Re:Java 24
by atomicnumber3 on Tuesday March 18, 2025 @11:37AM
(#43400717)
ALL? no. but based on my read of the article's linked bit on thread pinning, it seems like it has been substantially resolved, and remaining pinning scenarios are quite rare (they mention you can look for them in JFR).
I was wondering if native calls were solved, but I don't see how those can ever be solved
In the long term native calls will require a command line flag to be permitted. No matter whether they are done via FFI or JNA. By then, libraries better figure out a way to document them or to reduce using them as much as possible. A lot of off-heap memory trickery should now be possible via (unsafe of course) functions from the FFI library.
Have ALL previous pinning scenarios been resolved, including native calls?