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  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date:   Wed, 24 Oct 2018 08:52:08 -0400
From:   Rik van Riel <riel@...riel.com>
To:     Ingo Molnar <mingo@...nel.org>
Cc:     linux-kernel@...r.kernel.org, peterz@...radead.org,
        kernel-team@...com, songliubraving@...com, will.deacon@....com,
        hpa@...or.com, luto@...nel.org, npiggin@...il.com,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        Peter Zijlstra <a.p.zijlstra@...llo.nl>,
        Borislav Petkov <bp@...en8.de>
Subject: Re: [PATCH 0/7] x86/mm/tlb: make lazy TLB mode even lazier

On Wed, 2018-10-24 at 07:53 +0200, Ingo Molnar wrote:
> * Rik van Riel <riel@...riel.com> wrote:
> 
> > The big thing remaining is the reference count overhead of
> > the lazy TLB mm_struct, but getting rid of that is rather a
> > lot of code for a small performance gain. Not quite what
> > Linus asked for :)
> 
> BTW., what would be the plan to improve scalability there,
> is it even possible?

One thing I looked at was shooting down the lazy TLB
MM from the exit code path, that way the MM would no
longer be referenced from any CPU by the time the
memory gets freed.

Once we do that, we no longer need to do any refcounting
for the lazy TLB MM.

-- 
All Rights Reversed.

Download attachment "signature.asc" of type "application/pgp-signature" (489 bytes)

Powered by blists - more mailing lists