Re: many crashes of X server, what should I investigate?

[Frédéric Bron]

> The first thing that springs to mind is:  Had the first overheating
> cooked the CPU heatsink goo, and now the chip can't cool down properly.

I manage to go back to normal by unplugging the power unit and
removing the battery pack. It is probably a bug in the BIOS. I am
looking for an update.
However the problem with the X server remains...
Thanks,
Frédéric
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: many crashes of X server, what should I investigate?

[Tim]

Allegedly, on or about 11 January 2013, Frédéric Bron sent:
> Yesterday, the X server crashed after closing a game (Extreme Tux
> racer in full screen mode) -> black screen + CPU at 100%. -> had to
> hard power off.
> Now the fan is blocked at maximum speed and I do not know how to go
> back to normal state.

The first thing that springs to mind is:  Had the first overheating
cooked the CPU heatsink goo, and now the chip can't cool down properly.

-- 
[tim@localhost ~]$ uname -rsvp
Linux 3.6.10-2.fc17.x86_64 #1 SMP Tue Dec 11 18:07:34 UTC 2012 x86_64

All mail to my mailbox is automatically deleted, there is no point
trying to privately email me, I will only read messages posted to the
public lists.



-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: many crashes of X server, what should I investigate?

[Frédéric Bron]

I have looked at the BIOS and found no option to activate the Intel
video card instead of the NVidia one.
Yesterday, the X server crashed after closing a game (Extreme Tux
racer in full screen mode) -> black screen + CPU at 100%. -> had to
hard power off.
Now the fan is blocked at maximum speed and I do not know how to go
back to normal state.
It is really really painfull and I do not know what to look at to help
solve the issue.

Frédéric
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: many crashes of X server, what should I investigate?

[Kernel Guardian]

On 2 January 2013 08:42, Frédéric Bron  wrote:

> >> I have many crashes of the X server:
> >> - freezing when switching between user
> >> - freezing when starting a session
> >
> > What kind of machine are you running F17 on?  What processor?  How much
> > memory?  What video card?
>
> I have a laptop with:
> Processor Intel® Core I7 - I7-2760QM - 2.40 GHz / 3.50 GHz (Max Turbo
> Freq.) - (Quad-core mobile - Socket FCPGA988 - 6 Mo cache L2 - 32 nm -
> 45 W - 64-bitsVirtualisation - Hyperthreading (8 threads) - integrated
> * HD Graphics 3000)*
> LED screen 17.3" Full-HD 1920 x 1080 TFT
> 16 Go SO-DIMM RAM DDR-3 / 1333 MHz (4 x 4 Go)
> * Video controler nVidia GeForce GTX 560M with 1.5 Go dedicated VRAM*


Looks like your laptop equipped with Optimus chipset. If you are have
option in BIOS to disable nVidia that would be one of the solutions.
More info you can find
there
. I
recently read that it is possible to use Optimus with Bumblebee, but I'm
not sure if it was an F17 or F18.


>
> DDR-5 (Support PCIe x16, DirectX 11, nVidia PhysX, HDMI 1.4, HDCP,
> GeForce CUDA)
> Chipset Intel HM67 2nd generation
> Technology PCI-Express
> Main bus (FSB) 1333 MHz
> Technology Intel® for processors Intel® Core I7 and Core I7 Extreme
> Edition Sandy Bridge 2nd generation (64-bits mobile)
> Support PCI-Express x16 et DirectX 11
> Graphical accelerator 2D/3D at 128 bits, decoding MPEG I & II, DVD-playback
>
> I have installed F17 64 bits: Linux mymachine 3.6.10-2.fc17.x86_64 #1
> SMP Tue Dec 11 18:07:34 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux
>
> >  Do you have it overclocked?
>
> No
>
> > Are you using the
> > default video driver that comes with F17 or did you install the
> proprietary
> > video driver issued by the video chip manufacturer?
>
> I am using the nvidia proprietary driver (I used kmod-nvidia and
> switched to akmod-nvidia recently because the 3.6.10 driver did not
> come but this did not change anything).
> I have read a thread that reports the same issue on gdm bug tracker
> and someone tried both nvidia and nouveau with no difference but I
> have not done it myself.
>
> > Do you have another machine networked to this machine?  If you do you can
> > open a terminal session on this machine using ssh.  You might find some
> > clues in dmesg
>
> I will try that. I have just tested that I can connect by ssh and it works
> fine.
> Halas, I could not reproduce the freeze but I got only a small freeze
> (30 sec.) and dmesg during that freeze gave:
>
> [62645.848785] SELinux: initialized (dev fuse, type fuse), uses
> genfs_contexts
> [62906.805823] SELinux: initialized (dev fuse, type fuse), uses
> genfs_contexts
> [62965.449065] nepomukservices[26117]: segfault at 0 ip
> (null) sp 7fff6096e428 error 14 in nepomukservicestub[40+7000]
> [63022.645742] NVRM: GPU at :01:00:
> GPU-a26ca184-dbb8-cb76-a255-3c1db789ce86
> [63022.645758] NVRM: Xid (:01:00): 31, Ch 0002, engmask
> 0101, intr 1000
> [63023.160005] NVRM: Xid (:01:00): 8, Channel 0004
>
> I see also that after trying to open lots of sessions and close them,
> my 16 Gb memory is full and I had difficulties openning a new terminal
> because it said "bash fork retry no child processes". I do not know if
> it is linked but I read somewhere that when X freezes, it may be due
> to memory leaks...
>
> I will come back later with more info if a new crash happen.
>
> Thanks,
>
> Frédéric
> --
> users mailing list
> users@lists.fedoraproject.org
> To unsubscribe or change subscription options:
> https://admin.fedoraproject.org/mailman/listinfo/users
> Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
> Have a question? Ask away: http://ask.fedoraproject.org
>
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: many crashes of X server, what should I investigate?

[Roberto Ragusa]

On 01/05/2013 09:23 PM, Frédéric Bron wrote:
>> You also have the integrated Intel gfx stuff.
>> Did you try using that instead of nvidia? There could be an option
>> in your BIOS to modify.
>>
>> (hardware similar to yours, everything fine for me on F16 and Intel gfx)
> 
> You mean that I should not use the nvidia card but the Intel one
> integrated in the CPU, is that true? You mean I have in fact two cards
> and maybe the simple Intel one could be sufficient?
> I will have a look in the BIOS to see if I can select the active video
> card but it would be a bit sad to have nvidia and not use it, wouldn't
> it?
> Thanks for you help,
> Frédéric

Yes, you said you have a
  integrated HD Graphics 3000
which is an additional gfx hardware.

In some laptops the BIOS has a switch (it could be called "nvidia optimus").

The Intel card is less powerful (while still fast enough for non gfx critical
applications), but it could possibly solve your problems which appear to be
related to nouveau and the nvidia binary driver.

And you could gain some battery duration too, as nvidia will be turned off.

-- 
   Roberto Ragusamail at robertoragusa.it
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: many crashes of X server, what should I investigate?

[Frédéric Bron]

> You also have the integrated Intel gfx stuff.
> Did you try using that instead of nvidia? There could be an option
> in your BIOS to modify.
>
> (hardware similar to yours, everything fine for me on F16 and Intel gfx)

You mean that I should not use the nvidia card but the Intel one
integrated in the CPU, is that true? You mean I have in fact two cards
and maybe the simple Intel one could be sufficient?
I will have a look in the BIOS to see if I can select the active video
card but it would be a bit sad to have nvidia and not use it, wouldn't
it?
Thanks for you help,
Frédéric
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: many crashes of X server, what should I investigate?

[Roberto Ragusa]

On 01/02/2013 08:42 AM, Frédéric Bron wrote:

> I am using the nvidia proprietary driver (I used kmod-nvidia and
> switched to akmod-nvidia recently because the 3.6.10 driver did not
> come but this did not change anything).
> I have read a thread that reports the same issue on gdm bug tracker
> and someone tried both nvidia and nouveau with no difference but I
> have not done it myself.

You also have the integrated Intel gfx stuff.
Did you try using that instead of nvidia? There could be an option
in your BIOS to modify.

(hardware similar to yours, everything fine for me on F16 and Intel gfx)

-- 
   Roberto Ragusamail at robertoragusa.it
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: many crashes of X server, what should I investigate?

[Frédéric Bron]

It just froze again after user switch. I could connect from another
computer via ssh.
What happens is that the X server is taking 100% of CPU while doing
nothing. Here is the result of top:

top - 10:24:48 up  2:23,  5 users,  load average: 0.96, 0.63, 0.44
Tasks: 296 total,   2 running, 294 sleeping,   0 stopped,   0 zombie
Cpu(s): 12.6%us,  0.0%sy,  0.0%ni, 87.2%id,  0.0%wa,  0.0%hi,  0.0%si,  0.0%st
Mem:  16412940k total,  4394816k used, 12018124k free,   353460k buffers
Swap: 20971516k total,0k used, 20971516k free,  1703748k cached

  PID USER  PR  NI  VIRT  RES  SHR S %CPU %MEMTIME+  COMMAND
 2509 root  20   0  197m  81m  20m R 100.1  0.5   3:39.16 X

Here is the end of dmesg (full log attached):

[9.767452] RPC: Registered tcp NFSv4.1 backchannel transport module.
[   11.095246] NVRM: Your system is not currently configured to drive
a VGA console
[   11.095249] NVRM: on the primary VGA device. The NVIDIA Linux graphics driver
[   11.095250] NVRM: requires the use of a text-mode VGA console. Use
of other console
[   11.095251] NVRM: drivers including, but not limited to, vesafb,
may result in
[   11.095252] NVRM: corruption and stability problems, and is not supported.
[   11.272563] jme :03:00.0: p1p1: Link is up at ANed: 100 Mbps,
Full-Duplex, MDI-X
[   12.163589] IPv6: ADDRCONF(NETDEV_CHANGE): p1p1: link becomes ready
[   18.928156] fuse init (API version 7.20)
[   18.937985] SELinux: initialized (dev fuse, type fuse), uses genfs_contexts
[   18.958490] SELinux: initialized (dev fusectl, type fusectl), uses
genfs_contexts
[  397.734049] TCP: lp registered
[  807.384111] SELinux: initialized (dev fuse, type fuse), uses genfs_contexts
[  900.445624] usb 2-1.2: usbfs: interface 1 claimed by usblp while
'xsane' sets config #1
[  911.880596] usb 2-1.2: usbfs: interface 1 claimed by usblp while
'xsane' sets config #1
[ 1514.963116] SELinux: initialized (dev fuse, type fuse), uses genfs_contexts
[ 5379.310344] nepomukservices[4370]: segfault at 0 ip
(null) sp 7fff23a794e8 error 14 in nepomukservicestub[40+7000]
[ 8423.252973] usb 2-1.2: usbfs: interface 0 claimed by usbfs while
'xsane' sets config #1
[ 8428.030674] usb 2-1.2: usbfs: interface 0 claimed by usbfs while
'xsane' sets config #1
[ 8481.203974] NVRM: GPU at :01:00: GPU-a26ca184-dbb8-cb76-a255-3c1db789ce86
[ 8481.203978] NVRM: Xid (:01:00): 38, 0002 902d 
 0100 
[ 8483.259787] NVRM: os_schedule: Attempted to yield the CPU while in
atomic or interrupt context
[ 8485.256822] NVRM: os_schedule: Attempted to yield the CPU while in
atomic or interrupt context
[ 8486.657367] NVRM: Xid (:01:00): 38, 0002  
  
[ 8488.098020] NVRM: Xid (:01:00): 38,   
  

Does this help?
Frédéric


dmesg.log.bz2
Description: BZip2 compressed data
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: many crashes of X server, what should I investigate?

[Frédéric Bron]

>> I have many crashes of the X server:
>> - freezing when switching between user
>> - freezing when starting a session
>
> What kind of machine are you running F17 on?  What processor?  How much
> memory?  What video card?

I have a laptop with:
Processor Intel® Core I7 - I7-2760QM - 2.40 GHz / 3.50 GHz (Max Turbo
Freq.) - (Quad-core mobile - Socket FCPGA988 - 6 Mo cache L2 - 32 nm -
45 W - 64-bitsVirtualisation - Hyperthreading (8 threads) - integrated
HD Graphics 3000)
LED screen 17.3" Full-HD 1920 x 1080 TFT
16 Go SO-DIMM RAM DDR-3 / 1333 MHz (4 x 4 Go)
Video controler nVidia GeForce GTX 560M with 1.5 Go dedicated VRAM
DDR-5 (Support PCIe x16, DirectX 11, nVidia PhysX, HDMI 1.4, HDCP,
GeForce CUDA)
Chipset Intel HM67 2nd generation
Technology PCI-Express
Main bus (FSB) 1333 MHz
Technology Intel® for processors Intel® Core I7 and Core I7 Extreme
Edition Sandy Bridge 2nd generation (64-bits mobile)
Support PCI-Express x16 et DirectX 11
Graphical accelerator 2D/3D at 128 bits, decoding MPEG I & II, DVD-playback

I have installed F17 64 bits: Linux mymachine 3.6.10-2.fc17.x86_64 #1
SMP Tue Dec 11 18:07:34 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux

>  Do you have it overclocked?

No

> Are you using the
> default video driver that comes with F17 or did you install the proprietary
> video driver issued by the video chip manufacturer?

I am using the nvidia proprietary driver (I used kmod-nvidia and
switched to akmod-nvidia recently because the 3.6.10 driver did not
come but this did not change anything).
I have read a thread that reports the same issue on gdm bug tracker
and someone tried both nvidia and nouveau with no difference but I
have not done it myself.

> Do you have another machine networked to this machine?  If you do you can
> open a terminal session on this machine using ssh.  You might find some
> clues in dmesg

I will try that. I have just tested that I can connect by ssh and it works fine.
Halas, I could not reproduce the freeze but I got only a small freeze
(30 sec.) and dmesg during that freeze gave:

[62645.848785] SELinux: initialized (dev fuse, type fuse), uses genfs_contexts
[62906.805823] SELinux: initialized (dev fuse, type fuse), uses genfs_contexts
[62965.449065] nepomukservices[26117]: segfault at 0 ip
(null) sp 7fff6096e428 error 14 in nepomukservicestub[40+7000]
[63022.645742] NVRM: GPU at :01:00: GPU-a26ca184-dbb8-cb76-a255-3c1db789ce86
[63022.645758] NVRM: Xid (:01:00): 31, Ch 0002, engmask
0101, intr 1000
[63023.160005] NVRM: Xid (:01:00): 8, Channel 0004

I see also that after trying to open lots of sessions and close them,
my 16 Gb memory is full and I had difficulties openning a new terminal
because it said "bash fork retry no child processes". I do not know if
it is linked but I read somewhere that when X freezes, it may be due
to memory leaks...

I will come back later with more info if a new crash happen.

Thanks,

Frédéric
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: many crashes of X server, what should I investigate?

[Mark LaPierre]

On 01/01/2013 04:42 PM, Frédéric Bron wrote:

I run F17 with regular updates.
I have many crashes of the X server:
- freezing when switching between user
- freezing when starting a session
Many times, Ctrl+Alt+Backspace does not work, also Ctrl+Alt+F[1-9]
does not work either so that I cannot even start a terminal session to
see what happens. I find it difficult to justify my choice of linux
over windows in front of the familiy...
I would like to investigate but I do not know what to look at. Could
you help? Are there some logs that I can look at after the crash?

I used to run GDM + Gnome sessions. I changed to KDM + Kde but still
have issues.

Thanks for helping,

Regards,

Frédéric


Hey Frédéric,

What kind of machine are you running F17 on?  What processor?  How much 
memory?  What video card?  Do you have it overclocked?  Are you using 
the default video driver that comes with F17 or did you install the 
proprietary video driver issued by the video chip manufacturer?


There are a lot of things that can impact video system performance.

Do you have another machine networked to this machine?  If you do you 
can open a terminal session on this machine using ssh.  You might find 
some clues in dmesg


$ dmesg | tail

You want to set up and test ssh before time so that you will know that 
it works when you need it.  If your machine is hard crashing so that you 
can't log on through ssh, if the keyboard locks up so that caps lock and 
number lock don't work, then life gets more difficult when it comes time 
to diagnose lockups.


--
_
   °v°
  /(_)\
   ^ ^  Mark LaPierre
Registerd Linux user No #267004
https://linuxcounter.net/

--
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

Re: many crashes of X server, what should I investigate?

[Joe Zeff]

On 01/01/2013 01:42 PM, Frédéric Bron wrote:

Many times, Ctrl+Alt+Backspace does not work, also Ctrl+Alt+F[1-9]
does not work either so that I cannot even start a terminal session to
see what happens.


I've been having the same issue, using F16 and Xfce.  I have the Magic 
SysRq keys active, but when this happens they don't work.  I have an 
older system, with the MoBo maxed out at 1 GB RAM, and I've been working 
under the assumption that it's getting a tad too cramped.  Alas, I can't 
currently afford to upgrade so I'm doing the best with what I have.  If 
anybody has any ideas for either of us, please let us know!

--
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org

many crashes of X server, what should I investigate?

[Frédéric Bron]

I run F17 with regular updates.
I have many crashes of the X server:
- freezing when switching between user
- freezing when starting a session
Many times, Ctrl+Alt+Backspace does not work, also Ctrl+Alt+F[1-9]
does not work either so that I cannot even start a terminal session to
see what happens. I find it difficult to justify my choice of linux
over windows in front of the familiy...
I would like to investigate but I do not know what to look at. Could
you help? Are there some logs that I can look at after the crash?

I used to run GDM + Gnome sessions. I changed to KDM + Kde but still
have issues.

Thanks for helping,

Regards,

Frédéric
-- 
users mailing list
users@lists.fedoraproject.org
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org