netdev - Re: [PATCH net-next 0/5] net/sched: Retire some tc qdiscs and classifiers

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

Message-ID: <CAM0EoM=FseCqxJD1X5vJHw3PHHSJxGAaFvV4n6njCmPq2cxP6g@mail.gmail.com>
Date:   Wed, 15 Feb 2023 11:37:37 -0500
From:   Jamal Hadi Salim <jhs@...atatu.com>
To:     Pedro Tammela <pctammela@...atatu.com>
Cc:     Jiri Pirko <jiri@...nulli.us>, Jakub Kicinski <kuba@...nel.org>,
        netdev@...r.kernel.org, xiyou.wangcong@...il.com,
        davem@...emloft.net, edumazet@...gle.com, pabeni@...hat.com,
        stephen@...workplumber.org, dsahern@...il.com
Subject: Re: [PATCH net-next 0/5] net/sched: Retire some tc qdiscs and classifiers

On Wed, Feb 15, 2023 at 8:42 AM Pedro Tammela <pctammela@...atatu.com> wrote:
>
> On 15/02/2023 04:50, Jiri Pirko wrote:
> > Wed, Feb 15, 2023 at 12:05:23AM CET, jhs@...atatu.com wrote:
> >> On Tue, Feb 14, 2023 at 4:41 PM Jakub Kicinski <kuba@...nel.org> wrote:
> >>>
> >>> On Tue, 14 Feb 2023 13:40:13 -0800 Jakub Kicinski wrote:
> >>>>> I think we have to let the UAPI there to rot in order not to break
> >>>>> compilation of apps that use those (no relation to iproute2).
> >>>>
> >>>> Yeah, I was hoping there's no other users but this is the first match
> >>>> on GitHub:
> >>>>
> >>>> https://github.com/t2mune/nield/blob/0c0848d1d1e6c006185465ee96aeb5a13a1589e6/src/tcmsg_qdisc_dsmark.c
> >>>>
> >>>> :(
> >>>
> >>> I mean that in the context of deleting the uAPI, not the support,
> >>> to be clear.
> >>
> >> Looking at that code - the user is keeping their own copy of the uapi
> >> and listening to generated events from the kernel.
> >> With new kernels those events will never come, so they wont be able to
> >> print them. IOW, there's no dependency on the uapi being in the kernel
> >> - but maybe worth keeping.
> >> Note, even if they were to send queries - they will just get a failure back.
> >
> > Guys. I don't think that matters. There might be some non-public code
> > using this headers and we don't want to break them either. I believe we
> > have to stick with it. But perhaps we can include some note there.
> >
>
> I don't know the compilation requirements for uapi, but it could
> leverage the deprecated attribute, when available as a compiler
> extension (it's queued for C23 as well[0]).
>
> E.g. https://godbolt.org/z/d535jh9j8
>
> Essentially anytime someone uses one of the Netlink attributes they will
> get a harmless warning like:
> <source>:22:18: warning: 'TCA_TCINDEX_CLASSID' is deprecated
> [-Wdeprecated-declarations]
>
> [0] https://en.cppreference.com/w/c/language/attributes/deprecated

That would work well actually. So iproute2 pulling these headers from
the kernel will get warned.
I think the outstanding question is still: How long do we keep the
headers around?

cheers,
jamal