mbox series

[net-next,0/2] inet: factorize sk_wmem_alloc updates

Message ID 20180331201626.186789-1-edumazet@google.com
Headers show
Series inet: factorize sk_wmem_alloc updates | expand

Message

Eric Dumazet March 31, 2018, 8:16 p.m. UTC 
While testing my inet defrag changes, I found that senders
could spend ~20% of cpu cycles in skb_set_owner_w() updating
sk->sk_wmem_alloc for every fragment they cook, competing
with TX completion of prior skbs possibly happening on another cpus.

One solution to this problem is to use alloc_skb() instead
of sock_wmalloc() and manually perform a single sk_wmem_alloc change.

This greatly increases speed for applications sending big UDP datagrams.

Eric Dumazet (2):
  ipv4: factorize sk_wmem_alloc updates done by __ip_append_data()
  ipv6: factorize sk_wmem_alloc updates done by __ip6_append_data()

 net/ipv4/ip_output.c  | 17 ++++++++++++-----
 net/ipv6/ip6_output.c | 17 ++++++++++++-----
 2 files changed, 24 insertions(+), 10 deletions(-)

Comments

David Miller April 1, 2018, 6:09 p.m. UTC | #1 
From: Eric Dumazet <edumazet@google.com>
Date: Sat, 31 Mar 2018 13:16:24 -0700

> While testing my inet defrag changes, I found that senders
> could spend ~20% of cpu cycles in skb_set_owner_w() updating
> sk->sk_wmem_alloc for every fragment they cook, competing
> with TX completion of prior skbs possibly happening on another cpus.
> 
> One solution to this problem is to use alloc_skb() instead
> of sock_wmalloc() and manually perform a single sk_wmem_alloc change.
> 
> This greatly increases speed for applications sending big UDP datagrams.

Looks good, series applied, thanks.