[Bug] d.mon wx0 crashes after a few seconds (Manjaro, Arch Linux / AUR)

[neteler]

Describe the bug

When starting a GRASS GIS monitor, it crashes after a few seconds upon mouse-over, rendering an important part of GRASS GIS unusable.

Originally reported here:

• https://lists.osgeo.org/pipermail/grass-user/2021-May/082415.html
• https://lists.osgeo.org/pipermail/grass-user/2021-May/082416.html

GRASS 7.8.5 (nc_spm_08_grass7):~ > ps aux | grep python
manjaro     1248  0.1  1.2 566708 74276 ?        Sl   08:37   0:01 python3 /usr/bin/manjaro-hello
manjaro     1273  0.0  0.9 478728 60332 ?        Sl   08:37   0:00 /usr/bin/python /usr/bin/blueman-applet
manjaro     1474  0.0  0.7 391516 47076 ?        Sl   08:37   0:00 /usr/bin/python /usr/bin/blueman-tray
manjaro   116960  0.0  0.2  19788 15532 pts/0    Ss   08:59   0:00 python3 /usr/bin/grass78
manjaro   117045  1.0  0.0      0     0 pts/0    Z    09:01   0:02 [python3] <defunct>
manjaro   117586  2.9  2.4 796652 151940 pts/0   Sl   09:04   0:00 /opt/grass/gui/wxpython/mapdisp/main.py /opt/grass/gui/wxpython/mapdisp/main.py wx0 /home/manjaro/grassdata/nc_spm_08_grass7/user1/.tmp/manjaro/MONITORS/wx0 720 480 0
manjaro   117647  0.0  0.0   6404  2376 pts/0    S+   09:05   0:00 grep pytho
GRASS 7.8.5 (nc_spm_08_grass7):~ > strace -p 117586
strace: attach: ptrace(PTRACE_SEIZE, 117586): Operation not permitted
GRASS 7.8.5 (nc_spm_08_grass7):~ > sudo strace -p 117586
strace: Process 117586 attached
restart_syscall(<... resuming interrupted read ...>) = 0
newfstatat(AT_FDCWD, "/home/manjaro/grassdata/nc_spm_08_grass7/user1/.tmp/manjaro/MONITORS/wx0/cmd", {st_mode=S_IFREG|0644, st_size=0, ...}, 0) = 0
recvmsg(3, {msg_namelen=0}, 0)          = -1 EAGAIN (Resource temporarily unavailable)
recvmsg(3, {msg_namelen=0}, 0)          = -1 EAGAIN (Resource temporarily unavailable)
poll([{fd=3, events=POLLIN}, {fd=4, events=POLLIN}, {fd=8, events=POLLIN}], 3, 0) = 0 (Timeout)
recvmsg(3, {msg_namelen=0}, 0)          = -1 EAGAIN (Resource temporarily unavailable)
recvmsg(3, {msg_namelen=0}, 0)          = -1 EAGAIN (Resource temporarily unavailable)

...

recvmsg(3, {msg_namelen=0}, 0)          = -1 EAGAIN (Resource temporarily unavailable)
recvmsg(3, {msg_namelen=0}, 0)          = -1 EAGAIN (Resource temporarily unavailable)
poll([{fd=3, events=POLLIN}, {fd=4, events=POLLIN}, {fd=8, events=POLLIN}], 3, 0) = 0 (Timeout)
poll([{fd=3, events=POLLIN}, {fd=4, events=POLLIN}, {fd=8, events=POLLIN}], 3, 0) = 0 (Timeout)
poll([{fd=3, events=POLLIN|POLLOUT}], 1, -1) = 1 ([{fd=3, revents=POLLOUT}])
writev(3, [{iov_base="\204(\3\0\211\0\200\4\2\0\200\4", iov_len=12}, {iov_base=NULL, iov_len=0}, {iov_base="", iov_len=0}], 3) = 12
poll([{fd=3, events=POLLIN}], 1, -1)    = 1 ([{fd=3, revents=POLLIN}])
recvmsg(3, {msg_name=NULL, msg_namelen=0, msg_iov=[{iov_base="\1(\226\17\7\0\0\0002\5\0\0\0\0\0\0\0\0/\3\0\0\316\2\0\0\326\0\0\0\264\1"..., iov_len=4096}], msg_iovlen=1, msg_controllen=0, msg_flags=0}, 0) = 60
poll([{fd=3, events=POLLIN|POLLOUT}], 1, -1) = 1 ([{fd=3, revents=POLLOUT}])
writev(3, [{iov_base="((\4\0\211\0\200\0042\5\0\0\1\0*\0", iov_len=16}, {iov_base=NULL, iov_len=0}, {iov_base="", iov_len=0}], 3) = 16
poll([{fd=3, events=POLLIN}], 1, -1)    = 1 ([{fd=3, revents=POLLIN}])
recvmsg(3, {msg_name=NULL, msg_namelen=0, msg_iov=[{iov_base="\1\1\227\17\0\0\0\0\226$\205\0Z\2D\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0", iov_len=4096}], msg_iovlen=1, msg_controllen=0, msg_flags=0}, 0) = 32
poll([{fd=3, events=POLLIN|POLLOUT}], 1, -1) = 1 ([{fd=3, revents=POLLOUT}])
writev(3, [{iov_base="+(\1\0", iov_len=4}, {iov_base=NULL, iov_len=0}, {iov_base="", iov_len=0}], 3) = 4
poll([{fd=3, events=POLLIN}], 1, -1)    = 1 ([{fd=3, revents=POLLIN}])
recvmsg(3, {msg_name=NULL, msg_namelen=0, msg_iov=[{iov_base="\1\2\230\17\0\0\0\0\212\0\200\4\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0", iov_len=4096}], msg_iovlen=1, msg_controllen=0, msg_flags=0}, 0) = 32
poll([{fd=3, events=POLLIN|POLLOUT}], 1, -1) = 1 ([{fd=3, revents=POLLOUT}])
writev(3, [{iov_base="5\30\4\0\330\2\200\4\211\0\200\4\212\1\30\0\214\4\6\0\331\2\200\4\330\2\200\4*\0\0\0"..., iov_len=1016}, {iov_base=NULL, iov_len=0}, {iov_base="", iov_len=0}], 3) = 1016
recvmsg(3, {msg_namelen=0}, 0)          = -1 EAGAIN (Resource temporarily unavailable)
recvmsg(3, {msg_namelen=0}, 0)          = -1 EAGAIN (Resource temporarily unavailable)
--- SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_MAPERR, si_addr=0x10} ---
+++ killed by SIGSEGV (core dumped) +++

I tried to debug it but I am not familiar enough with gdb debugging of a child process. I updated initial steps here:

https://grasswiki.osgeo.org/wiki/GRASS_Debugging#Attaching_to_child_process

and would appreciate support.

At system level, this message is seen:

sudo dmesg  | grep python
[ 1451.246029] python3[117148]: segfault at 10 ip 00007f9fc4d4f54d sp 00007ffdd42a37a0 error 4 in libpython3.9.so.1.0[7f9fc4c9a000+21a000]

To Reproduce
Steps to reproduce the behavior:

(I installed Manjaro in Virtualbox for this test)

1. Download https://manjaro.org/download/
2. Register the ISO in Virtualbox, start
3. Boot into Manjaro
4. add AUR repo to Manjaro: https://linuxconfig.org/how-to-install-a-package-from-aur-on-manjaro-linux (then just install grass after update of the package database)
5. run GRASS GIS, start d.mon wx0
6. move the mouse over the monitor and it will close after a few seconds (the segfault message isn't shown)

Expected behavior
Should not crash...

System description (please complete the following information):

• Operating System: Manjaro / AUR
• GRASS GIS version

g.version -rge
version=7.8.5
date=2020
revision=4e3f393
build_date=2021-05-15
build_platform=x86_64-pc-linux-gnu
build_off_t_size=8
libgis_revision=2021-05-15T06:53:09+00:00
libgis_date=2021-05-15T10:00:00+02:00
proj=8.0.0
gdal=3.2.2
geos=3.9.1
sqlite=3.35.5

python3 -c "import sys, wx; print(sys.version); print(wx.version())"
3.9.4 (default, Apr 20 2021, 15:51:38) 
[GCC 10.2.0]
4.0.7.post2 gtk3 (phoenix) wxWidgets 3.0.5

[petrasovaa]

Isn't this the same as #1123?

[tmszi]

Isn't this the same as #1123?

I think so, the indicator is a combination of Python version 3.9.4 (default, Apr 20 2021, 15:51:38) and wxPython version 4.0.7.post2 gtk3 (phoenix) wxWidgets 3.0.5 if it is unpatched.

[neteler]

Ah, I had forgotten about this old report.

@benducke, it seems Arch/Manjaro are missing this upstream fix for wxPseudoDC.FindObjects crash:

wxWidgets/Phoenix#1849

(in Fedora, it had been backported in https://src.fedoraproject.org/rpms/python-wxpython4/c/f5471fb86aaae46a686b85c654fcbb98516355e6?branch=rawhide)

I'll close this report since it is not directly a GRASS GIS problem but has to be fixed in wxWidgets.
Please contact the maintainer of wxWidgets to patch it accordingly.