[PATCH] Fix N^2 behavior in _bfd_dwarf2_find_symbol_bias

[Tom Tromey]

>>>>> "Alan" == Alan Modra <amodra@gmail.com> writes:

Alan> On Fri, Aug 16, 2019 at 07:19:41AM -0600, Tom Tromey wrote:
>> * dwarf2.c (_bfd_dwarf2_find_symbol_bias): Create hash table
>> holding symbols.

Alan> OK, but..

>> +  sym_hash = htab_create_alloc (10, hash_asymbol, eq_asymbol,
>> +				NULL, xcalloc, free);
>> +  /* FIXME: Do we need to scan the aranges looking for the lowest
>> +     pc value ?  */

Alan> the above comment doesn't make sense here.  It should stay where it
Alan> was in the func loop since it is referring to function aranges.

Thanks for the review.  I've moved the comment back and I'm going to
check this in now.

Tom