Francesco Nigro

> The padding is there to protect the loaded field from contention induced by unfortunate placing next to fast updating memory locations. Is that important enough? not sure. In this...

@hl845740757 Thanks for sharing: just sent a PR to address this correctly

@hl845740757 I still think that the peek/relaxedPeek issue you've found for other queues is present as well for the xadd qs

cChunk.prev should always be null at this point

The changes of https://github.com/JCTools/JCTools/pull/281 have shown the need to simplify the mpmc xadd poll. To summarize the findings for the future: before the changes of https://github.com/JCTools/JCTools/pull/281, with `chunkSize = 1`...

Probably `@PolymorphicSignature` is creating issues while compiling with 1.6, but with 1.7 is fine and, of course, if I use `MethodHandle::invokeWithArguments`, that doesn't use `@PolymorphicSignature`, is fine too

Here https://github.com/LMAX-Exchange/disruptor/issues/183#issuecomment-269059148 some more information on how we could bring this in with 1.6, if it worths doing it

@nitsanw > In its current state this will not fly with building on 8 with JDK6 compatibility. I know, I've yet to work on this to make it possible, but...

@nitsanw Nitsan, bud, I cannot add any label to this PR, can you do it for me? is "DON'T MERGE IT YET" in red; I will change it accordly when...

Do you have something specific in mind? Burst offset benchmarks are well covered on https://github.com/real-logic/benchmarks and I think that they would be nice instead of the traditional ones focusing on...