Hello,

I did send the following message to the users-list but did not get a reply. Therefore I try
here again. Meanwhile I found that somebody else filed a bug on bugzilla.kernel.org
(Bug 53071 Problems with gpe13 interrupt storm) about it. Is there something else I can
do? 
 
Thanks.
Eberhard

Original post at the users list:


I installed F18 on a Samsung NP700Z7C-S04 laptop. It is dual boot with 
WIN 8 and has UEFI enabled,
but secure boot is disabled. I'm seeing kworker using up around 80% CPU 
time constantly and the
laptop is running hot.
I found several similar reports on other lists. Suggestions found there 
like having acpi=noirq and others
didn't help.
There is a closed bugzilla (Bug 840862) suggesting to do the following 
to find the cause of the problem:

echo workqueue:workqueue_queue_work > /sys/kernel/debug/tracing/set_event
cat /sys/kernel/debug/tracing/trace_pipe > out.txt

In out.txt I find

CPU:3 [LOST 12401 EVENTS]
          <idle>-0     [003] d.h.  1010.943940: workqueue_queue_work: 
work struct=ffff880328a55910 function=acpi_os_execute_deferred 
workqueue=ffff88032d856b40 req_cpu=0 cpu=0
          <idle>-0     [003] d.h.  1010.943969: workqueue_queue_work: 
work struct=ffff880328a55ad0 function=acpi_os_execute_deferred 
workqueue=ffff88032d856b40 req_cpu=0 cpu=0
          <idle>-0     [003] d.h.  1010.944000: workqueue_queue_work: 
work struct=ffff880328a553d0 function=acpi_os_execute_deferred 
workqueue=ffff88032d856b40 req_cpu=0 cpu=0

The acpi_os_execute_deferred message is showing up endlessly.

Also suggested in the bugzilla is doing

cat /proc/2/stack
[<ffffffff81082425>] kthreadd+0x1e5/0x1f0
[<ffffffff8163dfac>] ret_from_fork+0x7c/0xb0
[<ffffffffffffffff>] 0xffffffffffffffff

What to do with this information?
I'm running kernel 3.7.4-204.fc18.x86_64 (the earlier F18 kernels showed 
the same behaviour).

Any help would be appreciated. Should I sent this to another list?

Thanks.
Eberhard