[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <82fb2eba56d84887772f9d533faa7bda9e3b2ee4.camel@coverfire.com>
Date: Sun, 03 Nov 2019 21:08:08 -0500
From: dan@...erfire.com
To: Jakub Kicinski <jakub.kicinski@...ronome.com>,
"Samudrala, Sridhar" <sridhar.samudrala@...el.com>
Cc: bjorn.topel@...il.com, alexei.starovoitov@...il.com,
bjorn.topel@...el.com, bpf@...r.kernel.org,
intel-wired-lan@...ts.osuosl.org, maciej.fijalkowski@...el.com,
magnus.karlsson@...el.com, netdev@...r.kernel.org, toke@...hat.com,
tom.herbert@...el.com, David Miller <davem@...emloft.net>
Subject: Re: [Intel-wired-lan] FW: [PATCH bpf-next 2/4] xsk: allow AF_XDP
sockets to receive packets directly from a queue
On Thu, 2019-10-31 at 17:21 -0700, Jakub Kicinski wrote:
>
> My concern was that we want the applications to encode fast path
> logic
> in BPF and load that into the kernel. So your patch works
> fundamentally
> against that goal:
I'm only one AF_XDP user but my main goal is to get packets to
userspace as fast as possible. I don't (currently) need a BPF program
in the path at all. I suspect that many other people that look at
AF_XDP as a DPDK replacement have a similar view.
Powered by blists - more mailing lists