Home » Mailing list archives » linux-netdev » 2008 » February » 21

[RFC,PATCH] loopback: calls netif_receive_skb() instead of netif_rx()

!MAILaRCHIVE_VOTE_RePLACE
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]
[view in full thread]
From: Eric Dumazet <dada1@...>
To: David S. Miller <davem@...>
Cc: <netdev@...>
Subject: [RFC,PATCH] loopback: calls netif_receive_skb() instead of netif_rx()
Date: Thursday, February 21, 2008 - 2:51 pm

Hi David

This is an RFC, based on net-2.6 for convenience only.

Thank you

[RFC,PATCH] loopback: calls netif_receive_skb() instead of netif_rx()

Loopback transmit function loopback_xmit() actually calls netif_rx() to 
queue
a skb to the softnet queue, and arms a softirq so that this skb can be 
handled later.

This has a cost on SMP, because we need to hold a reference on the 
device, and free this
reference when softirq dequeues packet.

Following patch directly calls netif_receive_skb() and avoids lot of 
atomic operations.
(atomic_inc(&dev->refcnt), set_and_set_bit(NAPI_STATE_SCHED, &n->state), ...
  atomic_dec(&dev->refcnt)...), cache line ping-pongs on device refcnt, 
but also softirq overhead.

This gives a nice boost on tbench for example (5 % on my machine)

Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]
Messages in current thread:
[RFC,PATCH] loopback: calls netif_receive_skb() instead of n..., Eric Dumazet, (Thu Feb 21, 2:51 pm)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., David Miller, (Tue Feb 26, 10:21 pm)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Eric Dumazet, (Sat Mar 1, 6:26 am)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., David Miller, (Sun Mar 23, 6:29 am)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Eric Dumazet, (Sun Mar 23, 2:48 pm)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Ingo Molnar, (Mon Mar 31, 5:48 am)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., David Miller, (Mon Mar 31, 6:08 am)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Ingo Molnar, (Mon Mar 31, 6:44 am)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., David Miller, (Mon Mar 31, 7:02 am)
poor network loopback performance and scalability (was: Re: ..., Ingo Molnar, (Mon Mar 31, 7:36 am)
Re: poor network loopback performance and scalability (was: ..., Herbert Xu, (Sun Apr 20, 11:24 pm)
Re: poor network loopback performance and scalability, David Miller, (Sun Apr 20, 11:38 pm)
Re: poor network loopback performance and scalability, Ingo Molnar, (Mon Apr 21, 4:11 am)
Re: poor network loopback performance and scalability, David Miller, (Mon Apr 21, 4:16 am)
Re: poor network loopback performance and scalability, Herbert Xu, (Mon Apr 21, 6:19 am)
Re: poor network loopback performance and scalability, David Miller, (Mon Apr 21, 6:22 am)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Eric Dumazet, (Mon Mar 31, 6:01 am)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Ingo Molnar, (Mon Mar 31, 6:12 am)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Eric Dumazet, (Tue Apr 1, 5:19 am)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Pavel Machek, (Thu Apr 3, 10:06 am)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Eric Dumazet, (Thu Apr 3, 12:19 pm)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., David Miller, (Fri Mar 28, 9:36 pm)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Eric Dumazet, (Sat Mar 29, 4:18 am)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., David Miller, (Sat Mar 29, 7:54 pm)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Eric Dumazet, (Mon Mar 31, 2:38 am)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Andi Kleen, (Sun Mar 23, 3:15 pm)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., David Miller, (Tue Mar 4, 12:55 am)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Eric Dumazet, (Tue Mar 4, 2:27 am)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Stephen Hemminger, (Tue Mar 4, 1:15 am)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Jarek Poplawski, (Wed Feb 27, 3:20 am)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., David Miller, (Wed Feb 27, 3:23 am)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Jarek Poplawski, (Wed Feb 27, 3:34 am)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Daniel Lezcano, (Thu Feb 21, 4:14 pm)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Eric Dumazet, (Thu Feb 21, 7:19 pm)
Re: [RFC,PATCH] loopback: calls netif_receive_skb() instead ..., Daniel Lezcano, (Fri Feb 22, 6:19 am)