Re: [X2Go-Dev] Bug#1376: Arithmetic exception error reading variable: Division by zero git20190114.1758
- On 3/14/19 2:20 PM, Robert Kudyba wrote:
I'm not seeing the Keyboard/Configure keyboard settings in a Mac like I do on Windows. Is it hidden in some sub-menu?
It's not available on macOS at all, since macOS is very special. The keyboard settings are synchronized using an xmodmap timer, the X.Org rules are not touched at all. Were they, stuff would just blow up.
This is not a limitation of X2Go, but the way the X11/Quartz keyboard synchronization works with XQuartz for some reason. It's always set to pc104/us/null/null (or some variant of this), even if the actual layout is something entirely different.
Mihai
On Thu, Mar 14, 2019 at 2:26 PM Mihai Moldovan <ionic@ionic.de> wrote:
- On 3/14/19 2:20 PM, Robert Kudyba wrote:
I'm not seeing the Keyboard/Configure keyboard settings in a Mac like I do on Windows. Is it hidden in some sub-menu?
It's not available on macOS at all, since macOS is very special. The keyboard settings are synchronized using an xmodmap timer, the X.Org rules are not touched at all. Were they, stuff would just blow up.
This is not a limitation of X2Go, but the way the X11/Quartz keyboard synchronization works with XQuartz for some reason. It's always set to pc104/us/null/null (or some variant of this), even if the actual layout is something entirely different.
So chances are I introduced a problem here with this nxagent patch 522106e207e7230da5e62fb88f2e47742fdf1b21. What do you think, Mihai? Can you reproduce the crash?
Uli
- On 3/14/19 3:56 PM, Ulrich Sibiller wrote:
So chances are I introduced a problem here with this nxagent patch 522106e207e7230da5e62fb88f2e47742fdf1b21. What do you think, Mihai? Can you reproduce the crash?
Turns out that I cannot test this right now.
libssh doesn't let me connect to my test server in question because the ssh-agent is too old and doesn't support the newer signature algorithms libssh and the remote OpenSSH server agree on. Looks like I have to fix libssh first...
Mihai
For testing you could try without ssh-agent or with the one of the current openssh.
Uli
Mihai Moldovan <ionic@ionic.de> schrieb am Mo., 25. März 2019, 13:59:
- On 3/14/19 3:56 PM, Ulrich Sibiller wrote:
So chances are I introduced a problem here with this nxagent patch 522106e207e7230da5e62fb88f2e47742fdf1b21. What do you think, Mihai? Can you reproduce the crash?
Turns out that I cannot test this right now.
libssh doesn't let me connect to my test server in question because the ssh-agent is too old and doesn't support the newer signature algorithms libssh and the remote OpenSSH server agree on. Looks like I have to fix libssh first...
Mihai
x2go-dev mailing list x2go-dev@lists.x2go.org https://lists.x2go.org/listinfo/x2go-dev
- On 3/25/19 2:10 PM, Ulrich Sibiller wrote:
For testing you could try without ssh-agent or with the one of the current openssh.
Sure, that's the quickest workaround... I just switched to no-agent auth for testing.
Naturally, though, I wasn't able to reproduce the problem with the current master version. Everything was working fine as far as the keyboard was concerned.
It might be related to that commit, but I don't know. ctrl->num_groups is obviously zero in XkbAdjustGroup()/xkbUtils.c, but I have no idea what that actually means. The code in question does look a bit shady (for instance, it checks for group >= ctrls->num_groups and later might set group %= ctrls->num_groups, which obviously is a bad idea if both group and ctrls->num_groups is zero, but I have the suspicion that that should never have been the case in the first place.
I don't know what "groups" are in this context neither.
I somehow doubt that commit caused the problem, though. If the clone mechanism works fine, that change shouldn't cause any problems. There's probably no harm in cloning the config, even though we later sync up using xmodmap (other than a race condition, i.e., there MIGHT be a problem if we first sync up using xmodmap and then nxagent changes the whole keyboard settings again).
Seems to be yet another weird problem that can only be reproduced with user accounts from NIS, but I have no idea why that might be. As far as I remember, Roberts home directories are not even coming via NFS so the home dir should be available at the time nxagent/x2goagent starts and - in that case - tries to block keyboard file creation by making a directory. However, nx-libs-3.5.99.18-0.0build1.0.git20190208.3237.heuler.fc29.x86_64 is quite old and I recently synced the master branch up with Arctica Project's master branch again, which means that it has 3.5.99.19 + the last two commits on top of it. Maybe it's worthwhile to retest with that.
I'm wary of releasing 3.5.99.19 to X2Go land, though, since I did notice a different problem in my test of an XFCE session on a Devuan Unstable machine: for some reason I didn't have xfce4-terminal installed and the default terminal application was set to qterminal. That started up, but only worked for a few seconds, after which it froze up. x2goagent/nxagent used 100 % of one CPU thread at that point and even after killing qterminal it continued to do so for a few minutes, after which it settled down with occasional 100 % CPU spikes (but it does that without having ever started qterminal, too, so I won't worry about it. My machine's test CPU is a pretty bad one - Intel Atom 1.8 GHs).
Hopefully that's not a general problem with Qt5 applications?
Mihai
Perhaps these logs and core dump GDB backtrace will help? Notice that nx-libs is now nx-libs-3.5.99.19-0.0build1.0.git20190325.3293.heuler.fc29.x86_64
Mar 26 09:22:17 ourwokstation /usr/bin/x2gostartagent[16213]: successfully started X2Go Agent session with ID agw-58-1553606535_stDMATE_dp32 Mar 26 09:22:19 ourwokstation /usr/bin/x2goruncommand[16676]: launching session with Xsession-x2go mechanism, using STARTUP="mate-session" Mar 26 09:22:19 ourwokstation /usr/bin/x2goruncommand[16677]: dbus wrapper available as /usr/bin/dbus-run-session Mar 26 09:22:24 ourwokstation gnome-keyring-daemon[16915]: couldn't access control socket: /run/user/1201/keyring/control: No such file or directory Mar 26 09:22:24 ourwokstation mate-session[16692]: WARNING: keycode1 not existent Mar 26 09:22:24 ourwokstation mate-session[16692]: WARNING: keycode2 not existent Mar 26 09:22:30 ourwokstation mate-session[16692]: WARNING: Could not launch application 'dropbox.desktop': Unable to start application: Failed to execute child process “dropbox” (No such file or directory) Mar 26 09:22:32 ourwokstation gnome-keyring-daemon[16915]: The SSH agent was already initialized Mar 26 09:22:33 ourwokstation gnome-keyring-daemon[16915]: The PKCS#11 component was already initialized Mar 26 09:22:33 ourwokstation pulseaudio[30825]: Invalid MIT-MAGIC-COOKIE-1 keyE: [pulseaudio] x11wrap.c: XOpenDisplay() failed Mar 26 09:22:33 ourwokstation pulseaudio[30825]: E: [pulseaudio] module.c: Failed to load module "module-x11-publish" (argument: "display=:58"): initialization failed. Mar 26 09:22:33 ourwokstation gnome-keyring-daemon[16915]: The Secret Service was already initialized Mar 26 09:22:34 ourwokstation kernel: traps: x2goagent[16202] trap divide error ip:55c23629da1c sp:7ffe0df096e0 error:0 in nxagent[55c2361af000+319000] Mar 26 09:22:34 ourwokstation systemd[1]: Started Process Core Dump (PID 17060/UID 0). Mar 26 09:22:36 ourwokstation spice-vdagent[17111]: Cannot access vdagent virtio channel /dev/virtio-ports/com.redhat.spice.0 Mar 26 09:22:38 ourwokstation systemd-coredump[17061]: Process 16202 (x2goagent) of user 1201 dumped core. #012#012Stack trace of thread 16202:#012#0 0x000055c23629da1c XkbAdjustGroup (nxagent) #012#1 0x000055c23629daf5 XkbComputeDerivedState (nxagent) #012#2 0x000055c236293ca6 ProcXkbLatchLockState (nxagent) #012#3 0x000055c2361cfd18 Dispatch (nxagent) #012#4 0x000055c2361b3821 main (nxagent) #012#5 0x00007f64fdece413 __libc_start_main (libc.so.6)#012#6 0x000055c2361b3b5e _start (nxagent) Mar 26 09:22:40 ourwokstation python3[17121]: detected unhandled Python exception in '/usr/bin/dnfdragora-updater' Mar 26 09:22:41 ourwokstation systemd-logind[11665]: Session 25456 logged out. Waiting for processes to exit. Mar 26 09:22:45 ourwokstation abrt-server[17301]: Package 'nxagent' isn't signed with proper key Mar 26 09:22:46 ourwokstation abrt-server[17301]: 'post-create' on '/var/spool/abrt/ccpp-2019-03-26-09:22:39.879918-16202' exited with 1 Mar 26 09:22:46 ourwokstation abrt-server[17301]: Deleting problem directory '/var/spool/abrt/ccpp-2019-03-26-09:22:39.879918-16202' Mar 26 09:22:46 ourwokstation abrt-server[17234]: Deleting problem directory Python3-2019-03-26-09:22:41-17121 (dup of Python3-2019-02-18-12:04:02-13003) Mar 26 09:22:47 ourwokstation abrt-notification[17366]: Process 13003 (dnfdragora-updater) of user 1201 encountered an uncaught gi.repository.GLib.GError exception
Core was generated by `x2goagent -nolisten tcp -nolisten tcp -dpi 96 -D -auth /u/ourwokstation/agw/.Xauthori'. Program terminated with signal SIGFPE, Arithmetic exception. #0 0x000055c23629da1c in XkbAdjustGroup (group=<error reading variable: Division by zero>, ctrls=ctrls@entry=0x55c236e90810) at xkbUtils.c:688 688 xkbUtils.c: No such file or directory. (gdb) bt full #0 0x000055c23629da1c in XkbAdjustGroup (group=<error reading variable: Division by zero>, ctrls=ctrls@entry=0x55c236e90810) at xkbUtils.c:688 act = 0 #1 0x000055c23629daf5 in XkbComputeDerivedState (xkbi=0x55c236ea5c10) at xkbUtils.c:714 state = 0x55c236ea5c22 ctrls = 0x55c236e90810 grp = <optimized out> #2 0x000055c236293ca6 in ProcXkbLatchLockState (client=0x55c2372372d0) at xkb.c:560 status = <optimized out> dev = 0x55c236e6f6b0 oldState = {group = 0 '\000', locked_group = 0 '\000', base_group = 0, latched_group = 0, mods = 0 '\000', base_mods = 0 '\000', latched_mods = 0 '\000', locked_mods = 0 '\000', compat_state = 0 '\000', grab_mods = 0 '\000', compat_grab_mods = 0 '\000', lookup_mods = 0 '\000', compat_lookup_mods = 0 '\000', ptr_buttons = 0} newState = 0x55c236ea5c22 changed = <optimized out> stuff = <optimized out> #3 0x000055c2361cfd18 in Dispatch () at NXdispatch.c:482 clientReady = 0x55c236e8b060 result = <optimized out> client = 0x55c2372372d0 nready = <optimized out> icheck = 0x55c2365fe6d0 <checkForInput> start_tick = 0 currentDispatch = <optimized out> #4 0x000055c2361b3821 in main (argc=15, argv=0x7ffe0df09938, envp=<optimized out>) at main.c:353 i = <optimized out> xauthfile = <optimized out> alwaysCheckForInput = {0, 1}
Core dump uploaded to: https://storm.cis.fordham.edu/~rkudyba/core.x2goagent.1201.e138540091ef4dfe9...
rpm -q nx-libs nx-libs-3.5.99.19-0.0build1.0.git20190325.3293.heuler.fc29.x86_64
rpm -q x2goserver x2goserver-4.1.0.4-0.0x2go1.0.git20190114.1758.heuler.fc29.x86_64
On Mar 26, 2019, at 5:13 AM, Mihai Moldovan <ionic@ionic.de> wrote:
- On 3/25/19 2:10 PM, Ulrich Sibiller wrote:
For testing you could try without ssh-agent or with the one of the current openssh.
Sure, that's the quickest workaround... I just switched to no-agent auth for testing.
Naturally, though, I wasn't able to reproduce the problem with the current master version. Everything was working fine as far as the keyboard was concerned.
It might be related to that commit, but I don't know. ctrl->num_groups is obviously zero in XkbAdjustGroup()/xkbUtils.c, but I have no idea what that actually means. The code in question does look a bit shady (for instance, it checks for group >= ctrls->num_groups and later might set group %= ctrls->num_groups, which obviously is a bad idea if both group and ctrls->num_groups is zero, but I have the suspicion that that should never have been the case in the first place.
I don't know what "groups" are in this context neither.
I somehow doubt that commit caused the problem, though. If the clone mechanism works fine, that change shouldn't cause any problems. There's probably no harm in cloning the config, even though we later sync up using xmodmap (other than a race condition, i.e., there MIGHT be a problem if we first sync up using xmodmap and then nxagent changes the whole keyboard settings again).
Seems to be yet another weird problem that can only be reproduced with user accounts from NIS, but I have no idea why that might be. As far as I remember, Roberts home directories are not even coming via NFS so the home dir should be available at the time nxagent/x2goagent starts and - in that case - tries to block keyboard file creation by making a directory. However, nx-libs-3.5.99.18-0.0build1.0.git20190208.3237.heuler.fc29.x86_64 is quite old and I recently synced the master branch up with Arctica Project's master branch again, which means that it has 3.5.99.19 + the last two commits on top of it. Maybe it's worthwhile to retest with that.
I'm wary of releasing 3.5.99.19 to X2Go land, though, since I did notice a different problem in my test of an XFCE session on a Devuan Unstable machine: for some reason I didn't have xfce4-terminal installed and the default terminal application was set to qterminal. That started up, but only worked for a few seconds, after which it froze up. x2goagent/nxagent used 100 % of one CPU thread at that point and even after killing qterminal it continued to do so for a few minutes, after which it settled down with occasional 100 % CPU spikes (but it does that without having ever started qterminal, too, so I won't worry about it. My machine's test CPU is a pretty bad one - Intel Atom 1.8 GHs).
Hopefully that's not a general problem with Qt5 applications?
Mihai
#2 0x000055c236293ca6 in ProcXkbLatchLockState (client=0x55c2372372d0) at xkb.c:560
-> Proc* is always a call from outside to an extension, in this case the XKB extension.
So do you have some setxkbmap call in your shell startup files? What are they doing? Does this also happen with other session types (not MATE)?
Uli
On Tue, Mar 26, 2019 at 2:48 PM Robert Kudyba <rkudyba@fordham.edu> wrote:
Perhaps these logs and core dump GDB backtrace will help? Notice that nx-libs is now nx-libs-3.5.99.19-0.0build1.0.git20190325.3293.heuler.fc29.x86_64
Mar 26 09:22:17 ourwokstation /usr/bin/x2gostartagent[16213]: successfully started X2Go Agent session with ID agw-58-1553606535_stDMATE_dp32 Mar 26 09:22:19 ourwokstation /usr/bin/x2goruncommand[16676]: launching session with Xsession-x2go mechanism, using STARTUP="mate-session" Mar 26 09:22:19 ourwokstation /usr/bin/x2goruncommand[16677]: dbus wrapper available as /usr/bin/dbus-run-session Mar 26 09:22:24 ourwokstation gnome-keyring-daemon[16915]: couldn't access control socket: /run/user/1201/keyring/control: No such file or directory Mar 26 09:22:24 ourwokstation mate-session[16692]: WARNING: keycode1 not existent Mar 26 09:22:24 ourwokstation mate-session[16692]: WARNING: keycode2 not existent Mar 26 09:22:30 ourwokstation mate-session[16692]: WARNING: Could not launch application 'dropbox.desktop': Unable to start application: Failed to execute child process “dropbox” (No such file or directory) Mar 26 09:22:32 ourwokstation gnome-keyring-daemon[16915]: The SSH agent was already initialized Mar 26 09:22:33 ourwokstation gnome-keyring-daemon[16915]: The PKCS#11 component was already initialized Mar 26 09:22:33 ourwokstation pulseaudio[30825]: Invalid MIT-MAGIC-COOKIE-1 keyE: [pulseaudio] x11wrap.c: XOpenDisplay() failed Mar 26 09:22:33 ourwokstation pulseaudio[30825]: E: [pulseaudio] module.c: Failed to load module "module-x11-publish" (argument: "display=:58"): initialization failed. Mar 26 09:22:33 ourwokstation gnome-keyring-daemon[16915]: The Secret Service was already initialized Mar 26 09:22:34 ourwokstation kernel: traps: x2goagent[16202] trap divide error ip:55c23629da1c sp:7ffe0df096e0 error:0 in nxagent[55c2361af000+319000] Mar 26 09:22:34 ourwokstation systemd[1]: Started Process Core Dump (PID 17060/UID 0). Mar 26 09:22:36 ourwokstation spice-vdagent[17111]: Cannot access vdagent virtio channel /dev/virtio-ports/com.redhat.spice.0 Mar 26 09:22:38 ourwokstation systemd-coredump[17061]: Process 16202 (x2goagent) of user 1201 dumped core. #012#012Stack trace of thread 16202:#012#0 0x000055c23629da1c XkbAdjustGroup (nxagent) #012#1 0x000055c23629daf5 XkbComputeDerivedState (nxagent) #012#2 0x000055c236293ca6 ProcXkbLatchLockState (nxagent) #012#3 0x000055c2361cfd18 Dispatch (nxagent) #012#4 0x000055c2361b3821 main (nxagent) #012#5 0x00007f64fdece413 __libc_start_main (libc.so.6)#012#6 0x000055c2361b3b5e _start (nxagent) Mar 26 09:22:40 ourwokstation python3[17121]: detected unhandled Python exception in '/usr/bin/dnfdragora-updater' Mar 26 09:22:41 ourwokstation systemd-logind[11665]: Session 25456 logged out. Waiting for processes to exit. Mar 26 09:22:45 ourwokstation abrt-server[17301]: Package 'nxagent' isn't signed with proper key Mar 26 09:22:46 ourwokstation abrt-server[17301]: 'post-create' on '/var/spool/abrt/ccpp-2019-03-26-09:22:39.879918-16202' exited with 1 Mar 26 09:22:46 ourwokstation abrt-server[17301]: Deleting problem directory '/var/spool/abrt/ccpp-2019-03-26-09:22:39.879918-16202' Mar 26 09:22:46 ourwokstation abrt-server[17234]: Deleting problem directory Python3-2019-03-26-09:22:41-17121 (dup of Python3-2019-02-18-12:04:02-13003) Mar 26 09:22:47 ourwokstation abrt-notification[17366]: Process 13003 (dnfdragora-updater) of user 1201 encountered an uncaught gi.repository.GLib.GError exception
Core was generated by `x2goagent -nolisten tcp -nolisten tcp -dpi 96 -D -auth /u/ourwokstation/agw/.Xauthori'. Program terminated with signal SIGFPE, Arithmetic exception. #0 0x000055c23629da1c in XkbAdjustGroup (group=<error reading variable: Division by zero>, ctrls=ctrls@entry=0x55c236e90810) at xkbUtils.c:688 688 xkbUtils.c: No such file or directory. (gdb) bt full #0 0x000055c23629da1c in XkbAdjustGroup (group=<error reading variable: Division by zero>, ctrls=ctrls@entry=0x55c236e90810) at xkbUtils.c:688 act = 0 #1 0x000055c23629daf5 in XkbComputeDerivedState (xkbi=0x55c236ea5c10) at xkbUtils.c:714 state = 0x55c236ea5c22 ctrls = 0x55c236e90810 grp = <optimized out> #2 0x000055c236293ca6 in ProcXkbLatchLockState (client=0x55c2372372d0) at xkb.c:560 status = <optimized out> dev = 0x55c236e6f6b0 oldState = {group = 0 '\000', locked_group = 0 '\000', base_group = 0, latched_group = 0, mods = 0 '\000', base_mods = 0 '\000', latched_mods = 0 '\000', locked_mods = 0 '\000', compat_state = 0 '\000', grab_mods = 0 '\000', compat_grab_mods = 0 '\000', lookup_mods = 0 '\000', compat_lookup_mods = 0 '\000', ptr_buttons = 0} newState = 0x55c236ea5c22 changed = <optimized out> stuff = <optimized out> #3 0x000055c2361cfd18 in Dispatch () at NXdispatch.c:482 clientReady = 0x55c236e8b060 result = <optimized out> client = 0x55c2372372d0 nready = <optimized out> icheck = 0x55c2365fe6d0 <checkForInput> start_tick = 0 currentDispatch = <optimized out> #4 0x000055c2361b3821 in main (argc=15, argv=0x7ffe0df09938, envp=<optimized out>) at main.c:353 i = <optimized out> xauthfile = <optimized out> alwaysCheckForInput = {0, 1}
Core dump uploaded to: https://storm.cis.fordham.edu/~rkudyba/core.x2goagent.1201.e138540091ef4dfe9...
rpm -q nx-libs nx-libs-3.5.99.19-0.0build1.0.git20190325.3293.heuler.fc29.x86_64
rpm -q x2goserver x2goserver-4.1.0.4-0.0x2go1.0.git20190114.1758.heuler.fc29.x86_64
On Mar 26, 2019, at 5:13 AM, Mihai Moldovan <ionic@ionic.de> wrote:
- On 3/25/19 2:10 PM, Ulrich Sibiller wrote:
For testing you could try without ssh-agent or with the one of the current openssh.
Sure, that's the quickest workaround... I just switched to no-agent auth for testing.
Naturally, though, I wasn't able to reproduce the problem with the current master version. Everything was working fine as far as the keyboard was concerned.
It might be related to that commit, but I don't know. ctrl->num_groups is obviously zero in XkbAdjustGroup()/xkbUtils.c, but I have no idea what that actually means. The code in question does look a bit shady (for instance, it checks for group >= ctrls->num_groups and later might set group %= ctrls->num_groups, which obviously is a bad idea if both group and ctrls->num_groups is zero, but I have the suspicion that that should never have been the case in the first place.
I don't know what "groups" are in this context neither.
I somehow doubt that commit caused the problem, though. If the clone mechanism works fine, that change shouldn't cause any problems. There's probably no harm in cloning the config, even though we later sync up using xmodmap (other than a race condition, i.e., there MIGHT be a problem if we first sync up using xmodmap and then nxagent changes the whole keyboard settings again).
Seems to be yet another weird problem that can only be reproduced with user accounts from NIS, but I have no idea why that might be. As far as I remember, Roberts home directories are not even coming via NFS so the home dir should be available at the time nxagent/x2goagent starts and - in that case - tries to block keyboard file creation by making a directory. However, nx-libs-3.5.99.18-0.0build1.0.git20190208.3237.heuler.fc29.x86_64 is quite old and I recently synced the master branch up with Arctica Project's master branch again, which means that it has 3.5.99.19 + the last two commits on top of it. Maybe it's worthwhile to retest with that.
I'm wary of releasing 3.5.99.19 to X2Go land, though, since I did notice a different problem in my test of an XFCE session on a Devuan Unstable machine: for some reason I didn't have xfce4-terminal installed and the default terminal application was set to qterminal. That started up, but only worked for a few seconds, after which it froze up. x2goagent/nxagent used 100 % of one CPU thread at that point and even after killing qterminal it continued to do so for a few minutes, after which it settled down with occasional 100 % CPU spikes (but it does that without having ever started qterminal, too, so I won't worry about it. My machine's test CPU is a pretty bad one - Intel Atom 1.8 GHs).
Hopefully that's not a general problem with Qt5 applications?
Mihai
On Tue, Mar 26, 2019 at 5:17 PM Ulrich Sibiller <uli42@gmx.de> wrote:
#2 0x000055c236293ca6 in ProcXkbLatchLockState (client=0x55c2372372d0) at xkb.c:560
-> Proc* is always a call from outside to an extension, in this case the XKB extension.
So do you have some setxkbmap call in your shell startup files? What are they doing? Does this also happen with other session types (not MATE)?
After reading http://soc.if.usp.br/manual/x11proto-kb-dev/xkbproto.html#Querying_and_Chang... and http://soc.if.usp.br/manual/x11proto-kb-dev/xkbproto.html#Locking_and_Latchi... I think this might be connected to one of the following: a) you are using a special keyboard (hardware) b) you are using a special keyboard configuration c) you have configured a special keyboard configuration somewhere in Mate d) you have configured caps lock/numlock being active on Mate startup (I don't know if this is configurable in Mate) e) you have caps lock/numlock activated on the x2go client side
Uli
On Mar 26, 2019, at 12:42 PM, Ulrich Sibiller <uli42@gmx.de> wrote:
On Tue, Mar 26, 2019 at 5:17 PM Ulrich Sibiller <uli42@gmx.de> wrote:
#2 0x000055c236293ca6 in ProcXkbLatchLockState (client=0x55c2372372d0) at xkb.c:560
-> Proc* is always a call from outside to an extension, in this case the XKB extension.
So do you have some setxkbmap call in your shell startup files? What are they doing? Does this also happen with other session types (not MATE)?
After reading https://urldefense.proofpoint.com/v2/url?u=http-3A__soc.if.usp.br_manual_x11proto-2Dkb-2Ddev_xkbproto.html-23Querying-5Fand-5FChanging-5FKeyboard-5FState&d=DwIBaQ&c=aqMfXOEvEJQh2iQMCb7Wy8l0sPnURkcqADc2guUW8IM&r=X0jL9y0sL4r4iU_qVtR3lLNo4tOL1ry_m7-psV3GejY&m=dLVKSi9prZLgr0ANz67KRW6KNG4wuMcyP9P3OoXQHso&s=qw-54H_gvD2qcXsBfZH1YCoVGQc5FdqkKLGR82JyjM4&e= and https://urldefense.proofpoint.com/v2/url?u=http-3A__soc.if.usp.br_manual_x11proto-2Dkb-2Ddev_xkbproto.html-23Locking-5Fand-5FLatching-5FModifiers-5Fand-5FGroups&d=DwIBaQ&c=aqMfXOEvEJQh2iQMCb7Wy8l0sPnURkcqADc2guUW8IM&r=X0jL9y0sL4r4iU_qVtR3lLNo4tOL1ry_m7-psV3GejY&m=dLVKSi9prZLgr0ANz67KRW6KNG4wuMcyP9P3OoXQHso&s=JsyWI_DEOnTQ_5h5JPocIU3sKsNaHYRaDn1-DUjfIDQ&e= I think this might be connected to one of the following: a) you are using a special keyboard (hardware) b) you are using a special keyboard configuration c) you have configured a special keyboard configuration somewhere in Mate d) you have configured caps lock/numlock being active on Mate startup (I don't know if this is configurable in Mate) e) you have caps lock/numlock activated on the x2go client side
Responses from our user:
a) you are using a special keyboard (hardware) No.
b) you are using a special keyboard configuration No.
c) you have configured a special keyboard configuration somewhere in Mate I don't recall doing so. The only thing I did is allow the Hebrew keyboard (selectable in the menu bar), but I had used a Hebrew keyboard prior to this problem.
d) you have configured caps lock/numlock being active on Mate startup (I don't know if this is configurable in Mate) I haven't configured either of these for Mate. In fact, I'd like to configure all keyboards to not use caps lock. If you know a way to do this (so that caps lock is inactive for me, whether I'm on the console or elsewhere), please let me know. Caps lock is an annoyance at best.
e) you have caps lock/numlock activated on the x2go client side I have configured caps lock to be inactive on my Mac. My Mac keyboard doesn't have a num lock key.
On Tue, Mar 26, 2019 at 7:51 PM Robert Kudyba <rkudyba@fordham.edu> wrote:
c) you have configured a special keyboard configuration somewhere in Mate I don't recall doing so. The only thing I did is allow the Hebrew keyboard (selectable in the menu bar), but I had used a Hebrew keyboard prior to this problem.
Ok, this sounds relevant. Can you please describe how this is set up? and can you try if it works without that?
d) you have configured caps lock/numlock being active on Mate startup (I don't know if this is configurable in Mate) I haven't configured either of these for Mate. In fact, I'd like to configure all keyboards to not use caps lock. If you know a way to do this (so that caps lock is inactive for me, whether I'm on the console or elsewhere), please let me know. Caps lock is an annoyance at best.
does this help? https://github.com/mate-desktop/mate-control-center/issues/29
e) you have caps lock/numlock activated on the x2go client side I have configured caps lock to be inactive on my Mac. My Mac keyboard doesn't have a num lock key.
Ok, please also describe where you set this and retry with that setting restored to the default.
Thx,
Uli
On Tue, Mar 26, 2019 at 7:51 PM Robert Kudyba <rkudyba@fordham.edu> wrote:
c) you have configured a special keyboard configuration somewhere in Mate I don't recall doing so. The only thing I did is allow the Hebrew keyboard (selectable in the menu bar), but I had used a Hebrew keyboard prior to this problem.
Ok, this sounds relevant. Can you please describe how this is set up? and can you try if it works without that?
In the very top right of the screen there is a pull down menu, and Hebrew is available and unchecked.
d) you have configured caps lock/numlock being active on Mate startup (I don't know if this is configurable in Mate) I haven't configured either of these for Mate. In fact, I'd like to configure all keyboards to not use caps lock. If you know a way to do this (so that caps lock is inactive for me, whether I'm on the console or elsewhere), please let me know. Caps lock is an annoyance at best.
does this help? https://github.com/mate-desktop/mate-control-center/issues/29
e) you have caps lock/numlock activated on the x2go client side I have configured caps lock to be inactive on my Mac. My Mac keyboard doesn't have a num lock key.
Ok, please also describe where you set this and retry with that setting restored to the default.
The default setting is still there, it says Disabled. The screen shot from github link is older and says Default. It appears to now say Disabled. Just to restate, "Caps Lock is Disabled" is NOT selected. And when we played with setting it, that option didn't make any difference, i.e., using Tweaks, when we chose "Caps Lock is Disabled", caps lock still worked. The other radio button options at least did make a difference.
But it appears at least XFCE works and does not crash.
Here's an update. MATE still crashes. rpm -q x2goserver x2goserver-4.1.0.4-0.0x2go1.0.git20190329.1759.heuler.fc29.x86_64 May 15 09:07:05 ourserver /usr/bin/x2gostartagent[11784]: successfully started X2Go Agent session with ID ouruser-67-1557925622_stDMATE_dp32 May 15 09:07:06 ourserver /usr/bin/x2goruncommand[12242]: launching session with Xsession-x2go mechanism, using STARTUP="mate-session" May 15 09:07:06 ourserver /usr/bin/x2goruncommand[12243]: dbus wrapper available as /usr/bin/dbus-run-session May 15 09:07:12 ourserver gnome-keyring-daemon[12586]: couldn't access control socket: /run/user/1201/keyring/control: No such file or directory May 15 09:07:12 ourserver mate-session[12246]: WARNING: keycode1 not existent May 15 09:07:12 ourserver mate-session[12246]: WARNING: keycode2 not existent May 15 09:07:16 ourserver systemd[1561]: Starting Sound Service... May 15 09:07:16 ourserver dbus-daemon[908]: [system] Activating via systemd: service name='org.freedesktop.hostname1' unit='dbus-org.freedesktop.hostname1.service' requested by ':1.4752' (uid=1201 pid=12592 comm="/usr/libexec/mate-settings-daemon ") May 15 09:07:16 ourserver systemd[1]: Starting Hostname Service... May 15 09:07:17 ourserver dbus-daemon[908]: [system] Successfully activated service 'org.freedesktop.hostname1' May 15 09:07:17 ourserver systemd[1]: Started Hostname Service. May 15 09:07:19 ourserver kernel: traps: x2goagent[11774] trap divide error ip:55c9cc08ea1c sp:7ffe23323ec0 error:0 in nxagent[55c9cbfa0000+319000] May 15 09:07:19 ourserver systemd[1]: Created slice system-systemd\x2dcoredump.slice. May 15 09:07:19 ourserver systemd[1]: Started Process Core Dump (PID 12633/UID 0). May 15 09:07:21 ourserver systemd-logind[1485]: Session 162 logged out. Waiting for processes to exit. May 15 09:07:22 ourserver systemd-coredump[12634]: Process 11774 (x2goagent) of user 1201 dumped core. Stack trace of thread 11774: #0 0x000055c9cc08ea1c XkbAdjustGroup (nxagent) #1 0x000055c9cc08eaf5 XkbComputeDerivedState (nxagent) #2 0x000055c9cc084ca6 ProcXkbLatchLockState (nxagent) #3 0x000055c9cbfc0d18 Dispatch (nxagent) #4 0x000055c9cbfa4821 main (nxagent) #5 0x00007f8cd60cd413 __libc_start_main (libc.so.6) #6 0x000055c9cbfa4b5e _start (nxagent) May 15 09:07:25 ourserver abrt-server[12868]: Package 'nxagent' isn't signed with proper key May 15 09:07:25 ourserver abrt-server[12868]: 'post-create' on '/var/spool/abrt/ccpp-2019-05-15-09:07:23.219326-11774' exited with 1 May 15 09:07:25 ourserver abrt-server[12868]: Deleting problem directory '/var/spool/abrt/ccpp-2019-05-15-09:07:23.219326-11774' May 15 09:07:25 ourserver dbus-daemon[908]: [system] Rejected send message, 2 matched rules; type="method_call", sender=":1.4792" (uid=1201 pid=12613 comm="/usr/bin/pulseaudio --daemonize=no ") interface="org.freedesktop.DBus.ObjectManager" member="GetManagedObjects" error name="(unset)" requested_reply="0" destination="org.bluez" (bus) May 15 09:07:26 ourserver systemd[1561]: Started Sound Service. May 15 09:07:26 ourserver pulseaudio[12613]: E: [pulseaudio] bluez5-util.c: GetManagedObjects() failed: org.freedesktop.DBus.Error.AccessDenied: Rejected send message, 2 matched rules; type="method_call", sender=":1.4792" (uid=1201 pid=12613 comm="/usr/bin/pulseaudio --daemonize=no ") interface="org.freedesktop.DBus.ObjectManager" member="GetManagedObjects" error name="(unset)" requested_reply="0" destination="org.bluez" (bus) May 15 09:07:32 ourserver systemd-logind[1485]: Removed session 162. May 15 09:09:12 ourserver systemd-logind[1485]: New session 163 of user ouruser. May 15 09:09:12 ourserver systemd[1]: Started Session 163 of user ouruser. May 15 09:09:21 ourserver /usr/bin/x2gostartagent[13428]: successfully started X2Go Agent session with ID ouruser-78-1557925759_stDKDE_dp32 May 15 09:09:22 ourserver /usr/bin/x2goruncommand[13822]: ERROR: command failed to execute May 15 09:09:30 ourserver systemd-logind[1485]: Session 163 logged out. Waiting for processes to exit. May 15 09:09:30 ourserver systemd-logind[1485]: Removed session 163. May 15 09:10:13 ourserver systemd[1]: Starting system activity accounting tool... May 15 09:10:13 ourserver systemd[1]: Started system activity accounting tool. May 15 09:10:13 ourserver systemd-logind[1485]: New session 164 of user ouruser. May 15 09:10:13 ourserver systemd[1]: Started Session 164 of user ouruser. May 15 09:10:13 ourserver automount[1483]: add_host_addrs: hostname lookup for .cargo failed: No address associated with hostname May 15 09:15:04 ourserver dbus-daemon[908]: [system] Activating via systemd: service name='net.reactivated.Fprint' unit='fprintd.service' requested by ':1.4832' (uid=0 pid=14472 comm="su - ") May 15 09:15:04 ourserver systemd[1]: Starting Fingerprint Authentication Daemon...
gdb from the coredump: Core was generated by `x2goagent -nolisten tcp -nolisten tcp -dpi 96 -D -auth /ouruser/.Xauthori'. Program terminated with signal SIGFPE, Arithmetic exception. #0 0x000055c9cc08ea1c in XkbAdjustGroup (group=<error reading variable: Division by zero>, ctrls=ctrls@entry=0x55c9cc8d1820) at xkbUtils.c:688 688 xkbUtils.c: No such file or directory. (gdb) bt full #0 0x000055c9cc08ea1c in XkbAdjustGroup (group=<error reading variable: Division by zero>, ctrls=ctrls@entry=0x55c9cc8d1820) at xkbUtils.c:688 act = 0 #1 0x000055c9cc08eaf5 in XkbComputeDerivedState (xkbi=0x55c9cc8e6c20) at xkbUtils.c:714 state = 0x55c9cc8e6c32 ctrls = 0x55c9cc8d1820 grp = <optimized out> #2 0x000055c9cc084ca6 in ProcXkbLatchLockState (client=0x55c9ccc78500) at xkb.c:560 status = <optimized out> dev = 0x55c9cc8b06c0 oldState = {group = 0 '\000', locked_group = 0 '\000', base_group = 0, latched_group = 0, mods = 0 '\000', base_mods = 0 '\000', latched_mods = 0 '\000', locked_mods = 0 '\000', compat_state = 0 '\000', grab_mods = 0 '\000', compat_grab_mods = 0 '\000', lookup_mods = 0 '\000', compat_lookup_mods = 0 '\000', ptr_buttons = 0} newState = 0x55c9cc8e6c32 changed = <optimized out> stuff = <optimized out> #3 0x000055c9cbfc0d18 in Dispatch () at NXdispatch.c:482 clientReady = 0x55c9cc8cc070 result = <optimized out> client = 0x55c9ccc78500 nready = <optimized out> icheck = 0x55c9cc3ef6d0 <checkForInput> start_tick = 0 currentDispatch = <optimized out> #4 0x000055c9cbfa4821 in main (argc=15, argv=0x7ffe23324118, envp=<optimized out>) at main.c:353 i = <optimized out> xauthfile = <optimized out> alwaysCheckForInput = {0, 1}
On Wed, Mar 27, 2019 at 10:10 AM Robert Kudyba <rkudyba@fordham.edu> wrote:
On Tue, Mar 26, 2019 at 7:51 PM Robert Kudyba <rkudyba@fordham.edu> wrote:
c) you have configured a special keyboard configuration somewhere in Mate I don't recall doing so. The only thing I did is allow the Hebrew keyboard (selectable in the menu bar), but I had used a Hebrew keyboard prior to this problem.
Ok, this sounds relevant. Can you please describe how this is set up? and can you try if it works without that?
In the very top right of the screen there is a pull down menu, and Hebrew is available and unchecked.
d) you have configured caps lock/numlock being active on Mate startup (I don't know if this is configurable in Mate) I haven't configured either of these for Mate. In fact, I'd like to configure all keyboards to not use caps lock. If you know a way to do this (so that caps lock is inactive for me, whether I'm on the console or elsewhere), please let me know. Caps lock is an annoyance at best.
does this help? https://github.com/mate-desktop/mate-control-center/issues/29
e) you have caps lock/numlock activated on the x2go client side I have configured caps lock to be inactive on my Mac. My Mac keyboard doesn't have a num lock key.
Ok, please also describe where you set this and retry with that setting restored to the default.
The default setting is still there, it says Disabled. The screen shot from github link is older and says Default. It appears to now say Disabled. Just to restate, "Caps Lock is Disabled" is NOT selected. And when we played with setting it, that option didn't make any difference, i.e., using Tweaks, when we chose "Caps Lock is Disabled", caps lock still worked. The other radio button options at least did make a difference.
But it appears at least XFCE works and does not crash.
On Wed, May 15, 2019 at 4:24 PM Robert Kudyba <rkudyba@fordham.edu> wrote:
Here's an update. MATE still crashes. rpm -q x2goserver
This is not astonishing as I still have not patched anything. I wanted to understand the problem first. But as I do not have access to (or knowledge about) MacOS I cannot test this.
I will try to implement a check that will prevent the crashm but I am not sure if your keyboard will work afterwards. Are you able to compile a new nxagent once I have patch available?
Uli
Are you able to compile a new nxagent once I have patch available?
Sure, just need to know the process.
On Wed, May 15, 2019 at 10:32 AM Ulrich Sibiller <uli42@gmx.de> wrote:
On Wed, May 15, 2019 at 4:24 PM Robert Kudyba <rkudyba@fordham.edu> wrote:
Here's an update. MATE still crashes. rpm -q x2goserver
This is not astonishing as I still have not patched anything. I wanted to understand the problem first. But as I do not have access to (or knowledge about) MacOS I cannot test this.
I will try to implement a check that will prevent the crashm but I am not sure if your keyboard will work afterwards. Are you able to compile a new nxagent once I have patch available?
Uli
see https://github.com/ArcticaProject/nx-libs/pull/809 for the patch.
Run the debian/ubuntu package building is currently a bit broken, so we'll
- git clone https://github.com/uli42/nx-libs.git
- cd nx-libs
- git co pr/fix/xkb_crash
- git revert b32a984f8f34d34fcc4114028712367c47c545d7 (due to this
undo it for the test) 5. debuild -i -us -uc -b 6. cd - 7. ls -l *deb 8. dpkg -i libnx-x11-6_3.5.99.20-0_amd64.deb 9. dpkg -i nxagent_3.5.99.20-0_amd64.deb
Uli
On Wed, May 15, 2019 at 4:54 PM Robert Kudyba <rkudyba@fordham.edu> wrote:
Are you able to compile a new nxagent once I have patch available?
Sure, just need to know the process.
On Wed, May 15, 2019 at 10:32 AM Ulrich Sibiller <uli42@gmx.de> wrote:
On Wed, May 15, 2019 at 4:24 PM Robert Kudyba <rkudyba@fordham.edu> wrote:
Here's an update. MATE still crashes. rpm -q x2goserver
This is not astonishing as I still have not patched anything. I wanted to understand the problem first. But as I do not have access to (or knowledge about) MacOS I cannot test this.
I will try to implement a check that will prevent the crashm but I am not sure if your keyboard will work afterwards. Are you able to compile a new nxagent once I have patch available?
Uli
- dpkg -i libnx-x11-6_3.5.99.20-0_amd64.deb
- dpkg -i nxagent_3.5.99.20-0_amd64.deb
We are on Fedora, don't think this will work.
On Wed, May 15, 2019 at 2:52 PM Ulrich Sibiller <uli42@gmx.de> wrote:
Run the debian/ubuntu package building is currently a bit broken, so we'll
- git clone https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_uli42_nx-2Dlibs.git&d=DwIBaQ&c=aqMfXOEvEJQh2iQMCb7Wy8l0sPnURkcqADc2guUW8IM&r=X0jL9y0sL4r4iU_qVtR3lLNo4tOL1ry_m7-psV3GejY&m=OnHueT2o7SfCj7XmKePe_A5dnMu_OufmnZ-QiEFseqg&s=a6_kZNyJsY-lCypHR2m2mZi0tsnaLci9USEdg83FVew&e=
- cd nx-libs
- git co pr/fix/xkb_crash
- git revert b32a984f8f34d34fcc4114028712367c47c545d7 (due to this
undo it for the test) 5. debuild -i -us -uc -b 6. cd - 7. ls -l *deb 8. dpkg -i libnx-x11-6_3.5.99.20-0_amd64.deb 9. dpkg -i nxagent_3.5.99.20-0_amd64.deb
Uli
On Wed, May 15, 2019 at 4:54 PM Robert Kudyba <rkudyba@fordham.edu> wrote:
Are you able to compile a new nxagent once I have patch available?
Sure, just need to know the process.
On Wed, May 15, 2019 at 10:32 AM Ulrich Sibiller <uli42@gmx.de> wrote:
On Wed, May 15, 2019 at 4:24 PM Robert Kudyba <rkudyba@fordham.edu>
wrote:
Here's an update. MATE still crashes. rpm -q x2goserver
This is not astonishing as I still have not patched anything. I wanted to understand the problem first. But as I do not have access to (or knowledge about) MacOS I cannot test this.
I will try to implement a check that will prevent the crashm but I am not sure if your keyboard will work afterwards. Are you able to compile a new nxagent once I have patch available?
Uli
On Mar 26, 2019, at 12:17 PM, Ulrich Sibiller <uli42@gmx.de> wrote:
#2 0x000055c236293ca6 in ProcXkbLatchLockState (client=0x55c2372372d0) at xkb.c:560
-> Proc* is always a call from outside to an extension, in this case the XKB extension.
So do you have some setxkbmap call in your shell startup files? What are they doing?
Not that I’m aware of. Would this be in something like .bash_rc?
Does this also happen with other session types (not MATE)?
With Terminal selected, no issues. I installed XFCE, and it works as expected. With MATE I see these logs, even though audio is unechecked:
Mar 26 13:50:57 sobolev systemd[1652]: pulseaudio.service: Scheduled restart job, restart counter is at 2. Mar 26 13:50:57 sobolev systemd[1652]: Stopped Sound Service. Mar 26 13:50:57 sobolev systemd[1652]: Starting Sound Service... Mar 26 13:50:58 sobolev pulseaudio[23356]: W: [pulseaudio] pid.c: Stale PID file, overwriting. Mar 26 13:50:58 sobolev dbus-daemon[842]: [system] Rejected send message, 2 matched rules; type="method_call", sender=":1.622566" (uid=6105 pid=23356 comm="/usr/bin/pulseaudio --daemonize=no ") interface="org.fre edesktop.DBus.ObjectManager" member="GetManagedObjects" error name="(unset)" requested_reply="0" destination="org.bluez" (bus) Mar 26 13:50:58 sobolev pulseaudio[23356]: E: [pulseaudio] bluez5-util.c: GetManagedObjects() failed: org.freedesktop.DBus.Error.AccessDenied: Rejected send message, 2 matched rules; type="method_call", sender=": 1.622566" (uid=6105 pid=23356 comm="/usr/bin/pulseaudio --daemonize=no ") interface="org.freedesktop.DBus.ObjectManager" member="GetManagedObjects" error name="(unset)" requested_reply="0" destination="org.bluez" (bus) Mar 26 13:50:58 sobolev systemd[1652]: Started Sound Service. Mar 26 13:51:02 sobolev /usr/sbin/x2gocleansessions[32250]: kudyba-58-1553622659_stRTERMINAL_dp32: state file for this session does not exist: /tmp/.x2go-kudyba/C-kudyba-58-1553622659_stRTERMINAL_dp32/state (this can be ignored during session startups) Mar 26 13:51:03 sobolev /usr/bin/x2gostartagent[23804]: Failed to start X2Go Agent session with ID kudyba-58-1553622659_stRTERMINAL_dp32. X2Go Agent terminated unexpectedly. Aborting session startup. Mar 26 13:51:05 sobolev /usr/sbin/x2gocleansessions[32250]: kudyba-58-1553622659_stRTERMINAL_dp32: state file for this session does not exist: /tmp/.x2go-kudyba/C-kudyba-58-1553622659_stRTERMINAL_dp32/state (this can be ignored during session startups) Mar 26 13:51:07 sobolev /usr/sbin/x2gocleansessions[32250]: kudyba-58-1553622659_stRTERMINAL_dp32: state file for this session does not exist: /tmp/.x2go-kudyba/C-kudyba-58-1553622659_stRTERMINAL_dp32/state (this can be ignored during session startups) Mar 26 13:51:09 sobolev /usr/sbin/x2gocleansessions[32250]: kudyba-58-1553622659_stRTERMINAL_dp32: state file for this session does not exist: /tmp/.x2go-kudyba/C-kudyba-58-1553622659_stRTERMINAL_dp32/state (this can be ignored during session startups) Mar 26 13:51:11 sobolev /usr/sbin/x2gocleansessions[32250]: kudyba-58-1553622659_stRTERMINAL_dp32: state file for this session does not exist: /tmp/.x2go-kudyba/C-kudyba-58-1553622659_stRTERMINAL_dp32/state (this can be ignored during session startups) Mar 26 13:51:14 sobolev /usr/sbin/x2gocleansessions[32250]: kudyba-58-1553622659_stRTERMINAL_dp32: state file for this session does not exist: /tmp/.x2go-kudyba/C-kudyba-58-1553622659_stRTERMINAL_dp32/state (this can be ignored during session startups) Mar 26 13:51:16 sobolev /usr/sbin/x2gocleansessions[32250]: kudyba-58-1553622659_stRTERMINAL_dp32: state file for this session does not exist: /tmp/.x2go-kudyba/C-kudyba-58-1553622659_stRTERMINAL_dp32/state (this can be ignored during session startups) Mar 26 13:51:18 sobolev /usr/sbin/x2gocleansessions[32250]: kudyba-58-1553622659_stRTERMINAL_dp32: state file for this session does not exist: /tmp/.x2go-kudyba/C-kudyba-58-1553622659_stRTERMINAL_dp32/state (this can be ignored during session startups) Mar 26 13:51:20 sobolev /usr/sbin/x2gocleansessions[32250]: kudyba-58-1553622659_stRTERMINAL_dp32: state file for this session does not exist: /tmp/.x2go-kudyba/C-kudyba-58-1553622659_stRTERMINAL_dp32/state (this can be ignored during session startups) Mar 26 13:51:23 sobolev /usr/sbin/x2gocleansessions[32250]: kudyba-58-1553622659_stRTERMINAL_dp32: state file for this session does not exist: /tmp/.x2go-kudyba/C-kudyba-58-1553622659_stRTERMINAL_dp32/state (this can be ignored during session startups) Mar 26 13:51:25 sobolev /usr/sbin/x2gocleansessions[32250]: kudyba-58-1553622659_stRTERMINAL_dp32: state file for this session does not exist: /tmp/.x2go-kudyba/C-kudyba-58-1553622659_stRTERMINAL_dp32/state (this can be ignored during session startups) Mar 26 13:51:27 sobolev /usr/sbin/x2gocleansessions[32250]: kudyba-58-1553622659_stRTERMINAL_dp32: state file for this session does not exist: /tmp/.x2go-kudyba/C-kudyba-58-1553622659_stRTERMINAL_dp32/state (this can be ignored during session startups) Mar 26 13:51:30 sobolev su[23961]: (to kudyba) root on none Mar 26 13:51:30 sobolev systemd[1]: Started Session c7 of user kudyba. Mar 26 13:51:30 sobolev /usr/sbin/x2gocleansessions[32250]: kudyba-58-1553622659_stRTERMINAL_dp32, pid 23794: cleaning up stale X11 socket file: /tmp/.X11-unix/X58 Mar 26 13:51:30 sobolev /usr/sbin/x2gocleansessions[32250]: kudyba-58-1553622659_stRTERMINAL_dp32, pid 23794: cleaning up stale X11 lock file: /tmp/.X58-lock Mar 26 13:51:30 sobolev su[23982]: (to kudyba) root on none
<<I think this might be connected to one of the following: a) you are using a special keyboard (hardware) b) you are using a special keyboard configuration c) you have configured a special keyboard configuration somewhere in Mate d) you have configured caps lock/numlock being active on Mate startup (I don't know if this is configurable in Mate) e) you have caps lock/numlock activated on the x2go client side<<
I will double check this with my colleague. I don’t have a num lock on my Mac keyboard.
On Tue, Mar 26, 2019 at 10:13 AM Mihai Moldovan <ionic@ionic.de> wrote:
- On 3/25/19 2:10 PM, Ulrich Sibiller wrote:
For testing you could try without ssh-agent or with the one of the current openssh.
Sure, that's the quickest workaround... I just switched to no-agent auth for testing.
Naturally, though, I wasn't able to reproduce the problem with the current master version. Everything was working fine as far as the keyboard was concerned.
It might be related to that commit, but I don't know. ctrl->num_groups is obviously zero in XkbAdjustGroup()/xkbUtils.c, but I have no idea what that actually means. The code in question does look a bit shady (for instance, it checks for group >= ctrls->num_groups and later might set group %= ctrls->num_groups, which obviously is a bad idea if both group and ctrls->num_groups is zero, but I have the suspicion that that should never have been the case in the first place.
I don't know what "groups" are in this context neither.
Keyboard groups, as described here: http://soc.if.usp.br/manual/x11proto-kb-dev/xkbproto.html#Keyboard_State
I somehow doubt that commit caused the problem, though. If the clone mechanism works fine, that change shouldn't cause any problems. There's probably no harm in cloning the config, even though we later sync up using xmodmap (other than a race condition, i.e., there MIGHT be a problem if we first sync up using xmodmap and then nxagent changes the whole keyboard settings again).
_maybe_ introducing a newer version of the xkb code also requires additional initialization (t.i. cloning more states from the real X server) which I might have missed somewhere.
Seems to be yet another weird problem that can only be reproduced with user accounts from NIS, but I have no idea why that might be. As far as I remember,
I cannot think of any reason why NIS users should behave different from local users (after all, NIS is just a nameservice like ldap or local files). The only thing that comes to mind is some central configuration that is only pulled in for NIS users.
Roberts home directories are not even coming via NFS so the home dir should be available at the time nxagent/x2goagent starts and - in that case - tries to block keyboard file creation by making a directory. However, nx-libs-3.5.99.18-0.0build1.0.git20190208.3237.heuler.fc29.x86_64 is quite old and I recently synced the master branch up with Arctica Project's master branch again, which means that it has 3.5.99.19 + the last two commits on top of it. Maybe it's worthwhile to retest with that.
I'm wary of releasing 3.5.99.19 to X2Go land, though, since I did notice a different problem in my test of an XFCE session on a Devuan Unstable machine: for some reason I didn't have xfce4-terminal installed and the default terminal application was set to qterminal. That started up, but only worked for a few seconds, after which it froze up. x2goagent/nxagent used 100 % of one CPU thread at that point and even after killing qterminal it continued to do so for a few minutes, after which it settled down with occasional 100 % CPU spikes (but it does that without having ever started qterminal, too, so I won't worry about it. My machine's test CPU is a pretty bad one - Intel Atom 1.8 GHs).
This looks like the app is using OpenGL. Does it work if you disable GLX?
Hopefully that's not a general problem with Qt5 applications?
Well, it depends on the standard (system) configuration for Qt apps. It Qt is configured to use OpenGL most apps will suffer from this.
Uli
-
Mihai Moldovan -
Robert Kudyba -
Ulrich Sibiller