nV News Forums

 
 

nV News Forums (http://www.nvnews.net/vbulletin/index.php)
-   NVIDIA Linux (http://www.nvnews.net/vbulletin/forumdisplay.php?f=14)
-   -   Lots of NVRM: os_raise_smp_barrier(), invalid context! in dmesg (http://www.nvnews.net/vbulletin/showthread.php?t=155176)

Rynor 09-16-10 09:40 AM

Lots of NVRM: os_raise_smp_barrier(), invalid context! in dmesg
 
Hello,

I'm getting a lot of NVRM: os_raise_smp_barrier(), invalid context! messages in dmesg, the time between the messages varies from 5 - 30 seconds apart.

I haven't noticed any problems, except my PC did freeze once completely, no idea why but as I'm running Ubuntu 10.10 beta it might've been something else entirely.

I'm currently using the x86_64 260.19.06 driver with a GeForce GT 240.

Has anyone else seen these messages or does anyone have a clue on what is causing them?

KnutJorgen 09-16-10 03:31 PM

Re: Lots of NVRM: os_raise_smp_barrier(), invalid context! in dmesg
 
I also get a lot of NVRM: os_raise_smp_barrier(), invalid context!
but I am using a cuda version of nvidia. I am trying to lern how to use cuda to accelerate vmd.

Shiryaev 09-17-10 12:18 PM

Re: Lots of NVRM: os_raise_smp_barrier(), invalid context! in dmesg
 
Also seen this lines:
Code:

NVRM: os_raise_smp_barrier(), invalid context!
At x86 260.19.04 and 260.19.06 versions with my GeForce gt240m.
I don't know what it is, but don't have any problems.

Extender 09-18-10 01:48 AM

Re: Lots of NVRM: os_raise_smp_barrier(), invalid context! in dmesg
 
I see that in X.log after such freeze
[ 4750.986] (II) Open ACPI successful (/var/run/acpid.socket)
[ 4751.125] (II) NVIDIA(0): Setting mode "1280x1024_75"
[ 4751.381] (II) Power Button: Device reopened after 1 attempts.
[ 4751.411] (II) Power Button: Device reopened after 1 attempts.
[ 4751.431] (II) Genius Optical Mouse: Device reopened after 1 attempts.
[ 4751.451] (II) AT Translated Set 2 keyboard: Device reopened after 1 attempts.
[ 4760.990] [mi] EQ overflowing. The server is probably stuck in an infinite loop.
[ 4760.990]
Backtrace:
[ 4761.047] 0: /usr/bin/X (xorg_backtrace+0x28) [0x4a0fa8]
[ 4761.047] 1: /usr/bin/X (mieqEnqueue+0x1f4) [0x4a0824]
[ 4761.047] 2: /usr/bin/X (xf86PostMotionEventP+0xc4) [0x47c194]
[ 4761.047] 3: /usr/lib/xorg/modules/input/evdev_drv.so (0x7f6cab594000+0x5399) [0x7f6cab599399]
[ 4761.047] 4: /usr/lib/xorg/modules/input/evdev_drv.so (0x7f6cab594000+0x5a8e) [0x7f6cab599a8e]
[ 4761.047] 5: /usr/bin/X (0x400000+0x6b837) [0x46b837]
[ 4761.048] 6: /usr/bin/X (0x400000+0x11f4e3) [0x51f4e3]
[ 4761.048] 7: /lib/libpthread.so.0 (0x7f6cb1735000+0xfb40) [0x7f6cb1744b40]
[ 4761.048] 8: /usr/lib/xorg/extra-modules/nvidia_drv.so (0x7f6cac0ad000+0xff002) [0x7f6cac1ac002]
[ 4761.048] 9: /usr/lib/xorg/extra-modules/nvidia_drv.so (0x7f6cac0ad000+0xff5c6) [0x7f6cac1ac5c6]
[ 4761.048] 10: /usr/lib/xorg/extra-modules/nvidia_drv.so (0x7f6cac0ad000+0x10bd13) [0x7f6cac1b8d13]
[ 4761.048] 11: /usr/lib/xorg/extra-modules/nvidia_drv.so (0x7f6cac0ad000+0x10ef24) [0x7f6cac1bbf24]
[ 4761.048] 12: /usr/lib/xorg/extra-modules/nvidia_drv.so (0x7f6cac0ad000+0x12e5ee) [0x7f6cac1db5ee]
[ 4761.048] 13: /usr/lib/xorg/extra-modules/nvidia_drv.so (0x7f6cac0ad000+0xfa3e4) [0x7f6cac1a73e4]
[ 4761.048] 14: /usr/lib/xorg/extra-modules/nvidia_drv.so (0x7f6cac0ad000+0x3a2959) [0x7f6cac44f959]
[ 4761.048] 15: /usr/lib/xorg/extra-modules/nvidia_drv.so (0x7f6cac0ad000+0x3ad411) [0x7f6cac45a411]
[ 4761.048] 16: /usr/bin/X (0x400000+0x2c2d9) [0x42c2d9]
[ 4761.048] 17: /usr/bin/X (0x400000+0x2184b) [0x42184b]
[ 4761.048] 18: /lib/libc.so.6 (__libc_start_main+0xfe) [0x7f6cb06a0d8e]
[ 4761.048] 19: /usr/bin/X (0x400000+0x213d9) [0x4213d9]
[ 4814.009] (II) NVIDIA(0): Setting mode "CRT-1:1280x1024_60@1280x1024+0+0"

Licaon 09-18-10 04:18 PM

Re: Lots of NVRM: os_raise_smp_barrier(), invalid context! in dmesg
 
Quote:

Originally Posted by Extender (Post 2320348)
I see that in X.log after such freeze

can you read http://www.nvnews.net/vbulletin/showthread.php?t=46678 and generate a good report with the debug info as instructed after the issue appears ?

/LE: ignore my erroneous driver mistake, nvidia_drv is the right name

Extender 09-19-10 01:17 AM

Re: Lots of NVRM: os_raise_smp_barrier(), invalid context! in dmesg
 
1 Attachment(s)
After virtual console switch (Ctrl-Alt-F1 => Ctrl-Alt-А7) that was added to syslog:
Sep 19 13:12:15 max-desktop kernel: [17797.690629] NVRM: os_schedule: Attempted to yield the CPU while in atomic or interrupt context
Sep 19 13:12:15 max-desktop kernel: [17797.690629] NVRM: os_schedule: Attempted to yield the CPU while in atomic or interrupt context
Sep 19 13:12:17 max-desktop kernel: [17799.690897] NVRM: os_schedule: Attempted to yield the CPU while in atomic or interrupt context
Sep 19 13:12:17 max-desktop kernel: [17799.690897] NVRM: os_schedule: Attempted to yield the CPU while in atomic or interrupt context
Sep 19 13:12:21 max-desktop login[1721]: pam_unix(login:session): session opened for user max by LOGIN(uid=0)
Sep 19 13:12:26 max-desktop kernel: [17808.570384] NVRM: os_raise_smp_barrier(), invalid context!
Sep 19 13:12:26 max-desktop kernel: [17808.570384] NVRM: os_raise_smp_barrier(), invalid context!
Sep 19 13:12:26 max-desktop kernel: [17808.578937] NVRM: os_raise_smp_barrier(), invalid context!
Sep 19 13:12:26 max-desktop kernel: [17808.578937] NVRM: os_raise_smp_barrier(), invalid context!
Sep 19 13:12:38 max-desktop acpid: client 1066[0:0] has disconnected
Sep 19 13:12:38 max-desktop acpid: client 1066[0:0] has disconnected
Sep 19 13:12:38 max-desktop acpid: client connected from 1066[0:0]
Sep 19 13:12:38 max-desktop acpid: 1 client rule loaded
Sep 19 13:12:38 max-desktop acpid: client connected from 1066[0:0]
Sep 19 13:12:38 max-desktop acpid: 1 client rule loaded
Sep 19 13:13:53 max-desktop kernel: [17895.586900] NVRM: os_raise_smp_barrier(), invalid context!
Sep 19 13:13:53 max-desktop kernel: [17895.586900] NVRM: os_raise_smp_barrier(), invalid context!
Sep 19 13:13:53 max-desktop kernel: [17895.602244] NVRM: os_raise_smp_barrier(), invalid context!
Sep 19 13:13:53 max-desktop kernel: [17895.602244] NVRM: os_raise_smp_barrier(), invalid context!


Logs attached.
After usual 30-sec freeze system is working correctly.
I use GT240 card and Ubuntu, driver from PPA maybe this is cause for such driver name (nvidia_drv)

aenertia 09-22-10 03:02 AM

Re: Lots of NVRM: os_raise_smp_barrier(), invalid context! in dmesg
 
1 Attachment(s)
Seeing same error with latest 10.10 updates and xorg-swat ppa

pvelkovski 09-25-10 01:00 AM

Re: Lots of NVRM: os_raise_smp_barrier(), invalid context! in dmesg
 
1 Attachment(s)
Same system, same driver, same messages! My graphic card is nVidia GT 240. I start getting this message, after resuming from suspended mode. This is what I got today after resuming from suspend in dmesg:

....
13535.067316] NVRM: Xid (0001:00): 53, CMDre 00000000 00000080 00000000 00000005 00000005
[13538.099188] NVRM: Xid (0001:00): 53, CMDre 00000001 00000080 00000000 00000005 0000000b
[13538.099550] NVRM: Xid (0001:00): 53, CMDre 00000001 00000080 00000000 00000005 0000000b
[13538.099572] NVRM: Xid (0001:00): 53, CMDre 00000000 00000080 00000000 00000005 00000005
[13549.123506] NVRM: Xid (0001:00): 53, CMDre 00000001 00000080 00000000 00000005 0000000b
[13561.616791] NVRM: os_raise_smp_barrier(), invalid context!
[13561.632968] NVRM: os_raise_smp_barrier(), invalid context!
[13568.793219] NVRM: os_schedule: Attempted to yield the CPU while in atomic or interrupt context
[13694.598487] NVRM: os_raise_smp_barrier(), invalid context!
[13694.601096] NVRM: os_raise_smp_barrier(), invalid context!

I tried reporting a bug using the bug reporting email, they never answered.

drdoom 09-25-10 06:36 AM

Re: Lots of NVRM: os_raise_smp_barrier(), invalid context! in dmesg
 
Until recently I got the same "NVRM: os_raise_smp_barrier(), invalid context!" messages,
but exclusively in a 2.6.35.x-kernel | x<=4 (2.6.34 works).

My setup:
GT220
Xorg 1.7.7

I tried several nvidia-drivers, but it always ended in a total X-freeze.

This week I compiled the latest 2.6.35.5-kernel (I also changed some kernelconfigs.)
and got no X-freezes so far.
(There is a drm-related commit in 2.6.35.5.)

drdoom 09-25-10 12:45 PM

Re: Lots of NVRM: os_raise_smp_barrier(), invalid context! in dmesg
 
Here is the actual commit to the bug.
http://git.kernel.org/?p=linux/kerne...03a89f7becf47c

Extender 09-26-10 02:59 AM

Re: Lots of NVRM: os_raise_smp_barrier(), invalid context! in dmesg
 
I have freeze and instability only after 1-2 suspends. After clean reboot, all is fine.

Will check with 2.6.35.5 from kernel PPA.

waperboy 09-27-10 01:29 AM

Re: Lots of NVRM: os_raise_smp_barrier(), invalid context! in dmesg
 
Quote:

Originally Posted by Extender (Post 2320348)
I see that in X.log after such freeze
[ 4760.990] [mi] EQ overflowing. The server is probably stuck in an infinite loop.
[ 4760.990]
Backtrace:
[ 4761.047] 0: /usr/bin/X (xorg_backtrace+0x28) [0x4a0fa8]
[ 4761.047] 1: /usr/bin/X (mieqEnqueue+0x1f4) [0x4a0824]
[ 4761.047] 2: /usr/bin/X (xf86PostMotionEventP+0xc4) [0x47c194]
[ 4761.047] 3: /usr/lib/xorg/modules/input/evdev_drv.so (0x7f6cab594000+0x5399) [0x7f6cab599399]
[ 4761.047] 4: /usr/lib/xorg/modules/input/evdev_drv.so (0x7f6cab594000+0x5a8e) [0x7f6cab599a8e]
[ 4761.047] 5: /usr/bin/X (0x400000+0x6b837) [0x46b837]
[ 4761.048] 6: /usr/bin/X (0x400000+0x11f4e3) [0x51f4e3]
[ 4761.048] 7: /lib/libpthread.so.0 (0x7f6cb1735000+0xfb40) [0x7f6cb1744b40]
[ 4761.048] 8: /usr/lib/xorg/extra-modules/nvidia_drv.so (0x7f6cac0ad000+0xff002)
...

I'm getting this too, but only occasionally now.
It has gone from perfect, to very unstable, to less unstable like this:

1. Perfect:
- Asus P5W DH Deluxe motherboard + Intel C2D E6600 + 8800GT
- Ubuntu 9.10, 2.6.31-20
- Nvidia driver: 185.18.36
- No gfx issues whatsoever

2. Very unstable:
- New motherboard Asus P6T SE + Intel i7 930 (same 8800GT as above)
- Same Ubuntu installation + driver on same disk as above
- Very unstable, running glxgears freezes up in a matter of seconds, otherwise system freezes in 5-10 minutes

3. Less unstable:
- Upgraded above installation to Ubuntu 10.04
- Now runs glxgears fine, but can freeze up with same symptoms after some hours, sooner if watching video

So the change of MB+CPU made this happen, suggesting some hardware incompatibility issues with nvidia driver, or (possibly) kernel/X.

This is a getting to be a showstopper, I spend long hours working on the computer, and having the system randomly freeze up is unacceptable.

EDIT: Strangely in backtrace, it looks like the nvidia driver causing X mouse-move events (or perhaps it's the other way around, not used to reading these backtraces)


All times are GMT -5. The time now is 02:08 PM.

Powered by vBulletin® Version 3.7.1
Copyright ©2000 - 2015, Jelsoft Enterprises Ltd.
Copyright 1998 - 2014, nV News.