GNU bug report logs - #14831
cl-member warnings during bootstrap

Previous Next

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 14831 in the body.
You can then email your comments to 14831 AT debbugs.gnu.org in the normal way.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: Juanma Barranquero <lekktu <at> gmail.com>
To: Bug-Gnu-Emacs <bug-gnu-emacs <at> gnu.org>
Subject: cl-member warnings during bootstrap
Date: Tue, 9 Jul 2013 16:42:03 +0200

Package: emacs
Version: 24.3.50

During bootstrap, there are two cl-member related warnings:

In end of data:
international/mule-cmds.el:2981:1:Warning: the function `cl-member' is not
    known to be defined.

In end of data:
replace.el:2338:1:Warning: the function `cl-member' is not known to be
    defined.

and indeed, mule-cmds.elc and replace.elc contain references to
cl-member. Recompiling these files after bootstrap removes these
references.

Message #8 received at 14831 <at> debbugs.gnu.org (full text, mbox):

From: Glenn Morris <rgm <at> gnu.org>
To: Juanma Barranquero <lekktu <at> gmail.com>
Cc: 14831 <at> debbugs.gnu.org
Subject: Re: bug#14831: cl-member warnings during bootstrap
Date: Tue, 09 Jul 2013 11:54:29 -0400

Juanma Barranquero wrote:

> In end of data:
> international/mule-cmds.el:2981:1:Warning: the function `cl-member' is not
>     known to be defined.
>
> In end of data:
> replace.el:2338:1:Warning: the function `cl-member' is not known to be
>     defined.

It is not a real problem. It is due to

(or (load "cl-loaddefs" 'noerror 'quiet)
    ;; When bootstrapping, cl-loaddefs hasn't been built yet!
    (require 'cl-macs))

Loading cl-seq as well as cl-macs would make it go away.

Message #11 received at 14831 <at> debbugs.gnu.org (full text, mbox):

From: Juanma Barranquero <lekktu <at> gmail.com>
To: Glenn Morris <rgm <at> gnu.org>
Cc: 14831 <at> debbugs.gnu.org
Subject: Re: bug#14831: cl-member warnings during bootstrap
Date: Tue, 9 Jul 2013 17:56:03 +0200

On Tue, Jul 9, 2013 at 5:54 PM, Glenn Morris <rgm <at> gnu.org> wrote:

> It is not a real problem. It is due to

OK.

> Loading cl-seq as well as cl-macs would make it go away.

Any reason not to do so?

Message #14 received at 14831 <at> debbugs.gnu.org (full text, mbox):

From: Juanma Barranquero <lekktu <at> gmail.com>
To: Glenn Morris <rgm <at> gnu.org>
Cc: 14831 <at> debbugs.gnu.org
Subject: Re: bug#14831: cl-member warnings during bootstrap
Date: Mon, 15 Jul 2013 13:36:17 +0200

On Tue, Jul 9, 2013 at 5:54 PM, Glenn Morris <rgm <at> gnu.org> wrote:

> It is not a real problem. It is due to
>
> (or (load "cl-loaddefs" 'noerror 'quiet)
>     ;; When bootstrapping, cl-loaddefs hasn't been built yet!
>     (require 'cl-macs))
>
> Loading cl-seq as well as cl-macs would make it go away.

Hmm. I added (require 'cl-seq), bootstrapped, and the cl-member
warnings didn't go away.

Message #17 received at 14831 <at> debbugs.gnu.org (full text, mbox):

From: Glenn Morris <rgm <at> gnu.org>
To: Juanma Barranquero <lekktu <at> gmail.com>
Cc: 14831 <at> debbugs.gnu.org
Subject: Re: bug#14831: cl-member warnings during bootstrap
Date: Mon, 15 Jul 2013 12:10:08 -0400

Juanma Barranquero wrote:

>> Loading cl-seq as well as cl-macs would make it go away.
>
> Hmm. I added (require 'cl-seq), bootstrapped, and the cl-member
> warnings didn't go away.

Works for me (after changing cl-seq to provide itself).

Message #20 received at 14831 <at> debbugs.gnu.org (full text, mbox):

From: Juanma Barranquero <lekktu <at> gmail.com>
To: Glenn Morris <rgm <at> gnu.org>
Cc: 14831 <at> debbugs.gnu.org
Subject: Re: bug#14831: cl-member warnings during bootstrap
Date: Mon, 15 Jul 2013 18:23:45 +0200

On Mon, Jul 15, 2013 at 6:10 PM, Glenn Morris <rgm <at> gnu.org> wrote:

> Works for me (after changing cl-seq to provide itself).

Well, are you going to install the change, then, or is there any downside?

   J

Message #23 received at 14831 <at> debbugs.gnu.org (full text, mbox):

From: Glenn Morris <rgm <at> gnu.org>
To: Juanma Barranquero <lekktu <at> gmail.com>
Cc: 14831 <at> debbugs.gnu.org
Subject: Re: bug#14831: cl-member warnings during bootstrap
Date: Mon, 15 Jul 2013 15:48:08 -0400

Juanma Barranquero wrote:

> Well, are you going to install the change, then, or is there any downside?

I think it's up to Stefan, who authored the current state.
AFAIK, the warnings are harmless. Loading extra files slows down the
build by some fractional amount (that's the only downside I can see).
So maybe he prefers to leave it as-is.

Message #26 received at 14831 <at> debbugs.gnu.org (full text, mbox):

From: Juanma Barranquero <lekktu <at> gmail.com>
To: Glenn Morris <rgm <at> gnu.org>
Cc: 14831 <at> debbugs.gnu.org
Subject: Re: bug#14831: cl-member warnings during bootstrap
Date: Mon, 15 Jul 2013 21:50:21 +0200

> I think it's up to Stefan, who authored the current state.
> AFAIK, the warnings are harmless. Loading extra files slows down the
> build by some fractional amount (that's the only downside I can see).
> So maybe he prefers to leave it as-is.

OK, thanks.

Message #29 received at 14831 <at> debbugs.gnu.org (full text, mbox):

From: Stefan Monnier <monnier <at> IRO.UMontreal.CA>
To: Glenn Morris <rgm <at> gnu.org>
Cc: Juanma Barranquero <lekktu <at> gmail.com>, 14831 <at> debbugs.gnu.org
Subject: Re: bug#14831: cl-member warnings during bootstrap
Date: Mon, 15 Jul 2013 20:19:34 -0400

>> Well, are you going to install the change, then, or is there any downside?
> I think it's up to Stefan, who authored the current state.
> AFAIK, the warnings are harmless. Loading extra files slows down the
> build by some fractional amount (that's the only downside I can see).
> So maybe he prefers to leave it as-is.

It might be worth checking the impact on the bootstrap time.
I'd like to understand exactly where those `cl-member's come from.
I suspect the missing part is the (put 'cl-member 'compiler-macro
#'cl--compiler-macro-member) which is normally in cl-loaddefs.el.

Those residual cl-member calls may actually be a real problem, so maybe
this is not just a minor cosmetic fix.


        Stefan "who can't dig into it further now"

Message #32 received at 14831 <at> debbugs.gnu.org (full text, mbox):

From: Glenn Morris <rgm <at> gnu.org>
To: Stefan Monnier <monnier <at> IRO.UMontreal.CA>
Cc: Juanma Barranquero <lekktu <at> gmail.com>, 14831 <at> debbugs.gnu.org
Subject: Re: bug#14831: cl-member warnings during bootstrap
Date: Mon, 15 Jul 2013 21:37:44 -0400

Stefan Monnier wrote:

> I'd like to understand exactly where those `cl-member's come from.
> I suspect the missing part is the (put 'cl-member 'compiler-macro
> #'cl--compiler-macro-member) which is normally in cl-loaddefs.el.
>
> Those residual cl-member calls may actually be a real problem, so maybe
> this is not just a minor cosmetic fix.

Deleting cl-loadefs.el and batch byte-compiling a file with sole contents:

(defvar foo-bar nil)
(add-to-list 'foo-bar 1)

produces byte-code containing cl-member. Makes no sense to me...

You are correct that adding

(eval-and-compile (put 'cl-member 'compiler-macro#'cl--compiler-macro-member))

to the start of the file seems to fix it.

Message #35 received at 14831 <at> debbugs.gnu.org (full text, mbox):

From: Stefan Monnier <monnier <at> IRO.UMontreal.CA>
To: Glenn Morris <rgm <at> gnu.org>
Cc: Juanma Barranquero <lekktu <at> gmail.com>, 14831 <at> debbugs.gnu.org
Subject: Re: bug#14831: cl-member warnings during bootstrap
Date: Tue, 16 Jul 2013 04:57:42 -0400

>> I'd like to understand exactly where those `cl-member's come from.
>> I suspect the missing part is the (put 'cl-member 'compiler-macro
>> #'cl--compiler-macro-member) which is normally in cl-loaddefs.el.
>> 
>> Those residual cl-member calls may actually be a real problem, so maybe
>> this is not just a minor cosmetic fix.

> Deleting cl-loadefs.el and batch byte-compiling a file with sole contents:

> (defvar foo-bar nil)
> (add-to-list 'foo-bar 1)

> produces byte-code containing cl-member. Makes no sense to me...

Aha!  So it's the compiler-macro of add-to-list which turns an
add-to-list call into a cl-pushnew call, which is then macro-expanded to
something that uses cl-member, which should be turned into a plain
`member' by `cl--compiler-macro-member'.  So it can be fixed by changing
the compiler-macro of add-to-list not to use cl-pushnew.


        Stefan

Message #38 received at 14831 <at> debbugs.gnu.org (full text, mbox):

From: Glenn Morris <rgm <at> gnu.org>
To: Stefan Monnier <monnier <at> IRO.UMontreal.CA>
Cc: Juanma Barranquero <lekktu <at> gmail.com>, 14831 <at> debbugs.gnu.org
Subject: Re: bug#14831: cl-member warnings during bootstrap
Date: Tue, 16 Jul 2013 13:12:56 -0400

Stefan Monnier wrote:

> Aha!  So it's the compiler-macro of add-to-list which turns an
> add-to-list call into a cl-pushnew call, which is then macro-expanded to
> something that uses cl-member, which should be turned into a plain
> `member' by `cl--compiler-macro-member'.  So it can be fixed by changing
> the compiler-macro of add-to-list not to use cl-pushnew.

Or move the

(put 'cl-member 'compiler-macro #'cl--compiler-macro-member)

from cl-seq to cl-macs (and probably the same for cl-assoc).
That seems fine to me, since that is where cl--compiler-macro-member is
defined, but I guess you might want to keep that statement where
cl-member is defined instead.

Message #43 received at 14831-done <at> debbugs.gnu.org (full text, mbox):

From: Stefan Monnier <monnier <at> iro.umontreal.ca>
To: Glenn Morris <rgm <at> gnu.org>
Cc: Juanma Barranquero <lekktu <at> gmail.com>, 14831-done <at> debbugs.gnu.org
Subject: Re: bug#14831: cl-member warnings during bootstrap
Date: Mon, 22 Jul 2013 16:40:27 -0400

>> Aha!  So it's the compiler-macro of add-to-list which turns an
>> add-to-list call into a cl-pushnew call, which is then macro-expanded to
>> something that uses cl-member, which should be turned into a plain
>> `member' by `cl--compiler-macro-member'.  So it can be fixed by changing
>> the compiler-macro of add-to-list not to use cl-pushnew.

I installed the patch below which tightens things up a bit.


        Stefan


=== modified file 'lisp/subr.el'
--- lisp/subr.el	2013-07-19 12:18:16 +0000
+++ lisp/subr.el	2013-07-22 16:46:55 +0000
@@ -1498,9 +1498,10 @@
       ;; FIXME: Something like this could be used for `set' as well.
       (if (or (not (eq 'quote (car-safe list-var)))
               (special-variable-p (cadr list-var))
-              (and append compare-fn))
+              (not (macroexp-const-p append)))
           exp
         (let* ((sym (cadr list-var))
+               (append (eval append))
                (msg (format "`add-to-list' can't use lexical var `%s'; use `push' or `cl-pushnew'"
                             sym))
                ;; Big ugly hack so we only output a warning during
@@ -1513,13 +1514,17 @@
                           (when (assq sym byte-compile--lexical-environment)
                             (byte-compile-log-warning msg t :error))))
                (code
-                (if append
                     (macroexp-let2 macroexp-copyable-p x element
-                    `(unless (member ,x ,sym)
-                       (setq ,sym (append ,sym (list ,x)))))
+                  `(unless ,(if compare-fn
+                                (progn
                   (require 'cl-lib)
-                  `(cl-pushnew ,element ,sym
-                               :test ,(or compare-fn '#'equal)))))
+                                  `(cl-member ,x ,sym :test ,compare-fn))
+                              ;; For bootstrapping reasons, don't rely on
+                              ;; cl--compiler-macro-member for the base case.
+                              `(member ,x ,sym))
+                     ,(if append
+                          `(setq ,sym (append ,sym (list ,x)))
+                        `(push ,x ,sym))))))
           (if (not (macroexp--compiling-p))
               code
             `(progn

Previous Next

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