All of lore.kernel.org
 help / color / mirror / Atom feed
From: Eric Dumazet <eric.dumazet@gmail.com>
To: David Laight <David.Laight@ACULAB.COM>
Cc: "'Arnd Bergmann'" <arnd@arndb.de>,
	"linux-arm-kernel@lists.infradead.org" 
	<linux-arm-kernel@lists.infradead.org>,
	Jonas Jensen <jonas.jensen@gmail.com>,
	"netdev@vger.kernel.org" <netdev@vger.kernel.org>,
	"f.fainelli@gmail.com" <f.fainelli@gmail.com>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"mirqus@gmail.com" <mirqus@gmail.com>,
	"davem@davemloft.net" <davem@davemloft.net>
Subject: RE: [PATCH v6 2/4] net: moxa: replace build_skb() with netdev_alloc_skb_ip_align() / memcpy()
Date: Tue, 26 Aug 2014 03:55:30 -0700	[thread overview]
Message-ID: <1409050530.3173.12.camel@edumazet-glaptop2.roam.corp.google.com> (raw)
In-Reply-To: <063D6719AE5E284EB5DD2968C1650D6D1748126F@AcuExch.aculab.com>

On Tue, 2014-08-26 at 09:10 +0000, David Laight wrote:
> From: Arnd Bergmann

> > While this seems correct, I wonder why you don't do the normal approach of
> > dequeuing the skb from the chain and adding a newly allocated skb to it to
> > save the memcpy.
> 
> Because the receive buffer area isn't made of skbs.
> Post-allocating the skb also reduces the 'true size' of the skb.

This strategy assumes this is not a 10Gbe NIC.

We try to avoid copies because they are generally not needed.

Wifi devices are usually slow, and packet losses are more frequent, so
the copybreak gives better chance to not doing the collapses [1] later
in the TCP stack.

[1] collapses : reducing skb overhead (skb->len / skb->truesize ratio)



WARNING: multiple messages have this Message-ID (diff)
From: eric.dumazet@gmail.com (Eric Dumazet)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH v6 2/4] net: moxa: replace build_skb() with netdev_alloc_skb_ip_align() / memcpy()
Date: Tue, 26 Aug 2014 03:55:30 -0700	[thread overview]
Message-ID: <1409050530.3173.12.camel@edumazet-glaptop2.roam.corp.google.com> (raw)
In-Reply-To: <063D6719AE5E284EB5DD2968C1650D6D1748126F@AcuExch.aculab.com>

On Tue, 2014-08-26 at 09:10 +0000, David Laight wrote:
> From: Arnd Bergmann

> > While this seems correct, I wonder why you don't do the normal approach of
> > dequeuing the skb from the chain and adding a newly allocated skb to it to
> > save the memcpy.
> 
> Because the receive buffer area isn't made of skbs.
> Post-allocating the skb also reduces the 'true size' of the skb.

This strategy assumes this is not a 10Gbe NIC.

We try to avoid copies because they are generally not needed.

Wifi devices are usually slow, and packet losses are more frequent, so
the copybreak gives better chance to not doing the collapses [1] later
in the TCP stack.

[1] collapses : reducing skb overhead (skb->len / skb->truesize ratio)

  reply	other threads:[~2014-08-26 10:55 UTC|newest]

Thread overview: 41+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-08-19 14:49 [PATCH v4 1/2] net: moxa: clear TX descriptor length bits Jonas Jensen
2014-08-19 14:49 ` Jonas Jensen
2014-08-19 14:49 ` [PATCH v4 2/2] net: moxa: replace build_skb() with netdev_alloc_skb_ip_align() / memcpy() Jonas Jensen
2014-08-19 14:49   ` Jonas Jensen
2014-08-19 18:31   ` Eric Dumazet
2014-08-19 18:31     ` Eric Dumazet
2014-08-20 14:24     ` Jonas Jensen
2014-08-20 14:24       ` Jonas Jensen
2014-08-20 14:24       ` Jonas Jensen
2014-08-20 14:19   ` [PATCH v5 " Jonas Jensen
2014-08-20 14:19     ` Jonas Jensen
2014-08-25 14:22     ` [PATCH v6 2/4] " Jonas Jensen
2014-08-25 14:22       ` Jonas Jensen
2014-08-26  0:26       ` David Miller
2014-08-26  0:26         ` David Miller
2014-08-26  9:04       ` Arnd Bergmann
2014-08-26  9:04         ` Arnd Bergmann
2014-08-26  9:10         ` David Laight
2014-08-26  9:10           ` David Laight
2014-08-26  9:10           ` David Laight
2014-08-26 10:55           ` Eric Dumazet [this message]
2014-08-26 10:55             ` Eric Dumazet
2014-08-26 10:55             ` Eric Dumazet
2014-08-21 21:43   ` [PATCH v4 2/2] " Michał Mirosław
2014-08-21 21:43     ` Michał Mirosław
2014-08-21 21:43     ` Michał Mirosław
2014-08-25 14:23     ` Jonas Jensen
2014-08-25 14:23       ` Jonas Jensen
2014-08-25 14:23       ` Jonas Jensen
2014-08-20 14:18 ` [PATCH v5 1/2] net: moxa: clear TX descriptor length bits Jonas Jensen
2014-08-20 14:18   ` Jonas Jensen
2014-08-20 17:10   ` Eric Dumazet
2014-08-20 17:10     ` Eric Dumazet
2014-08-25 14:23     ` Jonas Jensen
2014-08-25 14:23       ` Jonas Jensen
2014-08-22  4:39   ` David Miller
2014-08-22  4:39     ` David Miller
2014-08-25 14:22   ` [PATCH v6 1/4] net: moxa: clear DESC1 on ndo_start_xmit() Jonas Jensen
2014-08-25 14:22     ` Jonas Jensen
2014-08-26  0:25     ` David Miller
2014-08-26  0:25       ` 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=1409050530.3173.12.camel@edumazet-glaptop2.roam.corp.google.com \
    --to=eric.dumazet@gmail.com \
    --cc=David.Laight@ACULAB.COM \
    --cc=arnd@arndb.de \
    --cc=davem@davemloft.net \
    --cc=f.fainelli@gmail.com \
    --cc=jonas.jensen@gmail.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mirqus@gmail.com \
    --cc=netdev@vger.kernel.org \
    /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.