RFR 8080623 CPU overhead in FJ due to spinning in awaitWork
dl at cs.oswego.edu
Wed May 20 13:29:58 UTC 2015
On 05/20/2015 09:19 AM, David Holmes wrote:
>> Not sure there was much context to begin with :-) best to look at the
>> implementation: a spin value, initialized to SPINS, is (if > 0) decremented if
>> the next value from a PRNG (Marsaglia-like) is non-negative.
> Not sure I grok why we would randomize the spin but ...
To better spread out delays to reduce future contention.
It's analogous to randomized back-offs.
More information about the core-libs-dev