constant call trace in log.

[Julie Sullivan]

(oops, sending to the mailing list as well this time, sorry, folks.)

Hi Max,

Maybe this might be fixed in a more recent kernel. Have you tried one?
See here for a range of reasons why a CPU stall warning might happen:

http://www.kernel.org/doc/Documentation/RCU/stallwarn.txt

So the report time after the RCU grace period starts is controlled by
CONFIG_RCU_CPU_STALL_TIMEOUT, which is to be found under 'Kernel
hacking  ---> RCU CPU stall timeout in seconds' from menuconfig.

Failing that at runtime you can change the values of
rcu_cpu_stall_suppress and rcu_cpu_stall_timeout in
/sys/module/rcutree/parameters/.

Cheers
Julie