Re: missing GC protection in Fbyte_code

[Richard Stallman]

    The comment for Bvarbind is a bit more explicit and mentions that it's
    because the code can call Fsignal which itself can GC.  I think the reason
    why it can GC is because of Vsignal_hook_function (used by edebug).

Fsignal doesn't return, so the local variables of all frames within
the innermost error handler or Lisp call frame are not going to be
used again.  Therefore, there is no need to protect any of them on
Fsignal's account.

If it is really necessary to gcpro because you call a function that
calls Fsignal, we would need to add lots of gcpros all around.
So I think it would be better to change the mechanism to make
this unnecessary.