Re: [patch] smart pointer support

[sami wagiaalla <swagiaal at redhat dot com>]

On 08/05/2010 06:57 PM, Ulrich Weigand wrote:
> sami wagiaalla wrote:
>
> > 	* infcall.c (call_function_by_hand): Set VALUE_LVAL of the
> > 	returned value to lval_memory.
>
>
> > @@ -1026,6 +1026,7 @@ When the function is done executing, GDB will silently stop."),
> >   	    retval = allocate_value (values_type);
> >   	    gdbarch_return_value (gdbarch, value_type (function), values_type,
> >   				  retbuf, value_contents_raw (retval), NULL);
> > +	    VALUE_LVAL (retval) = lval_memory;
> >   	    break;
>
> This part looks broken.  Setting VALUE_LVAL to lval_memory implies the
> value actually *has* location information pointing to memory, that is
> in particular, a memory address.
>
> Since this value doesn't have that information, and in general it may
> not even exist (function return values may be passed in registers), I
> don't see how this can be valid here.

Hmm I see. I was under the impression that here the value is copied from 
the inferior's return location to gdb memory. As the patch stands (with 
the kfail) this can be removed, but I am actually looking for guidance 
on how to ensure that this value is available for gdb to perform further 
function calls that utilize it, especially when the return value is a 
copy not a reference. As is described in this bug: 
http://sourceware.org/bugzilla/show_bug.cgi?id=11606