GNU bug report logs - #5191
replace ELCFILES in lisp/Makefile.in

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 5191 in the body.
You can then email your comments to 5191 AT debbugs.gnu.org in the normal way.

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

Message #5 received at submit <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Harald Maier <Harald <at> Maierh.de>
To: emacs-pretest-bug <at> gnu.org
Subject: 23.1.90; Make target compile-main fails under GNU/Linux Kernel 2.6.16
Date: Sat, 12 Dec 2009 11:28:48 +0100 (CET)

This bug report will be sent to the Free Software Foundation,
not to your local site managers!
Please write in English if possible, because the Emacs maintainers
usually do not have translators to read other languages for them.

Your bug report will be posted to the emacs-pretest-bug <at> gnu.org mailing list,
and to the gnu.emacs.bug news group.

Please describe exactly what actions triggered the bug
and the precise symptoms of the bug.  If you can, give
a recipe starting from `emacs -Q':

With the checkout today the build failed on a GNU/Linux Kernel
2.6.16. The problem ist the make target "compile-main" in the lisp 
directory.

$ LANG=en_US make compile-main
make[1]: Entering directory `/home/maierh/build/cvs/emacs/lisp'
make[1]: execvp: echo: Argument list too long
make[1]: *** [/home/maierh/build/cvs/emacs/lisp/mh-e/mh-alias.elc] Error 127
make[1]: Leaving directory `/home/maierh/build/cvs/emacs/lisp'
make: *** [compile-main] Error 2

Harald

If Emacs crashed, and you have the Emacs process in the gdb debugger,
please include the output from the following gdb commands:
    `bt full' and `xbacktrace'.
For information about debugging Emacs, please read the file
/usr/local/share/emacs/23.1.90/etc/DEBUG.


In GNU Emacs 23.1.90.1 (i686-pc-linux-gnu, GTK+ Version 2.8.11)
 of 2009-12-12 on nyx-s10
Windowing system distributor `The X.Org Foundation', version 11.0.60900000
Important settings:
  value of $LC_ALL: nil
  value of $LC_COLLATE: nil
  value of $LC_CTYPE: nil
  value of $LC_MESSAGES: nil
  value of $LC_MONETARY: nil
  value of $LC_NUMERIC: nil
  value of $LC_TIME: nil
  value of $LANG: de_DE.UTF-8
  value of $XMODIFIERS: @im=local
  locale-coding-system: utf-8-unix
  default enable-multibyte-characters: t

Message #8 received at 5191 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Glenn Morris <rgm <at> gnu.org>
To: 5191 <at> debbugs.gnu.org
Subject: Re: bug#5191: 23.1.90; Make target compile-main fails under GNU/Linux Kernel 2.6.16
Date: Sat, 12 Dec 2009 14:45:42 -0500

merge 5167 5191
severity 5167 serious
stop

Harald Maier wrote:

> $ LANG=en_US make compile-main
> make[1]: Entering directory `/home/maierh/build/cvs/emacs/lisp'
> make[1]: execvp: echo: Argument list too long

This answers my question from bug#5167.

Message #17 received at 5191 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Chong Yidong <cyd <at> stupidchicken.com>
To: Harald Maier <Harald <at> Maierh.de>
Cc: Glenn Morris <rgm <at> gnu.org>, 5191 <at> debbugs.gnu.org
Subject: Re: bug#5191: 23.1.90; Make target compile-main fails under GNU/Linux Kernel 2.6.16
Date: Sat, 12 Dec 2009 15:24:25 -0500

Does this patch fix it?  Instead of passing ELCFILES on the command
line, it saves it to a .elcfiles file, and the recursive Make includes
that.  (I am assuming the "-include" directive is supported, but we seem
to use it in src/Makefile already anyhow.)

*** emacs/lisp/Makefile.in.~1.215.~	2009-12-09 09:21:04.000000000 -0500
--- emacs/lisp/Makefile.in	2009-12-12 15:13:02.000000000 -0500
***************
*** 211,216 ****
--- 211,218 ----
  
  compile-first: $(COMPILE_FIRST)
  
+ -include $(lisp)/.elcfiles
+ 
  # ELCFILES is set dynamically in the recursive call from `compile-main'.
  compile-elcfiles: $(ELCFILES)
  
***************
*** 219,230 ****
  compile-main:
  	@wd=$(lisp); $(setwins); \
  	els=`echo "$$wins " | sed -e 's|/\./|/|g' -e 's|/\. | |g' -e 's| |/*.el |g'`; \
  	for el in $$els; do \
  	  test -f $$el || continue; \
  	  test ! -f $${el}c && GREP_OPTIONS= grep '^;.*no-byte-compile: t' $$el > /dev/null && continue; \
! 	  files="$$files $${el}c"; \
  	done; \
! 	$(MAKE) $(MFLAGS) compile-elcfiles EMACS=$(EMACS) ELCFILES="$$files"
  
  # Erase left-over .elc files that do not have a corresponding .el file.
  compile-clean:
--- 221,233 ----
  compile-main:
  	@wd=$(lisp); $(setwins); \
  	els=`echo "$$wins " | sed -e 's|/\./|/|g' -e 's|/\. | |g' -e 's| |/*.el |g'`; \
+ 	echo -n "ELCFILES=" > $(lisp)/.elcfiles; \
  	for el in $$els; do \
  	  test -f $$el || continue; \
  	  test ! -f $${el}c && GREP_OPTIONS= grep '^;.*no-byte-compile: t' $$el > /dev/null && continue; \
! 	  echo -n "$${el}c " >> $(lisp)/.elcfiles; \
  	done; \
! 	$(MAKE) $(MFLAGS) compile-elcfiles EMACS=$(EMACS)
  
  # Erase left-over .elc files that do not have a corresponding .el file.
  compile-clean:

Message #22 received at 5191 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Chong Yidong <cyd <at> stupidchicken.com>
To: Stefan Monnier  <monnier <at> iro.umontreal.ca>
Cc: Harald Maier <Harald <at> Maierh.de>, Glenn Morris <rgm <at> gnu.org>,
        5191 <at> debbugs.gnu.org
Subject: Re: bug#5191: 23.1.90; Make target compile-main fails under GNU/Linux Kernel 2.6.16
Date: Sat, 12 Dec 2009 16:00:45 -0500

> Does this patch fix it?  Instead of passing ELCFILES on the command
> line, it saves it to a .elcfiles file, and the recursive Make includes
> that.  (I am assuming the "-include" directive is supported, but we seem
> to use it in src/Makefile already anyhow.)

On second thought, I don't think this is a good idea.  Looks like the
BSDs use ".sinclude" for silent inclusion, and older systems don't even
support that.

If no better plan is forthcoming that can simultaneously cope with
ARG_MAX limitations and cross-Unix incompatibilities, I think it's
better to revert Makefile.in to the "ELCFILES abomination".  We can live
with it for a while longer.

Message #25 received at 5191 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Glenn Morris <rgm <at> gnu.org>
To: Chong Yidong <cyd <at> stupidchicken.com>
Cc: Stefan Monnier <monnier <at> iro.umontreal.ca>, Harald Maier <Harald <at> Maierh.de>,
        5191 <at> debbugs.gnu.org
Subject: Re: bug#5191: 23.1.90; Make target compile-main fails under GNU/Linux Kernel 2.6.16
Date: Sat, 12 Dec 2009 16:07:08 -0500

Chong Yidong wrote:

> On second thought, I don't think this is a good idea.  Looks like the
> BSDs use ".sinclude" for silent inclusion, and older systems don't even
> support that.

I think you are right that include is not portable. It's currently
only used in GNUstep builds. I don't think "echo -n" is portable
either. "printf" might be better, I'm not sure.

Message #30 received at 5191 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Jan Djärv <jan.h.d <at> swipnet.se>
To: Chong Yidong <cyd <at> stupidchicken.com>, 5191 <at> debbugs.gnu.org
Cc: Harald Maier <Harald <at> Maierh.de>
Subject: Re: bug#5191: 23.1.90;	Make target compile-main fails under GNU/Linux
 Kernel 2.6.16
Date: Sat, 12 Dec 2009 22:40:54 +0100

Chong Yidong skrev 2009-12-12 21.24:
> Does this patch fix it?  Instead of passing ELCFILES on the command
> line, it saves it to a .elcfiles file, and the recursive Make includes
> that.  (I am assuming the "-include" directive is supported, but we seem
> to use it in src/Makefile already anyhow.)

Note that in src/Makefile we only use it if we know we are using GNU Make.
Configure tests for GNU Make.

	Jan D.

>
> *** emacs/lisp/Makefile.in.~1.215.~	2009-12-09 09:21:04.000000000 -0500
> --- emacs/lisp/Makefile.in	2009-12-12 15:13:02.000000000 -0500
> ***************
> *** 211,216 ****
> --- 211,218 ----
>
>    compile-first: $(COMPILE_FIRST)
>
> + -include $(lisp)/.elcfiles
> +
>    # ELCFILES is set dynamically in the recursive call from `compile-main'.
>    compile-elcfiles: $(ELCFILES)
>
> ***************
> *** 219,230 ****
>    compile-main:
>    	@wd=$(lisp); $(setwins); \
>    	els=`echo "$$wins " | sed -e 's|/\./|/|g' -e 's|/\. | |g' -e 's| |/*.el |g'`; \
>    	for el in $$els; do \
>    	  test -f $$el || continue; \
>    	  test ! -f $${el}c&&  GREP_OPTIONS= grep '^;.*no-byte-compile: t' $$el>  /dev/null&&  continue; \
> ! 	  files="$$files $${el}c"; \
>    	done; \
> ! 	$(MAKE) $(MFLAGS) compile-elcfiles EMACS=$(EMACS) ELCFILES="$$files"
>
>    # Erase left-over .elc files that do not have a corresponding .el file.
>    compile-clean:
> --- 221,233 ----
>    compile-main:
>    	@wd=$(lisp); $(setwins); \
>    	els=`echo "$$wins " | sed -e 's|/\./|/|g' -e 's|/\. | |g' -e 's| |/*.el |g'`; \
> + 	echo -n "ELCFILES=">  $(lisp)/.elcfiles; \
>    	for el in $$els; do \
>    	  test -f $$el || continue; \
>    	  test ! -f $${el}c&&  GREP_OPTIONS= grep '^;.*no-byte-compile: t' $$el>  /dev/null&&  continue; \
> ! 	  echo -n "$${el}c ">>  $(lisp)/.elcfiles; \
>    	done; \
> ! 	$(MAKE) $(MFLAGS) compile-elcfiles EMACS=$(EMACS)
>
>    # Erase left-over .elc files that do not have a corresponding .el file.
>    compile-clean:
>
>

Message #35 received at 5191 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Chong Yidong <cyd <at> stupidchicken.com>
To: Glenn Morris <rgm <at> gnu.org>
Cc: Stefan Monnier <monnier <at> iro.umontreal.ca>, Harald Maier <Harald <at> Maierh.de>,
        5191 <at> debbugs.gnu.org
Subject: Re: bug#5191: 23.1.90; Make target compile-main fails under GNU/Linux Kernel 2.6.16
Date: Mon, 14 Dec 2009 18:29:24 -0500

Glenn Morris <rgm <at> gnu.org> writes:

> Chong Yidong wrote:
>
>> On second thought, I don't think this is a good idea.  Looks like the
>> BSDs use ".sinclude" for silent inclusion, and older systems don't even
>> support that.
>
> I think you are right that include is not portable. It's currently
> only used in GNUstep builds. I don't think "echo -n" is portable
> either. "printf" might be better, I'm not sure.

Another idea: how about autogenerating the ELCFILES Makefile using
configure?  Then Makefile.in could just have a line that says

ELCFILES=@elcfiles@

where @elcfiles@ is generated by configure.

Message #38 received at 5191 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Dan Nicolaescu <dann <at> ics.uci.edu>
To: Chong Yidong <cyd <at> stupidchicken.com>
Cc: 5191 <at> debbugs.gnu.org, Glenn Morris <rgm <at> gnu.org>,
        Harald Maier <Harald <at> Maierh.de>
Subject: Re: bug#5191: 23.1.90; Make target compile-main fails under GNU/Linux Kernel 2.6.16
Date: Mon, 14 Dec 2009 16:28:40 -0800 (PST)

Chong Yidong <cyd <at> stupidchicken.com> writes:

  > Glenn Morris <rgm <at> gnu.org> writes:
  > 
  > > Chong Yidong wrote:
  > >
  > >> On second thought, I don't think this is a good idea.  Looks like the
  > >> BSDs use ".sinclude" for silent inclusion, and older systems don't even
  > >> support that.
  > >
  > > I think you are right that include is not portable. It's currently
  > > only used in GNUstep builds. I don't think "echo -n" is portable
  > > either. "printf" might be better, I'm not sure.
  > 
  > Another idea: how about autogenerating the ELCFILES Makefile using
  > configure?  Then Makefile.in could just have a line that says
  > 
  > ELCFILES=@elcfiles@
  > 
  > where @elcfiles@ is generated by configure.

Yet another idea: how about we require GNU make?
Other GNU projects require GNU make already, like GCC.
GNU make is widely available on all free systems, and even on
proprietary unixes, I've personally never seen a machine used for
software development that did not have GNU make. 

We can float the idea in the 23.2 release, and make it mandatory a few
months after if we don't get significant pushback.

Message #43 received at 5191 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Chong Yidong <cyd <at> stupidchicken.com>
To: Dan Nicolaescu <dann <at> ics.uci.edu>
Cc: 5191 <at> debbugs.gnu.org, Glenn Morris <rgm <at> gnu.org>,
        Harald Maier <Harald <at> Maierh.de>
Subject: Re: bug#5191: 23.1.90; Make target compile-main fails under GNU/Linux Kernel 2.6.16
Date: Mon, 14 Dec 2009 21:26:36 -0500

Dan Nicolaescu <dann <at> ics.uci.edu> writes:

> Yet another idea: how about we require GNU make?  Other GNU projects
> require GNU make already, like GCC.  GNU make is widely available on
> all free systems, and even on proprietary unixes, I've personally
> never seen a machine used for software development that did not have
> GNU make.

What about the BSDs?  My impression is that they still use their own
Make version, no?

Message #46 received at 5191 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Dan Nicolaescu <dann <at> ics.uci.edu>
To: Chong Yidong <cyd <at> stupidchicken.com>
Cc: 5191 <at> debbugs.gnu.org, Glenn Morris <rgm <at> gnu.org>,
        Harald Maier <Harald <at> Maierh.de>
Subject: Re: bug#5191: 23.1.90; Make target compile-main fails under GNU/Linux Kernel 2.6.16
Date: Mon, 14 Dec 2009 19:11:48 -0800 (PST)

Chong Yidong <cyd <at> stupidchicken.com> writes:

  > Dan Nicolaescu <dann <at> ics.uci.edu> writes:
  > 
  > > Yet another idea: how about we require GNU make?  Other GNU projects
  > > require GNU make already, like GCC.  GNU make is widely available on
  > > all free systems, and even on proprietary unixes, I've personally
  > > never seen a machine used for software development that did not have
  > > GNU make.
  > 
  > What about the BSDs?  My impression is that they still use their own
  > Make version, no?

They do, but GNU make is still readily available.

Message #51 received at 5191 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Stefan Monnier <monnier <at> IRO.UMontreal.CA>
To: Chong Yidong <cyd <at> stupidchicken.com>
Cc: 5191 <at> debbugs.gnu.org, Harald Maier <Harald <at> Maierh.de>
Subject: Re: bug#5191: 23.1.90; Make target compile-main fails under GNU/Linux Kernel 2.6.16
Date: Mon, 14 Dec 2009 23:21:23 -0500

> + -include $(lisp)/.elcfiles

This is sadly not portable.
We could indeed put the list of elcfiles in a file, and then use

  xargs make <.elcfiles


-- Stefan

Message #60 received at 5191 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Stefan Monnier <monnier <at> IRO.UMontreal.CA>, 5191 <at> debbugs.gnu.org
Cc: cyd <at> stupidchicken.com, Harald <at> Maierh.de, 5191 <at> debbugs.gnu.org
Subject: Re: bug#5191: 23.1.90;
	Make target compile-main fails under GNU/Linux Kernel 2.6.16
Date: Tue, 15 Dec 2009 00:36:01 -0500

> From: Stefan Monnier <monnier <at> IRO.UMontreal.CA>
> Date: Mon, 14 Dec 2009 23:21:23 -0500
> Cc: Harald Maier <Harald <at> Maierh.de>, 5191 <at> emacsbugs.donarmstrong.com
> Reply-To: Stefan Monnier <monnier <at> IRO.UMontreal.CA>,
> 	5191 <at> emacsbugs.donarmstrong.com
> 
> > + -include $(lisp)/.elcfiles
> 
> This is sadly not portable.
> We could indeed put the list of elcfiles in a file, and then use
> 
>   xargs make <.elcfiles

If we eventually decide to go this way, please don't use a file name
that begins with a dot: apart of being disallowed on DOS filesystems,
such a file is hidden from `ls' and other commands, which in this case
does not seem to be justified, since this is a file generated by the
build.

Message #67 received at 5191-done <at> debbugs.gnu.org (full text, mbox):

From: Glenn Morris <rgm <at> gnu.org>
To: 5191-done <at> debbugs.gnu.org
Subject: Re: bug#5191: 23.1.90;
	Make target compile-main fails under GNU/Linux Kernel 2.6.16
Date: Wed, 09 Mar 2011 20:49:29 -0500

Version: 24.1

ELCFILES was removed, so this kind of thing won't happen in future.

Previous Next

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