From: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
To: Arnd Hannemann <arnd@arndnet.de>
Cc: linux-kernel@vger.kernel.org, netdev@vger.kernel.org, olel@ans.pl
Subject: Re: problems with e1000 and jumboframes
Date: Thu, 3 Aug 2006 22:29:11 +0400 [thread overview]
Message-ID: <20060803182911.GA8692@2ka.mipt.ru> (raw)
In-Reply-To: <44D23BC3.7040707@arndnet.de>
On Thu, Aug 03, 2006 at 08:09:07PM +0200, Arnd Hannemann (arnd@arndnet.de) wrote:
> Evgeniy Polyakov schrieb:
> > On Thu, Aug 03, 2006 at 07:16:31PM +0400, Evgeniy Polyakov (johnpol@2ka.mipt.ru) wrote:
> >>>> then skb_alloc adds a little
> >>>> (sizeof(struct skb_shared_info)) at the end, and this ends up
> >>>> in 32k request just for 9k jumbo frame.
> >>> Strange, why this skb_shared_info cannon be added before first alignment?
> >>> And what about smaller frames like 1500, does this driver behave similar
> >>> (first align then add)?
> >> It can be.
> >> Could attached (completely untested) patch help?
> >
> > Actually this patch will not help, this new one could.
> >
>
> I applied the attached pachted. And got this output:
>
> > Intel(R) PRO/1000 Network Driver - bufsz 13762
> > Intel(R) PRO/1000 Network Driver - bufsz 16222
> > Intel(R) PRO/1000 Network Driver - bufsz 16058
> > Intel(R) PRO/1000 Network Driver - bufsz 15894
> > Intel(R) PRO/1000 Network Driver - bufsz 15730
> > Intel(R) PRO/1000 Network Driver - bufsz 15566
> > Intel(R) PRO/1000 Network Driver - bufsz 15402
> > Intel(R) PRO/1000 Network Driver - bufsz 15238
> > Intel(R) PRO/1000 Network Driver - bufsz 15074
> > Intel(R) PRO/1000 Network Driver - bufsz 14910
> > Intel(R) PRO/1000 Network Driver - bufsz 14746
> > Intel(R) PRO/1000 Network Driver - bufsz 14582
> > Intel(R) PRO/1000 Network Driver - bufsz 14418
> > Intel(R) PRO/1000 Network Driver - bufsz 14254
> > Intel(R) PRO/1000 Network Driver - bufsz 14090
> > Intel(R) PRO/1000 Network Driver - bufsz 13926
> > Intel(R) PRO/1000 Network Driver - bufsz 13762
> > Intel(R) PRO/1000 Network Driver - bufsz 16222
> > Intel(R) PRO/1000 Network Driver - bufsz 16058
> > Intel(R) PRO/1000 Network Driver - bufsz 15894
> > Intel(R) PRO/1000 Network Driver - bufsz 15730
> > Intel(R) PRO/1000 Network Driver - bufsz 15566
> > Intel(R) PRO/1000 Network Driver - bufsz 15402
> > Intel(R) PRO/1000 Network Driver - bufsz 15238
> > Intel(R) PRO/1000 Network Driver - bufsz 15074
> > Intel(R) PRO/1000 Network Driver - bufsz 14910
> > Intel(R) PRO/1000 Network Driver - bufsz 14746
> > Intel(R) PRO/1000 Network Driver - bufsz 14582
> > Intel(R) PRO/1000 Network Driver - bufsz 14418
> > Intel(R) PRO/1000 Network Driver - bufsz 16222
> > Intel(R) PRO/1000 Network Driver - bufsz 16222
> > Intel(R) PRO/1000 Network Driver - bufsz 16222
> > Intel(R) PRO/1000 Network Driver - bufsz 16222
> > Intel(R) PRO/1000 Network Driver - bufsz 16222
> > Intel(R) PRO/1000 Network Driver - bufsz 16222
> > Intel(R) PRO/1000 Network Driver - bufsz 16222
> > Intel(R) PRO/1000 Network Driver - bufsz 16222
> > Intel(R) PRO/1000 Network Driver - bufsz 16222
> > Intel(R) PRO/1000 Network Driver - bufsz 16222
> > Intel(R) PRO/1000 Network Driver - bufsz 16222
>
> I'm a bit puzzled that there are so much allocations. However the patch
> seems to work. (at least not obviously breaks things for me yet)
Very strange output actually - comments in the code say that frame size
can not exceed 0x3f00, but in this log it is much more than 16128 and
that is after sizeof(struct skb_shared_info) has been removed...
Could you please remove debug output and run some network stress test in
parallel with high disk/memory activity to check if that does not break
your system and watch /proc/slabinfo for 16k and 32k sized pools.
> Best regards
> Arnd
--
Evgeniy Polyakov
next prev parent reply other threads:[~2006-08-03 18:29 UTC|newest]
Thread overview: 38+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-08-03 13:48 problems with e1000 and jumboframes Arnd Hannemann
2006-08-03 13:59 ` Evgeniy Polyakov
2006-08-03 14:37 ` Arnd Hannemann
2006-08-03 15:03 ` Evgeniy Polyakov
2006-08-03 15:08 ` Krzysztof Oledzki
2006-08-03 15:16 ` Evgeniy Polyakov
2006-08-03 15:37 ` Arnd Hannemann
2006-08-03 15:43 ` Evgeniy Polyakov
2006-08-03 15:41 ` Evgeniy Polyakov
2006-08-03 18:09 ` Arnd Hannemann
2006-08-03 18:29 ` Evgeniy Polyakov [this message]
2006-08-03 21:40 ` Arnd Hannemann
2006-08-03 15:57 ` Chris Leech
2006-08-03 16:10 ` Evgeniy Polyakov
2006-08-03 20:32 ` Chris Leech
2006-08-04 6:20 ` Evgeniy Polyakov
2006-08-04 15:16 ` Chris Leech
2006-08-03 16:24 ` Arnd Hannemann
2006-08-03 20:34 ` Chris Leech
2006-08-04 5:59 ` Herbert Xu
2006-08-04 6:15 ` Evgeniy Polyakov
2006-08-04 15:34 ` Chris Leech
2006-08-04 19:42 ` Evgeniy Polyakov
2006-08-04 21:02 ` Jesse Brandeburg
2006-08-05 9:58 ` Evgeniy Polyakov
2006-08-05 10:09 ` Herbert Xu
2006-08-05 10:24 ` Evgeniy Polyakov
2006-08-05 10:33 ` Herbert Xu
2006-08-05 10:41 ` Evgeniy Polyakov
2006-08-03 15:23 ` Evgeniy Polyakov
2006-08-04 5:52 ` Herbert Xu
2006-08-04 5:55 ` David Miller
2006-08-04 5:58 ` Evgeniy Polyakov
2006-08-03 14:24 ` Benjamin LaHaise
2006-08-03 14:49 ` Krzysztof Oledzki
2006-08-03 14:52 ` Benjamin LaHaise
2006-08-03 15:04 ` Krzysztof Oledzki
2006-08-03 15:32 ` Arnd Hannemann
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=20060803182911.GA8692@2ka.mipt.ru \
--to=johnpol@2ka.mipt.ru \
--cc=arnd@arndnet.de \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=olel@ans.pl \
/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).