The relax finalize pass has broken relax for h8300-coff
Alan Modra
amodra@bigpond.net.au
Mon Dec 1 23:18:00 GMT 2003
On Wed, Nov 26, 2003 at 01:38:35PM -0800, H. J. Lu wrote:
> * bout.c (b_out_bfd_relax_section): Skip the relax finalize
> pass.
> * coff-mips.c (mips_relax_section): Likewise.
> * coff-sh.c (sh_relax_section): Likewise.
> * elf-m10200.c (mn10200_elf_relax_section): Likewise.
> * elf-m10300.c (mn10300_elf_relax_section): Likewise.
> * elf32-h8300.c (elf32_h8_relax_section): Likewise.
> * elf32-ip2k.c (ip2k_elf_relax_section): Likewise.
> * elf32-m68hc11.c (m68hc11_elf_relax_section): Likewise.
> * elf32-ppc.c (ppc_elf_relax_section): Likewise.
> * elf32-sh.c (sh_elf_relax_section): Likewise.
> * elf32-v850.c (v850_elf_relax_section): Likewise.
> * elf32-xstormy16.c (xstormy16_elf_relax_section): Likewise.
> * elf32-xtensa.c (elf_xtensa_relax_section): Likewise.
> * elf64-alpha.c (elf64_alpha_relax_section): Likewise.
> * elf64-mmix.c (x_elf_relax_section): Likewise.
> * elfxx-mips.c (_bfd_mips_relax_section): Likewise.
> * reloc16.c (bfd_coff_reloc16_relax_section): Likewise.
Instead of this, I suggest adding a need_relax_finalize flag to
link_info, set by the ia64 backend and tested in lang_process.
--
Alan Modra
IBM OzLabs - Linux Technology Centre
More information about the Binutils
mailing list