[Bug breakpoints/12803] Regression: const/volatile method functionslinespec `break'

["keiths at redhat dot com" <sourceware-bugzilla at sourceware dot org>]

http://sourceware.org/bugzilla/show_bug.cgi?id=12803

Keith Seitz <keiths at redhat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
         AssignedTo|unassigned at sourceware    |keiths at redhat dot com
                   |dot org                     |

--- Comment #1 from Keith Seitz <keiths at redhat dot com> 2011-05-24 16:35:54 UTC ---
(In reply to comment #0)
> (gdb) b 'C::m() const' 
> the class C does not have any method named m() const
> Hint: try 'C::m() const<TAB> or 'C::m() const<ESC-?>
> (Note leading single quote.)
> (gdb) b 'C::n() volatile' 
> the class C does not have any method named n()
> Hint: try 'C::n() volatile<TAB> or 'C::n() volatile<ESC-?>
> (Note leading single quote.)

"break 'C::m() const'" does work, but you must "set lang c++" first. This is
because keep_name_info in linespec.c is only called when the current language
is set to c++. We could change this to unconditionally call this; from
decode_compound, it should be relatively safe (assertion untested). What do you
think?

The volatile case is a bug: keep_name_info does not include it. When it is
added, then it works. I can prepare a patch and some tests for this.

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.