When starting LibreOffice AKA LO with KDE4 backend on a remote server, there is a high chance LO starts and gets stuck with refresh problems (no refresh at all) on startup. I can always deadlock refresh of LO by opening any dialog of LO. It's a little better with the Gtk+2 backend, where the application starts, but can still be deadlocked with opening a dialog, easiest done with the complex extension manager. To make it worse the whole screen basically breaks. On any mouse movement parts of the screen flickers and is repainted wrong. At this point you can just hope to switch to the konsole and Ctrl+C LO, which restores the screen. Luckly the window switcher (Alt + Tab) let you identify the konsole, even if the image of the window is completely broken, but it still resembles the konsole. As I need this connection for my LO development, I just switched to WindowMaker - no more problems. I found #363500, but this looks much less severe. Reproducible: Always Steps to Reproduce: 1. ssh -X 2. start LibreOffice preferably with KDE4 backend Actual Results: Whole screen is a mess, any interaction flickers. Expected Results: Remote X "just works".
Harald made me aware of this bug. A few questions: * does that happen with compositing enabled/disabled? * can you provide the output of "qdbus org.kde.KWin /KWin supportInformation" when the problem happens? * is this happening on all systems, or maybe only with a specific type of hardware (e.g. Intel driver broken on Xorg level, something in that direction).
Might it be that LO needs glx and that's disabled in Xorg now?
any interesting output in /var/log/Xorg.0.log when it happens?
Created attachment 100204 [details] KWin support information
Created attachment 100205 [details] Linux kernel log (dmesg)
Created attachment 100206 [details] Flickering / broken desktop from ksnapshot This time the effect was even more extreme. Last time I had to do some interaction, like moving the mouse, to get the flickering (the desktop is nevertheless always broken in a way). This time I didn't even had to move the mouse and it was constantly flickering, albeit slower compared to the mouse movement.
Created attachment 100208 [details] xsession-errors from my working session It's not from the broken session, but still has a lot of troublesome looking messages, like: QXcbConnection: XCB error: 9 (BadDrawable), sequence: 64271, resource id: 0, major code: 14 (GetGeometry), minor code: 0
Sorry that I didn't reply earlier. I tried to reproduce the problem, but failed until today (well I didn't try to actively reproduce it, but continued my LibreOffice (LO) development with Plasma 2 WM). The freezeing of LO seem to happen, when I use our LO extension, which is build with Java 7, but LO uses the Java 6 VM. This is independent of the flickering. Now today it happened again. I was able to take a screenshot and got the dmesg, which showed a lot of errors from the radeon driver, as the Linux console worked fine: line 1300: [26610.825302] radeon 0000:01:00.0: r600_check_texture_resource:1566 tex pitch (792, 0x40, 1) invalid [26610.825349] [drm:radeon_cs_ioctl [radeon]] *ERROR* Invalid command stream ! ... line 2484 [26898.242109] radeon 0000:01:00.0: r600_check_texture_resource:1566 tex pitch (792, 0x40, 1) invalid [26898.242152] [drm:radeon_cs_ioctl [radeon]] *ERROR* Invalid command stream ! It might be a coincidence, but this probably started after I installed todays Neon + Ubuntu updates. Don't know if this happened last time too. Lot of packages changed from: 4:5.7.1-0neon+16.04+* => 4:5.7.2-0neon+16.04+* Still for me the flickering never happened with Neon 5.6, which I used before. I forgot to copy the old .xsesion-errors and Xorg.0.log, so these are gone now :-( This also happened because of an other problem: I often have to login twice, otherwise primary and secondary screen are inverted. Workaround is just to logout and then login again. It's annoying, that it also moves the desktop icons between the screens, but I'll open a 2nd bug, as it doesn't seem related. I'll update the bug with .xsession-errors and Xorg.0.log, if it happens again.
oh my the looks horrible. What surprises me most is that you are able to snapshot it with ksnapshot. Quite often such problems won't show in that case. What's interesting is the kernel log giving us radeon errors. Which makes me wonder: do you also get this with compositing disabled (Alt+Shift+f12)? And on (if you have access to) other hardware?
Created attachment 100400 [details] Even more broken snapshot
Created attachment 100401 [details] xsession errors from a flickering desktop
I had some more of these problems, but it's still not reproducible. My feeling is it's always at the end of the day, which might suggest leaking resources in the radeon driver. I don't really know, how to debug this. As you can see in the xsession-errors log, it contains a lot of > radeon: The kernel rejected CS, see dmesg for more information. so the kernel driver rejected the command stream from the Xorg driver, which matches the entries in the dmesg, literally thousand of lines of [40490.073706] radeon 0000:01:00.0: r600_check_texture_resource:1566 tex pitch (1752, 0x40, 1) invalid [40490.073763] [drm:radeon_cs_ioctl [radeon]] *ERROR* Invalid command stream ! in a few seconds / a minute. Now an application shouldn't be able to create a broken command stream. This looks more like a problem deeper in the stack. I'm not sure I remembered the right key combo, but any keys I hit didn't change the settings (I finally created a physical note, so I can active lookup the no-compositor-combo). I have access to other HW, but that's not my working system. Other people here running KDE5 with Intel HW (on Arch AFAIK) don't have this problem.
Can you change the Qt debug message pattern to get the application included? I would like to know whether the message with the invalid command stream is from KWin.
So this time I can confirm, that "Alt + Shift + F12" stops the compositing and the flickering. Guess I hit the wrong key combo last time, or I had a different problem, but xsession-errors and dmesg show almost the same messages: Today: [18199.442243] radeon 0000:01:00.0: r600_check_texture_resource:1566 tex pitch (888, 0x40, 1) invalid [18199.442334] [drm:radeon_cs_ioctl [radeon]] *ERROR* Invalid command stream ! Previous: [26610.825302] radeon 0000:01:00.0: r600_check_texture_resource:1566 tex pitch (792, 0x40, 1) invalid [26610.825349] [drm:radeon_cs_ioctl [radeon]] *ERROR* Invalid command stream ! What's the preferred way to set a different QT_MESSAGE_PATTERN? Edit startkde? Guess the example from the Qt docs is ok for you: QT_MESSAGE_PATTERN="[%{type}] %{appname} (%{file}:%{line}) - %{message}" This proves my theory about the "late at day" is wrong, as I neither installed new software, nor intensively used the system. Actually this time it happened almost instantly after the first start of LO, when opening a dialog.
> What's the preferred way to set a different QT_MESSAGE_PATTERN? Edit startkde? startkde is an option, though maybe overkill. Another possibility is to edit any environment file, e.g. /etc/environment or the local users overwrite. The example for the pattern looks good.
Created attachment 100430 [details] xsession-errors with extended error messages via QT_MESSAGE_PATTERN
Not sure if this helps. Searching for CS just shows the old lines, but now the previous lines list (almost) the same amount of warnings: [warning] plasmashell (unknown:0) - QXcbConnection: XCB error: 2 (BadValue), sequence: 50142, resource id: 119537670, major code: 142 (Unknown), minor code: 3 [warning] kwin (unknown:0) - QXcbConnection: XCB error: 9 (BadDrawable), sequence: 14337, resource id: 0, major code: 14 (GetGeometry), minor code: 0 [warning] kwin (unknown:0) - QXcbConnection: XCB error: 9 (BadDrawable), sequence: 14338, resource id: 0, major code: 14 (GetGeometry), minor code: 0 [warning] kwin (unknown:0) - QXcbConnection: XCB error: 9 (BadDrawable), sequence: 14340, resource id: 0, major code: 14 (GetGeometry), minor code: 0 [warning] kwin (unknown:0) - QXcbConnection: XCB error: 9 (BadDrawable), sequence: 14378, resource id: 0, major code: 14 (GetGeometry), minor code: 0 [warning] kwin (unknown:0) - QXcbConnection: XCB error: 3 (BadWindow), sequence: 17791, resource id: 132120671, major code: 15 (QueryTree), minor code: 0 [warning] kwin (unknown:0) - QXcbConnection: XCB error: 3 (BadWindow), sequence: 17840, resource id: 132120901, major code: 15 (QueryTree), minor code: 0 [warning] kwin (unknown:0) - QXcbConnection: XCB error: 3 (BadWindow), sequence: 17888, resource id: 132263908, major code: 15 (QueryTree), minor code: 0 [warning] kwin (unknown:0) - QXcbConnection: XCB error: 3 (BadWindow), sequence: 18528, resource id: 121634817, major code: 15 (QueryTree), minor code: 0 [warning] kwin (unknown:0) - QXcbConnection: XCB error: 3 (BadWindow), sequence: 20598, resource id: 121637445, major code: 15 (QueryTree), minor code: 0 radeon: The kernel rejected CS, see dmesg for more information. radeon: The kernel rejected CS, see dmesg for more information. radeon: The kernel rejected CS, see dmesg for more information. radeon: The kernel rejected CS, see dmesg for more information. radeon: The kernel rejected CS, see dmesg for more information. radeon: The kernel rejected CS, see dmesg for more information. radeon: The kernel rejected CS, see dmesg for more information. radeon: The kernel rejected CS, see dmesg for more information. radeon: The kernel rejected CS, see dmesg for more information. radeon: The kernel rejected CS, see dmesg for more information. [warning] plasmashell (file:///usr/lib/x86_64-linux-gnu/qt5/qml/QtQuick/Controls/Button.qml:100) - file:///usr/lib/x86_64-linux-gnu/qt5/qml/QtQuick/Controls/Button.qml:100: TypeError: Cannot read property of null [warning] plasmashell (file:///usr/share/plasma/plasmoids/org.kde.plasma.taskmanager/contents/ui/ToolTipDelegate.qml:83) - file:///usr/share/plasma/plasmoids/org.kde.plasma.taskmanager/contents/ui/ToolTipDelegate.qml:83: TypeError: Cannot read property 'DesktopEntry' of undefined [warning] plasmashell (unknown:0) - QXcbConnection: XCB error: 2 (BadValue), sequence: 53586, resource id: 115343366, major code: 142 (Unknown), minor code: 3 radeon: The kernel rejected CS, see dmesg for more information. radeon: The kernel rejected CS, see dmesg for more information.
Any news? I'm experiencing it on my laptop as well and see the same output as Jan-Marek Glogowski.
Thank you for reporting this bug in KDE software. As it has been a while since this issue was reported, can we please ask you to see if you can reproduce the issue with a recent software version? If you can reproduce the issue, please change the status to "REPORTED" when replying. Thank you!
Dear Bug Submitter, This bug has been in NEEDSINFO status with no change for at least 15 days. Please provide the requested information as soon as possible and set the bug status as REPORTED. Due to regular bug tracker maintenance, if the bug is still in NEEDSINFO status with no change in 30 days the bug will be closed as RESOLVED > WORKSFORME due to lack of needed information. For more information about our bug triaging procedures please read the wiki located here: https://community.kde.org/Guidelines_and_HOWTOs/Bug_triaging If you have already provided the requested information, please mark the bug as REPORTED so that the KDE team knows that the bug is ready to be confirmed. Thank you for helping us make KDE software even better for everyone!
This bug has been in NEEDSINFO status with no change for at least 30 days. The bug is now closed as RESOLVED > WORKSFORME due to lack of needed information. For more information about our bug triaging procedures please read the wiki located here: https://community.kde.org/Guidelines_and_HOWTOs/Bug_triaging Thank you for helping us make KDE software even better for everyone!