mbox series

[net-next,v11,0/3] skbuff: Optimize SKB coalescing for page pool

Message ID 20231215033011.12107-1-liangchen.linux@gmail.com (mailing list archive)
Headers show
Series skbuff: Optimize SKB coalescing for page pool | expand

Message

Liang Chen Dec. 15, 2023, 3:30 a.m. UTC
The combination of the following condition was excluded from skb coalescing:

from->pp_recycle = 1
from->cloned = 1
to->pp_recycle = 1

With page pool in use, this combination can be quite common(ex.
NetworkMananger may lead to the additional packet_type being registered,
thus the cloning). In scenarios with a higher number of small packets, it
can significantly affect the success rate of coalescing.

This patchset aims to optimize this scenario and enable coalescing of this
particular combination. That also involves supporting multiple users
referencing the same fragment of a pp page to accomondate the need to
increment the "from" SKB page's pp page reference count.


Changes from v10:
- re-number patches to 1/3, 2/3, 3/3

Changes from v9:
- patch 1 was already applied
- imporve description for patch 2
- make sure skb_pp_frag_ref only work for pp aware skbs



Liang Chen (3):
  page_pool: halve BIAS_MAX for multiple user references of a fragment
  skbuff: Add a function to check if a page belongs to page_pool
  skbuff: Optimization of SKB coalescing for page pool

 include/net/page_pool/helpers.h |  5 +++
 net/core/page_pool.c            |  2 +-
 net/core/skbuff.c               | 59 +++++++++++++++++++++++++--------
 3 files changed, 52 insertions(+), 14 deletions(-)

Comments

patchwork-bot+netdevbpf@kernel.org Dec. 17, 2023, 11:01 a.m. UTC | #1
Hello:

This series was applied to netdev/net-next.git (main)
by David S. Miller <davem@davemloft.net>:

On Fri, 15 Dec 2023 11:30:08 +0800 you wrote:
> The combination of the following condition was excluded from skb coalescing:
> 
> from->pp_recycle = 1
> from->cloned = 1
> to->pp_recycle = 1
> 
> With page pool in use, this combination can be quite common(ex.
> NetworkMananger may lead to the additional packet_type being registered,
> thus the cloning). In scenarios with a higher number of small packets, it
> can significantly affect the success rate of coalescing.
> 
> [...]

Here is the summary with links:
  - [net-next,v11,1/3] page_pool: halve BIAS_MAX for multiple user references of a fragment
    https://git.kernel.org/netdev/net-next/c/aaf153aecef1
  - [net-next,v11,2/3] skbuff: Add a function to check if a page belongs to page_pool
    https://git.kernel.org/netdev/net-next/c/8cfa2dee325f
  - [net-next,v11,3/3] skbuff: Optimization of SKB coalescing for page pool
    https://git.kernel.org/netdev/net-next/c/f7dc3248dcfb

You are awesome, thank you!