On Wed, 9 Apr 2014, Viresh Kumar wrote:
This patch tries to fix this by registering cpu notifiers from clocksource core, only when we start clocksource-watchdog. And if on the CPU_DEAD notification it is found that dying CPU was the CPU on which this timer is queued on, then it is removed from that CPU and queued to next CPU.
Gah, no. We realy don't want more notifier crap.
It's perfectly fine for the watchdog timer to be moved around on cpu down. And the timer itself is not pinned at all. add_timer_on() does not set the pinned bit.
Thanks,
tglx