This is the mail archive of the
binutils@sources.redhat.com
mailing list for the binutils project.
RE: [PATCH] MIPS/ld: Calculate DT_RELSZ correctly.
- From: "David Daney" <ddaney at avtrex dot com>
- To: "Daniel Jacobowitz" <drow at false dot org>,"Thiemo Seufer" <ths at networkno dot de>
- Cc: <binutils at sources dot redhat dot com>
- Date: Sat, 28 May 2005 15:11:28 -0700
- Subject: RE: [PATCH] MIPS/ld: Calculate DT_RELSZ correctly.
I think I do have write Binutils write access, but I have never tried it.
I am fine with Daniel committing this now, or I can try to commit on Tuesday.
Your call.
Sorry for the screwed up mail format, but I am using a web client that is broken.
David Daney.
-----Original Message-----
From: Daniel Jacobowitz [mailto:drow@false.org]
Sent: Sat 5/28/2005 2:55 PM
To: Thiemo Seufer
Cc: David Daney; binutils@sources.redhat.com
Subject: Re: [PATCH] MIPS/ld: Calculate DT_RELSZ correctly.
On Sat, May 28, 2005 at 02:42:53PM +0200, Thiemo Seufer wrote:
> David Daney wrote:
> > As outlined in this thread:
> >
> > http://sourceware.org/ml/binutils/2005-05/msg00624.html
> >
> > The 2.16 ld cannot correctly link the libgcj component of gcc 4.0.0.
> >
> > The problem is that DT_RELSZ was being calculated before all of the
> > relocations had been added to the dynamic relocations section. This
> > resulted in a value of DT_RELSZ that was too small. The result was that
> > the runtime loader did not do all the relocations that were needed.
> >
> >
> > 2005-05-23 David Daney <ddaney@avtrex.com>
> >
> > * elfxx-mips.c (_bfd_mips_elf_finish_dynamic_sections): Move
> > calculation of DT_RELSZ to occur after all dynamic relocations
> > are created.
> >
> >
> > Tested on the CVS HEAD mipsel-linux cross running on i686-pc-linux-gnu
> > with make -k check showing regressions.
> >
> > O.K. to commit? Should this go on the 2.16 branch as well?
>
> Ok.
And 2.16. I don't think David has write access; I'll take care of it.
--
Daniel Jacobowitz
CodeSourcery, LLC