jwm -p option segfaults

[01micko]

release 2.4.1.

I can't see anything in the commit log that would fix it since February release.

~$ dmesg | tail -n 6
[ 1554.005169] jwm[31592]: segfault at 10 ip 000000000040f218 sp 00007ffed5ac9620 error 4 in jwm[406000+25000]
[ 1554.005187] Code: 01 00 00 41 89 df 8d 83 00 01 00 00 44 39 e8 76 ce eb b9 0f 1f 80 00 00 00 00 31 c0 31 d2 e9 3b ff ff ff 48 8b 05 28 b8 02 00 <8b> 78 10 e8 d0 7d ff ff 8b 7c 24 18 e8 c7 7d ff ff 8b 7c 24 1c be
[ 1710.878069] jwm[31993]: segfault at 10 ip 000000000040f218 sp 00007fff8049c770 error 4 in jwm[406000+25000]
[ 1710.878086] Code: 01 00 00 41 89 df 8d 83 00 01 00 00 44 39 e8 76 ce eb b9 0f 1f 80 00 00 00 00 31 c0 31 d2 e9 3b ff ff ff 48 8b 05 28 b8 02 00 <8b> 78 10 e8 d0 7d ff ff 8b 7c 24 18 e8 c7 7d ff ff 8b 7c 24 1c be
[ 1730.998025] jwm[32036]: segfault at 10 ip 000000000040f218 sp 00007fffdbee86f0 error 4 in jwm[406000+25000]
[ 1730.998042] Code: 01 00 00 41 89 df 8d 83 00 01 00 00 44 39 e8 76 ce eb b9 0f 1f 80 00 00 00 00 31 c0 31 d2 e9 3b ff ff ff 48 8b 05 28 b8 02 00 <8b> 78 10 e8 d0 7d ff ff 8b 7c 24 18 e8 c7 7d ff ff 8b 7c 24 1c be

Every time I run jwm -p it hangs. I have to ctrl-C to bail and I see a new dmesg error each time as noted above. Do you need a debug build test?

Thanks for looking.

[joewing]

Unfortunately, I don't think there's enough information in that for me to do anything with. jwm -p runs ok for me, so I'm guessing it's something specific to your configuration maybe? If you could attach a debugger and get me a backtrace from a debug build that would be helpful. Or if you could somehow find a minimal change from the example configuration that causes the issue.

[01micko]

Hi @joewing and thank you for the response.

There definitely is things wrong with my configuration as I just upgraded from 2.3.7 to 2.4.1. However I don't think there is too much to cause a segfault. The whole point of me running jwm -p was to have a quick look at errors in the config and fix them. I still think it's a bug

I don't think there's enough information in that for me to do anything with.

Of course. When I get time I'll do a debug build and a backtrace with gdb. I actually just learnt how to do this remotely debugging a segfault in labwc (issue 194)

[joewing]

Thanks for looking into it.

It's almost certainly a JWM bug as it shouldn't be possible to cause a crash. What I meant to say is that being configuration related, it might be possible to provide me with a self-contained configuration so that I can reproduce the problem locally.

[01micko]

So this is really interesting (interesting = strange)

I'll use dotpoints

• kernel says there's a segfault (dmesg)

[  935.677788] Code: c0 0f 8e 2c 01 00 00 41 89 df 8d 83 00 01 00 00 44 39 e8 76 ce eb af 0f 1f 00 31 c0 31 d2 e9 2b ff ff ff 48 8b 05 40 f8 02 00 <8b> 78 10 e8 30 6d ff ff 8b 7c 24 18 e8 27 6d ff ff 8b 7c 24 1c be
[  995.738476] jwm[15202]: segfault at 10 ip 00000000004102f8 sp 00007fff76869170 error 4 in jwm[406000+2a000]
[  995.738495] Code: c0 0f 8e 2c 01 00 00 41 89 df 8d 83 00 01 00 00 44 39 e8 76 ce eb af 0f 1f 00 31 c0 31 d2 e9 2b ff ff ff 48 8b 05 40 f8 02 00 <8b> 78 10 e8 30 6d ff ff 8b 7c 24 18 e8 27 6d ff ff 8b 7c 24 1c be
[ 1055.798961] jwm[15325]: segfault at 10 ip 00000000004102f8 sp 00007fff76869170 error 4 in jwm[406000+2a000]
[ 1055.798979] Code: c0 0f 8e 2c 01 00 00 41 89 df 8d 83 00 01 00 00 44 39 e8 76 ce eb af 0f 1f 00 31 c0 31 d2 e9 2b ff ff ff 48 8b 05 40 f8 02 00 <8b> 78 10 e8 30 6d ff ff 8b 7c 24 18 e8 27 6d ff ff 8b 7c 24 1c be
[ 1115.859932] jwm[15446]: segfault at 10 ip 00000000004102f8 sp 00007fff76869170 error 4 in jwm[406000+2a000]
[ 1115.859955] Code: c0 0f 8e 2c 01 00 00 41 89 df 8d 83 00 01 00 00 44 39 e8 76 ce eb af 0f 1f 00 31 c0 31 d2 e9 2b ff ff ff 48 8b 05 40 f8 02 00 <8b> 78 10 e8 30 6d ff ff 8b 7c 24 18 e8 27 6d ff ff 8b 7c 24 1c be
[ 1175.897096] jwm[15746]: segfault at 10 ip 00000000004102f8 sp 00007fff76869170 error 4 in jwm[406000+2a000]
[ 1175.897117] Code: c0 0f 8e 2c 01 00 00 41 89 df 8d 83 00 01 00 00 44 39 e8 76 ce eb af 0f 1f 00 31 c0 31 d2 e9 2b ff ff ff 48 8b 05 40 f8 02 00 <8b> 78 10 e8 30 6d ff ff 8b 7c 24 18 e8 27 6d ff ff 8b 7c 24 1c be

• jwm does not crash - kernel says there's a segfault
• jwm is compiled with debuging and attaching gdb to the running process says so
• notice how these little 'crashes' occur almost every 60 seconds; ~150 milliseconds over

~$ gdb -p `pidof jwm`  
GNU gdb (GDB) 11.2
Copyright (C) 2022 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.
Type "show copying" and "show warranty" for details.
This GDB was configured as "x86_64-slackware-linux".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<https://www.gnu.org/software/gdb/bugs/>.
Find the GDB manual and other documentation resources online at:
    <http://www.gnu.org/software/gdb/documentation/>.

For help, type "help".
Type "apropos word" to search for commands related to "word".
Attaching to process 9286
Reading symbols from /usr/bin/jwm...
Reading symbols from /usr/lib64/libX11.so.6...
(No debugging symbols found in /usr/lib64/libX11.so.6)
Reading symbols from /usr/lib64/libpng16.so.16...
.. and so on

• and output of jwm -restart ; jwm -p

~$ jwm -restart
DEBUG: main.c[116]: debug mode started
DEBUG: main.c[235]: debug mode stopped
DEBUG: MEMORY: no memory leaks
~$ jwm -p
DEBUG: main.c[116]: debug mode started
JWM: warning: timeout: /root/.jwm/root_menus/menu1 did not complete in 60000 milliseconds
JWM: warning: $HOME/.jwmrc[15]: invalid include: exec:$HOME/.jwm/root_menus/menu1
JWM: warning: timeout: /root/.jwm/root_menus/menu2 did not complete in 60000 milliseconds
JWM: warning: $HOME/.jwmrc[16]: invalid include: exec:$HOME/.jwm/root_menus/menu2
JWM: warning: timeout: /root/.jwm/root_menus/menu3 did not complete in 60000 milliseconds
JWM: warning: $HOME/.jwmrc[17]: invalid include: exec:$HOME/.jwm/root_menus/menu3
JWM: warning: timeout: /root/.jwm/root_menus/menu4 did not complete in 60000 milliseconds
JWM: warning: $HOME/.jwmrc[18]: invalid include: exec:$HOME/.jwm/root_menus/menu4
JWM: warning: timeout: /root/.jwm/root_menus/menu5 did not complete in 60000 milliseconds
JWM: warning: $HOME/.jwmrc[19]: invalid include: exec:$HOME/.jwm/root_menus/menu5
JWM: warning: timeout: /usr/local/jwm_config/menu_build_bookmarks did not complete in 60000 milliseconds
JWM: warning: $HOME/.jwmrc[22]: invalid include: exec:/usr/local/jwm_config/menu_build_bookmarks
JWM: warning: timeout: /usr/local/jwm_config/menu_build_clock did not complete in 60000 milliseconds
JWM: warning: $HOME/.jwmrc[23]: invalid include: exec:/usr/local/jwm_config/menu_build_clock
JWM: warning: timeout: /usr/local/jwm_config/menu_build_places did not complete in 60000 milliseconds
JWM: warning: $HOME/.jwmrc[24]: invalid include: exec:/usr/local/jwm_config/menu_build_places
DEBUG: main.c[235]: debug mode stopped
DEBUG: MEMORY: no memory leaks

I'm typing this from the "bad" install right now, a pristine puppy, like I have been building for years on bare metal. (ryzen, amd graphics lots of RAM - usb install). ALL the menus work as expected!

Most (if not all) of these are dynamic. I noticed some trouble in previous jwm issues in this repo with dynamic menus.

As I can't cause a crash I can't give a backtrace. Been running near an hour now. 😕

Thanks for taking a look Joe.

[joewing]

I think what's happening is that JWM is killing subprocesses which the kernel is reporting as a segfault in JWM. As of 2.4, JWM gives subprocesses used for menu/config generation a timeout so they don't stall forever (it looks like JWM is just going through and killing a bunch of different menu generation programs).
So it could be that something is wrong with they way that works for jwm -p or that the menu generation scripts don't return in this case, what happens if you run the menu generation scripts directly?

[01micko]

Each program completes in a matter milliseconds, all with start and end <JWM>...</JWM> tags.

Only one shows a bad config (which can be easy fixed). You can see the scripts here.

/usr/local/jwm_config$ ./menu_build_none 
<?xml version="1.0"?>
<JWM></JWM>
/usr/local/jwm_config$ ./menu_build_places

<JWM>
	<Program label="Home" icon="/usr/local/lib/X11/pixmaps/home48.png">rox $HOME</Program>
	<Program label="System" icon="/usr/local/lib/X11/pixmaps/pc48.png">rox /</Program>
	<Program label="Network" icon="/usr/local/lib/X11/pixmaps/connect48.png">rox $HOME/network</Program>
<Program label="Downloads" icon="/usr/local/lib/X11/pixmaps/folder48.png">rox $HOME/Downloads</Program>
<Separator/>
	<Dynamic label="Drives" icon="/usr/local/lib/X11/pixmaps/drive48.png">exec:/usr/local/jwm_config/menu_build_places_drives</Dynamic>
	<Dynamic label="Recently used" icon="file.svg">exec:/usr/local/jwm_config/menu_build_recent_docs</Dynamic>
	<Separator/>
	<Program label="Web" icon="/usr/local/lib/X11/pixmaps/www48.png">defaultbrowser</Program>
	<Program label="Help" icon="/usr/local/lib/X11/pixmaps/help48.png">/usr/sbin/puppyhelp</Program>
</JWM>
/usr/local/jwm_config$ ./menu_build_places_drives 
<JWM>
grep: /root/.usb-drive-log-probedisk: No such file or directory
<Program label="nvme0n1p1   ext4   WDC WDS500G2B0C-00PXH0                  " icon="/usr/local/lib/X11/pixmaps/drive48.png">/tmp/pup_event_frontend/drive_nvme0n1p1/AppRun</Program>
grep: /root/.usb-drive-log-probedisk: No such file or directory
<Program label="sda1   vfat   ST1000DM010-2EP1" icon="/usr/local/lib/X11/pixmaps/drive48.png">/tmp/pup_event_frontend/drive_sda1/AppRun</Program>
grep: /root/.usb-drive-log-probedisk: No such file or directory
<Program label="sda2   ext4   ST1000DM010-2EP1" icon="/usr/local/lib/X11/pixmaps/drive48.png">/tmp/pup_event_frontend/drive_sda2/AppRun</Program>
grep: /root/.usb-drive-log-probedisk: No such file or directory
<Program label="sda3   ext4   ST1000DM010-2EP1" icon="/usr/local/lib/X11/pixmaps/drive48.png">/tmp/pup_event_frontend/drive_sda3/AppRun</Program>
grep: /root/.usb-drive-log-probedisk: No such file or directory
<Program label="sda4   ext4   ST1000DM010-2EP1" icon="/usr/local/lib/X11/pixmaps/drive48.png">/tmp/pup_event_frontend/drive_sda4/AppRun</Program>
grep: /root/.usb-drive-log-probedisk: No such file or directory
<Program label="sda5   ext4   ST1000DM010-2EP1" icon="/usr/local/lib/X11/pixmaps/drive48.png">/tmp/pup_event_frontend/drive_sda5/AppRun</Program>
grep: /root/.usb-drive-log-probedisk: No such file or directory
<Program label="sda6   ext4   ST1000DM010-2EP1" icon="/usr/local/lib/X11/pixmaps/drive48.png">/tmp/pup_event_frontend/drive_sda6/AppRun</Program>
grep: /root/.usb-drive-log-probedisk: No such file or directory
<Program label="sdc1   ntfs   Expansion Desk  " icon="/usr/local/lib/X11/pixmaps/drive48.png">/tmp/pup_event_frontend/drive_sdc1/AppRun</Program>
grep: /root/.usb-drive-log-probedisk: No such file or directory
<Program label="sdc2   ext4   Expansion Desk  " icon="/usr/local/lib/X11/pixmaps/drive48.png">/tmp/pup_event_frontend/drive_sdc2/AppRun</Program>
grep: /root/.usb-drive-log-probedisk: No such file or directory
</JWM>
/usr/local/jwm_config$ ./menu_build_bookmarks 
<?xml version="1.0"?>
<JWM>

<Program label="/mnt/home homedrive" icon="file.svg">exec rox "/mnt/home homedrive"</Program>
<Separator/>
<Program icon="bookmark.svg" label="Edit Bookmarks">exec /usr/local/jwm_config/menu -bookmarks</Program>
</JWM>
/usr/local/jwm_config$ ./menu_build_clock

<JWM>
	<Program label="Set time and date" icon="/usr/share/pixmaps/puppy/date_time.svg">set-time-for-puppy</Program>
	<Program label="Clock format" icon="/usr/share/pixmaps/puppy/clock.svg">/usr/local/jwm_config/tray -clock</Program>
	<Program label="Set timezone" icon="/usr/share/pixmaps/puppy/country_timezone.svg">timezone-set</Program>
	<Separator/>
<Program label="Calendar" icon="/usr/share/pixmaps/puppy/calendar.svg">minixcal</Program>
<Program label="Task sheduler" icon="/usr/share/pixmaps/puppy/clock_interval.svg">pschedule</Program>
</JWM>

[joewing]

Thanks for helping me narrow it down! I think I'm able to reproduce what you're seeing. I just pushed a commit that appears to fix it for me.
EDIT: that commit just breaks it in a different way... still working on a fix

[joewing]

Hopefully that last commit fixes it.

[01micko]

Great! I'll do a build and test. Thanks!

EDIT: had to disable nls because of syntax errors in pt_BR translation and my build system flat refused to ignore!

[01micko]

Well, thanks Joe! All fixed and as it happens, my configs are perfect 😄