netdev - Re: [Patch bpf-next v2 2/5] skmsg: get rid of struct sk_psock

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <CAM_iQpUkNA0D8vokqvjLgH3bv6vkKvNoz4Jg149dDnDjFH+qCA@mail.gmail.com>
Date:   Fri, 12 Feb 2021 17:58:03 -0800
From:   Cong Wang <xiyou.wangcong@...il.com>
To:     Lorenz Bauer <lmb@...udflare.com>
Cc:     Networking <netdev@...r.kernel.org>, bpf <bpf@...r.kernel.org>,
        duanxiongchun@...edance.com,
        Dongdong Wang <wangdongdong.6@...edance.com>,
        jiang.wang@...edance.com, Cong Wang <cong.wang@...edance.com>,
        John Fastabend <john.fastabend@...il.com>,
        Daniel Borkmann <daniel@...earbox.net>,
        Jakub Sitnicki <jakub@...udflare.com>
Subject: Re: [Patch bpf-next v2 2/5] skmsg: get rid of struct sk_psock_parser

On Fri, Feb 12, 2021 at 11:09 AM Cong Wang <xiyou.wangcong@...il.com> wrote:
>
> On Fri, Feb 12, 2021 at 2:56 AM Lorenz Bauer <lmb@...udflare.com> wrote:
> >
> > On Wed, 10 Feb 2021 at 02:21, Cong Wang <xiyou.wangcong@...il.com> wrote:
> > >
> > > From: Cong Wang <cong.wang@...edance.com>
> > >
> > > struct sk_psock_parser is embedded in sk_psock, it is
> > > unnecessary as skb verdict also uses ->saved_data_ready.
> > > We can simply fold these fields into sk_psock, and get rid
> > > of ->enabled.
> >
> > Looks nice, can you use sk_psock_strp_enabled() more? There are a
> > couple places in sock_map.c which test psock->saved_data_ready
> > directly.
>
> Its name tells it is for stream parser, so not suitable for others.
>
> Are you suggesting to rename it to sk_psock_enabled() and use
> it? Note it still has an additional !psock test, but I think that is fine
> for slow paths.

Well, I think it is a bug, sk_psock_strp_enabled() probably means
to check whether progs.stream_verdict is running, not whether
any of these progs is running. So, I'd leave this untouched in this
patchset and if needed a separate bug fix should be sent to -net.

Thanks.