[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <87bkyodoni.fsf@toke.dk>
Date: Wed, 02 Mar 2022 22:34:41 +0100
From: Toke Høiland-Jørgensen <toke@...hat.com>
To: Alexei Starovoitov <alexei.starovoitov@...il.com>
Cc: Alexei Starovoitov <ast@...nel.org>,
Daniel Borkmann <daniel@...earbox.net>,
Andrii Nakryiko <andrii@...nel.org>,
Martin KaFai Lau <kafai@...com>,
Song Liu <songliubraving@...com>, Yonghong Song <yhs@...com>,
John Fastabend <john.fastabend@...il.com>,
KP Singh <kpsingh@...nel.org>,
"David S. Miller" <davem@...emloft.net>,
Jakub Kicinski <kuba@...nel.org>,
Jesper Dangaard Brouer <hawk@...nel.org>,
Jonathan Corbet <corbet@....net>, linux-doc@...r.kernel.org,
netdev@...r.kernel.org, bpf@...r.kernel.org
Subject: Re: [PATCH bpf-next v8 2/5] Documentation/bpf: Add documentation
for BPF_PROG_RUN
Alexei Starovoitov <alexei.starovoitov@...il.com> writes:
> On Fri, Feb 18, 2022 at 06:50:26PM +0100, Toke Høiland-Jørgensen wrote:
>> This adds documentation for the BPF_PROG_RUN command; a short overview of
>> the command itself, and a more verbose description of the "live packet"
>> mode for XDP introduced in the previous commit.
>
> Overall the patch set looks great. The doc really helps.
Great, thanks!
> One nit below.
>
>> +- When running the program with multiple repetitions, the execution will happen
>> + in batches, where the program is executed multiple times in a loop, the result
>> + is saved, and other actions (like redirecting the packet or passing it to the
>> + networking stack) will happen for the whole batch after the execution. This is
>> + similar to how execution happens in driver-mode XDP for each hardware NAPI
>> + cycle. The batch size defaults to 64 packets (which is same as the NAPI batch
>> + size), but the batch size can be specified by userspace through the
>> + ``batch_size`` parameter, up to a maximum of 256 packets.
>
> This paragraph is a bit confusing.
> I've read it as the program can do only one kind of result per batch and
> it will apply to the whole batch.
> But the program can do XDP_PASS/REDIRECT in any order.
> Can you make "the result is saved" a bit more clear?
Yeah, re-reading it now, I see what you mean; will try to make it
clearer...
-Toke
Powered by blists - more mailing lists