linux-kernel - Re: [announce] vfs-scale git tree update

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:	Fri, 7 Jan 2011 11:09:53 +1100
From:	Nick Piggin <npiggin@...il.com>
To:	Jongman Heo <jongman.heo@...il.com>
Cc:	linux-kernel@...r.kernel.org, Chris Ball <cjb@...top.org>,
	Linus Torvalds <torvalds@...ux-foundation.org>
Subject: Re: [announce] vfs-scale git tree update

On Thu, Jan 6, 2011 at 1:12 PM, Jongman Heo <jongman.heo@...il.com> wrote:
> Nick Piggin <npiggin <at> kernel.dk> writes:
>
>>
>> Git tree is here:
>>
>>   git://git.kernel.org/pub/scm/linux/kernel/git/npiggin/linux-npiggin.git
>>
>> Branch is:
>>
>>     vfs-scale-working
>>
>
> I've observed issue with this vfs-scaling patch.
>
> It makes the boot time of my Fedora14 on vmware much slower, while Ubuntu
> (native, not on vmware) has no problem.

Chris also noticed the problem on native hardware, so it looks like it might be
specific to redhat.

http://marc.info/?l=linux-kernel&m=129287788825563&w=2

It's interesting because the rcu-walk path lookup failure even in the worst case
that everything blows up in our face, should be a barely measurable hit -- no
more than about 2x slowdown on something as fast as a path lookup.

So I suspect it is a bug in the sysfs .d_revalidate_rcu conversion. It is a bit
nasty and needs some cleanup anyway.

So what I will do is split those conversions properly so the core code is done
first, and the filesystem conversions done individually after that for better
bisecting. I'll also drop the sysfs conversion (it was mainly useful for my own
testing anyway, and not so performance critical).

I might ping you again to test the sysfs patch if I get keen, after the dust of
this merge window settles :)

Thanks,
Nick
--
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/