[concurrency-interest] @Contended (JEP-142)
radhakrishnan.mohan at gmail.com
Fri Nov 30 03:25:23 EST 2012
Isn't this specific to the processor ? The JVM will know how to calculate
cache line sizes dynamically and then pad ?
On Fri, Nov 30, 2012 at 6:13 AM, Vitaly Davidovich <vitalyd at gmail.com>wrote:
> What use case is that? If we're saying that this annotation is purely a VM
> hint, then I don't think it needs to surface at all in java code.
> How much overhead is there for having these be runtime retained? I don't
> imagine many classes will have themselves or fields annotated with it, so
> it may not be much of a concern. However, making it runtime from the get
> go puts us into a compatibility corner.
> Sent from my phone
> On Nov 29, 2012 7:35 PM, "Remi Forax" <forax at univ-mlv.fr> wrote:
>> On 11/30/2012 01:06 AM, Vitaly Davidovich wrote:
>>> Oh I see - you're saying CLASS retention is sufficient. I guess that
>>> depends on whether java level tooling (or generally, some java code period)
>>> should be able to grab this info or not.
>>> Sent from my phone
>> if the rentention policy is CLASS, I believe that there is no way using
>> the reflection API to get a runtime failure if the annotation class is not
>> present in the classpath.
>> Concurrency-interest mailing list
>> Concurrency-interest at cs.**oswego.edu <Concurrency-interest at cs.oswego.edu>
> Concurrency-interest mailing list
> Concurrency-interest at cs.oswego.edu
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Concurrency-interest