But it requires changing the virtual memory mapping tables, i.e. it requires a context switch to the kernel and back. Therefore it decreases effectiveness of TL

But it requires changing the virtual memory mapping tables, i.e. it requires a context switch to the kernel and back. Therefore it decreases effectiveness of TLAB caches. I.e. there is a performance cost. One that is hard to quantify for very short lived threads and only slightly better quantifiable for longer lived threads. And every argument that OS threads are good enough is in stark contrast to the length to which people go to avoid using them. I am talking about all the async libraries/patterns/language features people use to avoid using OS-level threads.