[Mageia-discuss] MGA2: Gnome session & system freeze issues

[Alejandro López]

On Fri, Jul 6, 2012 at 9:56 AM, Wolfgang Bornath <molch.b at googlemail.com>wrote:

> MGA2 x86_64, Gnome
> Nvidia graphics, latest proprietary driver from the MGA non-free repo
> All updates.
>
> 1. Gnome starts new session out of the blue
>
> Today I was editing a document in gedit. Other open applications were
> chromium-browser and gkrellm. Suddenly out of the blue in the middle
> of writing a word the screen changed to a new session, screen with GDM
> showing). Logging in again no application was saved, work lost as well
> as 2 tabs in the browser where I did not save the URLs yet. syslog
> shows a Gnome session I/O failure before the session switch:
>
>  - Jul  6 07:28:49 marvin-1 gnome-session[2301]: Gdk-WARNING:
> gnome-session: Fatal IO error 11 (ressource temporarily not available)
> on X server :0.
>
> After session restart a long list of "DEBUG" messages followed. No
> errors reported in that list.
>

For this part of your problem, I guess you are hitting this bug:

https://bugs.mageia.org/show_bug.cgi?id=2585

Alejandro.


>
> 2. Occasional system deep freeze
>
> Starting some time ago and occurring last time yesterday I get a deep
> freeze (only SysRq keys for forced reboot working). Inquiring I found
> this message in syslog:
>
>  - Jul  5 13:27:42 marvin-1 kernel: [10565.322087] NVRM: GPU at
> 0000:01:00.0 has fallen off the bus.
>
> man-draker in the German forum pointed me to
>
> http://www.cyberciti.biz/faq/debian-ubuntu-rhel-fedora-linux-nvidia-nvrm-gpu-fallen-off-bus/
> which talks about persistence mode of nvidia-smi. The persistence mode
> is disabled by default. Switched it to "enabled" but I am not sure if
> this is the true culprit - the same error should have come up more
> often and on other people's machines as well if this setting was the
> culprit, right?
>
> Another idea was offered by doktor5000:
> "It could be some malfunction of systemd and you could ask Colin about
> it". :)
>
> Additional information: the freeze is not caused by faulty memory, I
> let memtest run over nicht for 8 hours, no errors.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: </pipermail/mageia-discuss/attachments/20120706/b8768072/attachment.html>