Bug#909267: library-not-linked-against-libc: downgrade from error
- To: Guillem Jover <guillem@debian.org>, 909267@bugs.debian.org, Russ Allbery <rra@debian.org>
- Cc: Jeremy Bicha <jbicha@debian.org>, Simon McVittie <smcv@debian.org>, naesten@gmail.com
- Subject: Bug#909267: library-not-linked-against-libc: downgrade from error
- From: Chris Lamb <lamby@debian.org>
- Date: Mon, 14 Jan 2019 23:49:58 +0000
- Message-id: <[🔎] 1547509798.127265.1634646504.3E17D27B@webmail.messagingengine.com>
- Reply-to: Chris Lamb <lamby@debian.org>, 909267@bugs.debian.org
- In-reply-to: <1539105437.1388583.1536176088.2C5C5876@webmail.messagingengine.com>
- References: <1537461750.2004232.1514996608.71C30A1F@webmail.messagingengine.com> <20180920203714.GA10668@espresso.pseudorandom.co.uk> <CAAajCMb1+2fU47vrP-ZLxBGVrWb4HJGEx3aanNnPZ=2nDoZ-1A@mail.gmail.com> <1537478889.2097863.1515316640.769120E9@webmail.messagingengine.com> <874lejyfs7.fsf@hope.eyrie.org> <CAAajCMZtt1CvdL16z-29f8qiVazfOBVuWK5iU_H6Lx-=mfi4XQ@mail.gmail.com> <CAAajCMb1+2fU47vrP-ZLxBGVrWb4HJGEx3aanNnPZ=2nDoZ-1A@mail.gmail.com> <87pnx71z49.fsf@hope.eyrie.org> <CAAajCMb1+2fU47vrP-ZLxBGVrWb4HJGEx3aanNnPZ=2nDoZ-1A@mail.gmail.com> <20181006153318.GA30351@gaara.hadrons.org> <1539105437.1388583.1536176088.2C5C5876@webmail.messagingengine.com> <CAAajCMb1+2fU47vrP-ZLxBGVrWb4HJGEx3aanNnPZ=2nDoZ-1A@mail.gmail.com>
Chris Lamb wrote:
> > > I wonder if we would get all of the utility out of the tag if instead it
> > > looked for shared libraries with no NEEDED metadata. I think it's only
> > > catching libraries that aren't linked with anything else, so maybe just
> > > check for that explicitly?
> >
> > Yeah probably better than the status-quo. Any kind of plugin would need
> > to be excluded though, because it might simply be using symbols from the
> > loading binary (via -rdynamic). It would still emit false-positives for
> > any library that implements language run-times or does syscall wrapping.
> […]
> > So, I'd say the trade-off is worth it, as there's definitely going to
> > be way less false-positives on language run-time libraries, than the
> > current false-positives.
>
> Sounds reasonable. Can someone retitle this bug to match? (I doubt I
> personally have enough shared library foo to implement this myself,
> alas.)
Gentle ping on this, folks? :)
Regards,
--
,''`.
: :' : Chris Lamb
`. `'` lamby@debian.org 🍥 chris-lamb.co.uk
`-
Reply to: