netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "David S. Miller" <davem@davemloft.net>
To: rusty@rustcorp.com.au
Cc: caitlinb@broadcom.com, johnpol@2ka.mipt.ru, kelly@au1.ibm.com,
	netdev@vger.kernel.org
Subject: Re: [PATCH 1/3] Rough VJ Channel Implementation - vj_core.patch
Date: Fri, 28 Apr 2006 12:21:02 -0700 (PDT)	[thread overview]
Message-ID: <20060428.122102.76435590.davem@davemloft.net> (raw)
In-Reply-To: <1146212660.8029.38.camel@localhost.localdomain>

From: Rusty Russell <rusty@rustcorp.com.au>
Date: Fri, 28 Apr 2006 18:24:08 +1000

> Note that the problem space AFAICT includes strange advanced routing
> setups, ingress qos and possibly others, not just netfilter.  But
> perhaps the same solutions apply, so I'll concentrate on nf.

Yes, this hasn't been mentioned explicitly yet.

The big problem is that we don't want the classifier to become
overly complex.

One scheme I'm thinking about right now is an ordered lookup
that looks like:

1) Check for established sockets, they trump everything else.

2) Check for classifier rules, ie. netfilter and packet scheduler
   stuff

3) Check for listening sockets

4) default channel

#2 is still an unsolved problem, we don't want this big complex
classifier to be required in the hardware implementations.
However, using just IP addresses and ports does not map well to
what netfilter and co. want.

> Ah, this is a different problem.  Our idea was to have a syscall which
> would check & sanitize the buffers for output.  To do this, you need the
> ability to chain buffers (a simple next entry in the header, for us).
> 
> Sanitization would copy the header into a global buffer (ie. not one
> reachable by userspace), check the flowid, and chain on the rest of the
> user buffer.  After it had sanitized the buffers, it would activate the
> NIC, which would only send out buffers which started with a kernel
> buffer.
> 
> Of course, the first step (CAP_NET_RAW-only) wouldn't need this.  And,
> if the "sanitize_and_send" syscall were PF_VJCHAN's write(), then the
> contents of the write() could actually be the header: userspace would
> never deal with chained buffers.

I am not sure any of this is anything more than overhead.

If we just pop the buffers directly into the user mmap()'d ring
buffer, headers and all, and give an offset+length pair so the
user knows where the data starts and how much data is there, it
should all just work out.  Where to put the offset+length is
just a detail.

  reply	other threads:[~2006-04-28 19:22 UTC|newest]

Thread overview: 78+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-04-27 21:12 [PATCH 1/3] Rough VJ Channel Implementation - vj_core.patch Caitlin Bestler
2006-04-28  6:10 ` Evgeniy Polyakov
2006-04-28  7:20   ` David S. Miller
2006-04-28  7:32     ` Evgeniy Polyakov
2006-04-28 18:20       ` David S. Miller
2006-04-28  8:24 ` Rusty Russell
2006-04-28 19:21   ` David S. Miller [this message]
2006-04-28 22:04     ` Rusty Russell
2006-04-28 22:38       ` David S. Miller
2006-04-29  0:10         ` Rusty Russell
  -- strict thread matches above, loose matches on Subject: below --
2006-04-28 23:45 Caitlin Bestler
2006-04-28 17:55 Caitlin Bestler
2006-04-28 22:17 ` Rusty Russell
2006-04-28 22:40   ` David S. Miller
2006-04-29  0:22     ` Rusty Russell
2006-04-29  6:46       ` David S. Miller
2006-04-28 17:02 Caitlin Bestler
2006-04-28 17:18 ` Stephen Hemminger
2006-04-28 17:29   ` Evgeniy Polyakov
2006-04-28 17:41     ` Stephen Hemminger
2006-04-28 17:55       ` Evgeniy Polyakov
2006-04-28 19:16         ` David S. Miller
2006-04-28 19:49           ` Stephen Hemminger
2006-04-28 19:59             ` Evgeniy Polyakov
2006-04-28 22:00               ` David S. Miller
2006-04-29 13:54                 ` Evgeniy Polyakov
     [not found]                 ` <20060429124451.GA19810@2ka.mipt.ru>
2006-05-01 21:32                   ` David S. Miller
2006-05-02  7:08                     ` Evgeniy Polyakov
2006-04-28 19:52           ` Evgeniy Polyakov
2006-04-28 19:10   ` David S. Miller
2006-04-28 20:46     ` Brent Cook
2006-04-28 17:25 ` Evgeniy Polyakov
2006-04-28 19:14   ` David S. Miller
2006-04-28 15:59 Caitlin Bestler
2006-04-28 16:12 ` Evgeniy Polyakov
2006-04-28 19:09   ` David S. Miller
2006-04-27  1:02 Caitlin Bestler
2006-04-27  6:08 ` David S. Miller
2006-04-27  6:17   ` Andi Kleen
2006-04-27  6:27     ` David S. Miller
2006-04-27  6:41       ` Andi Kleen
2006-04-27  7:52         ` David S. Miller
2006-04-26 22:53 Caitlin Bestler
2006-04-26 22:59 ` David S. Miller
2006-04-26 20:20 Caitlin Bestler
2006-04-26 22:35 ` David S. Miller
2006-04-26 19:30 Caitlin Bestler
2006-04-26 19:46 ` Jeff Garzik
2006-04-26 22:40   ` David S. Miller
2006-04-27  3:40 ` Rusty Russell
2006-04-27  4:58   ` James Morris
2006-04-27  6:16     ` David S. Miller
2006-04-27  6:17   ` David S. Miller
2006-04-26 16:57 Caitlin Bestler
2006-04-26 19:23 ` David S. Miller
2006-04-26 11:47 Kelly Daly
2006-04-26  7:33 ` David S. Miller
2006-04-27  3:31   ` Kelly Daly
2006-04-27  6:25     ` David S. Miller
2006-04-27 11:51       ` Evgeniy Polyakov
2006-04-27 20:09         ` David S. Miller
2006-04-28  6:05           ` Evgeniy Polyakov
2006-05-04  2:59       ` Kelly Daly
2006-05-04 23:22         ` David S. Miller
2006-05-05  1:31           ` Rusty Russell
2006-04-26  7:59 ` David S. Miller
2006-05-04  7:28   ` Kelly Daly
2006-05-04 23:11     ` David S. Miller
2006-05-05  2:48       ` Kelly Daly
2006-05-16  1:02         ` Kelly Daly
2006-05-16  1:05           ` David S. Miller
2006-05-16  1:15             ` Kelly Daly
2006-05-16  5:16           ` David S. Miller
2006-06-22  2:05             ` Kelly Daly
2006-06-22  3:58               ` James Morris
2006-06-22  4:31                 ` Arnaldo Carvalho de Melo
2006-06-22  4:36                 ` YOSHIFUJI Hideaki / 吉藤英明
2006-07-08  0:05               ` David Miller

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=20060428.122102.76435590.davem@davemloft.net \
    --to=davem@davemloft.net \
    --cc=caitlinb@broadcom.com \
    --cc=johnpol@2ka.mipt.ru \
    --cc=kelly@au1.ibm.com \
    --cc=netdev@vger.kernel.org \
    --cc=rusty@rustcorp.com.au \
    /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).