netdev - Re: [RFC PATCH bpf-next 0/2] bpf_redirect_map() tail call detection and xdp_do

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] [day] [month] [year] [list]

Message-ID: <20200610142110.25fa5a14@carbon>
Date:   Wed, 10 Jun 2020 14:21:10 +0200
From:   Jesper Dangaard Brouer <brouer@...hat.com>
To:     Toke Høiland-Jørgensen <toke@...hat.com>
Cc:     Björn Töpel <bjorn.topel@...il.com>,
        ast@...nel.org, daniel@...earbox.net, bpf@...r.kernel.org,
        john.fastabend@...il.com, magnus.karlsson@...el.com,
        netdev@...r.kernel.org, maciej.fijalkowski@...el.com,
        bjorn.topel@...el.com, brouer@...hat.com
Subject: Re: [RFC PATCH bpf-next 0/2] bpf_redirect_map() tail call detection
 and xdp_do_redirect() avoidance


> Björn Töpel <bjorn.topel@...il.com> writes:
> 
> > Is this a good idea? I have only measured for AF_XDP redirects, but
> > all XDP_REDIRECT targets should benefit. For AF_XDP the rxdrop
> > scenario went from 21.5 to 23.2 Mpps on my machine.  

Do remember that you are reporting saving 3.4 nanosec (from 21.5 to
23.2 Mpps).  For comparison a function call cost around 1.2 ns, and I
think you have avoided two function calls xdp_do_redirect() and
dev_map_enqueue() (the rest should be inlined by compiler).  Thus, that
alone account for 2.4 ns.

-- 
Best regards,
  Jesper Dangaard Brouer
  MSc.CS, Principal Kernel Engineer at Red Hat
  LinkedIn: http://www.linkedin.com/in/brouer