netdev - Re: [PATCH bpf v3] libbpf: handle symbol versioning properly for libbpf.a

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: <30ee93ab-a3f7-e92d-a33e-477e74f1849a@fb.com>
Date:   Mon, 30 Sep 2019 18:18:35 +0000
From:   Alexei Starovoitov <ast@...com>
To:     Yonghong Song <yhs@...com>,
        "bpf@...r.kernel.org" <bpf@...r.kernel.org>,
        "netdev@...r.kernel.org" <netdev@...r.kernel.org>
CC:     Daniel Borkmann <daniel@...earbox.net>,
        Kernel Team <Kernel-team@...com>,
        Kevin Laatz <kevin.laatz@...el.com>,
        Arnaldo Carvalho de Melo <acme@...hat.com>,
        Andrii Nakryiko <andriin@...com>
Subject: Re: [PATCH bpf v3] libbpf: handle symbol versioning properly for
 libbpf.a

On 9/30/19 9:56 AM, Yonghong Song wrote:
> 
> 
> On 9/30/19 9:42 AM, Alexei Starovoitov wrote:
>> On 9/30/19 9:29 AM, Yonghong Song wrote:
>>> +OLD_VERSION(xsk_umem__create_v0_0_2, xsk_umem__create, LIBBPF_0.0.2)
>>> +NEW_VERSION(xsk_umem__create_v0_0_4, xsk_umem__create, LIBBPF_0.0.4)
>>
>> how this will look when yet another version of this function is
>> introduced, say in 0.0.6 ?
>>
>> OLD_VERSION(xsk_umem__create_v0_0_2, xsk_umem__create, LIBBPF_0.0.2)
>> OLD_VERSION(xsk_umem__create_v0_0_4, xsk_umem__create, LIBBPF_0.0.4)
>> NEW_VERSION(xsk_umem__create_v0_0_6, xsk_umem__create, LIBBPF_0.0.6)
> 
> Yes.
> 
>>
>> 0.0.4 will be renamed to OLD_ and the latest addition NEW_ ?
> 
> Right.
> 
>> The macro name feels a bit confusing. May be instead of NEW_
>> call it CURRENT_ ? or DEFAULT_ ?
>> NEW_ will become not so 'new' few months from now.
> 
> Right. After a few months, the version number may indeed be
> behind the libbpf versions.... "current" may not be current ....
> Let me use DEFAULT then. How about using
>      COMPAT_VERSION(...)
> for old versions, and using

COMPAT_VERSION sounds fine. I think OLD_VERSION was ok too.

>      DEFAULT_VERSION(...)
> for the new version?

sounds good.

Thanks!