This is the mail archive of the
binutils@sources.redhat.com
mailing list for the binutils project.
Re: prelink breaks 6 ld tests on ppclinux
- From: Daniel Jacobowitz <drow at mvista dot com>
- To: "H. J. Lu" <hjl at lucon dot org>
- Cc: Jack Howarth <howarth at bromo dot msbb dot uc dot edu>, binutils at sources dot redhat dot com,jakub at redhat dot com
- Date: Thu, 8 May 2003 21:42:03 -0400
- Subject: Re: prelink breaks 6 ld tests on ppclinux
- References: <200305090127.VAA60507@bromo.msbb.uc.edu> <20030508183110.A23001@lucon.org>
On Thu, May 08, 2003 at 06:31:10PM -0700, H. J. Lu wrote:
> On Thu, May 08, 2003 at 09:27:01PM -0400, Jack Howarth wrote:
> > I have tracked down the origin of the 6 unexpected
> > ld testsuite failures I was seeing on my debian ppc sid
> > box. It turns out that these tests fail if my box is
> > prelinked and don't if I unprelink. The tests in question
> > are...
> >
> > FAIL: ELF weak func first DSO
> > FAIL: ELF weak func last DSO
> > FAIL: ELF weak data first DSO
> > FAIL: ELF weak data last DSO
> > FAIL: ELF weak data first DSO common
> > FAIL: ELF weak data last DSO common
> >
> > I can reproduce this behavior (passes on unprelinked and
> > unexpected failures on prelinked) on both debian ppc sid
> > and YDL ppclinux (RedHat 8.0-based).
>
> It sounds a prelink bug to me. Jakub, how does prelink treat weak
> symbol?
Prelink asks ld.so what the symbol resolves to. Jack, have you tried
re-prelinking using a recent glibc? The search order changed recently.
I'm really surprised by this report, since the DSOs generated during
the binutils testsuite are not prelinked. It might be a bug in the
conflict resolution code in glibc but I can't see how that would
happen.
--
Daniel Jacobowitz
MontaVista Software Debian GNU/Linux Developer