System freeze in Ubuntu and Fedora

Bug #737429 reported by Diciannove
14
This bug affects 2 people
Affects Status Importance Assigned to Milestone
nwfermi
Confirmed
High
djp

Bug Description

In stress conditions i get system freeze in Ubuntu 10.10 and Fedora 14.
Now I have Ubuntu 10.10 with kernel 2.6.35-27 and testing Firefox 3.6.15 with "Grab and Drag" 2.7.6 Add-on on LXDE Desktop Environment.
No usefull messages.

Revision history for this message
Diciannove (tech-diciannove) wrote :
Revision history for this message
Diciannove (tech-diciannove) wrote :

I try with new kernel released (2.6.35-28) and the freeze still happen.
Nothing in messages or syslog :(

Revision history for this message
Nordheimer Michael (nordheimerm) wrote :

I have the same problem...

user@user-ET2010AG:~$ uname -a
Linux user-ET2010AG 2.6.35-28-generic-pae #49-Ubuntu SMP Tue Mar 1 14:58:06 UTC 2011 i686 GNU/Linux
user@user-ET2010AG:~$ lsusb
Bus 007 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 006 Device 002: ID 1926:0083 NextWindow 1950 HID Touchscreen
Bus 006 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 005 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 004 Device 002: ID 04f2:0860 Chicony Electronics Co., Ltd
Bus 004 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 003 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 002 Device 002: ID 1164:3edc YUAN High-Tech Development Co., Ltd
Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 001 Device 003: ID 0bda:5801 Realtek Semiconductor Corp.
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub

Revision history for this message
Diciannove (tech-diciannove) wrote :

Same with noapic, nolapic, acpi=off boot parameters and latest kernel 2.6.38.

Revision history for this message
Nordheimer Michael (nordheimerm) wrote :

i can add:
checked nwfermi-5.1 on kubuntu-10.10 amd64 freezes occurred more frequently

Revision history for this message
Diciannove (tech-diciannove) wrote :

djp,
can you help me to debug and give you some useful messages?
Thanks

Revision history for this message
WebNuLL (babciastefa) wrote :

I had X Freeze two times, but i dont know its caused by NWFermi.

My freezes looks: No mouse work, keyboard too, the screen was not refreshing - loging in by SSH and killing X with SIGKILL helped, restarted X and all worked fine.

Revision history for this message
Serge Lages (serge-lages) wrote :

I also experienced these freezes, but I modified 70-nw-fermi.conf by setting UseTouchHelp to 0 and setting big values to RightClickTimeout and DoubleClickTimeout (like 1000000000).

I didn't had enough time to really test these changes but my first test was encouraging. Can you test it ?

Revision history for this message
Nordheimer Michael (nordheimerm) wrote :

>I also experienced these freezes, but I modified 70-nw-fermi.conf by setting UseTouchHelp to 0 and setting big values to >RightClickTimeout and DoubleClickTimeout (like 1000000000).

>I didn't had enough time to really test these changes but my first test was encouraging. Can you test it ?

I've tested this settings - everything is perfect (2nd day of testings) except of the right click option which is disabled... Freezes are gone, for now at least...

Revision history for this message
Olaf Jansen-Olliges (olaf-orb-net) wrote :

Hi - i had theese freezes too with 0.50 and 0.51 the above (big values for RightClickTimeout and DoubleClickTimeout & UseTouchHelp = 0 ) seems to avoid the problem.

atm i'm running gentoo with kernel 2.6.37-rc1 since 2 days without any trouble. Is there Anything i should test or enable to debug this issue?
Freezes looked similar to WebNuLL: machine reacheable via ssh but no mouse/ no keyboard / no screen refresh and unfortunately nothing in any log

I'd really like to get the right mouse button back...

Olaf

BTW thx a lot for this driver :-)

Revision history for this message
djp (djpnewton) wrote :

Thank you for the feedback this should be fixed in the next release.

Changed in nwfermi:
importance: Undecided → High
assignee: nobody → djp (djpnewton)
status: New → Confirmed
Revision history for this message
Olaf Jansen-Olliges (olaf-orb-net) wrote :

Hi,

i've tested 0.52 without the RightClickTimeout/DoubleClickTimeOut & UseTouchHelp=0 set and the freezes are still there. Is there anything i can do to help debugging this issue?

Best
  Olaf

Revision history for this message
djp (djpnewton) wrote :

This issue may be fixed by the new 0.6 release

Revision history for this message
Olaf Jansen-Olliges (olaf-orb-net) wrote :

Hi Daniel,

i gave it a try and got a freeze after a few seconds of usage ...

with this as the last lines in Xorg.0.log:

[ 27.050] (II) config/udev: Adding input device Nextwindow Fermi Touchscreen (/dev/input/event5)
[ 27.050] (**) Nextwindow Fermi Touchscreen: Applying InputClass "evdev pointer catchall"
[ 27.050] (**) Nextwindow Fermi Touchscreen: Applying InputClass "nw fermi"
[ 27.050] (II) LoadModule: "nextwindow"
[ 27.050] (II) Loading /usr/lib64/xorg/modules/input/nextwindow_drv.so
[ 27.050] (II) Module nextwindow: vendor="X.Org Foundation"
[ 27.050] compiled for 1.9.4, module version = 0.0.2
[ 27.050] Module class: X.Org XInput Driver
[ 27.050] ABI class: X.Org XInput driver, version 11.0
[ 27.050] (**) Option "Device" "/dev/input/event5"
[ 27.050] (II) Nextwindow Fermi Touchscreen: Using device /dev/input/event5.
[ 27.051] (**) Option "UseTouchHelp" "1"
[ 27.051] (II) Nextwindow Fermi Touchscreen: Using touch help.
[ 27.051] (**) Option "DragThreshold" "10"
[ 27.051] (**) Option "RightClickTimeout" "1200"
[ 27.051] (**) Option "DoubleClickTimeout" "500"
[ 27.051] (**) Nextwindow Fermi Touchscreen: always reports core events
[ 27.060] (II) XINPUT: Adding extended input device "Nextwindow Fermi Touchscreen" (type: TOUCHSCREEN)
[ 27.060] (II) Nextwindow Fermi Touchscreen: On.
[ 27.064] (II) config/udev: Adding input device Nextwindow Fermi Touchscreen (/dev/input/mouse1)
[ 27.064] (II) No input driver/identifier specified (ignoring)

then i re-fetched the source for "nextwindow_drv.so" and recompiled it.

[ 26.996] (II) Module nextwindow: vendor="X.Org Foundation"
[ 26.996] compiled for 1.9.5, module version = 0.0.2
[ 26.996] Module class: X.Org XInput Driver
[ 26.996] ABI class: X.Org XInput driver, version 11.0

with this module i am running my machine now. WITHOUT freezes up to now. I hope this will stay stable... I'll report back after testing a little bit.

Many many thanks again...

Bests Olaf

Revision history for this message
Olaf Jansen-Olliges (olaf-orb-net) wrote :

Hi again,

machine is frozen again :-( - any ideas/suggestions

Bye
  Olaf

Revision history for this message
djp (djpnewton) wrote : Re: [Bug 737429] Re: System freeze in Ubuntu and Fedora

Perhaps we can start by making sure if the freeze is caused by
nextwindow_drv.so.

When the desktop freezes is the whole system frozen (ie can you still ssh in
and kill X)?

If you disable the X driver (say remove
/usr/share/X11/xorg.conf.d/70-nw-fermi.conf and restart X) do you still get
the freeze?

On Wed, Jun 8, 2011 at 3:56 AM, Olaf Jansen-Olliges <
<email address hidden>> wrote:

> Hi again,
>
> machine is frozen again :-( - any ideas/suggestions
>
> Bye
> Olaf
>
> --
> You received this bug notification because you are a bug assignee.
> https://bugs.launchpad.net/bugs/737429
>
> Title:
> System freeze in Ubuntu and Fedora
>
> Status in Nextwindow Fermi Driver:
> Confirmed
>
> Bug description:
> In stress conditions i get system freeze in Ubuntu 10.10 and Fedora 14.
> Now I have Ubuntu 10.10 with kernel 2.6.35-27 and testing Firefox 3.6.15
> with "Grab and Drag" 2.7.6 Add-on on LXDE Desktop Environment.
> No usefull messages.
>

Revision history for this message
Olaf Jansen-Olliges (olaf-orb-net) wrote :

Hi,

the machine is reachable via ssh. But it is not possible to kill X. I tried this several times the effect is somehow strange. When theese freezes occur I still can do things on the machine but if i try to shutdown the machine the whole system freezes. Then only the powerbutton helps. (SysRq: Sync or reBoot has no effect either) This "deep-freeze" can be forced by killing X prozesses ...

I will try disabling the nextwindow_drv this afternoon when i'm back home... - i'll switch from kde to a simple window manager (twm) to get rid of as many side effects as possible.

THX for the feedback
   Olaf

Revision history for this message
Olaf Jansen-Olliges (olaf-orb-net) wrote :

Hi again,

without the nextwindow_drv no freezes the whole afternoon. Stressed the machine with several glxgears + mplayer running.

Bests Olaf

Revision history for this message
Tii (tii) wrote :

Hello,

I'm experiencing Freeze also on my computer !

I have a Asus eeeTop ET2010 on Ubuntu 11.04 x64 with default ubuntu driver in 1600x900 resolution.

I tried nw-fermi 0.6 and 0.6.1 with xf86-input-nextwindow for natty and oneiric. (each with each > 4 tries) and always the same result :
The computer boot perfect, the touchscreen works perfect and is calibrated.
I can use the computer for hours without freeze if I don't touch the screen but when I start using the touchscreen, it freezes completely after 10 to 30 seconds, it always freezes when I'm touching it.
If I uninstall the nwfermi, the computer never freezes when touching.

The freeze is complete (screen not refreshing, no keyboard, no mouse) I have to hard shutdown the computer !

I thought at one point that it could be related to the fglrx ATI driver so I removed it but problem still occurs with the default driver (which works better than the proprietary driver by the way !)

hope I gave enough details, any suggestion ?

thank you !

Revision history for this message
Nordheimer Michael (nordheimerm) wrote :

For now there is only one workaround:

comment 8: https://bugs.launchpad.net/nwfermi/+bug/737429/comments/8
UseTouchHelp to 0 and setting big values to >RightClickTimeout and DoubleClickTimeout (like 1000000000).

Revision history for this message
Nordheimer Michael (nordheimerm) wrote :

it's been 4 months since last post
Looks like project is dead?..

will this bug get fixed?

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Bug attachments

Remote bug watches

Bug watches keep track of this bug in other bug trackers.