bug#16504: 24.3.50; emacs_backtrace.txt

bug#16504: 24.3.50; emacs_backtrace.txt

[Drew Adams]

Reproducible from emacs -Q.  Do `M-x set-frame-font', then enter
"-outline-Lucida Console-normal-normal-normal-mono-14-*-*-*-c-*-iso8859-1"

That is, including the double-quotes (I forgot to remove them).  Without
the double-quotes there is no crash.

Backtrace:
011fb4d1
011fb542
010efd72
01161c41
0122ed2b
01032d7a
01059096
0104de10
0104b408
010446df
0117ce48
0104466c
010436a2
0104171f
010f6324
011037b5
010f3e3d
0117cd35
010f3772
0117c2e2
010f372a
010f2ec0
010f307c
010f1274
010010f9
762c3366
776c9f6e
776c9f41


In GNU Emacs 24.3.50.1 (i686-pc-mingw32)
 of 2014-01-17 on ODIEONE
Bzr revision: 116059 michael.albinus@gmx.de-20140117182409-gz695p2c951idlee
Windowing system distributor `Microsoft Corp.', version 6.1.7601
Configured using:
 `configure --prefix=/c/Devel/emacs/binary --enable-checking=yes,glyphs
 'CFLAGS=-O0 -g3' LDFLAGS=-Lc:/Devel/emacs/lib
 CPPFLAGS=-Ic:/Devel/emacs/include'

bug#16504: 24.3.50; emacs_backtrace.txt

[Drew Adams]

Actually, in addition to the reproducible crash from emacs -Q,
I am also getting this same backtrace with my own setup (very
often), and when I do things not apparently related to setting
the frame font.  E.g., the last crash happened when I did C-a.

bug#16504: 24.3.50; emacs_backtrace.txt

[Juanma Barranquero]

??
??:0
w32_backtrace at w32fns.c:8383
emacs_abort at w32fns.c:8415
terminate_due_to_signal at emacs.c:378
die at alloc.c:6761
face_for_char at fontset.c:914
get_next_display_element at xdisp.c:6998
display_line at xdisp.c:19723
try_window at xdisp.c:16541
redisplay_window at xdisp.c:16058
redisplay_window_0 at xdisp.c:14054
internal_condition_case_1 at eval.c:1369
redisplay_windows at xdisp.c:14034
redisplay_internal at xdisp.c:13633
redisplay at xdisp.c:12919
read_char at keyboard.c:2563
read_key_sequence at keyboard.c:9071
command_loop_1 at keyboard.c:1445
internal_condition_case at eval.c:1345
command_loop_2 at keyboard.c:1170
internal_catch at eval.c:1109
command_loop at keyboard.c:1149
recursive_edit_1 at keyboard.c:777
Frecursive_edit at keyboard.c:841
main at emacs.c:1637
?? at crt1.c:0
??
??:0
??
??:0
??
??:0

bug#16504: 24.3.50; emacs_backtrace.txt

[Eli Zaretskii]

> From: Juanma Barranquero <lekktu@gmail.com>
> Date: Mon, 20 Jan 2014 12:58:50 +0100
> Cc: 16504@debbugs.gnu.org
> 
> w32_backtrace at w32fns.c:8383
> emacs_abort at w32fns.c:8415
> terminate_due_to_signal at emacs.c:378
> die at alloc.c:6761
> face_for_char at fontset.c:914

It aborts here:

  int
  face_for_char (struct frame *f, struct face *face, int c,
		 ptrdiff_t pos, Lisp_Object object)
  {
    Lisp_Object fontset, rfont_def, charset;
    int face_id;
    int id;

    eassert (fontset_id_valid_p (face->fontset));  <<<<<<<<<<<<<<

bug#16504: 24.3.50; emacs_backtrace.txt

[Juanma Barranquero]

[-- Attachment #1: Type: text/plain, Size: 368 bytes --]

On Mon, Jan 20, 2014 at 9:06 AM, Drew Adams <drew.adams@oracle.com> wrote:
> Reproducible from emacs -Q.  Do `M-x set-frame-font', then enter
> "-outline-Lucida Console-normal-normal-normal-mono-14-*-*-*-c-*-iso8859-1"

FWIW, with that same binary I don't get a crash, though there's
obviously a bug in that an invalid font shouldn't be accepted (see
attached image).

[-- Attachment #2: bug.png --]
[-- Type: image/png, Size: 63248 bytes --]

bug#16504: 24.3.50; emacs_backtrace.txt

[Eli Zaretskii]

> From: Juanma Barranquero <lekktu@gmail.com>
> Date: Mon, 20 Jan 2014 13:05:46 +0100
> Cc: 16504@debbugs.gnu.org
> 
> FWIW, with that same binary I don't get a crash

Neither do I with today's trunk.  I guess this has something to do
with which fonts are installed on the system.

> though there's obviously a bug in that an invalid font shouldn't be
> accepted (see attached image).

I think the font is valid (it does succeed to display the buffer),
it's just that it doesn't have a bold variant, which is needed to
display the buffer name on the mode line.

bug#16504: 24.3.50; emacs_backtrace.txt

[Juanma Barranquero]

On Mon, Jan 20, 2014 at 5:30 PM, Eli Zaretskii <eliz@gnu.org> wrote:

> it's just that it doesn't have a bold variant, which is needed to
> display the buffer name on the mode line.

Shouldn't Emacs in that case select another font?

bug#16504: 24.3.50; emacs_backtrace.txt

[Eli Zaretskii]

> From: Juanma Barranquero <lekktu@gmail.com>
> Date: Mon, 20 Jan 2014 17:33:29 +0100
> Cc: Drew Adams <drew.adams@oracle.com>, 16504@debbugs.gnu.org
> 
> On Mon, Jan 20, 2014 at 5:30 PM, Eli Zaretskii <eliz@gnu.org> wrote:
> 
> > it's just that it doesn't have a bold variant, which is needed to
> > display the buffer name on the mode line.
> 
> Shouldn't Emacs in that case select another font?

I don't know, sorry.  Fontsets and font selection were always a kind
of voodoo to me.

bug#16504: 24.3.50; emacs_backtrace.txt

[Juanma Barranquero]

On Mon, Jan 20, 2014 at 8:13 PM, Eli Zaretskii <eliz@gnu.org> wrote:

> I don't know, sorry.  Fontsets and font selection were always a kind
> of voodoo to me.

Perhaps we should ask Handa-san?

bug#16504: 24.3.50; emacs_backtrace.txt

[Eli Zaretskii]

> From: Juanma Barranquero <lekktu@gmail.com>
> Date: Tue, 21 Jan 2014 00:35:25 +0100
> Cc: Drew Adams <drew.adams@oracle.com>, 16504@debbugs.gnu.org
> 
> On Mon, Jan 20, 2014 at 8:13 PM, Eli Zaretskii <eliz@gnu.org> wrote:
> 
> > I don't know, sorry.  Fontsets and font selection were always a kind
> > of voodoo to me.
> 
> Perhaps we should ask Handa-san?

Either Handa-san or Jan, or anyone else who understands this stuff.

bug#16504: 24.3.50; emacs_backtrace.txt

[Eli Zaretskii]

> Date: Tue, 21 Jan 2014 05:40:08 +0200
> From: Eli Zaretskii <eliz@gnu.org>
> Cc: 16504@debbugs.gnu.org
> 
> > From: Juanma Barranquero <lekktu@gmail.com>
> > Date: Tue, 21 Jan 2014 00:35:25 +0100
> > Cc: Drew Adams <drew.adams@oracle.com>, 16504@debbugs.gnu.org
> > 
> > On Mon, Jan 20, 2014 at 8:13 PM, Eli Zaretskii <eliz@gnu.org> wrote:
> > 
> > > I don't know, sorry.  Fontsets and font selection were always a kind
> > > of voodoo to me.
> > 
> > Perhaps we should ask Handa-san?
> 
> Either Handa-san or Jan, or anyone else who understands this stuff.

Actually, this is further complicated by what the Windows build does
with fonts.  So perhaps Jason could chime in as well.

bug#16504: 24.3.50; emacs_backtrace.txt

[Drew Adams]

FWIW - I cannot use this build, because it crashes within a minute
or so of trying to use it.  It happens seemingly randomly.
(I understand that it is not random, and is apparently because of
a font-handling problem.)

bug#16504: 24.3.50; emacs_backtrace.txt

[Juanma Barranquero]

On Mon, Jan 20, 2014 at 10:29 PM, Drew Adams <drew.adams@oracle.com> wrote:

> FWIW - I cannot use this build, because it crashes within a minute
> or so of trying to use it.  It happens seemingly randomly.

Yep, I just filed two font-related bugs.

bug#16504: 24.3.50; emacs_backtrace.txt

[Eli Zaretskii]

> From: Juanma Barranquero <lekktu@gmail.com>
> Date: Tue, 21 Jan 2014 00:36:06 +0100
> Cc: Eli Zaretskii <eliz@gnu.org>, 16504@debbugs.gnu.org
> 
> On Mon, Jan 20, 2014 at 10:29 PM, Drew Adams <drew.adams@oracle.com> wrote:
> 
> > FWIW - I cannot use this build, because it crashes within a minute
> > or so of trying to use it.  It happens seemingly randomly.
> 
> Yep, I just filed two font-related bugs.

Thanks, but are we sure they are relevant to Drew's problems?  At
least one of those bugs is about trying to use non-existent fonts,
which I'm pretty sure is not what Drew tries to do as a matter of
routine.

Drew, can you please post a couple of backtraces from these frequent
crashes?

bug#16504: 24.3.50; emacs_backtrace.txt

[Juanma Barranquero]

On Tue, Jan 21, 2014 at 4:42 AM, Eli Zaretskii <eliz@gnu.org> wrote:

> Thanks, but are we sure they are relevant to Drew's problems?

No, but...

> At
> least one of those bugs is about trying to use non-existent fonts,

...I was thinking about the other one, which is a GC bug triggered
when viewing a file that causes lots of fonts to be loaded. Something
font-related *is* happening now that didn't happen a few days ago.
Ockham's razor suggests a single source is likely for these bugs.

bug#16504: 24.3.50; emacs_backtrace.txt

[Lars Ingebrigtsen]

Juanma Barranquero <lekktu@gmail.com> writes:

> On Mon, Jan 20, 2014 at 10:29 PM, Drew Adams <drew.adams@oracle.com> wrote:
>
>> FWIW - I cannot use this build, because it crashes within a minute
>> or so of trying to use it.  It happens seemingly randomly.
>
> Yep, I just filed two font-related bugs.

Was Drew's font issue here fixed by the other two font-related bugs?

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

bug#16504: 24.3.50; emacs_backtrace.txt

[Lars Ingebrigtsen]

Lars Ingebrigtsen <larsi@gnus.org> writes:

> Juanma Barranquero <lekktu@gmail.com> writes:
>
>> On Mon, Jan 20, 2014 at 10:29 PM, Drew Adams <drew.adams@oracle.com> wrote:
>>
>>> FWIW - I cannot use this build, because it crashes within a minute
>>> or so of trying to use it.  It happens seemingly randomly.
>>
>> Yep, I just filed two font-related bugs.
>
> Was Drew's font issue here fixed by the other two font-related bugs?

There was no response here a few months back, so I'm closing this
report.

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

bug#16504: 24.3.50; emacs_backtrace.txt

[Drew Adams]

> Drew, can you please post a couple of backtraces from these frequent
> crashes?

They are the same backtrace - this bug.

bug#16504: 24.3.50; emacs_backtrace.txt

[Eli Zaretskii]

> Date: Mon, 20 Jan 2014 19:57:48 -0800 (PST)
> From: Drew Adams <drew.adams@oracle.com>
> Cc: drew.adams@oracle.com, 16504@debbugs.gnu.org
> 
> > Drew, can you please post a couple of backtraces from these frequent
> > crashes?
> 
> They are the same backtrace - this bug.

I'm confused: you said that this bug was about specifying an incorrect
string for a font.  If so, the other crashes are most probably caused
by a different cause.

bug#16504: 24.3.50; emacs_backtrace.txt

[Drew Adams]

> > > Drew, can you please post a couple of backtraces from these frequent
> > > crashes?
> >
> > They are the same backtrace - this bug.
> 
> I'm confused: you said that this bug was about specifying an incorrect
> string for a font.  If so, the other crashes are most probably caused
> by a different cause.

What I provided in the original report for this bug is a way to reproduce the crash, starting from emacs -Q. It is also how I first encountered the crash (usually, I encounter a crash first in my own setup).

I later added that with my own setup I am frequently getting crashes that produce the same backtrace (exactly). And those appear to be random (do not appear to be associated with anything particular that I am doing at the time of the crash).

I also reported a different backtrace separately as a different bug. But for this bug all of the backtraces have been the same, and the original bug report gives you a way to reproduce it.

HTH.

bug#16504: 24.3.50; emacs_backtrace.txt

[Eli Zaretskii]

> Date: Tue, 21 Jan 2014 08:05:09 -0800 (PST)
> From: Drew Adams <drew.adams@oracle.com>
> Cc: lekktu@gmail.com, 16504@debbugs.gnu.org
> 
> What I provided in the original report for this bug is a way to reproduce the crash, starting from emacs -Q. It is also how I first encountered the crash (usually, I encounter a crash first in my own setup).
> 
> I later added that with my own setup I am frequently getting crashes that produce the same backtrace (exactly). And those appear to be random (do not appear to be associated with anything particular that I am doing at the time of the crash).

OK, thanks for explaining this.