> Do you have a citation (because I can't find one) specifying your assertion that original Java green threads were not analogous to Solaris user -> os -> hardware multiplexing?
I was writing from memory of second-hand after-the-fact recitations of the history. Doing some followup research prompted by your question, if I understand this document [0] correctly, Java initially had N:1 green threads on Solaris, then M:N green threads on Solaris with 1:1 native threads on Unix and Windows.
[0] https://docs.oracle.com/cd/E19455-01/806-3461/6jck06gqe/