Netdev List
 help / color / mirror / Atom feed
From: Eric Dumazet <eric.dumazet@gmail.com>
To: Alexander Duyck <alexander.h.duyck@intel.com>
Cc: "Alexander Duyck" <alexander.duyck@gmail.com>,
	"David Miller" <davem@davemloft.net>,
	netdev <netdev@vger.kernel.org>,
	"Neal Cardwell" <ncardwell@google.com>,
	"Tom Herbert" <therbert@google.com>,
	"Jeff Kirsher" <jeffrey.t.kirsher@intel.com>,
	"Michael Chan" <mchan@broadcom.com>,
	"Matt Carlson" <mcarlson@broadcom.com>,
	"Herbert Xu" <herbert@gondor.apana.org.au>,
	"Ben Hutchings" <bhutchings@solarflare.com>,
	"Ilpo Järvinen" <ilpo.jarvinen@helsinki.fi>,
	"Maciej Żenczykowski" <maze@google.com>
Subject: Re: [PATCH 3/4 v2 net-next] net: make GRO aware of skb->head_frag
Date: Wed, 02 May 2012 04:45:03 +0200	[thread overview]
Message-ID: <1335926703.22133.39.camel@edumazet-glaptop> (raw)
In-Reply-To: <4FA03D69.6060907@intel.com>

On Tue, 2012-05-01 at 12:45 -0700, Alexander Duyck wrote:

> I wasn't worried about the kfree vs put_page, I was worried about the
> coalesce case.  However, it looks like you are correct and I am not
> seeing any issues so everything seems to be working fine.
> 
> I have a hacked together ixgbe up and running now with the new build_skb
> logic and RSC/LRO disabled.  It looks like it is giving me a 5%
> performance boost for small packet routing, but I am using more CPU for
> netperf TCP receive tests and I was wondering if you had seen anything
> similar on the tg3 driver?
> 



Hmm, tg3 perf for small packets is not changed, because tg3 uses
copybreak anyway.

#define TG3_RX_COPY_THRESHOLD       256

For big packets, I have no changes in netperf because GRO cannot really
provide big enough skbs at 1Gb speed.

The changes are noticeable when BDP (Bandwidth*Delay Product) is large
and some packet losses happen : You notice lower latencies in TCP stack,
because of less collapses (big copies)

  reply	other threads:[~2012-05-02  2:45 UTC|newest]

Thread overview: 45+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-04-27 10:37 [PATCH 3/4 net-next] net: make GRO aware of skb->head_frag Eric Dumazet
2012-04-30 17:54 ` Eric Dumazet
2012-04-30 18:10 ` [PATCH 3/4 v2 " Eric Dumazet
2012-04-30 23:36   ` Alexander Duyck
2012-05-01  1:27     ` Eric Dumazet
2012-05-01  5:33       ` Alexander Duyck
2012-05-01  6:39         ` Eric Dumazet
2012-05-01 16:17           ` Alexander Duyck
2012-05-01 17:04             ` Eric Dumazet
2012-05-01 19:45               ` Alexander Duyck
2012-05-02  2:45                 ` Eric Dumazet [this message]
2012-05-02  8:24                 ` Eric Dumazet
2012-05-02 16:16                   ` Alexander Duyck
2012-05-02 16:19                     ` Eric Dumazet
2012-05-02 16:27                       ` Eric Dumazet
2012-05-02 17:04                         ` Alexander Duyck
2012-05-02 17:02                       ` Alexander Duyck
2012-05-02 17:16                   ` Rick Jones
2012-05-01 22:58               ` Alexander Duyck
2012-05-01 23:10                 ` Alexander Duyck
2012-05-02  2:47                   ` Eric Dumazet
2012-05-02  3:54                     ` Eric Dumazet
2012-05-02  8:13                     ` [PATCH net-next] net: take care of cloned skbs in tcp_try_coalesce() Eric Dumazet
2012-05-02 15:52                       ` Alexander Duyck
2012-05-02 16:12                         ` Eric Dumazet
2012-05-02 16:27                           ` Alexander Duyck
2012-05-02 16:46                             ` Eric Dumazet
2012-05-02 17:55                               ` [PATCH v2 " Eric Dumazet
2012-05-02 19:58                                 ` [PATCH net-next] net: implement tcp coalescing in tcp_queue_rcv() Eric Dumazet
2012-05-02 20:11                                   ` Joe Perches
2012-05-02 20:23                                     ` Eric Dumazet
2012-05-02 20:34                                       ` Joe Perches
2012-05-03  0:32                                       ` David Miller
2012-05-03  1:11                                   ` David Miller
2012-05-03  2:14                                     ` Eric Dumazet
2012-05-03  2:21                                       ` David Miller
2012-05-03  1:11                                 ` [PATCH v2 net-next] net: take care of cloned skbs in tcp_try_coalesce() David Miller
2012-05-02 18:05                               ` [PATCH " Alexander Duyck
2012-05-02 18:15                                 ` Eric Dumazet
2012-05-02 20:55                                   ` Alexander Duyck
2012-05-03  1:52                                     ` Eric Dumazet
2012-05-03  3:00                                       ` Alexander Duyck
2012-05-03  3:14                                         ` Eric Dumazet
2012-05-03  3:28                                           ` Alexander Duyck
2012-05-01  1:48   ` [PATCH 3/4 v2 net-next] net: make GRO aware of skb->head_frag David Miller

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=1335926703.22133.39.camel@edumazet-glaptop \
    --to=eric.dumazet@gmail.com \
    --cc=alexander.duyck@gmail.com \
    --cc=alexander.h.duyck@intel.com \
    --cc=bhutchings@solarflare.com \
    --cc=davem@davemloft.net \
    --cc=herbert@gondor.apana.org.au \
    --cc=ilpo.jarvinen@helsinki.fi \
    --cc=jeffrey.t.kirsher@intel.com \
    --cc=maze@google.com \
    --cc=mcarlson@broadcom.com \
    --cc=mchan@broadcom.com \
    --cc=ncardwell@google.com \
    --cc=netdev@vger.kernel.org \
    --cc=therbert@google.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