netdev - Re: [RFC net-next 9/9] xfrm: add a small xdst pcpu cache

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: <20170629131758.GE9307@breakpoint.cc>
Date:   Thu, 29 Jun 2017 15:17:58 +0200
From:   Florian Westphal <fw@...len.de>
To:     Ilan Tayari <ilant@...lanox.com>
Cc:     Florian Westphal <fw@...len.de>,
        "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        Yossi Kuperman <yossiku@...lanox.com>,
        Steffen Klassert <steffen.klassert@...unet.com>
Subject: Re: [RFC net-next 9/9] xfrm: add a small xdst pcpu cache

Ilan Tayari <ilant@...lanox.com> wrote:
> > -----Original Message-----
> > From: netdev-owner@...r.kernel.org [mailto:netdev-owner@...r.kernel.org]
> > Subject: [RFC net-next 9/9] xfrm: add a small xdst pcpu cache
> > 
> > retain last used xfrm_dst in a pcpu cache.
> > On next request, reuse this dst if the policies are the same.
> > 
> > UDP tests used 64byte packets, tests ran for one minute each,
> > value is average over ten iterations.
> 
> Hi Florian,
> 
> I want to give this a go with hw-offload and see the impact on performance.
> It may take us a few days to do that.

Sure, take your time, thanks for testing!

> > +void xfrm_policy_dev_unreg(void)
> 
> Maybe name it xfrm_policy_cache_flush() or something similar, and call it from some places where xfrm_garbage_collect() used to be called?
> 
> Such as from xfrm_policy_flush()
> And maybe even from xfrm_flush_sa() as well
> 
> This would allow to unload esp4 and/or esp4_offload (or other algo module) after 'ip x s f' (or the swan equivalent)

Good point.  I did not consider module unload just device removal.