[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
bug#34757: Invalid bytecode from byte compiler
From: |
Eli Zaretskii |
Subject: |
bug#34757: Invalid bytecode from byte compiler |
Date: |
Fri, 15 Mar 2019 10:08:12 +0200 |
> From: Pip Cet <pipcet@gmail.com>
> Date: Fri, 8 Mar 2019 21:13:37 +0000
> Cc: 34757@debbugs.gnu.org
>
> On Tue, Mar 5, 2019 at 3:30 PM chuntaro <chuntaro@sakura-games.jp> wrote:
> >
> > Invalid bytecode is output and error occurs when executed.
>
> If I'm looking at this correctly, the problem is that the byte code
> which is generated in an intermediate step:
>
> 0 constant 2
> 1 constant print
> 2 constant 1
> 3 call 1
> 4 discard
> 5 constant 3
> 6 return
>
> is considered a "trivial function" by byte-compile-out-toplevel, which
> assumes that all values on the stack are used by the call.
>
> We could fix byte-compile-out-toplevel to properly analyze how many
> stack arguments the call takes, but this patch simply treats forms
> like this as nontrivial:
>
> diff --git a/lisp/emacs-lisp/bytecomp.el b/lisp/emacs-lisp/bytecomp.el
> index 0b8f8824b4c..4e54e08ce14 100644
> --- a/lisp/emacs-lisp/bytecomp.el
> +++ b/lisp/emacs-lisp/bytecomp.el
> @@ -3025,6 +3025,7 @@ byte-compile-out-toplevel
> (or (null (cdr rest))
> (and (memq output-type '(file progn t))
> (cdr (cdr rest))
> + (eql (length body) (cdr (car rest)))
> (eq (car (nth 1 rest)) 'byte-discard)
> (progn (setq rest (cdr rest)) t))))
> (setq maycall nil) ; Only allow one real function call.
Stefan, any comments?