RFC: libiberty PATCH to disable demangling of ancient mangling schemes

[Jason Merrill <jason at redhat dot com>]

On Thu, Dec 6, 2018 at 11:14 AM Jason Merrill <jason@redhat.com> wrote:
>
> Looks good to me.  Independently, do you see a reason not to disable the
> old demangler entirely?

Like so.  Does anyone object to this?  These mangling schemes haven't
been relevant in decades.

commit 175323701ad923aa47f25e1e37fa1f3c487dc5ea
Author: Jason Merrill <jason@redhat.com>
Date:   Tue Nov 20 01:17:48 2018 -0500

            * cplus-dem.c (cplus_demangle): Turn off the old demangler.

diff --git a/libiberty/cplus-dem.c b/libiberty/cplus-dem.c
index 4f29d54d089..8ee23b2fe71 100644
--- a/libiberty/cplus-dem.c
+++ b/libiberty/cplus-dem.c
@@ -267,6 +267,7 @@ const struct demangler_engine libiberty_demanglers[] =
       "Automatic selection based on executable"
   }
   ,
+#ifdef OLD_DEMANGLERS
   {
     GNU_DEMANGLING_STYLE_STRING,
       gnu_demangling,
@@ -297,10 +298,11 @@ const struct demangler_engine libiberty_demanglers[] =
       "EDG style demangling"
   }
   ,
+#endif
   {
     GNU_V3_DEMANGLING_STYLE_STRING,
     gnu_v3_demangling,
-    "GNU (g++) V3 ABI-style demangling"
+    "GNU (g++) V3 (Itanium C++ ABI) style demangling"
   }
   ,
   {
@@ -915,8 +917,12 @@ cplus_demangle (const char *mangled, int options)
 	return ret;
     }
 
+#if OLD_DEMANGLERS
+  /* People have been busily breaking the old demangler with fuzzers
+     (CVE-2018-12641 etc), so let's turn it off.  */
   ret = internal_cplus_demangle (work, mangled);
   squangle_mop_up (work);
+#endif
   return (ret);
 }