Description of problem:Running MCC from the Live system is ok. after installation and reboot into the installd system, MCC will not open after the authentication dialogue is accepted. running from a user terminal, the dialogue box appears and accepts input, but after choosing *authenticate*, the dialogue closes with no GUI presented. from a root terminal, running mcc brings up the drakconf ncurses window. # hostnamectl set-hostname set_a_name and rebooting does not appear to fix this. Version-Release number of selected component (if applicable): PATH: Mageia-8-rc-Live-Xfce-i586 ISO: Mageia-8-rc-Live-Xfce-i586.iso DATE: Sun 31 Jan 2021 02:23:32 PM CET How reproducible: on this hardware - often Steps to Reproduce: 1.install from above .iso and reboot 2.attempt to open MCC 3.
output of terminal: $ mcc Too late to run INIT block at /usr/lib/perl5/vendor_perl/Glib/Object/Introspection.pm line 257. Ignore the following Glib::Object::Introspection & Gtk3 warnings No protocol specified Unable to init server: Could not connect: Connection refused Cannot be run in console mode.
Hardware: All => i586Keywords: (none) => 8rc1Target Milestone: --- => Mageia 8Whiteboard: (none) => round 5
hardware: # lspcidrake lpc_ich : Intel Corporation|Atom Processor Z36xxx/Z37xxx Series Power Control Unit [BRIDGE_ISA] (rev: 09) xhci_pci : Intel Corporation|Atom Processor Z36xxx/Z37xxx, Celeron N2000 Series USB xHCI [SERIAL_USB] (rev: 09) Card:Intel 810 and later: Intel Corporation|Atom Processor Z36xxx/Z37xxx Series Graphics & Display [DISPLAY_VGA] (rev: 09) mei_txe : Intel Corporation|Atom Processor Z36xxx/Z37xxx Series Trusted Execution Engine [CRYPT_OTHER] (rev: 09) unknown : Intel Corporation|Atom Processor Z36xxx/Z37xxx Series SoC Transaction Register [BRIDGE_HOST] (rev: 09) hub : Linux 5.10.12-desktop586-3.mga8 xhci-hcd|xHCI Host Controller [Hub|Unused|Full speed (or root) hub] usbhid : ASUSTek COMPUTER INC.|ASUS Base Station(T100) [Human Interface Device|Boot Interface Subclass|Keyboard] hub : Linux 5.10.12-desktop586-3.mga8 xhci-hcd|xHCI Host Controller [Hub|Unused|Full speed (or root) hub] hid_generic : ASUSTek COMPUTER INC. ASUS Base Station(T100) hid_generic : ASUSTek COMPUTER INC. ASUS Base Station(T100) hid_generic : ATML1000:00 03EB:8C0D hid_generic : ASUSTek COMPUTER INC. ASUS Base Station(T100)
CC: (none) => fri
heres a gotcha. after running launchGUIs -t 2 as root, it will now work. however, not permanently. issue returns after a reboot. some temporary change has been made?
Created attachment 12303 [details] launchGUIs script normally run as user. if running as root, disable internet, also would not normally be recommended.
So, from the error message in the terminal, this is definitely an X authentication failure. Can you show the output from hostnamectl after this happens. Does logging out and logging back in again allow it to work?
CC: (none) => mageia
(In reply to Martin Whitaker from comment #5) > So, from the error message in the terminal, this is definitely an X > authentication failure. Can you show the output from hostnamectl after this > happens. > [me@linux ~]$ hostnamectl Static hostname: localhost Transient hostname: linux.local Icon name: computer-laptop Chassis: laptop Machine ID: f1101e7bcd974a3eb74ce91801f15733 Boot ID: 1e4b8a00c0344481b946aefd165deb57 Operating System: Mageia 8 Kernel: Linux 5.10.12-desktop586-3.mga8 Architecture: x86 [me@linux ~]$ mcc Too late to run INIT block at /usr/lib/perl5/vendor_perl/Glib/Object/Introspection.pm line 257. Ignore the following Glib::Object::Introspection & Gtk3 warnings No protocol specified Unable to init server: Could not connect: Connection refused Cannot be run in console mode. ~ launchGUIs ~ $ mcc Too late to run INIT block at /usr/lib/perl5/vendor_perl/Glib/Object/Introspection.pm line 257. Ignore the following Glib::Object::Introspection & Gtk3 warnings Subroutine Gtk3::main redefined at /usr/share/perl5/vendor_perl/Gtk3.pm line 539. GLib-GObject-CRITICAL **: g_boxed_type_register_static: assertion 'g_type_from_name (name) == 0' failed at /usr/lib/perl5/DynaLoader.pm line 210. GLib-GObject-CRITICAL **: g_boxed_type_register_static: assertion 'g_type_from_name (name) == 0' failed at /usr/lib/perl5/DynaLoader.pm line 210. GLib-GObject-CRITICAL **: g_boxed_type_register_static: assertion 'g_type_from_name (name) == 0' failed at /usr/lib/perl5/DynaLoader.pm line 210. GLib-GObject-CRITICAL **: g_boxed_type_register_static: assertion 'g_type_from_name (name) == 0' failed at /usr/lib/perl5/DynaLoader.pm line 210. "cannot run /usr/sbin/isodumper" since it is not installed [Writing ISO] at /usr/libexec/drakconf line 833. Oops, secure memory pool already initialized Oops, secure memory pool already initialized (authenication dialogue presented, accepted and MCC GUI presented) $ hostnamectl Static hostname: localhost Transient hostname: linux.local Icon name: computer-laptop Chassis: laptop Machine ID: f1101e7bcd974a3eb74ce91801f15733 Boot ID: 1e4b8a00c0344481b946aefd165deb57 Operating System: Mageia 8 Kernel: Linux 5.10.12-desktop586-3.mga8 Architecture: x86 not sure what isodumper is doing in there. will install and see what happens > Does logging out and logging back in again allow it to work? will find out in a minute - logging out ..... no, after logging back in - problem persists. will see if adding isodumper changes thing
adding isodumper (gtk) no difference
When you log out and log back in again, is it disconnecting and reconnecting the network?
(In reply to Martin Whitaker from comment #8) > When you log out and log back in again, is it disconnecting and reconnecting > the network? yes so other stuff: [me@localhost Documents]$ mcc Too late to run INIT block at /usr/lib/perl5/vendor_perl/Glib/Object/Introspection.pm line 257. Ignore the following Glib::Object::Introspection & Gtk3 warnings No protocol specified Unable to init server: Could not connect: Connection refused Cannot be run in console mode. [me@localhost Documents]$ ./launchGUIs -s 63 -d 1 -t 2 Version 5 Input applications file = /home/me/guiApps.txt New Results file = /home/me/guiResults.txt New Failures file will = /home/me/guiFails.txt Running in Automatic mode Doing 1 tests 63-63 of 115, for 2 seconds The following lines will be run (unless excluded) [~ is space] (it would help to show their source line numbers...) /usr/bin/gparted~%f Is this OK? Just Enter to continue, anything else to quit: ======================================================== Trying line 63, test 1/1: '/usr/bin/gparted %f' localuser:root being added to access control list New process correctly found: 19067 gparted, killing it ./launchGUIs: line 326: 19067 Killed $COMMAND y/usr/bin/gparted %f seems to WORK ======================================================== 1/1 apparent successes, 0/1 apparent failures Press Enter to see the result summary (Q|q to quit 'less') mv: cannot stat '/home/me/guiFails.txt.19029': No such file or directory [me@localhost Documents]$ mcc Too late to run INIT block at /usr/lib/perl5/vendor_perl/Glib/Object/Introspection.pm line 257. Ignore the following Glib::Object::Introspection & Gtk3 warnings Subroutine Gtk3::main redefined at /usr/share/perl5/vendor_perl/Gtk3.pm line 539. GLib-GObject-CRITICAL **: g_boxed_type_register_static: assertion 'g_type_from_name (name) == 0' failed at /usr/lib/perl5/DynaLoader.pm line 210. GLib-GObject-CRITICAL **: g_boxed_type_register_static: assertion 'g_type_from_name (name) == 0' failed at /usr/lib/perl5/DynaLoader.pm line 210. GLib-GObject-CRITICAL **: g_boxed_type_register_static: assertion 'g_type_from_name (name) == 0' failed at /usr/lib/perl5/DynaLoader.pm line 210. GLib-GObject-CRITICAL **: g_boxed_type_register_static: assertion 'g_type_from_name (name) == 0' failed at /usr/lib/perl5/DynaLoader.pm line 210. Oops, secure memory pool already initialized (authenication dialogue presented, accepted and MCC GUI presented)
dont know if this is helpful: running gparted from terminal and ctrl+c to kill the dialogue box does allow mcc then to present GUI
heres a trick: connecting to the wifi via drakroam, then logging out or rebooting, MCC works. disconnection wifi via drakroam and reconnect with network manager = no MCC
(In reply to ben mcmonagle from comment #9) > (In reply to Martin Whitaker from comment #8) > > When you log out and log back in again, is it disconnecting and reconnecting > > the network? > > yes Ah, I suspect when you configured the network interface, you didn't tick the box to let all users share it. That would make a difference. Each time the connection is made, NetworkManager tries to discover and set the hostname from the network. linux.local is the default name returned by a reverse DNS lookup. > Trying line 63, test 1/1: '/usr/bin/gparted %f' > localuser:root being added to access control list So either launchGUIs or gparted is relaxing the X access permissions, which is why MCC starts afterwards. You can do the same by (as your normal user) xhost +si:localuser:root Setting a proper hostname should fix the problem. I don't know why 'hostnamectl set-hostname' didn't work for you. You could try doing it manually by editing the /etc/hostname file. The other possible fix is to append the line hostname-mode=none to the /etc/NetworkManager/NetworkManager.conf file, which will stop NetworkManager meddling with the hostname.
The problems are more widespread than first thought. Even on the Plasma Live, which doesn't use NM, running drakconnect to configure the network and leaving the host name as the default of localhost.localdomain causes subsequent attempts to start the MCC (or any drakx GUI tool) to fail. There has been some subtle change to the way host names are managed by the system. On Mageia 7, hostnamectl would report [root@localhost ~]# hostnamectl Static hostname: localhost.localdomain whilst on Mageia 8 it reports [root@localhost ~]# hostnamectl Static hostname: localhost.localdomain Transient hostname: localhost So I think the fix for all this is to do the same as we used to do for GNOME/Wayland, and have a startup script that executes xhost +si:localuser:root each time a user starts a graphical session.
I see this on Live Xfce 64 RC1 take 5 As an industrial machine designer i am stumped at why answering the phone in the office makes all control panels in manufacturing die... Confirming workaround: In Terminal window: $ xhost +si:localuser:root localuser:root being added to access control list
Severity: normal => majorHardware: i586 => AllPriority: Normal => release_blocker
forgot to say: ...and then MCC launches OK from both icon and Terminal.
Summary: Installed from Live Xfce32 MCC does not open after authentication entered => Drakxtools fail to launch in GUI mode after configuring network
(In reply to Martin Whitaker from comment #13) > The problems are more widespread than first thought. Even on the Plasma > Live, which doesn't use NM, running drakconnect to configure the network and > leaving the host name as the default of localhost.localdomain causes > subsequent attempts to start the MCC (or any drakx GUI tool) to fail. There > has been some subtle change to the way host names are managed by the system. > Well now this is weird. I have a Plasma install created by the Live iso just yesterday, and it's not showing the problem. The system is wifi-only, and the wifi was set up using the net_applet before the install, so maybe that did an end-run around drakconnect. After the boot the first thing I did was call up MCC, and everything is normal. I set up the Princeton mirror, and went after updates. When there were none, I used urpmi --auto-update. That got me several, but even after another boot MCC behaves normally. I did retain /home through the install, and as far as that goes this /home once belonged to a Mageia 7 install. I didn't do anything with the hostname - it's whatever it was set by the installer.
CC: (none) => andrewsfarm
Keywords: (none) => IN_ERRATA8
How about a script in /etc/X11/xinit.d/ with ... #!/bin/sh while inotifywait -e modify /proc/sys/kernel/hostname; do xhost +si:localuser:root done Untested at this point.
CC: (none) => davidwhodgins
(In reply to Martin Whitaker from comment #13) > The problems are more widespread than first thought. Even on the Plasma > Live, which doesn't use NM, running drakconnect to configure the network and > leaving the host name as the default of localhost.localdomain causes > subsequent attempts to start the MCC (or any drakx GUI tool) to fail. There > has been some subtle change to the way host names are managed by the system. > On Mageia 7, hostnamectl would report > > [root@localhost ~]# hostnamectl > Static hostname: localhost.localdomain > > whilst on Mageia 8 it reports > > [root@localhost ~]# hostnamectl > Static hostname: localhost.localdomain > Transient hostname: localhost > > So I think the fix for all this is to do the same as we used to do for > GNOME/Wayland, and have a startup script that executes > > xhost +si:localuser:root > > each time a user starts a graphical session. Seems no-one has a better solution for now so we should go for this one. Can you do it so we can get wide testing Then with a little dreaming hope to get it better fixed in mga9
Workaround has been pushed to repository to xinitrc package http://svnweb.mageia.org/packages?view=revision&revision=1675229 This need a Real test please. Thanks Martin W.
Status comment: (none) => workaround added to xinitrc package and pushed.CC: (none) => ouaurelien
Works as expected under Cinnamon, GNOME Wayland, GNOME, XFCE, icewm-session. Reducing as not a blocker one.
Target Milestone: Mageia 8 => Mageia 9Priority: release_blocker => High
Do the fix it not work in Plasma?
Is this related: Bug 28365 - Control Center does not start in Plasma Wayland
(In reply to Morgan Leijström from comment #22) > Is this related: > Bug 28365 - Control Center does not start in Plasma Wayland It's related in that it's an X authorisation problem, but it has a different cause. See bug 26101.
Now removed from Errata, including the mention of bug 27601.
Keywords: IN_ERRATA8 => (none)
(In reply to Aurelien Oudelet from comment #20) > Works as expected under Cinnamon, GNOME Wayland, GNOME, XFCE, icewm-session. > > Reducing as not a blocker one. Workaround set in xinitrc script. Assigning to Magiea Tools maintainers, against Cauldron.
Keywords: 8rc1 => (none)Severity: major => normalWhiteboard: round 5 => (none)Assignee: bugsquad => mageiatoolsVersion: 8 => Cauldron
fixed
Status: NEW => RESOLVEDResolution: (none) => FIXED