[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <56d32e37-afdd-342c-947d-dec329a504e5@linux.dev>
Date: Sun, 17 Sep 2023 18:10:11 +0100
From: Vadim Fedorenko <vadim.fedorenko@...ux.dev>
To: Coco Li <lixiaoyan@...gle.com>, Jakub Kicinski <kuba@...nel.org>,
Eric Dumazet <edumazet@...gle.com>, Neal Cardwell <ncardwell@...gle.com>,
Mubashir Adnan Qureshi <mubashirq@...gle.com>,
Paolo Abeni <pabeni@...hat.com>
Cc: netdev@...r.kernel.org, Chao Wu <wwchao@...gle.com>,
Wei Wang <weiwan@...gle.com>
Subject: Re: [PATCH v1 net-next 3/5] netns-ipv4: reorganize netns_ipv4 fast
path variables
On 16/09/2023 02:06, Coco Li wrote:
> Reorganize fast path variables on tx-txrx-rx order.
> Fastpath cacheline ends after sysctl_tcp_rmem.
> There are only read-only variables here. (write is on the control path
> and not considered in this case)
I believe udp sysctls can be aligned the same way. With HTTP/3 adoption
we should think about UDP traffic too, and looks like we do have some
space in hot-path cache lines for udp_early_demux and rmem/wmem.
And have you thought about cache-line boundary alignment for these values?
Powered by blists - more mailing lists