Sv: [PATCH] hrtimer: Update softirq_expires_next correctly after __hrtimer_get_next_event()

[Mikael Beckius]

> Micke, please test this patch - it is compile tested only...

Thanks Anna-Maria, it seems to work equally good or better than the change
I suggested in preventing interrupt storms. After running a test I created,
based on traces from a live system, several thousands of iterations
there are no reproductions. Normally both issues will reproduce within
seconds after only a few iterations.

Also by looking at the source code it seems like your patch should do the
trick.

Micke