Re: adding namespace support to GDB

[Daniel Berlin <dberlin at dberlin dot org>]

On Fri, 23 Aug 2002, Daniel Jacobowitz wrote:

> On Fri, Aug 23, 2002 at 02:48:46PM -0400, Daniel Berlin wrote:
> > On Fri, 23 Aug 2002, Daniel Jacobowitz wrote:
> > 
> > > On Fri, Aug 23, 2002 at 11:41:06AM -0700, David Carlton wrote:
> > > > In article <3D6677D3.6E84743C@caldera.com>, Petr Sorfa
> > > > <petrs@caldera.com> writes:
> > > > 
> > > > > Well to help things along I will be submitting a DWARF patch that
> > > > > will supported imported declarations which are essential for FORTRAN
> > > > > modules and C++ namespaces.
> > > > 
> > > > Great, I look forward to reading it.
> > > > 
> > > > > This is essential for the "using" commands in either language (of
> > > > > course the compiler needs to generate the correct DWARF
> > > > > ;o)).
> > > > 
> > > > Yes, well, there is always that.  Certainly it seems like a solution
> > > > for C++ will initially have to get recreate namespace info from
> > > > DW_AT_MIPS_linkage_name, and there's no way that we'll be able to
> > > > allow users to use symbol names as if all the appropriate using
> > > > directives were in effect, since that information simply isn't in the
> > > > debug information that GCC is currently producing.  (Though that's not
> > > > the end of the world: we should be able to do name lookup using C++'s
> > > > name resolution rules based on the enclosing function and its
> > > > arguments, presumably.)
> > > 
> > > C++ certainly does _not_ need to reconstruct from
> > > DW_AT_MIPS_linkage_name.  Everything we need should be there in
> > > DW_TAG_namespace and DW_TAG_structure/class markers; I have patches to
> > > use this information, with the caveats I noted in my other message.
> > 
> > If you hunt through the gcc patch archives, you'll find my patches to 
> > produce DW_AT_namespace, approved twice by Jason Merrill.
> > 
> > Should still apply cleanly.
> 
> Huh.  I could have sworn what we need was already there; am I on crack?

Must be, i never committed it.
Because GDB's wouldn't just ignore it, it would ignore all the children, 
so without a GDB that supports it, it severely affects the debuggability.
 
> It sure looks like it from reading this debug output; I'm going to have
> to find my notes... hope I still have them.
I guarantee you it doesn't produce DW_AT_namespace right now.
I'll resubmit the patch again today if you like.

> 
> (And is DW_AT_namespace a dwarf3 thing?  I don't see it in my v2 spec.)

Yup.

> 
>