netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Randy Dunlap <randy.dunlap@oracle.com>
To: Krishna Kumar <krkumar2@in.ibm.com>
Cc: johnpol@2ka.mipt.ru, jagana@us.ibm.com,
	herbert@gondor.apana.org.au, gaagaan@gmail.com,
	Robert.Olsson@data.slu.se, kumarkr@linux.ibm.com,
	rdreier@cisco.com, peter.p.waskiewicz.jr@intel.com,
	hadi@cyberus.ca, kaber@trash.net, jeff@garzik.org,
	general@lists.openfabrics.org, netdev@vger.kernel.org,
	tgraf@suug.ch, mcarlson@broadcom.com, sri@us.ibm.com,
	shemminger@linux-foundation.org, davem@davemloft.net,
	mchan@broadcom.com
Subject: [ofa-general] Re: [PATCH 1/10 Rev4] [Doc] HOWTO Documentation for batching
Date: Wed, 22 Aug 2007 08:50:13 -0700	[thread overview]
Message-ID: <20070822085013.d20e1fd4.randy.dunlap@oracle.com> (raw)
In-Reply-To: <20070822082858.11964.87377.sendpatchset@localhost.localdomain>

On Wed, 22 Aug 2007 13:58:58 +0530 Krishna Kumar wrote:

> Add Documentation describing batching skb xmit capability.
> 
> Signed-off-by: Krishna Kumar <krkumar2@in.ibm.com>
> ---
>  batching_skb_xmit.txt |   78 ++++++++++++++++++++++++++++++++++++++++++++++++++
>  1 files changed, 78 insertions(+)
> 
> diff -ruNp org/Documentation/networking/batching_skb_xmit.txt new/Documentation/networking/batching_skb_xmit.txt
> --- org/Documentation/networking/batching_skb_xmit.txt	1970-01-01 05:30:00.000000000 +0530
> +++ new/Documentation/networking/batching_skb_xmit.txt	2007-08-22 10:21:19.000000000 +0530
> @@ -0,0 +1,78 @@
> +		 HOWTO for batching skb xmit support
> +		 -----------------------------------
> +
> +Section 1: What is batching skb xmit
> +Section 2: How batching xmit works vs the regular xmit
> +Section 3: How drivers can support batching
> +Section 4: How users can work with batching
> +
> +
> +Introduction: Kernel support for batching skb
> +----------------------------------------------
> +
> +A new capability to support xmit of multiple skbs is provided in the netdevice
> +layer. Drivers which enable this capability should be able to process multiple
> +skbs in a single call to their xmit handler.
> +
> +
> +Section 1: What is batching skb xmit
> +-------------------------------------
> +
> +	This capability is optionally enabled by a driver by setting the
> +	NETIF_F_BATCH_SKBS bit in dev->features. The pre-requisite for a

                                                     prerequisite

> +	driver to use this capability is that it should have a reasonably

	I would say "reasonably-sized".

> +	sized hardware queue that can process multiple skbs.
> +
> +
> +Section 2: How batching xmit works vs the regular xmit
> +-------------------------------------------------------
> +
> +	The network stack gets called from upper layer protocols with a single
> +	skb to transmit. This skb is first enqueue'd and an attempt is made to

                                           enqueued

> +	transmit it immediately (via qdisc_run). However, events like tx lock
> +	contention, tx queue stopped, etc, can result in the skb not getting

                                      etc.,

> +	sent out and it remains in the queue. When the next xmit is called or
> +	when the queue is re-enabled, qdisc_run could potentially find
> +	multiple packets in the queue, and iteratively send them all out
> +	one-by-one.
> +
> +	Batching skb xmit is a mechanism to exploit this situation where all
> +	skbs can be passed in one shot to the device. This reduces driver
> +	processing, locking at the driver (or in stack for ~LLTX drivers)
> +	gets amortized over multiple skbs, and in case of specific drivers
> +	where every xmit results in a completion processing (like IPoIB) -
> +	optimizations can be made in the driver to request a completion for
> +	only the last skb that was sent which results in saving interrupts
> +	for every (but the last) skb that was sent in the same batch.
> +
> +	Batching can result in significant performance gains for systems that
> +	have multiple data stream paths over the same network interface card.
> +
> +
> +Section 3: How drivers can support batching
> +---------------------------------------------
> +
> +	Batching requires the driver to set the NETIF_F_BATCH_SKBS bit in
> +	dev->features.
> +
> +	The driver's xmit handler should be modified to process multiple skbs
> +	instead of one skb. The driver's xmit handler is called either with a

                                                                           an

> +	skb to transmit or NULL skb, where the latter case should be handled
> +	as a call to xmit multiple skbs. This is done by sending out all skbs
> +	in the dev->skb_blist list (where it was added by the core stack).
> +
> +
> +Section 4: How users can work with batching
> +---------------------------------------------
> +
> +	Batching can be disabled for a particular device, e.g. on desktop
> +	systems if only one stream of network activity for that device is
> +	taking place, since performance could be slightly affected due to
> +	extra processing that batching adds (unless packets are getting
> +	sent fast resulting in stopped queue's). Batching can be enabled if

                                       queues).

> +	more than one stream of network activity per device is being done,
> +	e.g. on servers; or even desktop usage with multiple browser, chat,
> +	file transfer sessions, etc.
> +
> +	Per device batching can be enabled/disabled by passing 'on' or 'off'
> +	respectively to ethtool.

	with what other parameter(s), e.g.,

	ethtool <dev> batching on/off ?

---
~Randy
*** Remember to use Documentation/SubmitChecklist when testing your code ***

  reply	other threads:[~2007-08-22 15:50 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-08-22  8:28 [PATCH 0/10 Rev4] Implement skb batching and support in IPoIB Krishna Kumar
2007-08-22  8:28 ` [PATCH 1/10 Rev4] [Doc] HOWTO Documentation for batching Krishna Kumar
2007-08-22 15:50   ` Randy Dunlap [this message]
2007-08-23  2:48     ` Krishna Kumar2
2007-08-22  8:29 ` [PATCH 2/10 Rev4] [core] Add skb_blist & support " Krishna Kumar
2007-08-22  8:29 ` [PATCH 3/10 Rev4] [sched] Modify qdisc_run to support batching Krishna Kumar
2007-08-22  8:29 ` [PATCH 4/10 Rev4] [ethtool] Add ethtool support Krishna Kumar
2007-08-22  8:30 ` [PATCH 5/10 Rev4] [IPoIB] Header file changes Krishna Kumar
2007-08-22  8:30 ` [PATCH 6/10 Rev4] [IPoIB] CM & Multicast changes Krishna Kumar
2007-08-22  8:30 ` [PATCH 7/10 Rev4] [IPoIB] Verbs changes Krishna Kumar
2007-08-22  8:31 ` [PATCH 8/10 Rev4] [IPoIB] Post and work completion handler changes Krishna Kumar
2007-08-22  8:31 ` [PATCH 9/10 Rev4] [IPoIB] Implement batching Krishna Kumar
2007-08-22  8:31 ` [PATCH 10/10 Rev4] [E1000] " Krishna Kumar
2007-08-22 14:39   ` [ofa-general] " Kok, Auke
2007-08-23  2:44     ` Krishna Kumar2

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=20070822085013.d20e1fd4.randy.dunlap@oracle.com \
    --to=randy.dunlap@oracle.com \
    --cc=Robert.Olsson@data.slu.se \
    --cc=davem@davemloft.net \
    --cc=gaagaan@gmail.com \
    --cc=general@lists.openfabrics.org \
    --cc=hadi@cyberus.ca \
    --cc=herbert@gondor.apana.org.au \
    --cc=jagana@us.ibm.com \
    --cc=jeff@garzik.org \
    --cc=johnpol@2ka.mipt.ru \
    --cc=kaber@trash.net \
    --cc=krkumar2@in.ibm.com \
    --cc=kumarkr@linux.ibm.com \
    --cc=mcarlson@broadcom.com \
    --cc=mchan@broadcom.com \
    --cc=netdev@vger.kernel.org \
    --cc=peter.p.waskiewicz.jr@intel.com \
    --cc=rdreier@cisco.com \
    --cc=shemminger@linux-foundation.org \
    --cc=sri@us.ibm.com \
    --cc=tgraf@suug.ch \
    /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).