I have experienced this problem with Chrome only (but for quite a while now, different versions) but the question is not about Chrome but about X (I guess). Sometimes (I am not aware of any pattern) an activity causes a part of the Chrome window to be redrawn (e.g. opening the context menu or the bookmark window) but this drawing (sometimes finished, sometimes an early stage, just a white or grey background) is dead then: Clicking it leads to no recognizable result.
If I move the main window then this dead part stays where it appeared (like with
position:fixed in CSS). Thus I (with nearly no X11 background knowledge) assume it must me an additional window which somehow got lost.
Chrome works normally afterwards (but the probability for this to happen again has increased a lot); the problem is: This dead window is the "top layer" on the display i.e. it covers everything.
I can kill this dead window using Ctrl-Alt-Esc but that (takes some time and) kills not only the dead window but the main window, too.
Is there a way to kill only the dead window? Maybe some kind of window handle which can be closed externally? I would not be surprised if there is none but I'll give it a try.
~/.xsession-errors but interesting lines in
/var/log/Xorg.0.log (which has an mtime which could be the moment of this event):
(EE) [mi] EQ overflowing. Additional events will be discarded until existing events are processed. (EE) (EE) Backtrace: (EE) 0: /usr/bin/Xorg (xorg_backtrace+0x3d) [0x585aad] (EE) 1: /usr/bin/Xorg (mieqEnqueue+0x22b) [0x5679db] (EE) 2: /usr/bin/Xorg (QueuePointerEvents+0x52) [0x4534a2] (EE) 3: /usr/lib64/xorg/modules/input/evdev_drv.so (0x7fb8059fa000+0x5b0d) [0x7fb8059ffb0d] (EE) 4: /usr/bin/Xorg (0x400000+0x797b8) [0x4797b8] (EE) 5: /usr/bin/Xorg (0x400000+0xa1e30) [0x4a1e30] (EE) 6: /lib64/libpthread.so.0 (0x7fb811996000+0xf9f0) [0x7fb8119a59f0] (EE) 7: /usr/bin/Xorg (0x400000+0x189c70) [0x589c70] (EE) 8: /lib64/libpthread.so.0 (0x7fb811996000+0xf9f0) [0x7fb8119a59f0] (EE) 9: /lib64/libc.so.6 (0x7fb8107e8000+0x7a154) [0x7fb810862154] (EE) 10: /usr/lib64/xorg/modules/libexa.so (0x7fb80ba81000+0x7d98) [0x7fb80ba88d98] (EE) 11: /usr/bin/Xorg (0x400000+0x110c61) [0x510c61] (EE) 12: /usr/bin/Xorg (0x400000+0xd7109) [0x4d7109] (EE) 13: /usr/bin/Xorg (FreePicture+0x138) [0x5048c8] (EE) 14: /usr/bin/Xorg (0x400000+0x5e962) [0x45e962] (EE) 15: /usr/bin/Xorg (FreeClientResources+0x6c) [0x45f98c] (EE) 16: /usr/bin/Xorg (CloseDownClient+0x5f) [0x43c10f] (EE) 17: /usr/bin/Xorg (0x400000+0x3c33f) [0x43c33f] (EE) 18: /usr/bin/Xorg (0x400000+0x3cc3e) [0x43cc3e] (EE) 19: /usr/bin/Xorg (0x400000+0x2c1ba) [0x42c1ba] (EE) 20: /lib64/libc.so.6 (__libc_start_main+0xf5) [0x7fb810809be5] (EE) 21: /usr/bin/Xorg (0x400000+0x2c501) [0x42c501] (EE) (EE) [mi] These backtraces from mieqEnqueue may point to a culprit higher up the stack. (EE) [mi] mieq is *NOT* the cause. It is a victim. [709482.747] [mi] Increasing EQ size to 1024 to prevent dropped events. [709482.750] [mi] EQ processing has resumed after 74 dropped events. [709482.750] [mi] EQ processing has resumed after 74 dropped events. [709482.750] [mi] This may be caused my a misbehaving driver monopolizing the server's resources.
graphics hardware and driver
I have an AMD E-450 (i.e. GPU integrated in the processor).
/var/log/Xorg.0.log there is this line:
[ 52.009] (--) RADEON(0): Chipset: "AMD Radeon HD 6300 Series Graphics" (ChipID = 0x9806)