All of lore.kernel.org
 help / color / mirror / Atom feed
From: Gilad Ben-Yossef <gilad@codefidence.com>
To: Eric Dumazet <eric.dumazet@gmail.com>
Cc: "Ilpo Järvinen" <ilpo.jarvinen@helsinki.fi>,
	Netdev <netdev@vger.kernel.org>,
	"Ori Finkalman" <ori@comsleep.com>
Subject: Re: [PATCH] [RFC] IPv4 TCP fails to send window scale option when window scale is zero
Date: Thu, 01 Oct 2009 11:39:43 +0200	[thread overview]
Message-ID: <4AC478DF.2060401@codefidence.com> (raw)
In-Reply-To: <4AC357D3.7080606@gmail.com>

Eric Dumazet wrote:

>
>>>
>>> Your version slows down the tcp_options_write() function, once per tx packet.
>>>       
>> Are you serious that anding would cost that much? :-/
>>     
>
> Not really :)
>   
LOL I was trying very hard to understand why you thought this was such 
an issue. My head was flying into all sorts of weird directions like 
cache effects and the like... ;-)

<snip>
> Yes, wscale 0 is RFC valid, but are we sure some equipment wont play funny games
> with such value ? At least sending "wscale 1-14" must be working...
>   
Well, there at least used to be routers that would actually zeroed the 
WS value in transit while leaving the option set, but this is another 
issue of course.

Anyway, I know Vista at least does set the window scale TCP option by 
default. One assumes they occasionally send a zero value scale. Not that 
Vista is such a good benchmark to compare Linux to but at least I tend 
to believe the issue would have popped up if it is common enough.

I can craft a patch to introduce a route table option to set TCP window 
scale minimum and maximum sizes, similar to window size route option, if 
you there is a need for that. Personally, I think it is just overkill.
>
> My quick&dirty patch was only for discussion, I have no strong opinion on it,
> only that was on one place to patch instead of two/three/four I dont know yet.
>
> So please Gilad & Ori send us a new patch :)
>
>   
Revised patch follows in next email.

Gilad

   

-- 
Gilad Ben-Yossef
Chief Coffee Drinker & CTO
Codefidence Ltd.

Web:   http://codefidence.com
Cell:  +972-52-8260388
Skype: gilad_codefidence
Tel:   +972-8-9316883 ext. 201
Fax:   +972-8-9316884
Email: gilad@codefidence.com

Check out our Open Source technology and training blog - http://tuxology.net

	"Now the world has gone to bed
	 Darkness won't engulf my head
	 I can see by infra-red
	 How I hate the night."


      reply	other threads:[~2009-10-01  9:39 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-09-29 15:05 [PATCH] [RFC] IPv4 TCP fails to send window scale option when window scale is zero Gilad Ben-Yossef
2009-09-29 17:19 ` Eric Dumazet
2009-09-30  6:28   ` Gilad Ben-Yossef
2009-09-30  7:16     ` Eric Dumazet
2009-09-30 11:42       ` Ilpo Järvinen
2009-09-30 13:06         ` Eric Dumazet
2009-10-01  9:39           ` Gilad Ben-Yossef [this message]

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=4AC478DF.2060401@codefidence.com \
    --to=gilad@codefidence.com \
    --cc=eric.dumazet@gmail.com \
    --cc=ilpo.jarvinen@helsinki.fi \
    --cc=netdev@vger.kernel.org \
    --cc=ori@comsleep.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 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.