compiz team mailing list archive

[Alex Munroe <737454@xxxxxxxxxxxxxxxxxx>]

I suspect I'm having the same problem.

I have an 8800 GTS and two 1680x1050 monitors running with TwinView.  I
upgraded my nvidia driver to 270.29 a few days ago (in an attempt to
resolve other Compiz issues, actually).  Since then, X will peg a core
for a few seconds, once a minute or so.  Happens even if I sit back and
watch an empty workspace.

What makes this particularly bad is that nothing redraws while X is
spinning its wheels; keyboard input is recognized after the few seconds
have passed, but the display is temporarily frozen and the cursor
doesn't move.

I see several instances of the following stack trace in Xorg.0.log:

[507156.487] [mi] EQ overflowing. The server is probably stuck in an infinite loop.
[507156.487] 
Backtrace:
[507156.487] 0: /usr/bin/X (xorg_backtrace+0x28) [0x45c5a8]
[507156.487] 1: /usr/bin/X (mieqEnqueue+0x1f4) [0x458ea4]
[507156.487] 2: /usr/bin/X (xf86PostMotionEventP+0xc4) [0x488d74]
[507156.487] 3: /usr/lib/xorg/modules/input/evdev_drv.so (0x7fe7d390d000+0x5319) [0x7fe7d3912319]
[507156.487] 4: /usr/lib/xorg/modules/input/evdev_drv.so (0x7fe7d390d000+0x5a0e) [0x7fe7d3912a0e]
[507156.487] 5: /usr/bin/X (0x400000+0x6d4e7) [0x46d4e7]
[507156.487] 6: /usr/bin/X (0x400000+0x1161a3) [0x5161a3]
[507156.487] 7: /lib/libpthread.so.0 (0x7fe7d9cd7000+0xfb40) [0x7fe7d9ce6b40]
[507156.487] 8: /usr/lib/xorg/extra-modules/nvidia_drv.so (0x7fe7d443b000+0xd9bb8) [0x7fe7d4514bb8]
[507156.487] 9: /usr/lib/xorg/modules/libwfb.so (wfbBlt+0x136) [0x7fe7d4001766]
[507156.487] 10: /usr/lib/xorg/modules/libwfb.so (wfbCopyNtoN+0x2bf) [0x7fe7d400621f]
[507156.487] 11: /usr/bin/X (miCopyRegion+0xa5) [0x55c2a5]
[507156.487] 12: /usr/bin/X (miDoCopy+0x43a) [0x55c93a]
[507156.487] 13: /usr/lib/xorg/modules/libwfb.so (wfbCopyArea+0x4c) [0x7fe7d400563c]
[507156.487] 14: /usr/lib/xorg/extra-modules/nvidia_drv.so (0x7fe7d443b000+0x40fb62) [0x7fe7d484ab62]
[507156.487] 15: /usr/bin/X (0x400000+0xdf664) [0x4df664]
[507156.487] 16: /usr/bin/X (0x400000+0x3e599) [0x43e599]
[507156.487] 17: /usr/bin/X (0x400000+0x3f979) [0x43f979]
[507156.487] 18: /usr/bin/X (0x400000+0x2187b) [0x42187b]
[507156.487] 19: /lib/libc.so.6 (__libc_start_main+0xfe) [0x7fe7d8c42d8e]
[507156.487] 20: /usr/bin/X (0x400000+0x21409) [0x421409]

...but there's definitely not such a crash for each temporary freeze.
I've had at least a dozen just typing this, and there are only five
traces since I switched to compiz.

10.10, compiz 0.8.6, default plugins plus ADD Helper, window previews,
and application switcher.  Nothing 3D-intensive running or been run
since boot; mainly terminals, gvim, Thunderbird 3.1, Firefox 4.  No such
problem when using metacity.

Going to give nouveau a shot.

-- 
You received this bug notification because you are a member of compiz
packagers, which is subscribed to compiz in Ubuntu.
https://bugs.launchpad.net/bugs/737454

Title:
  xorg high cpu usage with compiz while using top command; goes away
  when compiz is killed