[PATCH] Re: Don't relax undefined symbols on ppc.

[Douglas B Rupp]

Btw, the reason for my moving the branch_fixup struct to static was to 
be able to dump the fixups to the map file later on.  It was a customer 
enhancement request for us.

Do you think this is a worthwhile enhancement to merge?  I don't see how 
to do it easily without statically saving these fixups as they are created.

Suggestions welcomed.

--Doug



On 07/23/2018 10:40 PM, Alan Modra wrote:
> On Mon, Jul 23, 2018 at 08:59:49PM -0700, Douglas B Rupp wrote:
>> While I was working in gdb trying to figure this out I noted that
>> abfd->filename contained the name of the input object, so I believe your
>> contention that abfd is a constant cannot be correct.
>>
>> I'm assuming then that toff, which is indeed the symbol index, must be the
>> index w.r.t. the symbols in the input object, leading to the uniqueness
>> problem.
> 
> Yes, abfd can change between one invocation of ppc_elf_relax_section
> and the next, but the branch_fixups list starts from scratch each time
> too.
>