Hang when compiling the current bzr Emacs

Hang when compiling the current bzr Emacs

[Lars Magne Ingebrigtsen]

I've tried "make bootstrap" and "maintainer-clean", but no matter what I
do, Emacs hangs when it starts compiling auth-source.el:

make[2]: Entering directory `/home/larsi/src/emacs/trunk/lisp'
Compiling gnus/auth-source.el
^Cmake[2]: *** [gnus/auth-source.elc] Error 2
make[1]: *** [compile-main] Error 130
make: *** [lisp] Interrupt

Is there some other magical "make clean" incantation I don't know about
that's necessary?

-- 
(domestic pets only, the antidote for overdose, milk.)
  bloggy blog http://lars.ingebrigtsen.no/

Re: Hang when compiling the current bzr Emacs

[Juanma Barranquero]

On Fri, Jun 8, 2012 at 1:17 AM, Lars Magne Ingebrigtsen <larsi@gnus.org> wrote:

> I've tried "make bootstrap" and "maintainer-clean", but no matter what I
> do, Emacs hangs when it starts compiling auth-source.el:

I also get a hang during bootstrap; in my case, it happens while
compiling lisp/emacs-lisp/ert.el.

    Juanma

Re: Hang when compiling the current bzr Emacs

[Paul Eggert]

I am seeing similar problems.  Byte compilations are
reeeally slow and eventually hang.  Here's how I
reproduced it from a fresh checkout of the latest
revision (trunk bzr 108515).

./autogen.sh
./configure --enable-gcc-warnings
make -j5 bootstrap

on Fedora 15 x86-64 with GCC 4.7.0.

I do not see the problem with trunk bzr 108508
so most likely the problem was introduced in bzr 108509
through 108515.

Re: Hang when compiling the current bzr Emacs

[Christoph Scholtes]

On 6/7/2012 6:22 PM, Paul Eggert wrote:

> I do not see the problem with trunk bzr 108508
> so most likely the problem was introduced in bzr 108509
> through 108515.

Windows built fine this morning with bzr 108509. I have not tried 
anything later than that.

Re: Hang when compiling the current bzr Emacs

[Lars Magne Ingebrigtsen]

Christoph Scholtes <cschol2112@googlemail.com> writes:

> Windows built fine this morning with bzr 108509. I have not tried
> anything later than that.

I think the most suspicious-looking patches after that are the
cl*-related ones.  Especially since Emacs seems to infloop on
the especially cl(-macs)-heavy .el files...

-- 
(domestic pets only, the antidote for overdose, milk.)
  bloggy blog http://lars.ingebrigtsen.no/

Re: Hang when compiling the current bzr Emacs

[Paul Eggert]

On 06/07/2012 05:41 PM, Lars Magne Ingebrigtsen wrote:
> I think the most suspicious-looking patches after that are the
> cl*-related ones.

Bingo.  A bisect found that the problem was introduced in
trunk bzr 108513 or 108514, which are the cl*-related patches.
I can do a 'make bootstrap' with 108512.  Doing it with
108513 fails as follows:

cedet/srecode/insert.el:983:1:Error: Symbol's function definition is void: cl-expr-contains-any
make[3]: *** [cedet/srecode/insert.elc] Error 1

Doing it with 108514 loops when running batch-byte-compile on
various sources (for me, textmodes/rst.el, emacs-lisp/ert.el,
org/ob.el, progmodes/js.el, cedet/srecode/insert.el, but
possibly more since I was doing a make -j5).

Re: Hang when compiling the current bzr Emacs

[Stefan Monnier]

> I've tried "make bootstrap" and "maintainer-clean", but no matter what I
> do, Emacs hangs when it starts compiling auth-source.el:

I've installed earlier a patch to eval.c which should have fixed
this problem.


        Stefan

Re: Hang when compiling the current bzr Emacs

[Romain Francoise]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> Is there some other magical "make clean" incantation I don't know about
> that's necessary?

There's always `bzr clean-tree --force --ignored --detritus --unknown'...

Re: Hang when compiling the current bzr Emacs

[martin rudalics]

 >> I've tried "make bootstrap" and "maintainer-clean", but no matter what I
 >> do, Emacs hangs when it starts compiling auth-source.el:
 >
 > I've installed earlier a patch to eval.c which should have fixed
 > this problem.

After doing

bzr clean-tree --unknown --ignored --detritus --force

bootstrapping trunk on Windows XP currently starts complaining as


Generating autoloads for cal-bahai.el...
Autoload cookie error in cal-bahai.el:144 (void-function macroexp-let²)
Autoload cookie error in cal-bahai.el:153 (void-function macroexp-let²)
Autoload cookie error in cal-bahai.el:191 (void-function macroexp-let²)
Autoload cookie error in cal-bahai.el:317 (void-function macroexp-let²)
Autoload cookie error in cal-bahai.el:330 (void-function macroexp-let²)
Autoload cookie error in cal-bahai.el:343 (void-function macroexp-let²)


and eventually stops at


In toplevel form:
progmodes/cc-mode.el:106:21:Warning: `font-lock-syntactic-keywords' is an
     obsolete variable (as of 24.1); use `syntax-propertize-function' instead.
progmodes/cc-mode.el:121:1:Error: Symbol's function definition is void: macroexp-let²
progmodes/cc-mode.el:121:1:Error: Symbol's function definition is void: macroexp-let²
progmodes/cc-mode.el:121:1:Error: Symbol's function definition is void: macroexp-let²
progmodes/cc-mode.el:121:1:Error: Symbol's function definition is void: macroexp-let²
progmodes/cc-mode.el:121:1:Error: Symbol's function definition is void: macroexp-let²
progmodes/cc-mode.el:150:1:Error: Symbol's function definition is void: macroexp-let²

In toplevel form:
progmodes/cc-vars.el:74:1:Error: Symbol's function definition is void: macroexp-let²
make[1]: *** [compile-CMD] Error 1
make[1]: Leaving directory `c:/emacs/trunk/lisp'
make: *** [bootstrap-gmake] Error 2


where


              position: 430943 of 431084 (100%), column: 83
             character: Â (displayed as Â) (codepoint 194, #o302, #xc2)
     preferred charset: unicode-bmp (Unicode Basic Multilingual Plane (U+0000..U+FFFF))
code point in charset: 0xC2
                syntax: w 	which means: word
              category: .:Base, L:Left-to-right (strong), j:Japanese, l:Latin, v:Viet
           buffer code: #xC3 #x82
             file code: #xC3 #x82 (encoded by coding system utf-8-dos)
               display: by this font (glyph code)
     uniscribe:-outline-Courier New-normal-normal-normal-mono-13-*-*-*-c-*-iso8859-1 (#xC6)

Character code properties: customize what to show
   name: LATIN CAPITAL LETTER A WITH CIRCUMFLEX
   old-name: LATIN CAPITAL LETTER A CIRCUMFLEX
   general-category: Lu (Letter, Uppercase)
   decomposition: (65 770) ('A' '̂')


and


              position: 430944 of 431084 (100%), column: 84
             character: ² (displayed as ²) (codepoint 178, #o262, #xb2)
     preferred charset: unicode-bmp (Unicode Basic Multilingual Plane (U+0000..U+FFFF))
code point in charset: 0xB2
                syntax: _ 	which means: symbol
              category: .:Base, h:Korean, j:Japanese, l:Latin
           buffer code: #xC2 #xB2
             file code: #xC2 #xB2 (encoded by coding system utf-8-dos)
               display: by this font (glyph code)
     uniscribe:-outline-Courier New-normal-normal-normal-mono-13-*-*-*-c-*-iso8859-1 (#xF0)

Character code properties: customize what to show
   name: SUPERSCRIPT TWO
   old-name: SUPERSCRIPT DIGIT TWO
   general-category: No (Number, Other)
   decomposition: (super 50) (super '2')


Any ideas?

martin

Re: Hang when compiling the current bzr Emacs

[Stefan Monnier]

> bootstrapping trunk on Windows XP currently starts complaining as

I've just tried a bootstrap here (I try to avoid them now that I've
replaced all my desktops with netbook-type CPUs, making a bootstrap even
more painful than it was) and it worked.  Could you try to set
a "coding:utf-8" cookie on both pcase.el and macroexp.el to see if
that's the source of the problem?


        Stefan

Re: Hang when compiling the current bzr Emacs

[martin rudalics]

> I've just tried a bootstrap here (I try to avoid them now that I've
> replaced all my desktops with netbook-type CPUs, making a bootstrap even
> more painful than it was) and it worked.  Could you try to set
> a "coding:utf-8" cookie on both pcase.el and macroexp.el to see if
> that's the source of the problem?

Works now without any further fixing here.

Thanks, martin