Hello,

One liner:  I'm getting mysterious (to me), almost random hard freezes 
of the kernel running 2.6.10 and 2.6.11.
Kernel version: Linux version 2.6.11-gentoo-r3 (root@ender) (gcc version 
3.4.2 (Gentoo Linux 3.4.2-r2, ssp-3.4.1-1, pie-8.7.6.5))

Mark Nipper posted a message on March 5 regarding some mysterious kernel 
lockups which he didn't get a response to (I've contacted him about 
it).  Since I'm having what I think is the same problem, I thought I'd 
post a message so he's not just a single lonely voice in the dark.

Mark and I have similar set-ups.  We're both running x86_64 kernels, and 
ReiserFS3.  He's running Debian, I'm running Gentoo.  We haven't 
compared kernel config files yet; it might mean something to him, but to 
be honest, I barely know enough to compile my own kernels and wouldn't 
know where to begin to look for the problem.  Mark has only encountered 
this on 2.6.11, but I don't think he's tried any other kernel versions 
on x86_64; I get this problem on both 2.6.10 and 2.6.11.  I didn't see 
the problem on 2.6.9.

In both of our cases, the kernel is locking up, and requires a power 
cycle to get it back.  We're not able to SSH into our machines, and we 
get no response from any of the input devices.  Furthermore, even with 
full debugging turned on, there are no messages in the log file that 
appear to be related to the lockup.  In my logs, the last message before 
the crash is always (that I've noticed) an ACPI error:

    acpi_thermal-0400 [23] acpi_thermal_get_trip_: Invalid active 
threshold [0]

but this message appears a lot in my logs, so I think it is 
coincidence.  For Mark, the last message was some ReiserFS message.  
Mark feels like the error is ReiserFS related, and I was pretty sure it 
was swap related, until I turned off all swap partitions and the problem 
still occurred.  I *may* try converting all of my filesystems to 
something else if somebody knowledgeable thinks it could be the problem, 
but I'm guessing it is something deeper in; I've never seen a filesystem 
related problem that caused a lock-up like this.

I still feel that this may be memory related.  When I turn off swap, or 
when a drastically reduce my memory use, my laptop can run for hours, or 
even days with little use.  On the other hand, it can freeze up after 
five minutes, even before KDE has finished loading completely, with the 
swap on.  However, I haven't found a situation where it won't, 
eventually, lock up.  But I can't really pin it down, so I don't know 
where the problem is.

I haven't noticed the lockups without X, but I haven't run for any great 
length of time without X.  I'm running the ATI proprietary drivers, but 
I even when I revert to the XOrg ATI drivers (non-proprietary), I still 
get the lockups.

I'm really sorry that I can't provide more information; I'm usually not 
totally incompetent at narrowing down problems in software, but I have 
no idea where to even start looking for the problem here.  If there are 
any things I should try that might provide more information, please let 
me know.

I'm attaching my kernel config, plus all of the info from /proc that is 
suggested by the FAQ to be included.  I'll be happy to recompile my 
kernel with other options, if I can get some hints at starting points; I 
doubt my changing flags at random will help much.

Thanks, in advance.

Sean Russell