From: Pablo Neira <pablo@eurodev.net>
To: Herbert Xu <herbert@gondor.apana.org.au>
Cc: "David S. Miller" <davem@redhat.com>, jamal <hadi@cyberus.ca>,
netdev@oss.sgi.com
Subject: Re: [PATCH] Improve behaviour of Netlink Sockets
Date: Mon, 20 Sep 2004 02:31:32 +0200 [thread overview]
Message-ID: <414E24E4.1020104@eurodev.net> (raw)
In-Reply-To: <20040919234445.GC10124@gondor.apana.org.au>
Hi Herbert,
Herbert Xu wrote:
>On Mon, Sep 20, 2004 at 12:49:10AM +0200, Pablo Neira wrote:
>
>
>>Sorry, if i repeat things, I'm not willing to be annoying, just clarifying.
>>
>>
>
>You're alright. We're on the same side :)
>
>
that's fine :)
>There is no point for the kernel to wait at all. For unicast sockets
>used for sending commands to the kernel, you should never get overruns
>if you are reading your responses correctly and have set the queue size
>correctly.
>
>
sure, not for commands. When I talk about netlink sockets, I'm not
focused on commands because what they do currently is quite enough for
them I think. I know that commands are the main application of netlink
sockets but there are many others.
>>b) A module needs to send a lot of information from kernel space to user
>>space, if buffer gets full quickly, buffer overruns and
>>netlink_unicast/broadcast never wait, so they drop packets.
>>
>>
>
>I agree that this may be a problem for some, but I don't see how your
>patch addresses it. For the case of ip_queue the kernel simply cannot
>wait since it just creates a queue through the backdoor. You might as
>well just increase your receive queue length.
>
>
I agree, ip_queue case is kind of complex. Well, think about a logging
tool for packets in kernel space that sends messages via netlink when a
packet matches a condition. Now, we receive 300 packets in a very very
short period of time (with the default buffer size) and after that
everything' gets calm again, that is, there's nothing to send to user space.
In that case, netlink buffer will surely overrun, so those 300 messages
will be drop because kernel didn't wait a bit. This is what happens now,
I can reproduce this with my tool.
Well, if system work load is high (in term of sending netlink messages),
we don't have too many things to do...
regards,
Pablo
next prev parent reply other threads:[~2004-09-20 0:31 UTC|newest]
Thread overview: 96+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <412DF807.2040703@eurodev.net>
[not found] ` <20040826141407.38b56729.davem@redhat.com>
[not found] ` <412EB40A.6010100@eurodev.net>
[not found] ` <20040826214710.5e322f1a.davem@redhat.com>
[not found] ` <412F1269.8090303@eurodev.net>
[not found] ` <20040827172736.543dbd54.davem@redhat.com>
2004-08-30 0:37 ` [PATCH] Improve behaviour of Netlink Sockets Pablo Neira
2004-08-31 0:20 ` David S. Miller
2004-08-31 16:37 ` Pablo Neira
2004-08-31 20:16 ` David S. Miller
2004-09-18 10:25 ` Herbert Xu
2004-09-19 4:36 ` Pablo Neira
2004-09-19 5:18 ` Pablo Neira
2004-09-19 7:58 ` Herbert Xu
2004-09-19 12:02 ` Herbert Xu
2004-09-19 12:07 ` Herbert Xu
2004-09-19 20:50 ` Pablo Neira
2004-09-19 21:53 ` Herbert Xu
2004-09-19 22:49 ` Pablo Neira
2004-09-19 23:44 ` Herbert Xu
2004-09-20 0:31 ` Pablo Neira [this message]
2004-09-20 1:00 ` Herbert Xu
2004-09-19 20:50 ` Pablo Neira
2004-09-19 21:59 ` Herbert Xu
2004-09-19 22:39 ` jamal
2004-09-19 22:55 ` Pablo Neira
2004-09-19 23:04 ` jamal
2004-09-19 23:10 ` Herbert Xu
2004-09-19 23:17 ` Herbert Xu
2004-09-20 2:39 ` jamal
2004-09-20 2:58 ` Herbert Xu
2004-09-20 12:34 ` jamal
2004-09-20 18:14 ` Pablo Neira
2004-09-20 21:59 ` Herbert Xu
2004-09-21 11:47 ` jamal
2004-09-21 12:09 ` Herbert Xu
2004-09-22 0:05 ` Herbert Xu
2004-09-22 0:24 ` Pablo Neira
2004-09-22 2:48 ` Pablo Neira
2004-09-22 2:50 ` David S. Miller
2004-09-22 2:53 ` jamal
2004-09-22 3:46 ` Herbert Xu
2004-09-22 11:35 ` jamal
2004-09-23 12:05 ` Herbert Xu
2004-09-24 2:56 ` jamal
2004-09-24 3:20 ` Herbert Xu
2004-09-27 12:41 ` jamal
2004-09-22 4:52 ` Herbert Xu
2004-09-22 12:08 ` jamal
2004-09-22 17:52 ` David S. Miller
2004-09-23 15:40 ` Pablo Neira
2004-09-23 19:16 ` David S. Miller
2004-09-24 3:28 ` Herbert Xu
2004-09-24 5:39 ` David S. Miller
2004-09-24 6:26 ` Herbert Xu
2004-09-24 17:58 ` David S. Miller
2004-09-24 22:06 ` Herbert Xu
2004-09-24 22:28 ` Herbert Xu
2004-09-27 12:59 ` jamal
2004-09-27 12:53 ` jamal
2004-09-23 12:07 ` Herbert Xu
2004-09-24 1:19 ` Pablo Neira
2004-09-24 3:04 ` jamal
2004-09-24 3:24 ` Herbert Xu
2004-09-27 12:46 ` jamal
2004-09-27 21:36 ` Herbert Xu
2004-09-28 2:43 ` jamal
2004-09-28 2:46 ` Herbert Xu
2004-09-28 3:06 ` jamal
2004-09-28 3:23 ` Herbert Xu
2004-09-28 3:45 ` jamal
2004-09-28 3:59 ` Herbert Xu
2004-09-28 10:36 ` jamal
2004-09-28 11:11 ` Herbert Xu
2004-09-28 12:16 ` Herbert Xu
2004-09-28 12:39 ` On DaveMs congestion control algorithm WAS(Re: " jamal
2004-09-28 18:24 ` David S. Miller
2004-09-29 2:23 ` jamal
2004-09-28 12:32 ` jamal
2004-09-29 0:13 ` Herbert Xu
2004-09-29 2:52 ` jamal
2004-09-29 3:27 ` Herbert Xu
2004-09-29 4:02 ` David S. Miller
2004-09-29 10:50 ` jamal
2004-09-29 11:42 ` Herbert Xu
2004-09-29 13:55 ` jamal
2004-09-28 21:07 ` Pablo Neira
2004-09-28 23:19 ` Herbert Xu
2004-09-28 23:20 ` David S. Miller
2004-09-29 2:28 ` jamal
2004-09-29 2:30 ` Herbert Xu
2004-09-29 2:59 ` jamal
2004-09-22 2:57 ` jamal
2004-09-22 3:39 ` Herbert Xu
2004-09-19 20:47 ` Pablo Neira
2004-09-19 21:20 ` Herbert Xu
2004-09-19 22:14 ` Pablo Neira
2004-09-19 23:31 ` Herbert Xu
[not found] ` <E1C90Da-0001V7-00@gondolin.me.apana.org.au>
2004-09-19 12:00 ` Herbert Xu
2004-09-29 9:53 James Chapman
2004-09-29 9:59 ` Herbert Xu
-- strict thread matches above, loose matches on Subject: below --
2004-09-30 9:50 James Chapman
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=414E24E4.1020104@eurodev.net \
--to=pablo@eurodev.net \
--cc=davem@redhat.com \
--cc=hadi@cyberus.ca \
--cc=herbert@gondor.apana.org.au \
--cc=netdev@oss.sgi.com \
/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).