002-03-05 14:43 ` Jeff Dike
2002-03-05 16:37 ` H. Peter Anvin
2002-03-05 16:56 ` Wayne Whitney
2002-03-05 18:12 ` Jeff Dike
2002-03-05 18:30 ` Benjamin LaHaise
2002-03-06 14:59 ` Daniel Phillips
2002-03-06 15:24 ` Benjamin LaHaise
2002-03-06 15:24 ` Daniel Phillips
2002-03-06 16:36 ` Benjamin LaHaise
2002-03-06 23:14 ` Daniel Phillips
2002-03-06 23:20 ` Benjamin LaHaise
2002-03-06 23:26 ` Daniel Phillips
2002-03-06 23:33 ` H. Peter Anvin
2002-03-07 0:08 ` Daniel Phillips
2002-03-07 1:27 ` Jeff Dike
2002-03-07 1:52 ` Benjamin LaHaise
2002-03-08 19:17 ` Jeff Dike
2002-03-08 21:22 ` Benjamin LaHaise
2002-03-07 13:49 ` Alan Cox
2002-03-07 13:36 ` Daniel Phillips
2002-03-07 14:04 ` yodaiken
2002-03-07 14:21 ` Daniel Phillips
2002-03-07 14:38 ` yodaiken
2002-03-07 15:31 ` Daniel Phillips
2002-03-07 16:50 ` yodaiken
2002-03-07 18:07 ` Daniel Phillips
2002-03-07 18:15 ` yodaiken
2002-03-07 19:22 ` Alan Cox
2002-03-07 22:43 ` David Woodhouse
2002-03-07 23:09 ` Alan Cox
2002-03-07 22:57 ` David Woodhouse
2002-03-07 14:43 ` Alan Cox
2002-03-07 15:32 ` Daniel Phillips
2002-03-07 16:19 ` Alan Cox
2002-03-07 17:54 ` Daniel Phillips
2002-03-07 15:34 ` Daniel Phillips
2002-03-07 19:18 ` Andrew Morton
2002-03-07 20:10 ` Rik van Riel
2002-03-07 20:56 ` Andrew Morton [this message]
2002-03-07 21:23 ` Rik van Riel
2002-03-07 22:02 ` Andrew Morton
2002-03-07 22:10 ` Rik van Riel
2002-03-07 22:23 ` Andrew Morton
2002-03-07 22:27 ` Rik van Riel
2002-03-07 22:41 ` Andrew Morton
2002-03-07 22:42 ` David Lang
2002-03-06 16:03 ` Jesse Pollard
2002-03-06 17:08 ` Jeff Dike
2002-03-06 17:33 ` Alan Cox
2002-03-07 0:28 ` Jeff Dike
2002-03-07 0:44 ` Alan Cox
2002-03-05 18:46 ` H. Peter Anvin
2002-03-06 1:30 ` Alan Cox
2002-03-06 10:49 ` David Woodhouse
2002-03-06 14:26 ` Jeff Dike
2002-03-06 16:50 ` Alan Cox
2002-03-06 20:25 ` Jeff Dike
2002-03-06 20:54 ` Alan Cox
2002-03-06 21:27 ` Malcolm Beattie
2002-03-06 23:26 ` Jeff Dike
2002-03-06 21:27 ` David Woodhouse
2002-03-06 22:25 ` Joseph Malicki
2002-03-07 0:04 ` Richard Gooch
2002-03-07 0:28 ` Jeff Dike
2002-03-07 0:44 ` Alan Cox
2002-03-06 22:21 ` Pavel Machek
2002-03-07 11:30 ` Dave Jones
2002-03-07 18:21 ` H. Peter Anvin
2002-03-05 14:43 ` Jeff Dike
2002-03-05 16:57 ` H. Peter Anvin
2002-03-05 18:14 ` Jeff Dike
2002-03-05 18:45 ` H. Peter Anvin
2002-03-05 17:30 ` Jan Harkes
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=3C87D40C.603DE513@zip.com.au \
--to=akpm@zip.com.au \
--cc=alan@lxorguk.ukuu.org.uk \
--cc=bcrl@redhat.com \
--cc=hpa@zytor.com \
--cc=jdike@karaya.com \
--cc=linux-kernel@vger.kernel.org \
--cc=phillips@bonn-fries.net \
--cc=riel@conectiva.com.br \
--cc=yodaiken@fsmlabs.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