ACPI Kernel issue
Re: ACPI Kernel issue
Tried running memtest; it didn't report any RAM problems.
Re: ACPI Kernel issue
Solution!
I turned off hyperthreading and haven't had the problem since. My uptime is five days, whereas before this, it was usually less than 24 hours.
I turned off hyperthreading and haven't had the problem since. My uptime is five days, whereas before this, it was usually less than 24 hours.
Update for ACPI Kernel issue
On recent kernels, the old max_cstate parameter might have no effect.
I have been using processor.max_cstate=2 for some time. Now, on 2.6.37, I use
processor.max_cstate=2 processor_idle.max_cstate=3 intel_idle.max_cstate=3
(yes, this is overkill) This is related to a new config option that I selected.
However, powertop shows
Your CPU supports the following C-states : C1 C2 C4 C6
Your BIOS reports the following C-states : C1 C2
as if C3 is renamed to C4. (?)
Another hint, if you use a realtime kernel 2.6.33, (a realtime kernel was preinstalled), I recommend an update from git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git branch rt/2.6.33 This has commit e94445b78dc821e0e9de5a9bf5bd263f5ae2c471 with description:
I have been using processor.max_cstate=2 for some time. Now, on 2.6.37, I use
processor.max_cstate=2 processor_idle.max_cstate=3 intel_idle.max_cstate=3
(yes, this is overkill) This is related to a new config option that I selected.
However, powertop shows
Your CPU supports the following C-states : C1 C2 C4 C6
Your BIOS reports the following C-states : C1 C2
as if C3 is renamed to C4. (?)
Another hint, if you use a realtime kernel 2.6.33, (a realtime kernel was preinstalled), I recommend an update from git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git branch rt/2.6.33 This has commit e94445b78dc821e0e9de5a9bf5bd263f5ae2c471 with description:
(Moved from separate thread)In a low memory situation, kthreadd may call schedule_on_each_cpu() while trying to spawn a worker, and the system grinds to a halt as below.