harmony-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Rodrigo Kumpera <kump...@gmail.com>
Subject Re: [arch] Harmony VM & JSR166 integration
Date Fri, 25 Nov 2005 16:57:34 GMT
I think park/unpack can be implemented using java monitors without much trouble.

On 11/25/05, Andrey Chernyshev <a.y.chernyshev@gmail.com> wrote:
> I've been thinking a bit about the java.util.concurrent package, and
> how we might want to approach it here in Harmony. What kind of
> modifications one may need to do for VM in order to support
> util.concurrent package in the most efficient way? To my
> understanding, the original design of the JSR166 implied quite limited
> support at VM level. It basically included only 3 points of
> interaction:
> -          System.nanoTime() method (no other means to get nanosecond
> precision time);
> -          Park/unpark methods for
> java.util.concurrent.locks.LockSupport class (suspend/resume can't be
> used for this because of race condition issues);
> -          Field atomic update methods for java.util.concurrent.atomic
> classes (they might be too slow if implemented via conventional
> synchronized section)
> It seems the rest of the code can just be a pure Java. Do people on
> the list think the above set will be sufficient to achieve the right
> level of cooperation between VM and util.concurrent?
> Thank you,
> Andrey Chernyshev
> Intel Managed Runtime Division

View raw message