CVS HEAD fails to build

CVS HEAD fails to build

[Randal L. Schwartz]

% make clean bootstrap install
...
Compiling /Users/merlyn/MIRROR/emacs-CVS/lisp/./eshell/em-term.el
Loading /Users/merlyn/MIRROR/emacs-CVS/lisp/net/ange-ftp.el (source)...

In toplevel form:
eshell/em-term.el:48:1:Error: Wrong type argument: keymapp, nil
make[2]: *** [compile] Error 1
make[1]: *** [bootstrap-build] Error 2
make: *** [bootstrap] Error 2
make clean bootstrap install  168.80s user 32.79s system 67% cpu 4:59.70 total

-- 
Randal L. Schwartz - Stonehenge Consulting Services, Inc. - +1 503 777 0095
<merlyn@stonehenge.com> <URL:http://www.stonehenge.com/merlyn/>
Perl/Unix/security consulting, Technical writing, Comedy, etc. etc.
See PerlTraining.Stonehenge.com for onsite and open-enrollment Perl training!

Re: CVS HEAD fails to build

[Giorgos Keramidas]

On 2007-05-07 11:34, "Randal L. Schwartz" <merlyn@stonehenge.com> wrote:
>
> % make clean bootstrap install
> ...
> Compiling /Users/merlyn/MIRROR/emacs-CVS/lisp/./eshell/em-term.el
> Loading /Users/merlyn/MIRROR/emacs-CVS/lisp/net/ange-ftp.el (source)...
>
> In toplevel form:
> eshell/em-term.el:48:1:Error: Wrong type argument: keymapp, nil
> make[2]: *** [compile] Error 1
> make[1]: *** [bootstrap-build] Error 2
> make: *** [bootstrap] Error 2
> make clean bootstrap install  168.80s user 32.79s system 67% cpu 4:59.70 total

Same here.  I'm bisecting through commits to find the first change which
caused this.  Hold on for half an hour or so and I will have more details.

- Giorgos

Re: CVS HEAD fails to build

[Giorgos Keramidas]

On 2007-05-07 21:57, Giorgos Keramidas <keramida@ceid.upatras.gr> wrote:
>On 2007-05-07 11:34, "Randal L. Schwartz" <merlyn@stonehenge.com> wrote:
>>
>> % make clean bootstrap install
>> ...
>> Compiling /Users/merlyn/MIRROR/emacs-CVS/lisp/./eshell/em-term.el
>> Loading /Users/merlyn/MIRROR/emacs-CVS/lisp/net/ange-ftp.el (source)...
>>
>> In toplevel form:
>> eshell/em-term.el:48:1:Error: Wrong type argument: keymapp, nil
>> make[2]: *** [compile] Error 1
>> make[1]: *** [bootstrap-build] Error 2
>> make: *** [bootstrap] Error 2
>> make clean bootstrap install  168.80s user 32.79s system 67% cpu 4:59.70 total
>
> Same here.  I'm bisecting through commits to find the first change
> which caused this.  Hold on for half an hour or so and I will have
> more details.

The first bad revision is:
changeset:   81110:6207e966040e
user:        monnier
date:        Mon May 07 15:41:37 2007 +0000
summary:     (Fdefine_key, Flookup_key): Only do the 0x80->meta_modifier

Re: CVS HEAD fails to build

[Andreas Schwab]

Giorgos Keramidas <keramida@ceid.upatras.gr> writes:

> The first bad revision is:
> changeset:   81110:6207e966040e
> user:        monnier
> date:        Mon May 07 15:41:37 2007 +0000
> summary:     (Fdefine_key, Flookup_key): Only do the 0x80->meta_modifier

Does it still happen?

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux Products GmbH, Maxfeldstraße 5, 90409 Nürnberg, Germany
PGP key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: CVS HEAD fails to build

[Giorgos Keramidas]

On 2007-05-07 22:50, Giorgos Keramidas <keramida@ceid.upatras.gr> wrote:
> The first bad revision is:
> changeset:   81110:6207e966040e
> user:        monnier
> date:        Mon May 07 15:41:37 2007 +0000
> summary:     (Fdefine_key, Flookup_key): Only do the 0x80->meta_modifier

Hi Stefan and all,

this part of the change seems to be the one which is causing the
build problem:

% diff -r 62063c2fb043 -r 6207e966040e src/keymap.c
% --- a/src/keymap.c      Mon May 07 08:05:55 2007 +0000
% +++ b/src/keymap.c      Mon May 07 15:41:37 2007 +0000
% @@ -1312,7 +1313,7 @@ recognize the default bindings, just as
%         c = Fevent_convert_list (c);
%
%        /* Turn the 8th bit of string chars into a meta modifier.  */
% -      if (INTEGERP (c) && XINT (c) & 0x80 && STRINGP (key))
% +      if (STRINGP (key) && XINT (c) & 0x8 && !STRING_MULTIBYTE (key))
%          XSETINT (c, (XINT (c) | meta_modifier) & ~0x80);
%
%        /* Allow string since binding for `menu-bar-select-buffer'

I have locally reverted the condition to:

         if (INTEGERP (c) && XINT (c) & 0x80 && STRINGP (key) &&
             !STRING_MULTIBYTE (key))
           XSETINT (c, (XINT (c) | meta_modifier) & ~0x80);

but I am not sure if it matches what you originally intended to do,
Stefan.  Does the new check seem reasonable?

- Giorgos

Re: CVS HEAD fails to build

[Giorgos Keramidas]

On 2007-05-07 22:51, Andreas Schwab <schwab@suse.de> wrote:
> Giorgos Keramidas <keramida@ceid.upatras.gr> writes:
> > The first bad revision is:
> > changeset:   81110:6207e966040e
> > user:        monnier
> > date:        Mon May 07 15:41:37 2007 +0000
> > summary:     (Fdefine_key, Flookup_key): Only do the 0x80->meta_modifier
> 
> Does it still happen?

I just rsync'ed again and caught your change (-r 1.354 of
keymap.c), and Emacs builds fine again, thanks :)

Re: CVS HEAD fails to build

[Randal L. Schwartz]

>>>>> "Andreas" == Andreas Schwab <schwab@suse.de> writes:

Andreas> Does it still happen?

After the CVS refresh, builds fine for me.  Seems to launch too.

-- 
Randal L. Schwartz - Stonehenge Consulting Services, Inc. - +1 503 777 0095
<merlyn@stonehenge.com> <URL:http://www.stonehenge.com/merlyn/>
Perl/Unix/security consulting, Technical writing, Comedy, etc. etc.
See PerlTraining.Stonehenge.com for onsite and open-enrollment Perl training!

Re: CVS HEAD fails to build

[Stefan Monnier]

> % make clean bootstrap install
> ...
> Compiling /Users/merlyn/MIRROR/emacs-CVS/lisp/./eshell/em-term.el
> Loading /Users/merlyn/MIRROR/emacs-CVS/lisp/net/ange-ftp.el (source)...

> In toplevel form:
> eshell/em-term.el:48:1:Error: Wrong type argument: keymapp, nil
> make[2]: *** [compile] Error 1
> make[1]: *** [bootstrap-build] Error 2
> make: *** [bootstrap] Error 2
> make clean bootstrap install  168.80s user 32.79s system 67% cpu 4:59.70 total

I believe this was a typo, fixed by Andreas soon after I installed it.


        Stefan

Re: CVS HEAD fails to build

[Richard Stallman]

The change to deal with multibyte strings differently feels risky to
me.  It is a clever idea, and might offer a way out of the dilemma.
But we should try it when we have more time to fix whatever it might
break.

So I'd rather not have it in the 22.1 release.
Would you please take it out of there, for now?

Re: CVS HEAD fails to build

[Stefan Monnier]

> The change to deal with multibyte strings differently feels risky to
> me.  It is a clever idea, and might offer a way out of the dilemma.
> But we should try it when we have more time to fix whatever it might
> break.

> So I'd rather not have it in the 22.1 release.
> Would you please take it out of there, for now?

It's definitely not slated for inclusion in Emacs-22.1, of course.


        Stefan

Re: CVS HEAD fails to build

[Richard Stallman]

    It's definitely not slated for inclusion in Emacs-22.1, of course.

Oh, I see.  Thanks.

Re: CVS HEAD fails to build

[Randal L. Schwartz]

>>>>> "Richard" == Richard Stallman <rms@gnu.org> writes:

Richard>     It's definitely not slated for inclusion in Emacs-22.1, of course.
Richard> Oh, I see.  Thanks.

Yeah, that's why I said "CVS HEAD".  Not sure what to call that version. :)
Oh, it calls itself 22.1.50.  Sorry, next time I'll be more clear.
I hate CVS.

-- 
Randal L. Schwartz - Stonehenge Consulting Services, Inc. - +1 503 777 0095
<merlyn@stonehenge.com> <URL:http://www.stonehenge.com/merlyn/>
Perl/Unix/security consulting, Technical writing, Comedy, etc. etc.
See PerlTraining.Stonehenge.com for onsite and open-enrollment Perl training!

Re: CVS HEAD fails to build

[Stefan Monnier]

> Yeah, that's why I said "CVS HEAD".  Not sure what to call that version. :)
> Oh, it calls itself 22.1.50.  Sorry, next time I'll be more clear.

Every branch has a HEAD.  The "nameless" branch in CVS (which can be
selected via "update -A" or sometimes with "update -r 1") is called the
"trunk".  There's been patches floating around to make `cvs' accept the
special tag TRUNK to refer to the trunk, but I don't know if these will ever
be included (just like the patches that add a special ".BASE" suffix to
select the base of a branch).


        Stefan