C/P/N/Q par vs. seq break-even analysis with 10ms think time

Doug Lea dl at cs.oswego.edu
Wed Oct 17 07:53:07 PDT 2012


On 10/17/12 10:31, Aleksey Shipilev wrote:
> FYI, I had updated the fjp-trace to track UNPARK->UNPARKED edges. This
> allows to infer the cost of unparking the thread, please see updated
> chart at [4]. Notice the red edges there, their sources are at UNPARK
> request for a thread, and destinations are at UNPARKED, when thread has
> indeed woken up.

Thanks for the diagram! Much nicer than looking at
raw times and trying to infer anything about them,
which is what I usually do. I'll send some notes
with code to try out in possible followups off-list.

-Doug

>> [4]
>> http://shipilev.net/pub/jdk/lambda/20121003-fjpwakeup/forkjoin.trace.p24-subtrees.png
>>
>



More information about the lambda-dev mailing list