All of lore.kernel.org
 help / color / mirror / Atom feed
From: Pekka Riikonen <priikone@iki.fi>
To: David Miller <davem@davemloft.net>
Cc: eric.dumazet@gmail.com, eliezer.tamir@linux.intel.com,
	linux-kernel@vger.kernel.org, netdev@vger.kernel.org,
	jesse.brandeburg@intel.com, donald.c.skidmore@intel.com,
	e1000-devel@lists.sourceforge.net, willemb@google.com,
	andi@firstfloor.org, hpa@zytor.com, eliezer@tamir.org.il
Subject: Re: [PATCH v3 net-next 1/4] net: implement support for low latency socket polling
Date: Tue, 21 May 2013 21:31:51 +0200 (CEST)	[thread overview]
Message-ID: <alpine.GSO.2.00.1305212123120.15328@git.silcnet.org> (raw)
In-Reply-To: <20130521.114921.460497085402760002.davem@davemloft.net>

On Tue, 21 May 2013, David Miller wrote:

> From: Pekka Riikonen <priikone@iki.fi>
> Date: Tue, 21 May 2013 19:02:19 +0200 (CEST)
>
>> On Tue, 21 May 2013, Eric Dumazet wrote:
>>
>> : > > Alternatively, use a napi_id instead of a pointer.
>> : >
>> : > I'm not sure I understand what you propose.
>> :
>> : Oh well.
>> :
>> : To get a pointer to a struct net_device, we can use ifindex, and do a
>> : rcu lookup into a hash table to get the net_device. We do not need
>> : {pointer,ifindex} but {ifindex} is enough
>> :
>> : My suggestion is to not have skb->skb_ref but skb->napi_index : Its safe
>> : to copy its value from skb->napi_index to sk->napi_index without
>> : refcounting.
>> :
>> : All NAPI need to get a unique napi_index, and be inserted in a hash
>> : table for immediate/fast lookup.
>> :
>> Maybe even that's not needed.  Couldn't skb->queue_mapping give the
>> correct NAPI instance in multiqueue nics?  The NAPI instance could be made
>> easily available from skb->dev.  In any case an index is much better than
>> a new pointer.
>
> Queue mapping is more volatile, and consider layered devices.
>
Yes, true.  The napi_index then is probably the way to go.  Main thing for 
me is that it doesn't increase skb size when in union with dma_cookie (skb 
has been growing lately).

 	Pekka


WARNING: multiple messages have this Message-ID (diff)
From: Pekka Riikonen <priikone@iki.fi>
To: David Miller <davem@davemloft.net>
Cc: willemb@google.com, eliezer.tamir@linux.intel.com,
	e1000-devel@lists.sourceforge.net, netdev@vger.kernel.org,
	hpa@zytor.com, jesse.brandeburg@intel.com,
	linux-kernel@vger.kernel.org, andi@firstfloor.org,
	eliezer@tamir.org.il
Subject: Re: [PATCH v3 net-next 1/4] net: implement support for low latency socket polling
Date: Tue, 21 May 2013 21:31:51 +0200 (CEST)	[thread overview]
Message-ID: <alpine.GSO.2.00.1305212123120.15328@git.silcnet.org> (raw)
In-Reply-To: <20130521.114921.460497085402760002.davem@davemloft.net>

On Tue, 21 May 2013, David Miller wrote:

> From: Pekka Riikonen <priikone@iki.fi>
> Date: Tue, 21 May 2013 19:02:19 +0200 (CEST)
>
>> On Tue, 21 May 2013, Eric Dumazet wrote:
>>
>> : > > Alternatively, use a napi_id instead of a pointer.
>> : >
>> : > I'm not sure I understand what you propose.
>> :
>> : Oh well.
>> :
>> : To get a pointer to a struct net_device, we can use ifindex, and do a
>> : rcu lookup into a hash table to get the net_device. We do not need
>> : {pointer,ifindex} but {ifindex} is enough
>> :
>> : My suggestion is to not have skb->skb_ref but skb->napi_index : Its safe
>> : to copy its value from skb->napi_index to sk->napi_index without
>> : refcounting.
>> :
>> : All NAPI need to get a unique napi_index, and be inserted in a hash
>> : table for immediate/fast lookup.
>> :
>> Maybe even that's not needed.  Couldn't skb->queue_mapping give the
>> correct NAPI instance in multiqueue nics?  The NAPI instance could be made
>> easily available from skb->dev.  In any case an index is much better than
>> a new pointer.
>
> Queue mapping is more volatile, and consider layered devices.
>
Yes, true.  The napi_index then is probably the way to go.  Main thing for 
me is that it doesn't increase skb size when in union with dma_cookie (skb 
has been growing lately).

 	Pekka


------------------------------------------------------------------------------
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_may
_______________________________________________
E1000-devel mailing list
E1000-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/e1000-devel
To learn more about Intel&#174; Ethernet, visit http://communities.intel.com/community/wired

  reply	other threads:[~2013-05-21 19:32 UTC|newest]

Thread overview: 69+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-05-20 10:15 [PATCH v3 net-next 0/4] net: low latency Ethernet device polling Eliezer Tamir
2013-05-20 10:15 ` Eliezer Tamir
2013-05-20 10:16 ` [PATCH v3 net-next 1/4] net: implement support for low latency socket polling Eliezer Tamir
2013-05-20 10:16   ` Eliezer Tamir
2013-05-20 15:29   ` Eric Dumazet
2013-05-20 19:40     ` David Miller
2013-05-21  7:28     ` Eliezer Tamir
2013-05-21 13:28       ` Eric Dumazet
2013-05-21 17:02         ` Pekka Riikonen
2013-05-21 17:48           ` Eric Dumazet
2013-05-21 17:51             ` Eric Dumazet
2013-05-21 17:51               ` Eric Dumazet
2013-05-21 19:25               ` Eliezer Tamir
2013-05-21 19:25                 ` Eliezer Tamir
2013-05-21 19:29                 ` Eric Dumazet
2013-05-21 19:29                   ` Eric Dumazet
2013-05-21 18:49           ` David Miller
2013-05-21 19:31             ` Pekka Riikonen [this message]
2013-05-21 19:31               ` Pekka Riikonen
2013-05-20 10:16 ` [PATCH v3 net-next 2/4] tcp: add TCP support for low latency receive poll Eliezer Tamir
2013-05-20 10:16   ` Eliezer Tamir
2013-05-20 13:49   ` Eric Dumazet
2013-05-20 14:59     ` Eliezer Tamir
2013-05-20 14:59       ` Eliezer Tamir
2013-05-20 10:16 ` [PATCH v3 net-next 3/4] ixgbe: Add support for ndo_ll_poll Eliezer Tamir
2013-05-20 10:16   ` Eliezer Tamir
2013-05-20 20:20   ` Or Gerlitz
2013-05-20 20:20     ` Or Gerlitz
2013-05-20 20:33     ` Andi Kleen
2013-05-20 20:42       ` Or Gerlitz
2013-05-20 21:01         ` Andi Kleen
2013-05-21  6:23           ` Or Gerlitz
2013-05-21  6:23             ` Or Gerlitz
2013-05-21  6:54     ` Eliezer Tamir
2013-05-21  6:54       ` Eliezer Tamir
2013-05-21  7:06       ` Eilon Greenstein
2013-05-21  7:06         ` Eilon Greenstein
2013-05-21  7:14         ` David Miller
2013-05-21  8:24           ` Or Gerlitz
2013-05-21  8:24             ` Or Gerlitz
2013-05-21  8:31             ` Eliezer Tamir
2013-05-21  8:31               ` Eliezer Tamir
2013-05-21  8:39             ` David Miller
2013-05-21  8:42               ` Eliezer Tamir
2013-05-21  8:42                 ` Eliezer Tamir
2013-05-21  8:43               ` Or Gerlitz
2013-05-21  8:43                 ` Or Gerlitz
2013-05-21 10:27                 ` Eliezer Tamir
2013-05-21 10:41                   ` Or Gerlitz
2013-05-21  8:21         ` Or Gerlitz
2013-05-21  8:21           ` Or Gerlitz
2013-05-21  8:28           ` Eilon Greenstein
2013-05-21  8:28             ` Eilon Greenstein
2013-05-21 14:19       ` Willem de Bruijn
2013-05-21 14:19         ` Willem de Bruijn
2013-05-20 10:16 ` [PATCH v3 net-next 4/4] ixgbe: add extra stats " Eliezer Tamir
2013-05-20 10:16   ` Eliezer Tamir
2013-05-21 12:06 ` [PATCH v3 net-next 0/4] net: low latency Ethernet device polling Alex Rosenbaum
2013-05-21 12:29   ` Eliezer Tamir
2013-05-21 13:15     ` Alex Rosenbaum
2013-05-21 13:15       ` Alex Rosenbaum
2013-05-21 13:31       ` Eric Dumazet
2013-05-23 11:06       ` Eliezer Tamir
2013-05-23 11:45         ` Alex Rosenbaum
2013-05-21 14:30     ` Willem de Bruijn
2013-05-21 14:30       ` Willem de Bruijn
2013-05-21 18:15     ` Ben Hutchings
2013-05-21 18:15       ` Ben Hutchings
2013-05-22  9:35       ` Eliezer Tamir

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=alpine.GSO.2.00.1305212123120.15328@git.silcnet.org \
    --to=priikone@iki.fi \
    --cc=andi@firstfloor.org \
    --cc=davem@davemloft.net \
    --cc=donald.c.skidmore@intel.com \
    --cc=e1000-devel@lists.sourceforge.net \
    --cc=eliezer.tamir@linux.intel.com \
    --cc=eliezer@tamir.org.il \
    --cc=eric.dumazet@gmail.com \
    --cc=hpa@zytor.com \
    --cc=jesse.brandeburg@intel.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=willemb@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 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.