linux-kernel - Re: [PATCH] mm: evict streaming IO cache first

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Mon, 17 Nov 2008 17:19:31 +1100
From:	Nick Piggin <nickpiggin@...oo.com.au>
To:	Andrew Morton <akpm@...ux-foundation.org>
Cc:	Rik van Riel <riel@...hat.com>,
	Linus Torvalds <torvalds@...ux-foundation.org>,
	KOSAKI Motohiro <kosaki.motohiro@...fujitsu.com>,
	LKML <linux-kernel@...r.kernel.org>,
	linux-mm <linux-mm@...ck.org>,
	Gene Heskett <gene.heskett@...il.com>
Subject: Re: [PATCH] mm: evict streaming IO cache first

On Monday 17 November 2008 15:47, Andrew Morton wrote:
> On Sun, 16 Nov 2008 16:20:26 -0500 Rik van Riel <riel@...hat.com> wrote:

> > I will take a look at producing smoother self tuning behaviour
> > in get_scan_ratio(), with logic along these lines:
> > - the more file pages are inactive, the more eviction should
> >    focus on file pages, because we are not eating away at the
> >    working set yet
> > - the more file pages are active, the more there needs to be
> >    a balance between file and anon scanning, because we are
> >    starting to get to the working sets for both
>
> hm.  I wonder if it would be prohibitive to say "hey, we did the wrong
> thing in that scanning pass - rewind and try it again".  Probably it
> would be.
>
> Anyway, we need to do something.
>
> Shouldn't get_scan_ratio() be handling this case already?

I have a patch that was actually for the old vmscan logic that I
found really helps prevent working set get paged out. Actually
the old vmscan behaviour wasn't any good either at preventing
unmapped pagecache from being evicted, which is the main thing I
was trying to fix (eg. keep git tree in cache while doing other
use-once IO).

It ended up working really well, but I suspect it would still fall
over in the case where you were trying to populate your caches with
the git tree *while* the streaming IO is happening (if those pages
don't have a chance to get touched again for a while, they'll look
like use-once IO to the vm)... but still no worse than current
behaviour.

I need to forward port it to the new system, however...

But this would be a pretty big change and I can't see how it could
be appropriate for -rc6. Aren't we allergic to even single-line
changes in vmscan without seemingly multi-year "testing" phases? :)
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/