GNU bug report logs - #49507
28.0.50; macOS: Symbol’s value as variable is void: lock-file-name-transforms

Previous Next

Full log

Message #28 received at 49507 <at> debbugs.gnu.org (full text, mbox):

From: Naofumi Yasufuku <naofumi <at> yasufuku.dev>
To: Lars Ingebrigtsen <larsi <at> gnus.org>
Cc: 49507 <at> debbugs.gnu.org
Subject: Re: bug#49507: 28.0.50; macOS: Symbol’s value as variable is void:
 lock-file-name-transforms
Date: Sun, 11 Jul 2021 07:42:15 +0900

Lars Ingebrigtsen <larsi <at> gnus.org> writes:

> Naofumi Yasufuku <naofumi <at> yasufuku.dev> writes:
>
>> Precondition:
>> - Repository revision: 9ce6541ac9710933beca7f9944087fe4849d5ae9
>> - macOS
>> - $ mv ~/.emacs.d/eln-cache ~/.emacs.d/eln-cache.old
>>
>> Emacs doesn't start due to the following error:
>>   $ ./src/emacs -Q
>>   Symbol’s value as variable is void: lock-file-name-transforms
>>   $
>
> [...]
>
>> -  return call1 (Qmake_lock_file_name, Fexpand_file_name (fn, Qnil));
>> +  Lisp_Object func = intern ("make-lock-file-name");
>> +  if (NILP (Fboundp (func)))
>> +    return Qnil;
>> +  return call1 (func, Fexpand_file_name (fn, Qnil));
>
> Well, that code was buggy (as Michael pointed out) -- Fboundp doesn't
> work on interned symbols, apparently?
>
> But in any case, make-lock-file-name and lock-file-name-transforms are
> defined in the same file ("files.el"), so what you're seeing here just
> shouldn't be possible -- it should complain about make-lock-file-name
> being undefined, not the variable.
>

I found that the following simple `lock-file-name-transforms' change
can also avoid this error:

----
 lisp/files.el | 1 -
 1 file changed, 1 deletion(-)

diff --git a/lisp/files.el b/lisp/files.el
index 0dfcab8f89..894a06e6e7 100644
--- a/lisp/files.el
+++ b/lisp/files.el
@@ -424,7 +424,6 @@ lock-file-name-transforms
   :type '(repeat (list (regexp :tag "Regexp")
                        (string :tag "Replacement")
 		       (boolean :tag "Uniquify")))
-  :initialize 'custom-initialize-delay
   :version "28.1")
 
 (defvar auto-save--timer nil "Timer for `auto-save-visited-mode'.")
----


> So I think this points to there being something odd going on in your
> build tree.  Are you sure you have no private modifications in the tree?
> Do you still see this issue from a freshly checked out tree?

I've tested with fresh build of new checked out under both macOS and
linux (Ubuntu 21.04).  The result is same.

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.