At Wed, 12 May 2004 09:38:05 -0400,
Daniel Jacobowitz wrote:
> >     /usr/bin/objcopy: debian/libc6.1/usr/lib/debug//lib/ld-2.3.2.so: File truncated
> >     dh_strip: command returned error code 256
> >     148 blocks
> >     /usr/bin/objcopy: ../libc6.1-dbg/usr/lib/debug/./lib/libpthread-0.10.so: File 
> > truncated
> >     /usr/bin/objcopy: ./lib/stUl2lFk: File truncated
> > 
> > So it was failed to strip libc6.1 files.  Moreover, on alpha, objcopy
> > does not work with the current debian binutils.  See #234021.
> > 
> > I have not checked the latest binutils yet on alpha.  Daniel, did you
> > know this bug was already fixed in the latest binutils?  If not, is it
> > difficult to fix?
> 
> I don't know, but this looks like a bug I encountered on ARM.

Thanks for your suggestion.  Unfortunatelly 2.3.2.ds1-12 on ARM and
IA64 are still missing.  At this point, this problem affected only
alpha (i386, ppc, sparc, s390, hppa, mipsel, mips, m68k are ok).

> If it's the same bug it hasn't been fixed yet but I had a potential
> patch from Alan:

I checked the latest debian binutils with this patch.  The result is
quite positive.  It works objcopy with the debian-glibc's -R .section
option:


alpha:~/binutils/binutils-2.14.90.0.7/builddir-single/binutils> cp /lib/libc-2.3.2.so .
alpha:~/binutils/binutils-2.14.90.0.7/builddir-single/binutils> /usr/bin/objcopy 
libc-2.3.2.so
/usr/bin/objcopy: stZIeXU2: File truncated
alpha:~/binutils/binutils-2.14.90.0.7/builddir-single/binutils> ls -al libc-2.3.2.so
-rw-r--r--    1 gotom    gotom    18832508 May 12 23:34 libc-2.3.2.so
alpha:~/binutils/binutils-2.14.90.0.7/builddir-single/binutils> ./objcopy libc-2.3.2.so
alpha:~/binutils/binutils-2.14.90.0.7/builddir-single/binutils> ls -al libc-2.3.2.so
-rw-r--r--    1 gotom    gotom    18832516 May 12 23:35 libc-2.3.2.so
alpha:~/binutils/binutils-2.14.90.0.7/builddir-single/binutils> ./objcopy -R 
.debug_info -R .debug_aranges -R .debug_pubnames -R .debug_abbrev -R .debug_line -R 
.debug_str -R .debug_ranges -R .comment -R .note ./libc-2.3.2.so
alpha:~/binutils/binutils-2.14.90.0.7/builddir-single/binutils> ls -al libc-2.3.2.so
-rw-r--r--    1 gotom    gotom     2069652 May 12 23:35 libc-2.3.2.so
alpha:~/binutils/binutils-2.14.90.0.7/builddir-single/binutils> cp /lib/libc-2.3.2.so .
alpha:~/binutils/binutils-2.14.90.0.7/builddir-single/binutils> ./strip-new 
./libc-2.3.2.so
alpha:~/binutils/binutils-2.14.90.0.7/builddir-single/binutils> ls -al libc-2.3.2.so
-rw-r--r--    1 gotom    gotom     1622408 May 12 23:35 libc-2.3.2.so


> Index: bfd/elf.c
> ===================================================================
> RCS file: /cvs/src/src/bfd/elf.c,v
> retrieving revision 1.218
> diff -u -p -r1.218 elf.c
> --- bfd/elf.c 5 Mar 2004 11:26:04 -0000       1.218
> +++ bfd/elf.c 18 Mar 2004 10:14:34 -0000
> @@ -4255,9 +4255,7 @@ assign_file_positions_except_relocs (bfd
>             off = _bfd_elf_assign_file_position_for_section (hdr, off,
>                                                              FALSE);
>           }
> -       else if (hdr->sh_type == SHT_REL
> -                || hdr->sh_type == SHT_RELA
> -                || hdr == i_shdrpp[tdata->symtab_section]
> +       else if (hdr == i_shdrpp[tdata->symtab_section]
>                  || hdr == i_shdrpp[tdata->symtab_shndx_section]
>                  || hdr == i_shdrpp[tdata->strtab_section])
>           hdr->sh_offset = -1;

James, could you look at to apply this patch for alpha?  We need this
patch for libc6.1 on alpha (and maybe arm).

> but someone should try cvs binutils and/or that patch.

I tried the latest cvs binutils, but objcopy failed:

alpha:~/toolchain/build.binutils.040512.1/binutils> cp /lib/libc-2.3.2.so .
alpha:~/toolchain/build.binutils.040512.1/binutils> ./objcopy -R .debug_info -R 
.debug_aranges -R .debug_pubnames -R .debug_abbrev -R .debug_line -R .debug_str -R 
.debug_ranges -R .comment -R .note ./libc-2.3.2.so
./objcopy: ./stTA4Egz: File truncated


Regards,
-- gotom


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Reply via email to