your stabsread patch

your stabsread patch

[Daniel Jacobowitz]

...is not correct. 

Look at the debug output for a v3/stabs static method.  Note that the
method is in fact stubbed!  The physname is all there (which will be
handled in gdb_mangle_name after your last patch) and the return type
is there, but the arguments are missing.

Feel free to argue that this is GCC's bug instead of ours.  I think
both should be fixed.

[I would have replied, but I have technical difficulties with my mail
right now.]

-- 
Daniel Jacobowitz                           Carnegie Mellon University
MontaVista Software                         Debian GNU/Linux Developer

Re: your stabsread patch

[Peter.Schauer]

But my patch still sets the is_stub flag in this case, as the full physname
(which is output by gcc3) does not match the main_fn_name.
And together with the last gdb_mangle_name patch, check_stub_method will
do the right thing with v3/stabs.

Or am I missing something else ?

> ...is not correct. 
> 
> Look at the debug output for a v3/stabs static method.  Note that the
> method is in fact stubbed!  The physname is all there (which will be
> handled in gdb_mangle_name after your last patch) and the return type
> is there, but the arguments are missing.
> 
> Feel free to argue that this is GCC's bug instead of ours.  I think
> both should be fixed.
> 
> [I would have replied, but I have technical difficulties with my mail
> right now.]
> 
> -- 
> Daniel Jacobowitz                           Carnegie Mellon University
> MontaVista Software                         Debian GNU/Linux Developer
> 
> 


-- 
Peter Schauer			pes@regent.e-technik.tu-muenchen.de

Re: your stabsread patch

[Daniel Jacobowitz]

I am going to stop trying to review patches when I don't have time to
look at them properly.  You're quite correct.  The situation is
somewhat suboptimal, as static methods will now 'always' be marked
stubbed in v3; but we'll return to that if/when GCC fixes debug output. 
No further objection; I'm not sure I can approve C++-specific patches to
stabsread.c, though.

On Wed, Feb 27, 2002 at 10:05:26PM +0100, Peter.Schauer wrote:
> But my patch still sets the is_stub flag in this case, as the full physname
> (which is output by gcc3) does not match the main_fn_name.
> And together with the last gdb_mangle_name patch, check_stub_method will
> do the right thing with v3/stabs.
> 
> Or am I missing something else ?
> 
> > ...is not correct. 
> > 
> > Look at the debug output for a v3/stabs static method.  Note that the
> > method is in fact stubbed!  The physname is all there (which will be
> > handled in gdb_mangle_name after your last patch) and the return type
> > is there, but the arguments are missing.
> > 
> > Feel free to argue that this is GCC's bug instead of ours.  I think
> > both should be fixed.
> > 
> > [I would have replied, but I have technical difficulties with my mail
> > right now.]
> > 
> > -- 
> > Daniel Jacobowitz                           Carnegie Mellon University
> > MontaVista Software                         Debian GNU/Linux Developer
> > 
> > 
> 
> 
> -- 
> Peter Schauer			pes@regent.e-technik.tu-muenchen.de
> 

-- 
Daniel Jacobowitz                           Carnegie Mellon University
MontaVista Software                         Debian GNU/Linux Developer