Re: PATCH: PR gold/13505: apply_relocation always use 64bit little endian relocation

[Ian Lance Taylor <iant at google dot com>]

"H.J. Lu" <hongjiu.lu@intel.com> writes:

> This patch replaces <64, false> with <size, big_endian> in
> apply_relocation.  Tested on Linux/x86-64 and Linux/ia32.  OK to install?

Cary, this is approved if it looks OK to you.

Ian

> 2011-12-15  H.J. Lu  <hongjiu.lu@intel.com>
>
> 	PR gold/13505
> 	* target-reloc.h (apply_relocation): Replace <64, false> with
> 	<size, big_endian>.
>
> diff --git a/gold/target-reloc.h b/gold/target-reloc.h
> index 464a3fa..6efcdf4 100644
> --- a/gold/target-reloc.h
> +++ b/gold/target-reloc.h
> @@ -423,17 +423,17 @@ apply_relocation(const Relocate_info<size, big_endian>* relinfo,
>  		 section_size_type view_size)
>  {
>    // Construct the ELF relocation in a temporary buffer.
> -  const int reloc_size = elfcpp::Elf_sizes<64>::rela_size;
> +  const int reloc_size = elfcpp::Elf_sizes<size>::rela_size;
>    unsigned char relbuf[reloc_size];
> -  elfcpp::Rela<64, false> rel(relbuf);
> -  elfcpp::Rela_write<64, false> orel(relbuf);
> +  elfcpp::Rela<size, big_endian> rel(relbuf);
> +  elfcpp::Rela_write<size, big_endian> orel(relbuf);
>    orel.put_r_offset(r_offset);
> -  orel.put_r_info(elfcpp::elf_r_info<64>(0, r_type));
> +  orel.put_r_info(elfcpp::elf_r_info<size>(0, r_type));
>    orel.put_r_addend(r_addend);
>  
>    // Setup a Symbol_value for the global symbol.
> -  const Sized_symbol<64>* sym = static_cast<const Sized_symbol<64>*>(gsym);
> -  Symbol_value<64> symval;
> +  const Sized_symbol<size>* sym = static_cast<const Sized_symbol<size>*>(gsym);
> +  Symbol_value<size> symval;
>    gold_assert(sym->has_symtab_index() && sym->symtab_index() != -1U);
>    symval.set_output_symtab_index(sym->symtab_index());
>    symval.set_output_value(sym->value());