linux-kernel - Re: get_vmalloc_info() and /proc/meminfo insanely expensive

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]

Message-ID: <CA+55aFxzOAx7365Mx2o55TZOS+bZGh_Pfr=vVF3QGg0btsDumg@mail.gmail.com>
Date:	Thu, 13 Aug 2015 11:32:52 -0700
From:	Linus Torvalds <torvalds@...ux-foundation.org>
To:	Andrew Morton <akpm@...ux-foundation.org>
Cc:	Joonsoo Kim <iamjoonsoo.kim@....com>,
	Al Viro <viro@...iv.linux.org.uk>,
	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: get_vmalloc_info() and /proc/meminfo insanely expensive

On Wed, Aug 12, 2015 at 10:52 PM, Linus Torvalds
<torvalds@...ux-foundation.org> wrote:
>
> I get the feeling this file should be rewritten. But that's not going
> to happen. The "let's just cache the last value for one jiffy" seemed
> to be the minimal fixup to it.

Here's a totally untested patch (I'll reboot and test soon - it does
at least compile for me).

Notice the total lack of locking, which means that it's fundamentally
racy. I really can't find it inside myself to care. Introducing those
vmalloc fields was a mistake to begin with, any races here are "ok, we
get values that were valid at some point, but it might have been a
second ago".

And I also can't find it in myself to care about the "on 32-bit,
jiffies wraps in 49 days if HZ is 1000". If somebody carefully avoids
ever reading /proc/meminfo for 49 days, and then reads it in _just_
the right second, and gets a really stale value, I'm just going to do
my honey badger impression.

Because we really shouldn't have added the vmalloc information to
/proc/meminfo to begin with, and nobody ever cares about those values
anyway.

Comments?

                Linus

View attachment "vmalloc.patch" of type "text/x-patch" (2253 bytes)