Machine freezes locally, Xorg EQ overflow

[AG]

We observe spontaneous and not (yet) reproducible freezes on many of our heterogeneous PCs. A machine freezes such that it cannot be accessed locally anymore (keyboard dead, screen black or with last image, sometimes mouse can be moved but no action on clicks) but we have full access via ssh. There are no heavy processes or such running. Freeze happens sometime during usage but also during screen-lock without background activity.

We have observed the problem
 * on Ubuntu 12.04 and 14.04,
 * on PCs running a multitude of graphics adapters (nvidia, ati, intel)
 * on PCs running different versions of e.g. the nvidia driver

The only log file with some information related to this is /var/log/Xorg.0.log where we find

==>>===================================================================
(EE) [mi] EQ overflowing. Additional events will be discarded until existing
events are processed.
(EE)
(EE) Backtrace:
(EE) 0: /usr/bin/X (xorg_backtrace+0x48) [0x7f1bd1327848]
(EE) 1: /usr/bin/X (mieqEnqueue+0x22b) [0x7f1bd130983b]
(EE) 2: /usr/bin/X (QueuePointerEvents+0x52) [0x7f1bd11eaf72]
(EE) 3: /usr/lib/xorg/modules/input/evdev_drv.so (0x7f1bc8621000+0x57fd)
[0x7f1bc86267fd]
(EE) 4: /usr/bin/X (0x7f1bd117e000+0x93518) [0x7f1bd1211518]
(EE) 5: /usr/bin/X (0x7f1bd117e000+0xbbc90) [0x7f1bd1239c90]
(EE) 6: /lib/x86_64-linux-gnu/libpthread.so.0 (0x7f1bd027a000+0x10340)
[0x7f1bd028a340]
(EE) 7: /lib/x86_64-linux-gnu/libc.so.6 (__poll+0x10) [0x7f1bced86110]
(EE) 8: /lib/x86_64-linux-gnu/libc.so.6 (0x7f1bcec99000+0x1284a0) [0x7f1bcedc14a0]
(EE) 9: /lib/x86_64-linux-gnu/libnsl.so.1 (0x7f1bc7ff7000+0x515f) [0x7f1bc7ffc15f]
(EE) 10: /lib/x86_64-linux-gnu/libnsl.so.1 (yp_match+0xb2) [0x7f1bc7ffc6a2]
(EE) 11: /lib/x86_64-linux-gnu/libnss_nis.so.2 (_nss_nis_getpwnam_r+0x68)
[0x7f1bc7df0a98]
(EE) 12: /lib/x86_64-linux-gnu/libnss_compat.so.2 (0x7f1bc8211000+0x156c)
[0x7f1bc821256c]
(EE) 13: /lib/x86_64-linux-gnu/libnss_compat.so.2 (_nss_compat_getpwnam_r+0x3f7)
[0x7f1bc8215617]
(EE) 14: /lib/x86_64-linux-gnu/libc.so.6 (getpwnam_r+0xdd) [0x7f1bced5906d]
(EE) 15: /lib/x86_64-linux-gnu/libc.so.6 (getpwnam+0x7f) [0x7f1bced589ff]
(EE) 16: /usr/bin/X (0x7f1bd117e000+0x1a7ae5) [0x7f1bd1325ae5]
(EE) 17: /usr/bin/X (0x7f1bd117e000+0x1a8ac7) [0x7f1bd1326ac7]
(EE) 18: /usr/bin/X (InvalidHost+0x170) [0x7f1bd1327180]
(EE) 19: /usr/bin/X (ClientAuthorized+0x125) [0x7f1bd1328905]
(EE) 20: /usr/bin/X (0x7f1bd117e000+0x559ac) [0x7f1bd11d39ac]
(EE) 21: /usr/bin/X (0x7f1bd117e000+0x55f0e) [0x7f1bd11d3f0e]
(EE) 22: /usr/bin/X (0x7f1bd117e000+0x59d9a) [0x7f1bd11d7d9a]
(EE) 23: /lib/x86_64-linux-gnu/libc.so.6 (__libc_start_main+0xf5) [0x7f1bcecbaec5]
(EE) 24: /usr/bin/X (0x7f1bd117e000+0x451ee) [0x7f1bd11c31ee]
(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.
(EE) [mi] EQ overflow continuing. 100 events have been dropped.
==>>===================================================================

That then counts in steps of 100 up to 1000 and then the machine is frozen locally. Different to other bug reports on similar issues, it appears here that libnss_nis.so.2 and libnss_compat.so.2 are involved (user login is via yp).

Any workaround would be appreciated. I'm happy to provide more details if required.

[AG (alf-gerisch)]

Please note: The problem does NOT only apply to nvidia drivers, it might not even be related to a graphics driver at all!.

[Launchpad Janitor (janitor)]

Status changed to 'Confirmed' because the bug affects multiple users.

[Nick (q-nick)]

I'm running Ubuntu 16.04 with Nvidia 367.57, but still having this EQ overflow...

[Nick (q-nick)]

I don't agree the problem does not only apply to Nvidia drivers, because it seems that using Nouveau can't trigger this EQ overflow here.

[Nick (q-nick)]

https://bugs.launchpad.net/ubuntu/+source/nvidia-graphics-drivers-367/+bug/1665065