bug#18094: 24.4.50; emacs_backtrace.txt

bug#18094: 24.4.50; emacs_backtrace.txt

[Drew Adams]

Backtrace:
01201baf
01201c20
010f4223
0116762f
010eea46
01065a4e
01062e45
01063633
01062197
01061d32
0105044a
01047acd
0118283e
01047a5a
01046a90
01044b0d
010fa886
01107d88
010f8355
0118272b
010f7c8a
01181cd8
010f7c42
010f73d8
010f7594
010f5724
010010f9
76993386
771d9f6e
771d9f41




In GNU Emacs 24.4.50.1 (i686-pc-mingw32)
 of 2014-06-28 on ODIEONE
Bzr revision: 117431 rgm@gnu.org-20140628015517-eku6hj8mpgcvfnso
Windowing system distributor `Microsoft Corp.', version 6.1.7601
Configured using:
 `configure --prefix=/c/Devel/emacs/snapshot/trunk
 --enable-checking=yes,glyphs 'CFLAGS=-O0 -g3'
 LDFLAGS=-Lc:/Devel/emacs/lib 'CPPFLAGS=-DGC_MCHECK=1
 -Ic:/Devel/emacs/include''

bug#18094: 24.4.50; emacs_backtrace.txt

[Eli Zaretskii]

> Date: Wed, 23 Jul 2014 22:39:16 -0700 (PDT)
> From: Drew Adams <drew.adams@oracle.com>
> 
> Backtrace:
> 01201baf
> 01201c20
> 010f4223
> 0116762f
> 010eea46
> 01065a4e
> 01062e45
> 01063633
> 01062197
> 01061d32
> 0105044a
> 01047acd
> 0118283e
> 01047a5a
> 01046a90
> 01044b0d
> 010fa886
> 01107d88
> 010f8355
> 0118272b
> 010f7c8a
> 01181cd8
> 010f7c42
> 010f73d8
> 010f7594
> 010f5724
> 010010f9
> 76993386
> 771d9f6e
> 771d9f41
> 
> 
> 
> 
> In GNU Emacs 24.4.50.1 (i686-pc-mingw32)
>  of 2014-06-28 on ODIEONE
> Bzr revision: 117431 rgm@gnu.org-20140628015517-eku6hj8mpgcvfnso

Translation:

  w32_backtrace at c:\Devel\emacs\repo\binary\src/w32fns.c:8490
  emacs_abort at c:\Devel\emacs\repo\binary\src/w32fns.c:8522
  terminate_due_to_signal at c:\Devel\emacs\repo\binary\src/emacs.c:383
  die at c:\Devel\emacs\repo\binary\src/alloc.c:7044
  face_at_string_position at c:\Devel\emacs\repo\binary\src/xfaces.c:6163
  display_string at c:\Devel\emacs\repo\binary\src/xdisp.c:23098
  display_mode_element at c:\Devel\emacs\repo\binary\src/xdisp.c:21850
  display_mode_element at c:\Devel\emacs\repo\binary\src/xdisp.c:22022
  display_mode_line at c:\Devel\emacs\repo\binary\src/xdisp.c:21539
  display_mode_lines at c:\Devel\emacs\repo\binary\src/xdisp.c:21482
  redisplay_window at c:\Devel\emacs\repo\binary\src/xdisp.c:16628
  redisplay_window_0 at c:\Devel\emacs\repo\binary\src/xdisp.c:14225
  internal_condition_case_1 at c:\Devel\emacs\repo\binary\src/eval.c:1373
  redisplay_windows at c:\Devel\emacs\repo\binary\src/xdisp.c:14205
  redisplay_internal at c:\Devel\emacs\repo\binary\src/xdisp.c:13804
  redisplay at c:\Devel\emacs\repo\binary\src/xdisp.c:13090
  read_char at c:\Devel\emacs\repo\binary\src/keyboard.c:2561
  read_key_sequence at c:\Devel\emacs\repo\binary\src/keyboard.c:9085
  command_loop_1 at c:\Devel\emacs\repo\binary\src/keyboard.c:1439
  internal_condition_case at c:\Devel\emacs\repo\binary\src/eval.c:1349
  command_loop_2 at c:\Devel\emacs\repo\binary\src/keyboard.c:1170
  internal_catch at c:\Devel\emacs\repo\binary\src/eval.c:1113
  command_loop at c:\Devel\emacs\repo\binary\src/keyboard.c:1149
  recursive_edit_1 at c:\Devel\emacs\repo\binary\src/keyboard.c:770
  Frecursive_edit at c:\Devel\emacs\repo\binary\src/keyboard.c:841
  main at c:\Devel\emacs\repo\binary\src/emacs.c:1656

This makes no sense to me.  The crash is here:

  int
  face_at_string_position (struct window *w, Lisp_Object string,
			   ptrdiff_t pos, ptrdiff_t bufpos,
			   ptrdiff_t *endptr, enum face_id base_face_id,
			   int mouse_p)
  {
    [...]
    base_face = FACE_FROM_ID (f, base_face_id);
    eassert (base_face);  <<<<<<<<<<<<<<<<<<<<<<<<<<

The argument base_face_id comes from the current mode line face, which
must be a valid face.

What in the world are you doing with your mode line format??

bug#18094: 24.4.50; emacs_backtrace.txt

[Drew Adams]

> This makes no sense to me.  The crash is here:
>   base_face = FACE_FROM_ID (f, base_face_id);
>   eassert (base_face);  <<<<<<<<<<<<<<<<<<<<<<<<<<
> 
> The argument base_face_id comes from the current mode line face, which
> must be a valid face. What in the world are you doing with your mode
> line format??

I use these faces on parts of the mode line:

(defface modelinepos-column-warning '((t (:foreground "Red")))
  "*Face used to highlight the modeline column number.
This is used when the current column number is greater than
`modelinepos-column-limit'."
  :group 'Modeline :group 'Convenience :group 'Help :group 'faces)

(defface modelinepos-region '((t :inherit region))
  "*Face used to highlight the modeline position and size when
the region is active."
  :group 'Modeline :group 'Convenience :group 'Help :group 'faces)

(defface modelinepos-region-acting-on '((t (:inherit region
  :box (:line-width 3 :color "Red"))))
  "*Face for modeline position & size when a command acts on active region."
  :group 'Modeline :group 'Convenience :group 'Help :group 'faces)

As for what in the world I am doing with my mode line format, the code
is here: http://www.emacswiki.org/emacs-en/download/modeline-posn.el.

At least that is much of what in the world I am doing with my mode line
format.  And this is what I use for face `mode-line-buffer-id':
((t (:overline "red" :underline "red"))).

And another piece of what in the world I am doing with my mode line
format is here:
http://www.emacswiki.org/emacs-en/download/bindings%2b.el.

And another piece, for Dired buffers, is function
`diredp-nb-marked-in-mode-name', from here:
http://www.emacswiki.org/emacs-en/download/dired%2b.el

And I sometimes dynamically put different things in the mode-line,
using this face:

(defface icicle-mode-line-help
    '((((background dark)) (:foreground "#AC4AAC4A0000"))
      (t (:foreground "Blue")))
  "*Face used to highlight help shown in the mode-line."
  :group 'Icicles-Completions-Display :group 'Icicles-Miscellaneous
  :group 'faces)

See functions `icicle-display-candidates-in-Completions'
`icicle-show-help-in-mode-line', and `icicle-show-in-mode-line' in
this file, for the code that does that:
http://www.emacswiki.org/emacs-en/download/icicles-fn.el.  It is
done at various times, including on `post-command-hook'.

I also have code that dynamically changes one or more lighters.
For example, `icicle-(un)highlight-lighter', in icicles-fn.el.

I also make changes to the mode line in certain modes.

For example, in Info, I (optionally) insert breadcrumbs in the mode
line.  The relevant code is in `Info-insert-breadcrumbs-in-mode-line',
`Info-set-mode-line' and `Info-breadcrumbs-in-mode-line-mode', here:
http://www.emacswiki.org/emacs-en/download/info%2b.el

And in Isearch, I update the lighter to show case-sensitivity.  The
code is `isearchp-highlight-lighter', here:
http://www.emacswiki.org/emacs-en/download/isearch%2b.el

That's a start wrt what in the world I am doing with my mode line
format.  There are other places where I change things in the
mode-line, but they are likely to be even less relevant.

FWIW, I am not doing anything new with my mode-line.  Whatever in
the world I am doing with my mode-line I have been doing for years now.

bug#18094: 24.4.50; emacs_backtrace.txt

[Eli Zaretskii]

> Date: Fri, 25 Jul 2014 07:32:11 -0700 (PDT)
> From: Drew Adams <drew.adams@oracle.com>
> Cc: 18094@debbugs.gnu.org
> 
> > This makes no sense to me.  The crash is here:
> >   base_face = FACE_FROM_ID (f, base_face_id);
> >   eassert (base_face);  <<<<<<<<<<<<<<<<<<<<<<<<<<
> > 
> > The argument base_face_id comes from the current mode line face, which
> > must be a valid face. What in the world are you doing with your mode
> > line format??
> 
> I use these faces on parts of the mode line:

This is not relevant to the issue at hand.  The assertion violation
you've hit means that the base face of the mode line (known as
'mode-line') is somehow invalid.  Any additional faces you add to the
strings displayed on the mode line are irrelevant.

Do you have any face remappings for the mode-line face itself?  If so,
please show them, and please explain when and how are these remappings
changed.

bug#18094: 24.4.50; emacs_backtrace.txt

[Drew Adams]

> This is not relevant to the issue at hand.  The assertion violation
> you've hit means that the base face of the mode line (known as
> 'mode-line') is somehow invalid.  Any additional faces you add to the
> strings displayed on the mode line are irrelevant.
> 
> Do you have any face remappings for the mode-line face itself?

Nope.

bug#18094: 24.4.50; emacs_backtrace.txt

[Eli Zaretskii]

> Date: Fri, 25 Jul 2014 08:41:44 -0700 (PDT)
> From: Drew Adams <drew.adams@oracle.com>
> Cc: 18094@debbugs.gnu.org
> 
> > This is not relevant to the issue at hand.  The assertion violation
> > you've hit means that the base face of the mode line (known as
> > 'mode-line') is somehow invalid.  Any additional faces you add to the
> > strings displayed on the mode line are irrelevant.
> > 
> > Do you have any face remappings for the mode-line face itself?
> 
> Nope.

Then I have no idea how the mode-line face could become invalid.

bug#18094:

[Glenn Morris]

Nothing can be done with this information.