linux-kernel - Re: [RFC/fix] Re: libbpf build broken on musl libc (Alpine Linux)

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: <20180918041758.rruwsw52jxpxh6bk@ast-mbp>
Date:   Mon, 17 Sep 2018 21:18:00 -0700
From:   Alexei Starovoitov <alexei.starovoitov@...il.com>
To:     Jakub Kicinski <jakub.kicinski@...ronome.com>
Cc:     Arnaldo Carvalho de Melo <arnaldo.melo@...il.com>,
        Daniel Borkmann <daniel@...earbox.net>,
        Thomas Richter <tmricht@...ux.ibm.com>,
        Hendrik Brueckner <brueckner@...ux.ibm.com>,
        Jiri Olsa <jolsa@...nel.org>,
        Namhyung Kim <namhyung@...nel.org>,
        linux-kernel@...r.kernel.org, netdev@...r.kernel.org
Subject: Re: [RFC/fix] Re: libbpf build broken on musl libc (Alpine Linux)

On Mon, Sep 17, 2018 at 07:39:52PM -0700, Jakub Kicinski wrote:
> On Mon, 17 Sep 2018 17:52:59 -0700, Alexei Starovoitov wrote:
> > LGPL-2.1 in the above?
> 
> Could we possibly make it dual licensed LGPL + BSD?

that's a separate discussion. let's not mix it up with this fix.

> > The rest looks good to me.
> > Should we take it via bpf-next tree?
> > If you feel there is an urgency to fix musl build, we can take it via
> > bpf tree too.
> >
> > Jakub, thoughts? you've been messing with strerror last..
> 
> Sorry for the mess, I didn't know libc compat is such a pain.
> 
> I kind of knowingly ignored the existing str_error_r(), as it didn't
> seem excessively clean.  Yet, two months later I have no better
> ideas..  We could as well switch back to the XSI version, which we used
> before I needed access to reallocarray() (I mean change the str_error()
> to be a plain XSI wrapper).
> 
> Perhaps placing the new helper in libbpf_errno.c file would make
> sense?  The only reason for this separate file to exist is in fact to
> make use of XSI-compliant strerror_r().

that's a good idea. libbpf_errno.c sounds more appropriate to me as well.

> Also, I need to go relicense tools/include/tools/libc_compat.h ASAP
> too :S

argh. an oversight. pls send the patch.