From: John David Anglin <dave.anglin@bell.net>
To: Carlos O'Donell <carlos@redhat.com>
Cc: Andreas Schwab <schwab@linux-m68k.org>,
GNU C Library <libc-alpha@sourceware.org>,
Carlos O'Donell <carlos@systemhalted.org>
Subject: Re: [committed] Return to caller when _dl_fixup fails on hppa
Date: Sun, 23 Jul 2017 23:58:00 -0000 [thread overview]
Message-ID: <B509681B-F4FB-459D-950B-DEB8131115A4@bell.net> (raw)
In-Reply-To: <b5579ffc-6401-6511-96c6-0d7bcf076c6c@redhat.com>
[-- Attachment #1: Type: text/plain, Size: 521 bytes --]
On 2017-07-17, at 7:11 AM, Carlos O'Donell wrote:
> On 07/17/2017 03:10 AM, Andreas Schwab wrote:
>> This information should be added in a comment.
>
> Dave, Thanks for all the fixes. I agree with Andreas, this should be
> documented with a comment.
I committed the attached comment update. I tied it to the BZ report so it's clear what we
are trying to fix.
At this time, Debian has a local patch which removes __gmon_start__ from crtn.S on hppa.
Dave
--
John David Anglin dave.anglin@bell.net
[-- Attachment #2: dl-trampoline-v3.S.d.txt --]
[-- Type: text/plain, Size: 1497 bytes --]
2017-07-23 John David Anglin <danglin@gcc.gnu.org>
[BZ 19170]
* sysdeps/hppa/dl-trampoline.S (_dl_runtime_resolve): Revise comment.
diff --git a/sysdeps/hppa/dl-trampoline.S b/sysdeps/hppa/dl-trampoline.S
index f1294a931f..3165c6f0e2 100644
--- a/sysdeps/hppa/dl-trampoline.S
+++ b/sysdeps/hppa/dl-trampoline.S
@@ -82,9 +82,19 @@ _dl_runtime_resolve:
bl _dl_fixup,%rp
copy %r21,%r19 /* set fixup func ltp */
- /* Sometimes a final executable may attempt to call an undefined
- weak function (e.g., __gmon_start__). Return if the function
- was not resolved by _dl_fixup */
+ /* While the linker will set a function pointer to NULL when it
+ encounters an undefined weak function, we need to dynamically
+ detect removed weak functions. The issue arises because a weak
+ __gmon_start__ function was added to shared executables to work
+ around issues in _init that are now resolved. The presence of
+ __gmon_start__ in every shared library breaks the linker
+ `--as-needed' option. This __gmon_start__ function does nothing
+ but removal is tricky. Depending on the binding, removal can
+ cause an application using it to fault. The call to _dl_fixup
+ returns NULL when a function isn't resolved. In order to help
+ with __gmon_start__ removal, we return directly to the caller
+ when _dl_fixup returns NULL. This check could be removed when
+ BZ 19170 is fixed. */
comib,= 0,%r28,1f
/* Load up the returned func descriptor */
prev parent reply other threads:[~2017-07-23 17:01 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-07-16 17:50 John David Anglin
2017-07-16 18:53 ` Andreas Schwab
2017-07-16 19:46 ` John David Anglin
2017-07-16 21:31 ` Andreas Schwab
2017-07-16 22:02 ` John David Anglin
2017-07-17 7:10 ` Andreas Schwab
2017-07-17 11:11 ` Carlos O'Donell
2017-07-17 11:52 ` John David Anglin
2017-07-23 23:58 ` John David Anglin [this message]
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=B509681B-F4FB-459D-950B-DEB8131115A4@bell.net \
--to=dave.anglin@bell.net \
--cc=carlos@redhat.com \
--cc=carlos@systemhalted.org \
--cc=libc-alpha@sourceware.org \
--cc=schwab@linux-m68k.org \
/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).