From: Stephen Hemminger <shemminger@linux-foundation.org>
To: Ben Greear <greearb@candelatech.com>
Cc: netdev@vger.kernel.org, bridge@linux-foundation.org,
David Miller <davem@davemloft.net>
Subject: Re: [Bridge] [PATCH] bridge: avoid ptype_all packet handling
Date: Wed, 28 Feb 2007 19:56:11 -0800 [thread overview]
Message-ID: <20070228195611.24cf19ee@deepthought> (raw)
In-Reply-To: <45E62C29.7010203@candelatech.com>
On Wed, 28 Feb 2007 17:28:09 -0800
Ben Greear <greearb@candelatech.com> wrote:
> Stephen Hemminger wrote:
> > I was measuring bridging/routing performance and noticed this.
> >
> > The current code runs the "all packet" type handlers before calling
> > the bridge hook. If an application (like some DHCP clients) is
> > using AF_PACKET, this means that each received packet gets run
> > through the Berkeley Packet Filter code in sk_run_filter (slow).
> >
> > By moving the bridging hook to run first, the packets flowing
> > through the bridge get filtered out there. This results in a 14%
> > improvement in performance, but it does mean that some snooping
> > applications would miss packets if being used on a bridge. The
> > correct way to see all packets on a bridge is to set the bridge
> > pseudo-device to promiscuous mode.
>
> Seems it would be better to fix these clients to be more selective as
> to where they bind.
The problem is any use of BPF is a lose, if it has to be done to all
traffic.
> This breaks the case where you want to see packets on a particular
> interface, not just the entire bridge, right?
It might be possible to use promisc counter to handle this.
WARNING: multiple messages have this Message-ID (diff)
From: Stephen Hemminger <shemminger@linux-foundation.org>
To: Ben Greear <greearb@candelatech.com>
Cc: David Miller <davem@davemloft.net>,
bridge@linux-foundation.org, netdev@vger.kernel.org
Subject: Re: [PATCH] bridge: avoid ptype_all packet handling
Date: Wed, 28 Feb 2007 19:56:11 -0800 [thread overview]
Message-ID: <20070228195611.24cf19ee@deepthought> (raw)
In-Reply-To: <45E62C29.7010203@candelatech.com>
On Wed, 28 Feb 2007 17:28:09 -0800
Ben Greear <greearb@candelatech.com> wrote:
> Stephen Hemminger wrote:
> > I was measuring bridging/routing performance and noticed this.
> >
> > The current code runs the "all packet" type handlers before calling
> > the bridge hook. If an application (like some DHCP clients) is
> > using AF_PACKET, this means that each received packet gets run
> > through the Berkeley Packet Filter code in sk_run_filter (slow).
> >
> > By moving the bridging hook to run first, the packets flowing
> > through the bridge get filtered out there. This results in a 14%
> > improvement in performance, but it does mean that some snooping
> > applications would miss packets if being used on a bridge. The
> > correct way to see all packets on a bridge is to set the bridge
> > pseudo-device to promiscuous mode.
>
> Seems it would be better to fix these clients to be more selective as
> to where they bind.
The problem is any use of BPF is a lose, if it has to be done to all
traffic.
> This breaks the case where you want to see packets on a particular
> interface, not just the entire bridge, right?
It might be possible to use promisc counter to handle this.
next prev parent reply other threads:[~2007-03-01 3:56 UTC|newest]
Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-03-01 1:18 [Bridge] [PATCH] bridge: avoid ptype_all packet handling Stephen Hemminger
2007-03-01 1:18 ` Stephen Hemminger
2007-03-01 1:28 ` [Bridge] " Ben Greear
2007-03-01 1:28 ` Ben Greear
2007-03-01 3:56 ` Stephen Hemminger [this message]
2007-03-01 3:56 ` Stephen Hemminger
2007-03-01 4:05 ` [Bridge] " Ben Greear
2007-03-01 4:05 ` Ben Greear
2007-03-01 7:04 ` [Bridge] " Stephen Hemminger
2007-03-01 7:04 ` Stephen Hemminger
2007-03-01 7:22 ` [Bridge] " David Miller
2007-03-01 7:22 ` David Miller
2007-03-01 7:26 ` [Bridge] " Stephen Hemminger
2007-03-01 7:26 ` Stephen Hemminger
2007-03-01 7:30 ` [Bridge] " David Miller
2007-03-01 7:30 ` David Miller
2007-03-01 11:47 ` [Bridge] " jamal
2007-03-01 11:47 ` jamal
2007-03-03 2:14 ` [Bridge] " Andi Kleen
2007-03-03 2:14 ` Andi Kleen
2007-03-03 4:22 ` [Bridge] " David Miller
2007-03-03 4:22 ` David Miller
2007-03-03 7:09 ` [Bridge] " Stephen Hemminger
2007-03-03 7:09 ` Stephen Hemminger
2007-03-03 12:30 ` [Bridge] " Andi Kleen
2007-03-03 12:30 ` Andi Kleen
2007-03-02 21:26 ` [Bridge] " David Miller
2007-03-02 21:26 ` David Miller
2007-03-02 22:09 ` [Bridge] [RFC 1/2] " Stephen Hemminger
2007-03-02 22:09 ` Stephen Hemminger
2007-03-02 22:14 ` [Bridge] [RFC 2/2] bridge: per device promiscious taps Stephen Hemminger
2007-03-02 22:14 ` Stephen Hemminger
2007-03-02 22:48 ` [Bridge] [RFC 1/2] bridge: avoid ptype_all packet handling David Miller
2007-03-02 22:48 ` David Miller
2007-03-02 23:18 ` [Bridge] " David Miller
2007-03-02 23:18 ` David Miller
2007-03-02 23:34 ` [Bridge] " Stephen Hemminger
2007-03-02 23:34 ` Stephen Hemminger
2007-03-02 23:41 ` [Bridge] " David Miller
2007-03-02 23:41 ` David Miller
2007-03-03 5:38 ` [Bridge] " Herbert Xu
2007-03-03 5:38 ` Herbert Xu
2007-03-03 5:59 ` [Bridge] " David Miller
2007-03-03 5:59 ` David Miller
2007-03-03 6:42 ` [Bridge] " Herbert Xu
2007-03-03 6:42 ` Herbert Xu
2007-03-02 22:15 ` [Bridge] " Stephen Hemminger
2007-03-02 22:15 ` Stephen Hemminger
2007-03-03 12:04 ` [Bridge] [PATCH] " Stefan Rompf
2007-03-03 12:04 ` Stefan Rompf
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=20070228195611.24cf19ee@deepthought \
--to=shemminger@linux-foundation.org \
--cc=bridge@linux-foundation.org \
--cc=davem@davemloft.net \
--cc=greearb@candelatech.com \
--cc=netdev@vger.kernel.org \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.