From: Daniel Phillips <phillips@bonn-fries.net>
To: Andrea Arcangeli <andrea@suse.de>
Cc: Bill Davidsen <davidsen@tmr.com>,
Mike Fedyk <mfedyk@matchmail.com>,
linux-kernel@vger.kernel.org
Subject: Re: 2.4.19pre1aa1
Date: Mon, 4 Mar 2002 04:22:20 +0100 [thread overview]
Message-ID: <E16hj3N-0000SY-00@starship.berlin> (raw)
In-Reply-To: <20020301013056.GD2711@matchmail.com> <E16hhYV-0000Qz-00@starship.berlin> <20020304032535.F20606@dualathlon.random>
In-Reply-To: <20020304032535.F20606@dualathlon.random>
On March 4, 2002 03:25 am, Andrea Arcangeli wrote:
> On Mon, Mar 04, 2002 at 02:46:22AM +0100, Daniel Phillips wrote:
> > As soon as you have shared pages you start to benefit from rmap's ability
> > to unmap in one step, so the cost of creating the link is recovered by not
>
> we'd benefit also with unshared pages.
>
> BTW, for the map shared mappings we just collect the rmap information,
> we need it for vmtruncate, but it's not layed out for efficient
> browsing, it's only meant to make vmtruncate work.
Sorry, transmission error, what did you mean?
> > having to scan two page tables to unmap it. In theory. Do you see a hole
> > in that?
>
> Just the fact you never need the reverse lookup during lots of
> important production usages (first that cames to mind is when you have
> enough ram to do your job, all number crunching/fileserving, and most
> servers are setup that way). This is the whole point. Note that this
> has nothing to do with the "cache" part, this is only about the
> pageout/swapout stage, only a few servers really needs heavy swapout.
You always have to unmap the page at some point, so you win back the cost
of creating the pte_chain there, hopefully. You could argue that paying
the cost up front makes latency a little worse. You might have trouble
measuring that though.
> ...Another bit in the current design of round robin cycling
> over the whole VM clearing the accessed bitflag and activating physical
> pages if needed, can also be see also as a feature in some ways. It is
> much better at providing a kind of "clock based" aging to the accessed
> bit information, while the lru pass rmap aware, wouldn't really be fair
> with all the virtual pages the same way as we do now.
You get a perfectly good clock by scanning the lru list. It's not
totally fair because a page newly promoted from the cold end to the hot
end of the list will get scanned again after a much shorter delta-t,
but it's hard to see why that's bad.
--
Daniel
next prev parent reply other threads:[~2002-03-04 3:27 UTC|newest]
Thread overview: 77+ messages / expand[flat|nested] mbox.gz Atom feed top
2002-02-27 12:50 2.4.19pre1aa1 Andrea Arcangeli
2002-02-28 22:11 ` 2.4.19pre1aa1 Bill Davidsen
2002-03-01 1:30 ` 2.4.19pre1aa1 Mike Fedyk
2002-03-01 3:26 ` 2.4.19pre1aa1 Bill Davidsen
2002-03-01 3:46 ` 2.4.19pre1aa1 Mike Fedyk
2002-03-01 12:51 ` 2.4.19pre1aa1 Rik van Riel
2002-03-01 18:37 ` 2.4.19pre1aa1 Mike Fedyk
2002-03-01 10:17 ` 2.4.19pre1aa1 Marco Colombo
2002-03-01 11:37 ` 2.4.19pre1aa1 Alan Cox
2002-03-02 2:06 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-02 2:28 ` 2.4.19pre1aa1 Alan Cox
2002-03-02 3:30 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-03 21:38 ` 2.4.19pre1aa1 Daniel Phillips
2002-03-04 0:49 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-04 1:46 ` 2.4.19pre1aa1 Daniel Phillips
2002-03-04 2:25 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-04 3:22 ` Daniel Phillips [this message]
2002-03-04 12:41 ` 2.4.19pre1aa1 Rik van Riel
2002-03-04 14:05 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-04 14:23 ` 2.4.19pre1aa1 Rik van Riel
2002-03-04 16:10 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-04 16:28 ` 2.4.19pre1aa1 Rik van Riel
2002-03-04 16:59 ` 2.4.19pre1aa1 Martin J. Bligh
2002-03-04 18:19 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-04 21:36 ` 2.4.19pre1aa1 Rik van Riel
2002-03-04 23:01 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-04 23:11 ` 2.4.19pre1aa1 Rik van Riel
2002-03-04 23:52 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-05 0:01 ` 2.4.19pre1aa1 Rik van Riel
2002-03-05 1:05 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-05 1:26 ` 2.4.19pre1aa1 Rik van Riel
2002-03-05 1:40 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-05 12:22 ` 2.4.19pre1aa1 Rik van Riel
2002-03-05 15:01 ` 2.4.19pre1aa1 Andrea Arcangeli
[not found] ` <Pine.LNX.4.44L.0203050921510.1413-100000@duckman.distro.conecti va>
2002-03-05 15:29 ` 2.4.19pre1aa1 Martin J. Bligh
2002-03-05 15:43 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-05 1:55 ` 2.4.19pre1aa1 Martin J. Bligh
2002-03-05 5:16 ` 2.4.19pre1aa1 Samuel Ortiz
2002-03-05 5:47 ` 2.4.19pre1aa1 Martin J. Bligh
2002-03-05 6:33 ` 2.4.19pre1aa1 Samuel Ortiz
2002-03-05 3:05 ` 2.4.19pre1aa1 Bill Davidsen
2002-03-05 8:35 ` 2.4.19pre1aa1 arjan
2002-03-05 12:41 ` 2.4.19pre1aa1 Rik van Riel
2002-03-05 15:10 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-05 16:57 ` 2.4.19pre1aa1 Rik van Riel
2002-03-05 18:26 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-05 18:30 ` 2.4.19pre1aa1 Arjan van de Ven
2002-03-05 19:12 ` 2.4.19pre1aa1 Andrew Morton
2002-03-05 23:03 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-05 23:05 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-05 23:24 ` 2.4.19pre1aa1 Andrew Morton
2002-03-05 23:37 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-05 23:51 ` 2.4.19pre1aa1 Andrew Morton
2002-03-06 0:09 ` 2.4.19pre1aa1 Daniel Phillips
2002-03-05 14:55 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-05 5:38 ` 2.4.19pre1aa1 Martin J. Bligh
2002-03-05 6:45 ` 2.4.19pre1aa1 David Lang
2002-03-04 18:41 ` 2.4.19pre1aa1 Stephan von Krawczynski
2002-03-04 18:46 ` 2.4.19pre1aa1 Martin J. Bligh
2002-03-04 22:06 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-04 23:03 ` 2.4.19pre1aa1 Samuel Ortiz
2002-03-05 0:12 ` 2.4.19pre1aa1 Rik van Riel
2002-03-05 6:21 ` 2.4.19pre1aa1 Martin J. Bligh
2002-03-05 11:23 ` 2.4.19pre1aa1 Stephan von Krawczynski
2002-03-05 17:35 ` 2.4.19pre1aa1 Samuel Ortiz
2002-03-04 18:56 ` 2.4.19pre1aa1 Martin J. Bligh
2002-03-04 22:25 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-04 23:09 ` 2.4.19pre1aa1 Gerrit Huizenga
2002-03-05 0:19 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-05 2:00 ` 2.4.19pre1aa1 Gerrit Huizenga
2002-03-04 22:38 ` 2.4.19pre1aa1 Daniel Phillips
2002-03-04 18:18 ` 2.4.19pre1aa1 Stephan von Krawczynski
2002-03-04 21:37 ` 2.4.19pre1aa1 Rik van Riel
[not found] ` <200203021958.g22JwKq08818@Port.imtp.ilyichevsk.odessa.ua>
2002-03-02 20:47 ` 2.4.19pre1aa1 Andrea Arcangeli
2002-03-02 20:58 ` 2.4.19pre1aa1 Robert Love
2002-03-05 22:16 ` 2.4.19pre1aa1 Bill Davidsen
2002-02-28 2:57 2.4.19pre1aa1 rwhron
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=E16hj3N-0000SY-00@starship.berlin \
--to=phillips@bonn-fries.net \
--cc=andrea@suse.de \
--cc=davidsen@tmr.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mfedyk@matchmail.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
all inboxes | Powered by JetHome®