RFR: 8332455: Improve G1/ParallelGC tasks to not override array lengths

[Kim Barrett]

On Fri, 5 Jul 2024 15:04:01 GMT, Roman Kennke <rkennke at openjdk.org> wrote:

> About the address space concern, yes, I am aware of this. We're doing the exact same slicing in Shenandoah since forever, and so far never had a customer who would get to that limit. Our reasoning there was that if one uses this amount of Java heap, using 2x as much native memory for the GC task queues (using the fallback version of the task encoding) would probably don't matter all that much.

I would guess there aren't tons of users of Linux Large Virtual address space.

For fat tasks, I'm less concerned about the actual space impact than on the
additional memory transfer costs for the normal and dominant non-array case,
where we end up spending cycles copying around these completely unused words.

I'd much have allocated states or segregated queues than fat tasks.

-------------

PR Comment: https://git.openjdk.org/jdk/pull/19282#issuecomment-2217251786