[Bug c/68839] [6 regression] ICE in contains_point at diagnostic-show-locus.c:340

["dmalcolm at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68839

David Malcolm <dmalcolm at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |dmalcolm at gcc dot gnu.org

--- Comment #3 from David Malcolm <dmalcolm at gcc dot gnu.org> ---
Thanks; I agree that it's a dup.

The ranges in question are:

$2 = {locus = 201728, src_range = {m_start = 201568, m_finish = 2147483641},
data = 0x0}
(gdb) call inform (201728, "locus")
pr68473-3.c: In function âfâ:
pr68473-3.c:11:19: note: locus
     else if (p->q != NULL) { }
                   ^

(gdb) call inform (201568, "m_start")
pr68473-3.c:11:14: note: m_start
     else if (p->q != NULL) { }
              ^

(gdb) call inform (2147483641, "m_finish")
pr68473-3.c:6:21: note: m_finish
 #define NULL (void*)0
                     ^

pr68473-3.c:11:22: note: in expansion of macro âNULLâ
     else if (p->q != NULL) { }
                      ^~~~

With the patch, it prints:

pr68473-3.c: In function âfâ:
pr68473-3.c:11:19: warning: duplicated âifâ condition [-Wduplicated-cond]
     else if (p->q != NULL) { }
                   ^

pr68473-3.c:10:14: note: previously used here
     if (p->q != NULL) { }
              ^

i.e. no range information. I believe this is degrading gracefully to
the old "just carets,  no ranges" result.


Perhaps ideally it might print some form of the ranges as they pertain to the
primary caret within the location:

pr68473-3.c: In function âfâ:
pr68473-3.c:11:19: warning: duplicated âifâ condition [-Wduplicated-cond]
     else if (p->q != NULL) { }
              ~~~~~^~~~~~~

pr68473-3.c:10:14: note: previously used here
     if (p->q != NULL) { }
         ~~~~~^~~~~~~

but making it do that would be a much more involved patch.