unofficial mirror of bug-gnu-emacs@gnu.org 
 help / color / mirror / code / Atom feed
* bug#37786: 26.3; Emacs crashes when calling function to decode string
@ 2019-10-17  4:11 Allen Li
  2019-10-17  4:22 ` Lars Ingebrigtsen
  0 siblings, 1 reply; 15+ messages in thread
From: Allen Li @ 2019-10-17  4:11 UTC (permalink / raw)
  To: 37786


The following expression, when evaluated in emacs -Q, causes emacs to
crash and dump core:

(mail-decode-encoded-word-string #("=?UTF-8?B?4pyoUERYQ09OIFNBTEXinKggRW5qb3kgU3BlY2lhbCBPZmZlcnMg?= =?UTF-8?B?V2hpbGUgWW91IFdhaXQh?=" 0 64 (ws-butler-chg chg) 64 65 (ws-butler-chg chg) 65 97 (ws-butler-chg chg)))

In GNU Emacs 26.3 (build 1, x86_64-pc-linux-gnu, GTK+ Version 3.24.10)
 of 2019-08-29 built on juergen
Windowing system distributor 'The X.Org Foundation', version 11.0.12005000
System Description:	Arch Linux

Configured using:
 'configure --prefix=/usr --sysconfdir=/etc --libexecdir=/usr/lib
 --localstatedir=/var --with-x-toolkit=gtk3 --with-xft --with-modules
 'CFLAGS=-march=x86-64 -mtune=generic -O2 -pipe -fno-plt'
 CPPFLAGS=-D_FORTIFY_SOURCE=2
 LDFLAGS=-Wl,-O1,--sort-common,--as-needed,-z,relro,-z,now'

Configured features:
XPM JPEG TIFF GIF PNG RSVG IMAGEMAGICK SOUND GPM DBUS GSETTINGS GLIB
NOTIFY ACL GNUTLS LIBXML2 FREETYPE M17N_FLT LIBOTF XFT ZLIB
TOOLKIT_SCROLL_BARS GTK3 X11 XDBE XIM MODULES THREADS LIBSYSTEMD LCMS2

Important settings:
  value of $LANG: en_US.UTF-8
  locale-coding-system: utf-8-unix





^ permalink raw reply	[flat|nested] 15+ messages in thread

* bug#37786: 26.3; Emacs crashes when calling function to decode string
  2019-10-17  4:11 bug#37786: 26.3; Emacs crashes when calling function to decode string Allen Li
@ 2019-10-17  4:22 ` Lars Ingebrigtsen
  2019-10-17  9:53   ` Robert Pluim
  0 siblings, 1 reply; 15+ messages in thread
From: Lars Ingebrigtsen @ 2019-10-17  4:22 UTC (permalink / raw)
  To: Allen Li; +Cc: 37786

Allen Li <darkfeline@felesatra.moe> writes:

> The following expression, when evaluated in emacs -Q, causes emacs to
> crash and dump core:
>
> (mail-decode-encoded-word-string #("=?UTF-8?B?4pyoUERYQ09OIFNBTEXinKggRW5qb3kgU3BlY2lhbCBPZmZlcnMg?= =?UTF-8?B?V2hpbGUgWW91IFdhaXQh?=" 0 64 (ws-butler-chg chg) 64 65 (ws-butler-chg chg) 65 97 (ws-butler-chg chg)))

I'm unable to reproduce this in Emacs 27 or Emacs 26.3.

Could you do this under gdb and post the backtrace?

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





^ permalink raw reply	[flat|nested] 15+ messages in thread

* bug#37786: 26.3; Emacs crashes when calling function to decode string
  2019-10-17  4:22 ` Lars Ingebrigtsen
@ 2019-10-17  9:53   ` Robert Pluim
  2019-10-18  4:28     ` Allen Li
  0 siblings, 1 reply; 15+ messages in thread
From: Robert Pluim @ 2019-10-17  9:53 UTC (permalink / raw)
  To: Lars Ingebrigtsen; +Cc: 37786, Allen Li

>>>>> On Thu, 17 Oct 2019 06:22:10 +0200, Lars Ingebrigtsen <larsi@gnus.org> said:

    Lars> Allen Li <darkfeline@felesatra.moe> writes:
    >> The following expression, when evaluated in emacs -Q, causes emacs to
    >> crash and dump core:
    >> 
    >> (mail-decode-encoded-word-string
    >> #("=?UTF-8?B?4pyoUERYQ09OIFNBTEXinKggRW5qb3kgU3BlY2lhbCBPZmZlcnMg?=
    >> =?UTF-8?B?V2hpbGUgWW91IFdhaXQh?=" 0 64 (ws-butler-chg chg) 64 65
    >> (ws-butler-chg chg) 65 97 (ws-butler-chg chg)))

    Lars> I'm unable to reproduce this in Emacs 27 or Emacs 26.3.

    Lars> Could you do this under gdb and post the backtrace?

Probably font-related. I suspect "C-x 8 RET 2728" will cause the same
crash for OP.

Robert





^ permalink raw reply	[flat|nested] 15+ messages in thread

* bug#37786: 26.3; Emacs crashes when calling function to decode string
  2019-10-17  9:53   ` Robert Pluim
@ 2019-10-18  4:28     ` Allen Li
  2019-10-18  9:49       ` Robert Pluim
  0 siblings, 1 reply; 15+ messages in thread
From: Allen Li @ 2019-10-18  4:28 UTC (permalink / raw)
  To: Lars Ingebrigtsen; +Cc: 37786

Robert Pluim <rpluim@gmail.com> writes:

>>>>>> On Thu, 17 Oct 2019 06:22:10 +0200, Lars Ingebrigtsen <larsi@gnus.org> said:
>
>     Lars> Allen Li <darkfeline@felesatra.moe> writes:
>     >> The following expression, when evaluated in emacs -Q, causes emacs to
>     >> crash and dump core:
>     >> 
>     >> (mail-decode-encoded-word-string
>     >> #("=?UTF-8?B?4pyoUERYQ09OIFNBTEXinKggRW5qb3kgU3BlY2lhbCBPZmZlcnMg?=
>     >> =?UTF-8?B?V2hpbGUgWW91IFdhaXQh?=" 0 64 (ws-butler-chg chg) 64 65
>     >> (ws-butler-chg chg) 65 97 (ws-butler-chg chg)))
>
>     Lars> I'm unable to reproduce this in Emacs 27 or Emacs 26.3.
>
>     Lars> Could you do this under gdb and post the backtrace?

This is the backtrace using the simpler repro from Robert.

#0  0x0000000000565ae9 in terminate_due_to_signal (sig=6, backtrace_limit=40) at emacs.c:363
#1  0x000000000058ca83 in emacs_abort () at sysdep.c:2380
#2  0x000000000045c17e in redisplay_internal () at xdisp.c:13797
#3  0x000000000045e0b7 in redisplay_preserve_echo_area (from_where=13) at xdisp.c:14602
#4  0x0000000000667fbe in Fdelete_process (process=XIL(0x6aa3395)) at process.c:1054
#5  0x000000000067716b in kill_buffer_processes (buffer=XIL(0)) at process.c:7819
#6  0x00000000005683d0 in shut_down_emacs (sig=0, stuff=XIL(0)) at emacs.c:2096
#7  0x000000000052d854 in x_connection_closed (dpy=0x2b77f50, error_message=0x7fffffff6fc0 "X protocol error: BadLength (poly request too large or internal Xlib length error) on protocol request 139", ioerror=false) at xterm.c:9799
#8  0x000000000052da1b in x_error_quitter (display=0x2b77f50, event=0x7fffffff7160) at xterm.c:9893
#9  0x000000000052d966 in x_error_handler (display=0x2b77f50, event=0x7fffffff7160) at xterm.c:9863
#10 0x00007ffff6ce75db in _XError () at /usr/lib/libX11.so.6
#11 0x00007ffff6ce4388 in  () at /usr/lib/libX11.so.6
#12 0x00007ffff6ce4425 in  () at /usr/lib/libX11.so.6
#13 0x00007ffff6ce4d8a in _XEventsQueued () at /usr/lib/libX11.so.6
#14 0x00007ffff6cd6782 in XPending () at /usr/lib/libX11.so.6
#15 0x00007ffff7671a00 in  () at /usr/lib/libgdk-3.so.0
#16 0x00007ffff6e79a60 in g_main_context_prepare () at /usr/lib/libglib-2.0.so.0
#17 0x00007ffff6e7a0a6 in  () at /usr/lib/libglib-2.0.so.0
#18 0x00007ffff6e7a2fa in g_main_context_pending () at /usr/lib/libglib-2.0.so.0
#19 0x00007ffff79ae550 in gtk_events_pending () at /usr/lib/libgtk-3.so.0
#20 0x000000000052c185 in XTread_socket (terminal=0x1232e40 <bss_sbrk_buffer+7610880>, hold_quit=0x7fffffff74a0) at xterm.c:9120
#21 0x000000000057685f in gobble_input () at keyboard.c:6909
#22 0x0000000000576da8 in handle_async_input () at keyboard.c:7146
#23 0x0000000000576dc7 in process_pending_signals () at keyboard.c:7160
#24 0x0000000000576e07 in unblock_input_to (level=0) at keyboard.c:7175
#25 0x0000000000576e2b in unblock_input () at keyboard.c:7194
#26 0x00000000006ab340 in xftfont_open (f=0x1279c30 <bss_sbrk_buffer+7901168>, entity=XIL(0x5f805b5), pixel_size=15) at xftfont.c:391
#27 0x000000000063107b in font_open_entity (f=0x1279c30 <bss_sbrk_buffer+7901168>, entity=XIL(0x5f805b5), pixel_size=15) at font.c:2903
#28 0x0000000000632a50 in font_open_for_lface (f=0x1279c30 <bss_sbrk_buffer+7901168>, entity=XIL(0x5f805b5), attrs=0x32610d0, spec=XIL(0)) at font.c:3332
#29 0x00000000006aea8a in fontset_find_font (fontset=XIL(0x1470c35), c=10024, face=0x32610d0, charset_id=-1, fallback=true) at fontset.c:707
#30 0x00000000006aefb7 in fontset_font (fontset=XIL(0x1470c35), c=10024, face=0x32610d0, id=-1) at fontset.c:788
#31 0x00000000006af6bc in face_for_char (f=0x1279c30 <bss_sbrk_buffer+7901168>, face=0x32610d0, c=10024, pos=1, object=XIL(0)) at fontset.c:990
#32 0x0000000000563994 in FACE_FOR_CHAR (f=0x1279c30 <bss_sbrk_buffer+7901168>, face=0x32610d0, character=10024, pos=1, object=XIL(0)) at dispextern.h:1818
#33 0x000000000044b256 in get_next_display_element (it=0x7fffffff8ef0) at xdisp.c:7288
#34 0x00000000004730c0 in display_line (it=0x7fffffff8ef0, cursor_vpos=0) at xdisp.c:21337
#35 0x0000000000467f0f in try_window (window=XIL(0x127ac35), pos=..., flags=1) at xdisp.c:17592
#36 0x0000000000465a0d in redisplay_window (window=XIL(0x127ac35), just_this_one_p=false) at xdisp.c:17039
#37 0x000000000045e869 in redisplay_window_0 (window=XIL(0x127ac35)) at xdisp.c:14799
#38 0x000000000061316d in internal_condition_case_1 (bfun=0x45e827 <redisplay_window_0>, arg=XIL(0x127ac35), handlers=XIL(0xb5afb3), hfun=0x45e7ef <redisplay_window_error>) at eval.c:1356
#39 0x000000000045e7c3 in redisplay_windows (window=XIL(0x127ac35)) at xdisp.c:14779
#40 0x000000000045d608 in redisplay_internal () at xdisp.c:14268
#41 0x000000000045b4ee in redisplay () at xdisp.c:13488
#42 0x000000000056db90 in read_char (commandflag=1, map=XIL(0x6564303), prev_event=XIL(0), used_mouse_menu=0x7fffffffe1f1, end_time=0x0) at keyboard.c:2480
#43 0x000000000057b533 in read_key_sequence (keybuf=0x7fffffffe390, bufsize=30, prompt=XIL(0), dont_downcase_last=false, can_return_switch_frame=true, fix_current_buffer=true, prevent_redisplay=false) at keyboard.c:9147
#44 0x000000000056adfb in command_loop_1 () at keyboard.c:1368
#45 0x00000000006130c6 in internal_condition_case (bfun=0x56a9cf <command_loop_1>, handlers=XIL(0x5220), hfun=0x56a17a <cmd_error>) at eval.c:1332
#46 0x000000000056a6b0 in command_loop_2 (ignore=XIL(0)) at keyboard.c:1110
#47 0x0000000000612963 in internal_catch (tag=XIL(0xc6c0), func=0x56a683 <command_loop_2>, arg=XIL(0)) at eval.c:1097
#48 0x000000000056a64e in command_loop () at keyboard.c:1089
#49 0x0000000000569d4b in recursive_edit_1 () at keyboard.c:695
#50 0x0000000000569ecc in Frecursive_edit () at keyboard.c:766
#51 0x00000000005679ae in main (argc=1, argv=0x7fffffffe7f8) at emacs.c:1713


>
> Probably font-related. I suspect "C-x 8 RET 2728" will cause the same
> crash for OP.
>
> Robert

Yep, this also crashes for me.





^ permalink raw reply	[flat|nested] 15+ messages in thread

* bug#37786: 26.3; Emacs crashes when calling function to decode string
  2019-10-18  4:28     ` Allen Li
@ 2019-10-18  9:49       ` Robert Pluim
  2019-10-19  1:18         ` Allen Li
  0 siblings, 1 reply; 15+ messages in thread
From: Robert Pluim @ 2019-10-18  9:49 UTC (permalink / raw)
  To: Allen Li; +Cc: Lars Ingebrigtsen, 37786

>>>>> On Thu, 17 Oct 2019 21:28:35 -0700, Allen Li <darkfeline@felesatra.moe> said:

    Allen> This is the backtrace using the simpler repro from Robert.

    Allen> #0  0x0000000000565ae9 in terminate_due_to_signal (sig=6, backtrace_limit=40) at emacs.c:363
    Allen> #1  0x000000000058ca83 in emacs_abort () at sysdep.c:2380
    Allen> #2  0x000000000045c17e in redisplay_internal () at xdisp.c:13797
    Allen> #3  0x000000000045e0b7 in redisplay_preserve_echo_area (from_where=13) at xdisp.c:14602
    Allen> #4  0x0000000000667fbe in Fdelete_process (process=XIL(0x6aa3395)) at process.c:1054
    Allen> #5  0x000000000067716b in kill_buffer_processes (buffer=XIL(0)) at process.c:7819
    Allen> #6  0x00000000005683d0 in shut_down_emacs (sig=0, stuff=XIL(0)) at emacs.c:2096
    Allen> #7 0x000000000052d854 in x_connection_closed (dpy=0x2b77f50,
    Allen> error_message=0x7fffffff6fc0 "X protocol error: BadLength (poly
    Allen> request too large or internal Xlib length error) on protocol request
    Allen> 139", ioerror=false) at xterm.c:9799

I thought weʼd fixed these kinds of bugs, but obviously not. Can you
try your test in combination with the following from etc/DEBUG, it
should tell us which font is responsible:

    For X protocol errors related to displaying unusual characters or to
    font-related customizations, try invoking Emacs like this:

      XFT_DEBUG=16 emacs -xrm "emacs.synchronous: true"

    This should produce information from the libXft library which could
    give useful hints regarding font-related problems in that library.

Regards

Robert





^ permalink raw reply	[flat|nested] 15+ messages in thread

* bug#37786: 26.3; Emacs crashes when calling function to decode string
  2019-10-18  9:49       ` Robert Pluim
@ 2019-10-19  1:18         ` Allen Li
  2019-10-21 10:02           ` Robert Pluim
  0 siblings, 1 reply; 15+ messages in thread
From: Allen Li @ 2019-10-19  1:18 UTC (permalink / raw)
  To: Robert Pluim; +Cc: 37786

Robert Pluim <rpluim@gmail.com> writes:

>>>>>> On Thu, 17 Oct 2019 21:28:35 -0700, Allen Li <darkfeline@felesatra.moe> said:
>
>     Allen> This is the backtrace using the simpler repro from Robert.
>
>     Allen> #0  0x0000000000565ae9 in terminate_due_to_signal (sig=6, backtrace_limit=40) at emacs.c:363
>     Allen> #1  0x000000000058ca83 in emacs_abort () at sysdep.c:2380
>     Allen> #2  0x000000000045c17e in redisplay_internal () at xdisp.c:13797
>     Allen> #3  0x000000000045e0b7 in redisplay_preserve_echo_area (from_where=13) at xdisp.c:14602
>     Allen> #4  0x0000000000667fbe in Fdelete_process (process=XIL(0x6aa3395)) at process.c:1054
>     Allen> #5  0x000000000067716b in kill_buffer_processes (buffer=XIL(0)) at process.c:7819
>     Allen> #6  0x00000000005683d0 in shut_down_emacs (sig=0, stuff=XIL(0)) at emacs.c:2096
>     Allen> #7 0x000000000052d854 in x_connection_closed (dpy=0x2b77f50,
>     Allen> error_message=0x7fffffff6fc0 "X protocol error: BadLength (poly
>     Allen> request too large or internal Xlib length error) on protocol request
>     Allen> 139", ioerror=false) at xterm.c:9799
>
> I thought weʼd fixed these kinds of bugs, but obviously not. Can you
> try your test in combination with the following from etc/DEBUG, it
> should tell us which font is responsible:
>
>     For X protocol errors related to displaying unusual characters or to
>     font-related customizations, try invoking Emacs like this:
>
>       XFT_DEBUG=16 emacs -xrm "emacs.synchronous: true"
>
>     This should produce information from the libXft library which could
>     give useful hints regarding font-related problems in that library.

That gives me the following output:

Loading file /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Regular.otf/0
FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-It.otf/0 matches new
Loading file /usr/share/fonts/adobe-source-code-pro/SourceCodePro-It.otf/0
FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Light.otf/0 matches new
Loading file /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Light.otf/0
FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Bold.otf/0 matches new
Loading file /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Bold.otf/0
FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Regular.otf/0 matches existing (2)
FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Light.otf/0 matches existing (2)
FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Bold.otf/0 matches existing (2)
FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Regular.otf/0 matches existing (3)
FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Light.otf/0 matches existing (3)
FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Bold.otf/0 matches existing (3)
FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Regular.otf/0 matches existing (4)
FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Regular.otf/0 matches existing (4)
FontFile /usr/share/fonts/noto/NotoColorEmoji.ttf/0 matches new
Loading file /usr/share/fonts/noto/NotoColorEmoji.ttf/0

and then Emacs hangs at this point.

>
> Regards
>
> Robert





^ permalink raw reply	[flat|nested] 15+ messages in thread

* bug#37786: 26.3; Emacs crashes when calling function to decode string
  2019-10-19  1:18         ` Allen Li
@ 2019-10-21 10:02           ` Robert Pluim
  2019-10-22  3:44             ` Allen Li
  0 siblings, 1 reply; 15+ messages in thread
From: Robert Pluim @ 2019-10-21 10:02 UTC (permalink / raw)
  To: Allen Li; +Cc: 37786

>>>>> On Fri, 18 Oct 2019 18:18:34 -0700, Allen Li <darkfeline@felesatra.moe> said:
    Allen> That gives me the following output:

    Allen> Loading file /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Regular.otf/0
    Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-It.otf/0 matches new
    Allen> Loading file /usr/share/fonts/adobe-source-code-pro/SourceCodePro-It.otf/0
    Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Light.otf/0 matches new
    Allen> Loading file /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Light.otf/0
    Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Bold.otf/0 matches new
    Allen> Loading file /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Bold.otf/0
    Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Regular.otf/0 matches existing (2)
    Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Light.otf/0 matches existing (2)
    Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Bold.otf/0 matches existing (2)
    Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Regular.otf/0 matches existing (3)
    Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Light.otf/0 matches existing (3)
    Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Bold.otf/0 matches existing (3)
    Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Regular.otf/0 matches existing (4)
    Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Regular.otf/0 matches existing (4)
    Allen> FontFile /usr/share/fonts/noto/NotoColorEmoji.ttf/0 matches new
    Allen> Loading file /usr/share/fonts/noto/NotoColorEmoji.ttf/0

    Allen> and then Emacs hangs at this point.

Hmm, that configuration of emacs should not be loading NotoColorEmoji
(unless you've somehow arranged for xft-ignore-color-fonts to be
nil). What's your version of fontconfig (probably from
/usr/include/fontconfig/fontconfig.h)?

(note that building emacs-27 with cairo enabled should solve this:
that doesnʼt use XFT).

Robert





^ permalink raw reply	[flat|nested] 15+ messages in thread

* bug#37786: 26.3; Emacs crashes when calling function to decode string
  2019-10-21 10:02           ` Robert Pluim
@ 2019-10-22  3:44             ` Allen Li
  2019-10-22  8:18               ` Robert Pluim
  0 siblings, 1 reply; 15+ messages in thread
From: Allen Li @ 2019-10-22  3:44 UTC (permalink / raw)
  To: 37786

Robert Pluim <rpluim@gmail.com> writes:

>>>>>> On Fri, 18 Oct 2019 18:18:34 -0700, Allen Li <darkfeline@felesatra.moe> said:
>     Allen> That gives me the following output:
>
>     Allen> Loading file /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Regular.otf/0
>     Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-It.otf/0 matches new
>     Allen> Loading file /usr/share/fonts/adobe-source-code-pro/SourceCodePro-It.otf/0
>     Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Light.otf/0 matches new
>     Allen> Loading file /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Light.otf/0
>     Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Bold.otf/0 matches new
>     Allen> Loading file /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Bold.otf/0
>     Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Regular.otf/0 matches existing (2)
>     Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Light.otf/0 matches existing (2)
>     Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Bold.otf/0 matches existing (2)
>     Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Regular.otf/0 matches existing (3)
>     Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Light.otf/0 matches existing (3)
>     Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Bold.otf/0 matches existing (3)
>     Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Regular.otf/0 matches existing (4)
>     Allen> FontFile /usr/share/fonts/adobe-source-code-pro/SourceCodePro-Regular.otf/0 matches existing (4)
>     Allen> FontFile /usr/share/fonts/noto/NotoColorEmoji.ttf/0 matches new
>     Allen> Loading file /usr/share/fonts/noto/NotoColorEmoji.ttf/0
>
>     Allen> and then Emacs hangs at this point.
>
> Hmm, that configuration of emacs should not be loading NotoColorEmoji
> (unless you've somehow arranged for xft-ignore-color-fonts to be
> nil). What's your version of fontconfig (probably from
> /usr/include/fontconfig/fontconfig.h)?

I haven't touched xft-ignore-color-fonts (and it reproduces from emacs -Q)

#define FC_MAJOR	2
#define FC_MINOR	13
#define FC_REVISION	91

>
> (note that building emacs-27 with cairo enabled should solve this:
> that doesnʼt use XFT).
>
> Robert





^ permalink raw reply	[flat|nested] 15+ messages in thread

* bug#37786: 26.3; Emacs crashes when calling function to decode string
  2019-10-22  3:44             ` Allen Li
@ 2019-10-22  8:18               ` Robert Pluim
  2019-10-23  3:25                 ` Allen Li
  0 siblings, 1 reply; 15+ messages in thread
From: Robert Pluim @ 2019-10-22  8:18 UTC (permalink / raw)
  To: Allen Li; +Cc: 37786

>>>>> On Mon, 21 Oct 2019 20:44:09 -0700, Allen Li <darkfeline@felesatra.moe> said:

    >> Hmm, that configuration of emacs should not be loading NotoColorEmoji
    >> (unless you've somehow arranged for xft-ignore-color-fonts to be
    >> nil). What's your version of fontconfig (probably from
    >> /usr/include/fontconfig/fontconfig.h)?

    Allen> I haven't touched xft-ignore-color-fonts (and it reproduces from emacs -Q)

    Allen> #define FC_MAJOR	2
    Allen> #define FC_MINOR	13
    Allen> #define FC_REVISION	91

Now Iʼm definitely confused. Which distribution/version of GNU/Linux is this?

Robert





^ permalink raw reply	[flat|nested] 15+ messages in thread

* bug#37786: 26.3; Emacs crashes when calling function to decode string
  2019-10-22  8:18               ` Robert Pluim
@ 2019-10-23  3:25                 ` Allen Li
  2019-10-25 16:32                   ` Robert Pluim
  0 siblings, 1 reply; 15+ messages in thread
From: Allen Li @ 2019-10-23  3:25 UTC (permalink / raw)
  To: 37786

Robert Pluim <rpluim@gmail.com> writes:

>>>>>> On Mon, 21 Oct 2019 20:44:09 -0700, Allen Li <darkfeline@felesatra.moe> said:
>
>     >> Hmm, that configuration of emacs should not be loading NotoColorEmoji
>     >> (unless you've somehow arranged for xft-ignore-color-fonts to be
>     >> nil). What's your version of fontconfig (probably from
>     >> /usr/include/fontconfig/fontconfig.h)?
>
>     Allen> I haven't touched xft-ignore-color-fonts (and it reproduces from emacs -Q)
>
>     Allen> #define FC_MAJOR	2
>     Allen> #define FC_MINOR	13
>     Allen> #define FC_REVISION	91
>
> Now Iʼm definitely confused. Which distribution/version of GNU/Linux is this?

Arch Linux

>
> Robert





^ permalink raw reply	[flat|nested] 15+ messages in thread

* bug#37786: 26.3; Emacs crashes when calling function to decode string
  2019-10-23  3:25                 ` Allen Li
@ 2019-10-25 16:32                   ` Robert Pluim
  2019-10-26  9:44                     ` Allen Li
  0 siblings, 1 reply; 15+ messages in thread
From: Robert Pluim @ 2019-10-25 16:32 UTC (permalink / raw)
  To: Allen Li; +Cc: 37786

>>>>> On Tue, 22 Oct 2019 20:25:40 -0700, Allen Li <darkfeline@felesatra.moe> said:

    Allen> I haven't touched xft-ignore-color-fonts (and it reproduces from emacs -Q)
    >> 
    Allen> #define FC_MAJOR	2
    Allen> #define FC_MINOR	13
    Allen> #define FC_REVISION	91
    >> 

Thatʼs a very recent (and maybe even unreleased) version of
fontconfig.

    Allen> Arch Linux

Hmm. Could you try the following against emacs-26 and see if it fixes
your crash? (the character will almost certainly end up displayed
wrong). This looks more like a fontconfig bug than anything else: the
pattern we've supplied to FcFontList explicitly says "donʼt give me
color fonts".

diff --git i/lisp/international/fontset.el w/lisp/international/fontset.el
index c90d4f53bd..e80a1a87b9 100644
--- i/lisp/international/fontset.el
+++ w/lisp/international/fontset.el
@@ -804,7 +804,6 @@ setup-default-fontset
              #x2664
              (#x2667 . #x2669)
              (#x266C . #x26FF)
-             (#x2700 . #x27bF) ;; Dingbats
              (#x27C0 . #x27EF) ;; Misc Mathematical Symbols-A
              (#x27F0 . #x27FF) ;; Supplemental Arrows-A
              (#x2900 . #x297F) ;; Supplemental Arrows-B
diff --git i/src/ftfont.c w/src/ftfont.c
index 823fb2095c..017b349318 100644
--- i/src/ftfont.c
+++ w/src/ftfont.c
@@ -861,6 +861,9 @@ ftfont_list (struct frame *f, Lisp_Object spec)
 #endif /* FC_CAPABILITY */
 #ifdef FC_FONTFORMAT
                             FC_FONTFORMAT,
+#endif
+#ifdef FC_COLOR
+                             FC_COLOR,
 #endif
                             NULL);
   if (! objset)
@@ -902,6 +905,15 @@ ftfont_list (struct frame *f, Lisp_Object spec)
     {
       Lisp_Object entity;

+      {
+        FcBool b;
+        if (FcPatternGetBool (fontset->fonts[i], FC_COLOR, 0, &b)
+            == FcResultMatch && b)
+          {
+            fprintf (stderr, "Skipping Color font\n");
+            continue;
+          }
+      }
       if (spacing >= 0)
        {
          int this;





^ permalink raw reply related	[flat|nested] 15+ messages in thread

* bug#37786: 26.3; Emacs crashes when calling function to decode string
  2019-10-25 16:32                   ` Robert Pluim
@ 2019-10-26  9:44                     ` Allen Li
  2019-10-27  8:26                       ` Robert Pluim
  0 siblings, 1 reply; 15+ messages in thread
From: Allen Li @ 2019-10-26  9:44 UTC (permalink / raw)
  To: 37786

Robert Pluim <rpluim@gmail.com> writes:

>>>>>> On Tue, 22 Oct 2019 20:25:40 -0700, Allen Li <darkfeline@felesatra.moe> said:
>
>     Allen> I haven't touched xft-ignore-color-fonts (and it reproduces from emacs -Q)
>     >> 
>     Allen> #define FC_MAJOR	2
>     Allen> #define FC_MINOR	13
>     Allen> #define FC_REVISION	91
>     >> 
>
> Thatʼs a very recent (and maybe even unreleased) version of
> fontconfig.
>
>     Allen> Arch Linux
>
> Hmm. Could you try the following against emacs-26 and see if it fixes
> your crash? (the character will almost certainly end up displayed
> wrong). This looks more like a fontconfig bug than anything else: the
> pattern we've supplied to FcFontList explicitly says "donʼt give me
> color fonts".

Yep, this fixes the crash and I get the box with unicode codepoint
inside.  I don't know anything about fontconfig, so I can't comment on
whether this is a fontconfig bug.


>
> diff --git i/lisp/international/fontset.el w/lisp/international/fontset.el
> index c90d4f53bd..e80a1a87b9 100644
> --- i/lisp/international/fontset.el
> +++ w/lisp/international/fontset.el
> @@ -804,7 +804,6 @@ setup-default-fontset
>               #x2664
>               (#x2667 . #x2669)
>               (#x266C . #x26FF)
> -             (#x2700 . #x27bF) ;; Dingbats
>               (#x27C0 . #x27EF) ;; Misc Mathematical Symbols-A
>               (#x27F0 . #x27FF) ;; Supplemental Arrows-A
>               (#x2900 . #x297F) ;; Supplemental Arrows-B
> diff --git i/src/ftfont.c w/src/ftfont.c
> index 823fb2095c..017b349318 100644
> --- i/src/ftfont.c
> +++ w/src/ftfont.c
> @@ -861,6 +861,9 @@ ftfont_list (struct frame *f, Lisp_Object spec)
>  #endif /* FC_CAPABILITY */
>  #ifdef FC_FONTFORMAT
>                              FC_FONTFORMAT,
> +#endif
> +#ifdef FC_COLOR
> +                             FC_COLOR,
>  #endif
>                              NULL);
>    if (! objset)
> @@ -902,6 +905,15 @@ ftfont_list (struct frame *f, Lisp_Object spec)
>      {
>        Lisp_Object entity;
>
> +      {
> +        FcBool b;
> +        if (FcPatternGetBool (fontset->fonts[i], FC_COLOR, 0, &b)
> +            == FcResultMatch && b)
> +          {
> +            fprintf (stderr, "Skipping Color font\n");
> +            continue;
> +          }
> +      }
>        if (spacing >= 0)
>         {
>           int this;





^ permalink raw reply	[flat|nested] 15+ messages in thread

* bug#37786: 26.3; Emacs crashes when calling function to decode string
  2019-10-26  9:44                     ` Allen Li
@ 2019-10-27  8:26                       ` Robert Pluim
  2019-11-05 17:40                         ` Robert Pluim
  0 siblings, 1 reply; 15+ messages in thread
From: Robert Pluim @ 2019-10-27  8:26 UTC (permalink / raw)
  To: Allen Li; +Cc: 37786

>>>>> On Sat, 26 Oct 2019 02:44:52 -0700, Allen Li <darkfeline@felesatra.moe> said:

    >> Hmm. Could you try the following against emacs-26 and see if it fixes
    >> your crash? (the character will almost certainly end up displayed
    >> wrong). This looks more like a fontconfig bug than anything else: the
    >> pattern we've supplied to FcFontList explicitly says "donʼt give me
    >> color fonts".

    Allen> Yep, this fixes the crash and I get the box with unicode codepoint
    Allen> inside.  I don't know anything about fontconfig, so I can't comment on
    Allen> whether this is a fontconfig bug.

Thanks for testing. Iʼll have to write a standalone test case against
fontconfig before I can say whether itʼs a fontconfig bug or not, in
the meantime Iʼll clean up the patch.

Regards

Robert





^ permalink raw reply	[flat|nested] 15+ messages in thread

* bug#37786: 26.3; Emacs crashes when calling function to decode string
  2019-10-27  8:26                       ` Robert Pluim
@ 2019-11-05 17:40                         ` Robert Pluim
  2019-11-13 14:03                           ` Robert Pluim
  0 siblings, 1 reply; 15+ messages in thread
From: Robert Pluim @ 2019-11-05 17:40 UTC (permalink / raw)
  To: Allen Li; +Cc: 37786

[-- Attachment #1: Type: text/plain, Size: 1242 bytes --]

>>>>> On Sun, 27 Oct 2019 09:26:35 +0100, Robert Pluim <rpluim@gmail.com> said:

>>>>> On Sat, 26 Oct 2019 02:44:52 -0700, Allen Li <darkfeline@felesatra.moe> said:
    >>> Hmm. Could you try the following against emacs-26 and see if it fixes
    >>> your crash? (the character will almost certainly end up displayed
    >>> wrong). This looks more like a fontconfig bug than anything else: the
    >>> pattern we've supplied to FcFontList explicitly says "donʼt give me
    >>> color fonts".

    Allen> Yep, this fixes the crash and I get the box with unicode codepoint
    Allen> inside.  I don't know anything about fontconfig, so I can't comment on
    Allen> whether this is a fontconfig bug.

    Robert> Thanks for testing. Iʼll have to write a standalone test case against
    Robert> fontconfig before I can say whether itʼs a fontconfig bug or not, in
    Robert> the meantime Iʼll clean up the patch.

I think itʼs a fontconfig bug, but Iʼve received no response from the
fontconfig guys. Perhaps I need to use their new-fangled issue tracker
thing rather than good old email. In any case, both Arch and Fedora 31
have fontconfig packages with this issue, so we need to paper^Wfix the
issue our end.


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0001-Ignore-color-fonts-returned-from-FcFontList.patch --]
[-- Type: text/x-patch, Size: 1696 bytes --]

From aabd13bba282563410ef95764cad39f02ecb5d84 Mon Sep 17 00:00:00 2001
From: Robert Pluim <rpluim@gmail.com>
Date: Mon, 4 Nov 2019 17:44:57 +0100
Subject: [PATCH] Ignore color fonts returned from FcFontList
To: emacs-devel@gnu.org

* src/ftfont.c (ftfont_list): [HAVE_XFT && FC_COLOR]: Ask FcFontList
to return FC_COLOR attribute.  Check returned attribute for
non-FcFalse, since some color fonts have a color attribute that's
neither FcFalse nor FcTrue (Bug#37786).
---
 src/ftfont.c | 17 ++++++++++++++++-
 1 file changed, 16 insertions(+), 1 deletion(-)

diff --git a/src/ftfont.c b/src/ftfont.c
index 77a4cf5de5..b066f55a18 100644
--- a/src/ftfont.c
+++ b/src/ftfont.c
@@ -864,6 +864,9 @@ ftfont_list (struct frame *f, Lisp_Object spec)
 #endif	/* FC_CAPABILITY */
 #ifdef FC_FONTFORMAT
 			     FC_FONTFORMAT,
+#endif
+#if defined HAVE_XFT && defined FC_COLOR
+                             FC_COLOR,
 #endif
 			     NULL);
   if (! objset)
@@ -904,7 +907,19 @@ ftfont_list (struct frame *f, Lisp_Object spec)
   for (i = 0; i < fontset->nfont; i++)
     {
       Lisp_Object entity;
-
+#if defined HAVE_XFT && defined FC_COLOR
+      {
+        /* Some fonts, notably NotoColorEmoji, have an FC_COLOR value
+           that's neither FcTrue nor FcFalse, which means FcFontList
+           returns them even when it shouldn't really do so, so we
+           need to manually skip them here (Bug#37786).  */
+        FcBool b;
+        if (Vxft_ignore_color_fonts
+            && FcPatternGetBool (fontset->fonts[i], FC_COLOR, 0, &b)
+            == FcResultMatch && b != FcFalse)
+            continue;
+      }
+#endif
       if (spacing >= 0)
 	{
 	  int this;
-- 
2.19.1.816.gcd69ec8cde


^ permalink raw reply related	[flat|nested] 15+ messages in thread

* bug#37786: 26.3; Emacs crashes when calling function to decode string
  2019-11-05 17:40                         ` Robert Pluim
@ 2019-11-13 14:03                           ` Robert Pluim
  0 siblings, 0 replies; 15+ messages in thread
From: Robert Pluim @ 2019-11-13 14:03 UTC (permalink / raw)
  To: Allen Li; +Cc: 37786

tags 37786 fixed
close 37786 27.1
quit

Fixed by adding "Noto Color Emoji" to face-ignored-fonts
instead. Closing.

Committed as eae50e88ef





^ permalink raw reply	[flat|nested] 15+ messages in thread

end of thread, other threads:[~2019-11-13 14:03 UTC | newest]

Thread overview: 15+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2019-10-17  4:11 bug#37786: 26.3; Emacs crashes when calling function to decode string Allen Li
2019-10-17  4:22 ` Lars Ingebrigtsen
2019-10-17  9:53   ` Robert Pluim
2019-10-18  4:28     ` Allen Li
2019-10-18  9:49       ` Robert Pluim
2019-10-19  1:18         ` Allen Li
2019-10-21 10:02           ` Robert Pluim
2019-10-22  3:44             ` Allen Li
2019-10-22  8:18               ` Robert Pluim
2019-10-23  3:25                 ` Allen Li
2019-10-25 16:32                   ` Robert Pluim
2019-10-26  9:44                     ` Allen Li
2019-10-27  8:26                       ` Robert Pluim
2019-11-05 17:40                         ` Robert Pluim
2019-11-13 14:03                           ` Robert Pluim

Code repositories for project(s) associated with this public inbox

	https://git.savannah.gnu.org/cgit/emacs.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).