You are not logged in.
Hi all,
Since an update I did approximately last weekend to KDE 4.6.2-1 packages and nvidia driver 270.41.03-1, I've found that KDE will "freeze" for about 3-5 seconds at a time, seemingly randomly. I'm guessing it's more of a graphics problem as disabling compositing seems to prevent it.
Anyone else found any issues like this? I saw some other threads around KDE pausing when resizing a window, but mine seems to do it even when I'm *not* resizing!
I'm on x64, and am not running testing.
Last edited by spektr (2011-04-25 12:24:52)
Offline
Hi all,
No one having this issue? I can't believe my setup is anything unusual (core i7/930, ASUS p6x58D-E, Nvidia GTX460). Am getting what seems to be some common errors cropping up in logs:
dmesg:
NVRM: Xid (0003:00): 8, Channel 00000001
NVRM: Xid (0003:00): 8, Channel 00000001
NVRM: Xid (0003:00): 8, Channel 00000001
NVRM: Xid (0003:00): 13, 0001 00000000 00009097 00001614 00000000 0000000d
NVRM: Xid (0003:00): 8, Channel 00000001
NVRM: Xid (0003:00): 8, Channel 00000001
NVRM: Xid (0003:00): 8, Channel 00000001
and Xorg.log
33.788] (WW) NVIDIA(0): WAIT (0, 7, 0x8000, 0x0000d674, 0x0000d674)
[ 36.789] (WW) NVIDIA(0): WAIT (0, 7, 0x8000, 0x0000d6c0, 0x0000d6c0)
[ 39.790] (WW) NVIDIA(0): WAIT (2, 7, 0x8000, 0x0000d6c0, 0x0000d70c)
[ 53.586] (WW) NVIDIA(0): WAIT (2, 6, 0x8000, 0x00005df8, 0x0000d608)
[ 60.586] (WW) NVIDIA(0): WAIT (1, 6, 0x8000, 0x00005df8, 0x0000d608)
[ 63.587] (WW) NVIDIA(0): WAIT (2, 7, 0x8000, 0x00005df8, 0x0000e71c)
[ 82.785] (WW) NVIDIA(0): WAIT (0, 6, 0x8000, 0x00009ff0, 0x00009ff0)
[ 90.718] (WW) NVIDIA(0): WAIT (0, 6, 0x8000, 0x000076ec, 0x000076ec)
[ 93.719] (WW) NVIDIA(0): WAIT (2, 6, 0x8000, 0x000076ec, 0x00005f6c)
[ 847.259] (WW) NVIDIA(0): WAIT (2, 6, 0x8000, 0x000028f0, 0x00005728)
[ 854.259] (WW) NVIDIA(0): WAIT (1, 6, 0x8000, 0x000028f0, 0x00005728)
[ 860.912] (WW) NVIDIA(0): WAIT (2, 6, 0x8000, 0x00006180, 0x000066a0)
[ 867.912] (WW) NVIDIA(0): WAIT (1, 6, 0x8000, 0x00006180, 0x000066a0)
[ 870.915] (WW) NVIDIA(0): WAIT (2, 6, 0x8000, 0x00006180, 0x00003aac)
[ 903.116] [mi] EQ overflowing. The server is probably stuck in an infinite loop.
[ 903.116]
Backtrace:
[ 903.117] 0: /usr/bin/X (xorg_backtrace+0x28) [0x49f718]
[ 903.117] 1: /usr/bin/X (mieqEnqueue+0x1f4) [0x49eb14]
[ 903.117] 2: /usr/bin/X (xf86PostMotionEventP+0xc4) [0x47bab4]
[ 903.117] 3: /usr/lib/xorg/modules/input/evdev_drv.so (0x7f0457f28000+0x46aa) [0x7f0457f2c6aa]
[ 903.117] 4: /usr/bin/X (0x400000+0x698e7) [0x4698e7]
[ 903.117] 5: /usr/bin/X (0x400000+0x118a93) [0x518a93]
[ 903.117] 6: /lib/libpthread.so.0 (0x7f045dee9000+0xf750) [0x7f045def8750]
[ 903.117] 7: /usr/lib/xorg/modules/drivers/nvidia_drv.so (0x7f0458a58000+0x6c1f0) [0x7f0458ac41f0]
[ 903.117] 8: /usr/lib/xorg/modules/drivers/nvidia_drv.so (0x7f0458a58000+0x6da71) [0x7f0458ac5a71]
[ 903.118] 9: /usr/lib/xorg/modules/drivers/nvidia_drv.so (0x7f0458a58000+0x7b104) [0x7f0458ad3104]
[ 903.118] 10: /usr/lib/xorg/modules/drivers/nvidia_drv.so (0x7f0458a58000+0x81c47) [0x7f0458ad9c47]
[ 903.118] 11: /usr/lib/xorg/modules/drivers/nvidia_drv.so (0x7f0458a58000+0x3e3d07) [0x7f0458e3bd07]
[ 903.118] 12: /usr/bin/X (0x400000+0xd7073) [0x4d7073]
[ 903.118] 13: /usr/lib/xorg/modules/drivers/nvidia_drv.so (0x7f0458a58000+0x3e2b17) [0x7f0458e3ab17]
[ 903.118] 14: /usr/bin/X (0x400000+0xd2b41) [0x4d2b41]
[ 903.118] 15: /usr/bin/X (0x400000+0x2d201) [0x42d201]
[ 903.118] 16: /usr/bin/X (0x400000+0x2131e) [0x42131e]
[ 903.118] 17: /lib/libc.so.6 (__libc_start_main+0xed) [0x7f045ce6ff6d]
[ 903.118] 18: /usr/bin/X (0x400000+0x20ec9) [0x420ec9]
[ 921.560] (WW) NVIDIA(0): WAIT (2, 6, 0x8000, 0x0000ab40, 0x0000ae58)
[ 951.515] (WW) NVIDIA(0): WAIT (2, 6, 0x8000, 0x00004c48, 0x00006088)
Now, I've taken the advice of a few other threads and downgraded from Xorg 1.10 to 1.09, and downgraded my kernel/nvidia driver to 2.6.37/260.x, but for some reason it still doesn't seem to be co-operating? The only thing that works reliably is turning off compositing (and not using anything GL-related).
Ideas??
Offline
I have tis problem occasionally as well. Its most likely a bug with the nvidia driver.
Offline
I had similar problems, where X froze in my computer in 64 bit Arch Linux only (I can dual boot both 32bit and 64bit). Creating file /etc/X11/xorg.conf.d/use-events.conf with the following content solved the problem:
Section "Device"
Identifier "NVidia 9600 GT"
Option "UseEvents" "false"
EndSection
The hint for this was found in Fedora forum (http://forums.fedoraforum.org/archive/i … 53744.html).
Last edited by timosa (2011-06-26 08:33:17)
Offline
Yes, had it today.
After https://bugs.archlinux.org/task/23381, another bug again
Xserser had frozen for about 60s with CPU busy at 100%.
Then, I found this in Xorg.log :
[ 40010.513] [mi] EQ overflowing. The server is probably stuck in an infinite loop.
[ 40010.513]
Backtrace:
[ 40010.525] 0: /usr/bin/X (xorg_backtrace+0x37) [0x80e8bb7]
[ 40010.525] 1: /usr/bin/X (mieqEnqueue+0x1d1) [0x80e7f41]
[ 40010.525] 2: /usr/bin/X (xf86PostMotionEventM+0xb0) [0x80c3d40]
[ 40010.525] 3: /usr/bin/X (xf86PostMotionEventP+0x6f) [0x80c3e8f]
[ 40010.525] 4: /usr/lib/xorg/modules/input/evdev_drv.so (0xb4b09000+0x2acd) [0xb4b0bacd]
[ 40010.525] 5: /usr/lib/xorg/modules/input/evdev_drv.so (0xb4b09000+0x3908) [0xb4b0c908]
[ 40010.525] 6: /usr/bin/X (0x8048000+0x68de1) [0x80b0de1]
[ 40049.377] [mi] EQ overflowing. The server is probably stuck in an infinite loop.
[ 40049.377]
Backtrace:
[ 40049.377] 0: /usr/bin/X (xorg_backtrace+0x37) [0x80e8bb7]
[ 40049.377] 1: /usr/bin/X (mieqEnqueue+0x1d1) [0x80e7f41]
[ 40049.377] 2: /usr/bin/X (xf86PostMotionEventM+0xb0) [0x80c3d40]
[ 40049.377] 3: /usr/bin/X (xf86PostMotionEventP+0x6f) [0x80c3e8f]
[ 40049.377] 4: /usr/lib/xorg/modules/input/evdev_drv.so (0xb4b09000+0x2acd) [0xb4b0bacd]
[ 40049.377] 5: /usr/lib/xorg/modules/input/evdev_drv.so (0xb4b09000+0x3908) [0xb4b0c908]
[ 40049.377] 6: /usr/bin/X (0x8048000+0x68de1) [0x80b0de1]
I'm going to try 'Option "UseEvents" "false"' ...
Offline
I had similar problems, where X froze in my computer in 64 bit Arch Linux only (I can dual boot both 32bit and 64bit). Creating file /etc/X11/xorg.conf.d/use-events.conf with the following content solved the problem:
Section "Device" Identifier "NVidia 9600 GT" Option "UseEvents" "false" EndSection
The hint for this was found in Fedora forum (http://forums.fedoraforum.org/archive/i … 53744.html).
Thanks, we have the same setup, 9600GT on 64 bit. I've added the file, X reports:
(**) NVIDIA(0): Option "UseEvents" "false"
Let's hope nouveau matures soon so we can ditch the unstable pile of crap Nvidia calls linux drivers
Edit: interesting comment found, apparently UseEvents is now true by default: http://magazine.redhat.com/2007/11/20/t … mment-1926
Last edited by Wilco (2011-07-02 12:17:40)
Offline