ForkJoin wakeup update
Aleksey Shipilev
aleksey.shipilev at oracle.com
Thu Oct 18 14:13:28 PDT 2012
On 10/19/2012 12:36 AM, Dave Dice wrote:
> On 2012-10-18, at 3:26 PM, Aleksey Shipilev
> <aleksey.shipilev at oracle.com> wrote:
>>> * The main systems-level goal should be to find ways to make
>>> wakeups faster. Possibly entailing support of some sort of bulk
>>> signalling facilities?
>>
>> Yes, the ball is in Dave's court here.
>
> I'd still like to see the results of a hard spin with no yield().
> Yield is often a disaster on solaris and linux.
Substituting park in scan() with yield():
(as measured before)
150 +- 15 us/op
Eliminating park in scan() without yield:
(just measured)
120 +- 12 us/op
Haven't got the fjp-trace for that run, since it would look similar to
the chart we have with yield(). Can make some round to get that if you need.
-Aleksey.
More information about the lambda-dev
mailing list