RFR: 8293922: Extend barrier-less Java thread transitions to native transitions [v2]

[Robbin Ehn]

> As suggested by @TheRealMDoerr, avoid the fence for native trans when UseSystemMemoryBarrier is enabled.
> 
> Locally I can see methods like Thread.holdsLock() getting a 4 ns speed boost, (~20% faster).
> 
> Running t1-5 with UseSystemMemoryBarrier on.

Robbin Ehn has updated the pull request incrementally with one additional commit since the last revision:

  Use proper accessors for _thread_state

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

Changes:
  - all: https://git.openjdk.org/jdk/pull/10302/files
  - new: https://git.openjdk.org/jdk/pull/10302/files/d1f143ed..52bad66c

Webrevs:
 - full: https://webrevs.openjdk.org/?repo=jdk&pr=10302&range=01
 - incr: https://webrevs.openjdk.org/?repo=jdk&pr=10302&range=00-01

  Stats: 4 lines in 1 file changed: 0 ins; 0 del; 4 mod
  Patch: https://git.openjdk.org/jdk/pull/10302.diff
  Fetch: git fetch https://git.openjdk.org/jdk pull/10302/head:pull/10302

PR: https://git.openjdk.org/jdk/pull/10302