From: Joseph Myers <joseph@codesourcery.com>
To: "H.J. Lu" <hjl.tools@gmail.com>
Cc: "Fāng-ruì Sòng" <maskray@google.com>,
"GNU C Library" <libc-alpha@sourceware.org>,
Binutils <binutils@sourceware.org>
Subject: Re: [PATCH v2] elf: Support DT_RELR relative relocation format [BZ #27924]
Date: Mon, 18 Oct 2021 19:21:46 +0000 [thread overview]
Message-ID: <alpine.DEB.2.22.394.2110181915020.114463@digraph.polyomino.org.uk> (raw)
In-Reply-To: <CAMe9rOqFBxQN6v66fUTNwpB_69WUfWiX=uC7EWnog9Y4gEf48A@mail.gmail.com>
On Mon, 18 Oct 2021, H.J. Lu via Binutils wrote:
> For Linux targets, we need the glibc version dependency for
> DT_RELR support at run-time.
What we've supposedly previously done with new ELF features is use
EI_ABIVERSION (see libc-abis in glibc).
That requires the static linker to compute the right ABI version based on
which of the relevant features is used by the generated binary (note that
the ordering of different features, and thus the numbers assigned to them
as well as which of them counts as the newest feature to be stored in
EI_ABIVERSION, is *not* the same across glibc architectures). I'm not
sure the binutils side of things was ever implemented on most
architectures, or that the checks in glibc get properly applied to all
relevant ELF objects (dynamic linker, main program, loaded shared
libraries), but in principle the mechanism exists and avoids the linker
needing to know about specific glibc version numbers itself (at the
expense of needing to know the per-architecture version numbers and
ordering for different ELF features).
--
Joseph S. Myers
joseph@codesourcery.com
next prev parent reply other threads:[~2021-10-18 19:22 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-10-17 0:50 Fangrui Song
2021-10-18 14:42 ` H.J. Lu
2021-10-18 16:16 ` Fāng-ruì Sòng
2021-10-18 17:28 ` H.J. Lu
2021-10-18 18:15 ` Fāng-ruì Sòng
2021-10-18 18:27 ` H.J. Lu
2021-10-18 19:19 ` Fāng-ruì Sòng
2021-10-18 19:44 ` H.J. Lu
2021-10-18 20:11 ` Fāng-ruì Sòng
2021-10-18 21:10 ` Joseph Myers
2021-10-18 22:27 ` H.J. Lu
2021-10-18 22:30 ` Joseph Myers
2021-10-18 22:42 ` H.J. Lu
2021-10-18 23:00 ` Joseph Myers
2021-10-18 23:36 ` H.J. Lu
2021-10-18 23:44 ` Joseph Myers
2021-10-19 1:05 ` H.J. Lu
2021-10-18 19:21 ` Joseph Myers [this message]
2021-10-18 19:45 ` H.J. Lu
2021-10-29 18:21 ` Carlos O'Donell
2021-10-29 18:36 ` H.J. Lu
2021-10-29 19:15 ` Fangrui Song
2021-10-29 18:38 ` Fāng-ruì Sòng
2021-10-29 19:35 ` Carlos O'Donell
2021-10-29 19:53 ` Adhemerval Zanella
2021-11-01 4:50 ` Fāng-ruì Sòng
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=alpine.DEB.2.22.394.2110181915020.114463@digraph.polyomino.org.uk \
--to=joseph@codesourcery.com \
--cc=binutils@sourceware.org \
--cc=hjl.tools@gmail.com \
--cc=libc-alpha@sourceware.org \
--cc=maskray@google.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).