On Mon, May 9, 2022 at 2:00 AM Matthieu Baerts
<matthieu.baerts@xxxxxxxxxxxx> wrote:
>
> Hi Andrii,
>
> Thank you for the review!
>
> On 07/05/2022 00:26, Andrii Nakryiko wrote:
> > On Mon, May 2, 2022 at 2:12 PM Mat Martineau
> > <mathew.j.martineau@xxxxxxxxxxxxxxx> wrote:
>
> (...)
>
> >> diff --git a/MAINTAINERS b/MAINTAINERS
> >> index 359afc617b92..d48d3cb6abbc 100644
> >> --- a/MAINTAINERS
> >> +++ b/MAINTAINERS
> >> @@ -13780,6 +13780,7 @@ F: include/net/mptcp.h
> >> F: include/trace/events/mptcp.h
> >> F: include/uapi/linux/mptcp.h
> >> F: net/mptcp/
> >> +F: tools/testing/selftests/bpf/bpf_mptcp_helpers.h
> >> F: tools/testing/selftests/bpf/*/*mptcp*.c
> >> F: tools/testing/selftests/net/mptcp/
> >>
> >> diff --git a/tools/testing/selftests/bpf/bpf_mptcp_helpers.h b/tools/testing/selftests/bpf/bpf_mptcp_helpers.h
> >> new file mode 100644
> >> index 000000000000..18da4cc65e89
> >> --- /dev/null
> >> +++ b/tools/testing/selftests/bpf/bpf_mptcp_helpers.h
> >> @@ -0,0 +1,14 @@
> >> +/* SPDX-License-Identifier: GPL-2.0 */
> >> +/* Copyright (c) 2022, SUSE. */
> >> +
> >> +#ifndef __BPF_MPTCP_HELPERS_H
> >> +#define __BPF_MPTCP_HELPERS_H
> >> +
> >> +#include "bpf_tcp_helpers.h"
> >> +
> >> +struct mptcp_sock {
> >> + struct inet_connection_sock sk;
> >> +
> >> +} __attribute__((preserve_access_index));
> >
> > why can't all this live in bpf_tcp_helpers.h? why do we need extra header?
>
> The main reason is related to the maintenance: to have MPTCP ML being
> cc'd for all patches modifying this file.
>
> Do you prefer if all these specific MPTCP structures and macros and
> mixed with TCP ones?
>
These definitions don't even have to be 1:1 w/ whatever is kernel
defining in terms of having all the fields, or their order, etc. So I
think it won't require active maintenance and thus can be merged into
bpf_tcp_helpers.h to keep it in one place.
> Cheers,
> Matt
> --
> Tessares | Belgium | Hybrid Access Solutions
> www.tessares.net
