[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <51abec01-c4ce-434f-694a-f932e0e203ec@linux.dev>
Date: Wed, 25 Oct 2023 15:09:43 -0700
From: Martin KaFai Lau <martin.lau@...ux.dev>
To: Kui-Feng Lee <sinquersw@...il.com>
Cc: netdev@...r.kernel.org, razor@...ckwall.org, ast@...nel.org,
andrii@...nel.org, john.fastabend@...il.com, sdf@...gle.com,
toke@...nel.org, kuba@...nel.org, andrew@...n.ch,
Toke Høiland-Jørgensen <toke@...hat.com>,
Daniel Borkmann <daniel@...earbox.net>, bpf@...r.kernel.org
Subject: Re: [PATCH bpf-next v4 1/7] netkit, bpf: Add bpf programmable net
device
On 10/25/23 2:24 PM, Kui-Feng Lee wrote:
>
>
> On 10/24/23 14:48, Daniel Borkmann wrote:
>> This work adds a new, minimal BPF-programmable device called "netkit"
>> (former PoC code-name "meta") we recently presented at LSF/MM/BPF. The
>> core idea is that BPF programs are executed within the drivers xmit routine
>> and therefore e.g. in case of containers/Pods moving BPF processing closer
>> to the source.
>>
>
> Sorry for intruding into this discussion! Although it is too late to
> mentioned this since this patchset have been v4 already.
>
> I notice netkit has introduced a new attach type. I wonder if it
> possible to implement it as a new struct_ops type.
Could your elaborate more about what does this struct_ops type do and how is it
different from the SCHED_CLS bpf prog that the netkit is running?
Powered by blists - more mailing lists