Random Reboots after last Kernel Update

Sylvain Rouillard RouillardSy at yahoo.fr
Mon Apr 25 12:44:17 UTC 2005


1253 is just fine for me, stability-wise, as were all its predecessors. 
However, I use none of the applications you mentioned. The ones I use do not 
seem to produce any kind of oddities like you describe. As for 1258 and 1261, 
the instability that I observe does not seem to be related to any software, 
since the random reboots/freeze really seem to be *ahum*, random, i.e. not 
related to anything noticeable.

Le Lundi 25 Avril 2005 14:41, Ken Nordquist a écrit :
> Every kernel after 1240 gives me the udev error mentioned in other posts
> (at boot).  I have noticed other differences in the post-1240 kernel
> (for x86_64).  First, Yahoo! chat rooms are set up to have 50 people in
> there.  After starting Gaim and going into my favorite geek room, I
> noticed that the number kept rising...  so that while there were about
> 35 actually in the room (I asked one of the other people in the room),
> the number on Gaim was showing 60+.  I rolled the kernel back to 1240
> last night, put myself in a chat room and went to bed, now that I am up,
> the correct number is still showing on Gaim.
>
> Also, the higher the kernel number, the higher the odds that Firefox
> will hog CPU time and eventually freeze my computer.  The higher the
> kernel number, the less stable Jedit becomes (java-based IDE) - it will
> not even start with kernel 1261(segmentation faults...  even when
> starting the program as root).  Also, with the last kernel, Thunderbird
> does not recognize any settings...  while it does send and retrieve
> mail, I cannot make any changes to my accounts, etc...
>
> Is anyone else using x86_64 experiencing problems like this?
>
> Regards,
>
> Ken Nordquist
>
> Sylvain Rouillard wrote:
> >I started a similar discussion under
> >Re: missing dependency "libgnutls.so.11(GNUTLS_REL_1_0_9)"
> >
> >We can surely continue it here as this surely deserves its own thread.
> >
> >For the record, that is four 64b processor users hit by the problem so
> > far.
> >
> >Sylvain





More information about the test mailing list