X server hangs; reboot required

[shrek-m at gmx.de]

Kevin J. Cummings wrote:

> Vladimir G. Ivanovic wrote:
>
>> I'm disappointed that the random X server hangs are still present in
>> FC3. (I can't find the bug in Bugzilla; I know it's there because I
>> think I commented on it. I wasn't the original filer.)
>>
>> What happens is that X exits abruptly. The screen contains whatever
>> was on it at the time, but of course nothing works. It is possible to
>> log in and start up another X session, but when that session exits,
>> the screen still contains what was on it when X first exited.
>>
>> With FC2 and older kernels, this kind of hang happened a half to a
>> dozen times a day. Now the frequency is down to 1 every other day, or
>> so. 
>
> My biggest complaint with the FC2 X hangs is that the keyboard is 
> always "dead" (no response, can't change VC to a text window) and 
> *most* of the time the mouse is "dead" as well; though I have had a 
> couple of hangs in the middle of actively using X and the mouse motion 
> remains, mouse clicks are not processed; but most have occurred while 
> the system in idle and running a screensaver.  Its obvious because the 
> screensaver stop changing.  The rest of the sysytem continues to run 
> normally.  My http server still runs, I still get email, other 
> processes continue to run.  During the X hangs, I have to login 
> remotely, and top shows that the "X" process is chewing all of the 
> otherwise unused CPU time (usually > 85%), but is recoverable if the 
> "X" process is killed (usually requires a "kill -9").  My frequency 
> under FC2 is 1 hang every day or two.
>
> I'm interested in knowing how these symptoms change in FC3....


nvidia-driver ??

i am using  rh 9, fedora core 1|2|3  20/6/350 = 
hours_the_day/days_per_week/days in_year  without your described 
problems with the NV-driver
the last time i had such problems was in rh 8.0 with the NVIDIA-driver.

-- 
shrek-m