RFR: 8298059: Linked stack watermarks don't support nesting [v3]

[Erik Österlund]

On Tue, 6 Dec 2022 17:02:21 GMT, Patricio Chilano Mateo <pchilanomate at openjdk.org> wrote:

> Looks good, but why do we need the lock? Isn't each thread only accessing its own _linked_watermarks field?

Yeah we don't currently need the lock if this is used in the currently expected way. But I feel like the boundaries of what is the current thread vs effectively current thread such as when handshaking/suspending another thread and running code on it as if it was current, tends to blur the boundaries a bit. And I suppose I figured better safe than sorry so I don't have to fix this code again.

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

PR: https://git.openjdk.org/jdk/pull/11530