Bootstrapping failure

Bootstrapping failure

[Luc Teirlinck]

Recent changes to progmodes/mixal-mode.el break bootstrapping:

Compiling /home/teirllm/emacscvsdir/emacs/lisp/./progmodes/mixal-mode.el

End of make bootstrap output:

In toplevel form:
progmodes/mixal-mode.el:123:1:Error: Lisp nesting exceeds
`max-lisp-eval-depth'
make[1]: *** [compile] Error 1
make[1]: Leaving directory `/home/teirllm/emacscvsdir/emacs/lisp'
make: *** [bootstrap-build] Error 2

Re: Bootstrapping failure

[Stefan Monnier]

> Recent changes to progmodes/mixal-mode.el break bootstrapping:
> Compiling /home/teirllm/emacscvsdir/emacs/lisp/./progmodes/mixal-mode.el

> End of make bootstrap output:

> In toplevel form:
> progmodes/mixal-mode.el:123:1:Error: Lisp nesting exceeds
> `max-lisp-eval-depth'

I've just installed a change that should fix it,


        Stefan

Re: Bootstrapping failure

[Romain Francoise]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

> I've just installed a change that should fix it,

It doesn't.

-- 
Romain Francoise <romain@orebokech.com> | I used to think there is no
it's a miracle -- http://orebokech.com/ | future left at all.

Re: Bootstrapping failure

[Stefan Monnier]

>> I've just installed a change that should fix it,
> It doesn't.

Grrr!!!

OK, here is the problem: mixal-mode.el has now a big constant built with
a backquote.  Originally, it blew up the byte-compiler's stack because the
backquote macroexpands into a very deep expression.  So I wrapped the thing
in an eval-when-compile because there's really no point in byte-compiling
the big expression since it actually returns a constant value and can be
evaluated quickly once and for all without being byte-compiled.

But as it turns out, the definition of eval-when-compile (in
byte-compile-initial-macro-environment) is

  (byte-compile-eval (byte-compile-top-level (cons 'progn body)))

I.e. it byte-compiles before evaluating.  So my "fix" didn't fix anything
since the deep expression still gets byte-compiled.  I find this behavior to
be wrong, and indeed if we change it to just

  (byte-compile-eval (cons 'progn body))

the problem goes away.  I find it kind of silly to byte-compile the content
of an eval-when-compile since the bytecode will only be evaluated once, but
maybe I'm missing something?  Note that this behavior dates back to version
1.1 of bytecomp.el (Jul 92).

In any case if the patch below is not acceptable, I guess I'll
replace the big backquote with the result of its evaluation.


        Stefan


--- bytecomp.el	22 aoû 2005 10:22:50 -0400	2.178
+++ bytecomp.el	13 oct 2005 17:07:10 -0400	
@@ -429,8 +429,7 @@
 ;; 			       (apply 'byte-compiler-options-handler forms)))
     (eval-when-compile . (lambda (&rest body)
 			   (list 'quote
-				 (byte-compile-eval (byte-compile-top-level
-						     (cons 'progn body))))))
+				 (byte-compile-eval (cons 'progn body)))))
     (eval-and-compile . (lambda (&rest body)
 			  (byte-compile-eval-before-compile (cons 'progn body))
 			  (cons 'progn body))))

Re: Bootstrapping failure

[Kim F. Storm]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>>> I've just installed a change that should fix it,
>> It doesn't.
>
> In any case if the patch below is not acceptable, I guess I'll
> replace the big backquote with the result of its evaluation.

Which is a trivial change ... so why not do it anyway?

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: Bootstrapping failure

[Kenichi Handa]

In article <jwvmzldunur.fsf-monnier+emacs@gnu.org>, Stefan Monnier <monnier@iro.umontreal.ca> writes:

> In any case if the patch below is not acceptable, I guess I'll
> replace the big backquote with the result of its evaluation.

It seems that this (dirty) trick works.  But, I'm not sure
it's worth using such a trick.

(eval-when-compile
  (defvar orig-max-lisp-eval-depth max-lisp-eval-depth)
  (setq max-lisp-eval-depth 3000))

(defvar mixal-operation-codes-alist
  ;; FIXME: the codes FADD, FSUB, FMUL, FDIV, JRAD, and FCMP were in
  ;; mixal-operation-codes but not here.  They should probably be added here.
  ...)

(eval-when-compile
  (setq max-lisp-eval-depth orig-max-lisp-eval-depth))

---
Kenichi Handa
handa@m17n.org

Bootstrapping failure

[Leo Liu]

I can't seem to bootstrap the git master branch (on centos7):

  Loading emacs-lisp/byte-run (source)...
  Loading emacs-lisp/backquote (source)...
  Loading subr (source)...
  Loading version (source)...
  Loading widget (source)...
  Loading custom (source)...
  Loading emacs-lisp/map-ynp (source)...
  Loading international/mule (source)...
  Loading international/mule-conf (source)...
  Loading env (source)...
  Loading format (source)...
  Loading bindings (source)...
  Symbol's function definition is void: read-kbd-macro
  make[2]: *** [bootstrap-emacs] Error 1
  make[2]: Leaving directory `/home/shared/sources/emacs/src'
  make[1]: *** [src] Error 2
  make[1]: Leaving directory `/home/shared/sources/emacs'
  make: *** [bootstrap] Error 2

Re: Bootstrapping failure

[Eli Zaretskii]

> From: Leo Liu <sdl.web@gmail.com>
> Date: Fri, 14 Nov 2014 13:54:51 +0800
> 
> I can't seem to bootstrap the git master branch (on centos7):
> 
>   Loading emacs-lisp/byte-run (source)...
>   Loading emacs-lisp/backquote (source)...
>   Loading subr (source)...
>   Loading version (source)...
>   Loading widget (source)...
>   Loading custom (source)...
>   Loading emacs-lisp/map-ynp (source)...
>   Loading international/mule (source)...
>   Loading international/mule-conf (source)...
>   Loading env (source)...
>   Loading format (source)...
>   Loading bindings (source)...
>   Symbol's function definition is void: read-kbd-macro
>   make[2]: *** [bootstrap-emacs] Error 1
>   make[2]: Leaving directory `/home/shared/sources/emacs/src'
>   make[1]: *** [src] Error 2
>   make[1]: Leaving directory `/home/shared/sources/emacs'
>   make: *** [bootstrap] Error 2

Please try again.  (Lars should learn not to use the fancy 'kbd'
thingy in files that are pre-loaded.)

Re: Bootstrapping failure

[Leo Liu]

On 2014-11-14 15:36 +0800, Eli Zaretskii wrote:
> Please try again.  (Lars should learn not to use the fancy 'kbd'
> thingy in files that are pre-loaded.)

thanks! Leo