Yes, Azul's C4 paper states that they do use some "synchronisation" stop-the-world phases (although their previous GCs that had hardware/kernel read barriers mi

Yes, Azul's C4 paper states that they do use some "synchronisation" stop-the-world phases (although their previous GCs that had hardware/kernel read barriers might not have), but IIRC Doligez-Leroy GC doesn't (although I'm not sure it was ever properly implemented). On the other hand, Shenandoah has an explicit synchronisation phase in its algorithm (not sure if that's a fundamental property of the algorithm or just a JVM limitation though). From [0] (that was 2 years ago, things might have improved since, but I can't find a more recent version). > Pauses only long enough to scan root set [0]: https://rkennke.files.wordpress.com/2014/02/shenandoahtake4.pdf https://rkennke.files.wordpress.com/2014/02/shenandoahtake4....