Re: [PATCH 0/2] Demangler crash handler

[Gary Benson <gbenson@redhat.com>]

Pedro Alves wrote:
> On 05/13/14 20:16, Gary Benson wrote:
> > Pedro Alves wrote:
> > > Or you could just pass that list through c++filt, the binutils
> > > program, which also uses the same demangler.  That might be
> > > faster, though obviously won't catch gdb-specific bugs.
> > 
> > c++filt would be enough to catch the crashes we've seen recently.
> > It would be best to have c++filt built with the latest libiberty
> > from GCC upstream, as two fixes have gone in recently.
> 
> The fix should have been merged to the binutils-gdb repo too, then
> it's just a matter of building binutils ToT.

Ah, I didn't realise that's where c++filt came from.

> Has it not been yet?  We need to make sure that's done.

One fix has been merged, the other hasn't been committed to GCC yet.
I only just noticed the latter, but I've pinged it so hopefully it
can go in today.  That one's not so important though, it's a failure
to demangle a symbol that's using a completely different mangling
scheme.

Cheers,
Gary

-- 
http://gbenson.net/