Re: [PATCH] gdb: track parents of nameless DIEs in cooked indexer

Tom Tromey <[email protected]> writes:

>>>>>> "Andrew" == Andrew Burgess <[email protected]> writes:
>
> Andrew> I found this patch while looking at PR gdb/33447, Tom Tromey had
> Andrew> suggest this patch might fix that bug, so I took the patch from the
> Andrew> mailing list, rebased it to current HEAD, checked the test still
> Andrew> passed, and then checked to see if it fixed PR gdb/33447.
>
> Thanks for resurrecting this.  I should have reviewed it ages ago.
>
> Andrew> +      else if (this_parent_entry != nullptr)
> Andrew> +	{
> Andrew> +	  /* Record this DIE's parent in the map so that deferred
> Andrew> +	     parent lookups through DW_AT_specification chains can
> Andrew> +	     find the parent even for DIEs that have no name and thus
> Andrew> +	     no cooked_index_entry.  */
>
> I just really, truly loathe this part of DWARF.
>
> Approved-By: Tom Tromey <[email protected]>

Pushed.

Thanks,
Andrew