Re: [PATCH] bfd: RISC-V: relax to gp in more cases.

[Palmer Dabbelt <palmer at dabbelt dot com>]

On Wed, Feb 22, 2017 at 8:03 PM, Palmer Dabbelt <palmer@dabbelt.com> wrote:
> From: Andrew Waterman <andrew@sifive.com>
>
> bfd/ChangeLog:
>
> 2017-02-21  Andrew Waterman  <andrew@sifive.com>
>
>         Relax to gp in more cases.
>         * elfnn-riscv.c (GP_NAME): New macro.
>         (riscv_global_pointer_value): Use it.
>         (_bfd_riscv_relax_lui): If symbol and global pointer are in same
>         output section, consider only that section's alignment.
> ---
>  bfd/elfnn-riscv.c | 16 +++++++++++++++-
>  1 file changed, 15 insertions(+), 1 deletion(-)
>
> diff --git a/bfd/elfnn-riscv.c b/bfd/elfnn-riscv.c
> index 6c14313..2ccfc5f 100644
> --- a/bfd/elfnn-riscv.c
> +++ b/bfd/elfnn-riscv.c
> @@ -52,6 +52,10 @@
>  #define ELF_MAXPAGESIZE                        0x1000
>  #define ELF_COMMONPAGESIZE             0x1000
>
> +/* The global pointer's symbol name.  */
> +
> +#define GP_NAME "__global_pointer$"
> +
>  /* The RISC-V linker needs to keep track of the number of relocs that it
>     decides to copy as dynamic relocs in check_relocs for each symbol.
>     This is so that it can later discard them if they are found to be
> @@ -1463,7 +1467,7 @@ riscv_global_pointer_value (struct bfd_link_info *info)
>  {
>    struct bfd_link_hash_entry *h;
>
> -  h = bfd_link_hash_lookup (info->hash, "__global_pointer$", FALSE, FALSE, TRUE);
> +  h = bfd_link_hash_lookup (info->hash, GP_NAME, FALSE, FALSE, TRUE);
>    if (h == NULL || h->type != bfd_link_hash_defined)
>      return 0;
>
> @@ -2809,6 +2813,16 @@ _bfd_riscv_relax_lui (bfd *abfd,
>
>    BFD_ASSERT (rel->r_offset + 4 <= sec->size);
>
> +  if (gp)
> +    {
> +      /* If gp and the symbol are in the same output section, then
> +        consider only that section's alignment.  */
> +      struct bfd_link_hash_entry *h =
> +       bfd_link_hash_lookup (link_info->hash, GP_NAME, FALSE, FALSE, TRUE);
> +      if (h->u.def.section->output_section == sym_sec->output_section)
> +       max_alignment = (bfd_vma) 1 << sym_sec->output_section->alignment_power;
> +    }
> +
>    /* Is the reference in range of x0 or gp?
>       Valid gp range conservatively because of alignment issue.  */
>    if (VALID_ITYPE_IMM (symval)
> --
> 2.10.2
>

Committed.