public inbox for netdev@vger.kernel.org 
 help / color / mirror / Atom feed
From: Casey Schaufler <casey@schaufler-ca•com>
To: Eric Dumazet <eric.dumazet@gmail•com>
Cc: David Miller <davem@davemloft•net>,
	pmoore@redhat•com, netdev@vger•kernel.org, mvadkert@redhat•com,
	selinux@tycho•nsa.gov, linux-security-module@vger•kernel.org,
	Casey Schaufler <casey@schaufler-ca•com>
Subject: Re: [PATCH] tcp: assign the sock correctly to an outgoing SYNACK packet
Date: Mon, 08 Apr 2013 17:59:59 -0700	[thread overview]
Message-ID: <5163680F.4030200@schaufler-ca.com> (raw)
In-Reply-To: <1365467636.3887.67.camel@edumazet-glaptop>

On 4/8/2013 5:33 PM, Eric Dumazet wrote:
> On Mon, 2013-04-08 at 16:40 -0700, Casey Schaufler wrote:
>
>> OK, let's do the math.
>>
>> First off, it's 4 bytes, not 8. It replaces the secmark.
>> Your increased memory usage is going to be
>>
>> 	4 bytes/packet *  M packets/second * N seconds
>>
>> Where M is the rate at which you're processing packets and
>> N is the length of time it takes to process a packet.
>>
>> Let's pretend we have an embedded system that does nothing but send
>> 128 byte packets on a 10Gb port. That's 10M packets/second. If it
>> takes a full second to process a packet the overhead is 40MB for that
>> second. I have it on good authority that packets can be processed
>> in considerably less time than that. The real number is more like
>> 0.05 seconds. That means your actual overhead is more like 1MB.
>>
>> These are dumbed down calculations. I am not a memory usage expert.
>> I am convinced that "real" calculations are going to get similar
>> numbers. I am, of course, willing to be swayed by evidence that I
>> am wrong.
>>
>> Compare that to the overhead associated with using CIPSO on packets
>> that never leave the box.
> Maths are not that simple,

I am willing to accept that. I am willing to be
educated. I would be interested to see what the
"real" maths are, and how far off my dumb version
might actually be.

> and its not about size of sk_buff, since the
> number of in-flight skb should be quite small.

The reason we're told that there can't be a blob
pointer in the sk_buff is that it increases the size
of the sk_buff. Yes, it *is* about the size.

> Its the time to init this memory for _every_ packet.

Which is a function of the size, no?

> sizeof(sk_buff) is 0xf8, very close to cross the 256 bytes limit.

0xf8 + 0x4 = 0xfc
248 + 4 = 252

> Add a single _byte_ and it becomes a matter of adding a _cache_ line,
> and thats 25 % cost, assuming 64bytes cache lines.

I don't see that with adding 4 bytes. Again, I'm willing to be
educated if I'm wrong.

> So instead of processing 10M packets per second, we would process 9M
> packets per second, or maybe less.
>
> Yes, 256 bytes per sk_buff, this is the current insane situation.
> (Not counting the struct skb_shared_info, adding at least one additional
> cache line)

Sorry, but I don't see what's insane with either the
current 248 byte sk_buff or a 252 byte sk_buff.

As always, I am willing to be educated.

Thank you.


  reply	other threads:[~2013-04-09  0:59 UTC|newest]

Thread overview: 64+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-04-08 15:45 [PATCH] tcp: assign the sock correctly to an outgoing SYNACK packet Paul Moore
2013-04-08 16:14 ` David Miller
2013-04-08 17:22   ` Paul Moore
2013-04-08 17:36     ` Eric Dumazet
2013-04-08 17:40       ` Paul Moore
2013-04-08 17:47         ` Eric Dumazet
2013-04-08 18:01           ` Eric Dumazet
2013-04-08 18:12           ` Paul Moore
2013-04-08 18:21             ` Eric Dumazet
2013-04-08 18:26               ` Paul Moore
2013-04-08 18:34                 ` Eric Dumazet
2013-04-08 18:30               ` Eric Dumazet
2013-04-08 20:37                 ` Paul Moore
2013-04-08 20:44                   ` David Miller
2013-04-08 20:53                     ` Paul Moore
2013-04-08 20:55                   ` Eric Dumazet
2013-04-08 21:09                     ` Paul Moore
2013-04-08 21:14                       ` David Miller
2013-04-08 21:17                       ` Eric Dumazet
2013-04-09  3:58                       ` [PATCH] selinux: add a skb_owned_by() hook Eric Dumazet
2013-04-09  4:29                         ` Casey Schaufler
2013-04-09  4:41                           ` David Miller
2013-04-09  5:14                             ` Casey Schaufler
2013-04-09 11:39                             ` Paul Moore
2013-04-09  6:24                           ` Eric Dumazet
2013-04-09 11:45                           ` Paul Moore
2013-04-09  7:38                         ` James Morris
2013-04-09 12:06                         ` Paul Moore
2013-04-09 17:23                         ` David Miller
2013-04-08 18:32             ` [PATCH] tcp: assign the sock correctly to an outgoing SYNACK packet Paul Moore
2013-04-08 21:10               ` Paul Moore
2013-04-08 21:15                 ` David Miller
2013-04-08 21:24                   ` Paul Moore
2013-04-08 21:33                     ` David Miller
2013-04-08 22:01                       ` Paul Moore
2013-04-08 22:08                         ` David Miller
2013-04-08 23:40                       ` Casey Schaufler
2013-04-09  0:33                         ` Eric Dumazet
2013-04-09  0:59                           ` Casey Schaufler [this message]
2013-04-09  1:09                             ` Eric Dumazet
2013-04-09  1:24                               ` Casey Schaufler
2013-04-09 13:19                                 ` Paul Moore
2013-04-09 13:33                                   ` Paul Moore
2013-04-09 14:00                                   ` Eric Dumazet
2013-04-09 14:19                                     ` Paul Moore
2013-04-09 14:31                                       ` Eric Dumazet
2013-04-09 14:52                                         ` Paul Moore
2013-04-09 15:05                                           ` Paul Moore
2013-04-09 15:07                                           ` Eric Dumazet
2013-04-09 15:17                                             ` Paul Moore
2013-04-09 15:32                                               ` Eric Dumazet
2013-04-09 15:57                                                 ` Paul Moore
2013-04-09 16:11                                                 ` Casey Schaufler
2013-04-09 16:56                                                 ` David Miller
2013-04-09 17:00                                                   ` Paul Moore
2013-04-09 17:09                                                     ` David Miller
2013-04-09 17:10                                                       ` David Miller
2013-04-09 14:05                                   ` Ben Hutchings
2013-04-09 14:10                                     ` Paul Moore
2013-04-08 21:34                     ` Ben Hutchings
2013-04-08 19:25     ` David Miller
2013-04-08 16:19 ` Eric Dumazet
2013-04-08 18:03 ` Sergei Shtylyov
2013-04-08 18:12   ` Paul Moore

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=5163680F.4030200@schaufler-ca.com \
    --to=casey@schaufler-ca$(echo .)com \
    --cc=davem@davemloft$(echo .)net \
    --cc=eric.dumazet@gmail$(echo .)com \
    --cc=linux-security-module@vger$(echo .)kernel.org \
    --cc=mvadkert@redhat$(echo .)com \
    --cc=netdev@vger$(echo .)kernel.org \
    --cc=pmoore@redhat$(echo .)com \
    --cc=selinux@tycho$(echo .)nsa.gov \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html
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