Re: [PATCH 1/2] addr2line: Iterate scopes for inline's parent function

[Mark Wielaard <mjw at redhat dot com>]

On Wed, 2014-12-10 at 11:41 -0800, Josh Stone wrote:
> The function which contains an inline might not be the immediate next
> die scope.  For instance, there may be a lexical scope in between.
> Instead, iterate the remaining scopes until an appropriate tag is found.
>
> +2014-12-10  Josh Stone  <jistone@redhat.com>
> +
> +	* addr2line.c (handle_address): Find the proper inline parents.
>
> --- a/src/addr2line.c
> +++ b/src/addr2line.c
> @@ -672,7 +672,22 @@ handle_address (const char *string, Dwfl *dwfl)
>  			continue;
>  
>  		      if (show_functions)
> -			print_diesym (&scopes[i + 1]);
> +			{
> +			  /* Search for the parent inline or function.  It
> +			     might not be directly above this inline -- e.g.
> +			     there could be a lexical_block in between.  */
> +			  for (int j = i + 1; j < nscopes; j++)
> +			    {
> +			      Dwarf_Die *parent = &scopes[j];
> +			      int tag = dwarf_tag (parent);
> +			      if (tag == DW_TAG_inlined_subroutine
> +				  || tag == DW_TAG_subprogram)

I believe technically you also want to match DW_TAG_entry_point. That
would match what eu-stack does. Admittedly GCC doesn't emit
DW_TAG_entry_point (the code is commented out in dwarf2out.c), but other
(fortran) compilers might. DWARF describes subprograms and entry_points
almost identically, both have a name and possibly a linkage_name, which
is what we care about here.

> +				{
> +				  print_diesym (parent);
> +				  break;
> +				}
> +			    }
> +			}
>  
>  		      src = NULL;
>  		      lineno = 0;

Looks good, but please add DW_TAG_entry_point above before pushing this
commit.

Thanks,

Mark