From: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
To: Johann Baudy <johaahn@gmail.com>
Cc: netdev@vger.kernel.org, "Ulisses Alonso Camaró" <uaca@alumni.uv.es>
Subject: Re: Packet mmap: TX RING and zero copy
Date: Tue, 2 Sep 2008 23:46:04 +0400 [thread overview]
Message-ID: <20080902194603.GA2825@2ka.mipt.ru> (raw)
In-Reply-To: <7e0dd21a0809021127n35cd38d7m98be5ee5d7158788@mail.gmail.com>
Hi Johann.
On Tue, Sep 02, 2008 at 08:27:36PM +0200, Johann Baudy (johaahn@gmail.com) wrote:
> I've made lot of tests, playing with jumbo frames, raw sockets, ...
> I've never exceeded ~25Mbytes/s. So I've decided to analyze deeply the
> packet socket transmission process.
>
> The main blocking point was the memcpy_fromiovec() function that is
> located in the packet_sendmsg() of af_packet.c.
Can you saturate the link with usual tcp/udp socket?
> But, I would like to get as much criticism as possible in order to
> start a discussion with experts about a conceivable way to mix
> zero-copy, sk_buff management and packet socket.
> Which is perhaps impossible with current network kernel flow ...
Did you try vmsplice and splice?
It is the preferred way to do a zero-copy.
--
Evgeniy Polyakov
next prev parent reply other threads:[~2008-09-02 19:49 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-09-02 18:27 Packet mmap: TX RING and zero copy Johann Baudy
2008-09-02 19:46 ` Evgeniy Polyakov [this message]
2008-09-03 7:56 ` Johann Baudy
2008-09-03 10:38 ` Johann Baudy
2008-09-03 11:06 ` David Miller
2008-09-03 13:05 ` Johann Baudy
2008-09-03 13:27 ` Evgeniy Polyakov
2008-09-03 14:57 ` Christoph Lameter
2008-09-03 15:00 ` Johann Baudy
2008-09-03 15:13 ` Evgeniy Polyakov
2008-09-03 15:58 ` Johann Baudy
2008-09-03 16:43 ` Evgeniy Polyakov
2008-09-03 20:30 ` Johann Baudy
2008-09-03 22:03 ` Evgeniy Polyakov
2008-09-04 14:44 ` Johann Baudy
2008-09-05 7:17 ` Evgeniy Polyakov
[not found] ` <7e0dd21a0809050216r65b8f08fm1ad0630790a13a54@mail.gmail.com>
2008-09-05 9:17 ` Fwd: " Johann Baudy
2008-09-05 11:31 ` Evgeniy Polyakov
2008-09-05 12:44 ` Johann Baudy
2008-09-05 13:16 ` Evgeniy Polyakov
2008-09-05 13:29 ` Johann Baudy
2008-09-05 13:37 ` Evgeniy Polyakov
2008-09-05 13:55 ` Johann Baudy
2008-09-05 14:19 ` Evgeniy Polyakov
2008-09-05 14:45 ` Johann Baudy
2008-09-05 14:59 ` Evgeniy Polyakov
2008-09-05 15:30 ` Johann Baudy
2008-09-05 15:38 ` Evgeniy Polyakov
2008-09-05 16:01 ` Johann Baudy
2008-09-05 16:34 ` Evgeniy Polyakov
2008-09-08 10:21 ` Johann Baudy
2008-09-08 11:26 ` Evgeniy Polyakov
2008-09-08 13:01 ` Johann Baudy
2008-09-08 15:28 ` Evgeniy Polyakov
2008-09-08 15:38 ` Evgeniy Polyakov
2008-09-09 23:11 ` Johann Baudy
2008-09-10 6:09 ` Evgeniy Polyakov
2008-09-05 10:28 ` Robert Iakobashvili
2008-09-05 13:06 ` Johann Baudy
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20080902194603.GA2825@2ka.mipt.ru \
--to=johnpol@2ka.mipt.ru \
--cc=johaahn@gmail.com \
--cc=netdev@vger.kernel.org \
--cc=uaca@alumni.uv.es \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).