[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <4ae938a0-6331-f5d6-baa5-62eb8b07e63f@kernel.org>
Date: Tue, 12 Sep 2023 10:17:13 -0600
From: David Ahern <dsahern@...nel.org>
To: Aurelien Aptel <aaptel@...dia.com>, linux-nvme@...ts.infradead.org,
netdev@...r.kernel.org, sagi@...mberg.me, hch@....de, kbusch@...nel.org,
axboe@...com, chaitanyak@...dia.com, davem@...emloft.net, kuba@...nel.org
Cc: Boris Pismenny <borisp@...dia.com>, aurelien.aptel@...il.com,
smalin@...dia.com, malin1024@...il.com, ogerlitz@...dia.com,
yorayz@...dia.com, galshalom@...dia.com, mgurtovoy@...dia.com,
edumazet@...gle.com, pabeni@...hat.com, imagedong@...cent.com
Subject: Re: [PATCH v15 01/20] net: Introduce direct data placement tcp
offload
On 9/12/23 3:59 AM, Aurelien Aptel wrote:
> As a result, data from page frags should not be copied out to
> the linear part. To avoid needless copies, such as when using
> skb_condense, we mark the skb->ulp_ddp bit.
Not consolidating frags into linear is a requirement for any Rx ZC
solution (e.g., the gpu devmem RFC set and accompanying io_uring set for
userspace memory). Meaning, the bit can be more generic and less tied to
ulp_ddp.
Powered by blists - more mailing lists