On 20/06/17 17:12, Prarit Bhargava wrote: >>> Hmm ... odd that I haven't seen this. We're running a pretty wide >>> variety of systems here. Do you have a reproducer? I'd like to see >>> this occur on production HW. "Production" is where this patch was born and still lives right now: https://chromium-review.googlesource.com/c/506327/ >> It only happens on a few specific CPU SKUs with a very wide Turbo range. > > Which ones? > The ones with turbo mode > 2.5 x TSC_MHz when you stress them hard and long enough. Simple maths: just compare the soft and hard timers in the code. The factor 3 moves the condition to: turbo > 7.5 x TSC_MHz.