RFR: 8325937: runtime/handshake/HandshakeDirectTest.java causes "monitor end should be strictly below the frame pointer" assertion failure on AArch64

[David Holmes]

On Tue, 1 Oct 2024 18:38:33 GMT, Coleen Phillimore <coleenp at openjdk.org> wrote:

> Add missing StoreLoad fences in handshaking code to match safepoint code.  Thanks to @pchilano for finding this bug.
> 
> Tested with tier1-4 and tier8 which has Kitchensink in it.

FWIW Posix simply  refers to "synchronize memory with respect to other threads" with no details and no strong vs. relaxed variants - so that should reasonably be inferred as providing a "fence" as we would describe it. Of course there is no guarantee the implementation doesn't weaken the actual barrier used because it thinks it is safe to do so.

And it could still be that this is an issue with the Arm Neoverse N2 hardware.

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

PR Comment: https://git.openjdk.org/jdk/pull/21295#issuecomment-2387311318
PR Comment: https://git.openjdk.org/jdk/pull/21295#issuecomment-2387312516