What's New notes for 0.32

[doveye]

First draft (live content for previous release is at https://www.eclipse.org/openj9/news). Note: this isn't an exhaustive list of all the changes, just the main ones that we want to tell users about:

Eclipse OpenJ9 version 0.32.0 released April 2022

We're pleased to announce the availability of Eclipse OpenJ9 v0.32.0.

This release supports OpenJDK version 8, 11, 17, and 18. For more information about supported platforms and OpenJDK versions, see Supported environments.

This release includes an early access build for Apple Silicon (AArch64) macOS. You can use this build for evaluation purposes; it is not yet suitable for use in production environments.

Other updates in this release include the following:

• The JITServer technology feature is now fully supported on Linux for IBM Z® (64-bit)
• In Java 11 and later, the default value of the -XX:MaxDirectMemorySize option, which limits the amount of heap memory that is used for direct byte buffers, is now the same as the maximum heap size. Previously, the limit was 87.5% of the maximum heap size.
• You can now use the SharedClassStatistics API to get the name, path, and directory of a shared classes cache. Depending on the operating system, you can also get the number of attached VMs for a non-persistent cache. This information is available through the following new methods: cacheDir(), cacheName(), cachePath(), and numberAttached(). For more information, see the API documentation.

To read more about these and other changes, see the OpenJ9 user documentation.

Other performance highlights include exploitation of the new IBM® z16™ hardware platform, for example:

• External decimal conversions acceleration with the Vector Packed Decimal Facility in the data access accelerator library
• Acceleration of the Integer.toString(), Long.toString(), and BigDecimal.toString() methods

[pshipton]

@vijaysun-omr @zl-wang @0xdaryl @joransiu any other highlights you want to call out?

[joransiu]

IBM z16 exploitation:

• External Decimal conversions acceleration with Vector Packed Decimal in Data Access Accelerator Library
• Integer.toString(), Long.toString(), BigDecimal.toString() acceleration

JITServer technology is released on Linux on Z / LinuxONE.

/cc: @dchopra001 @mpirvu @r30shah

[vijaysun-omr]

@0xdaryl should we mention some of the improvement in method handle performance that Nazim/Devin have (I believe) merged ?

@zl-wang it may also be worth summarizing Jimmy/Younes work on general throughput in a sentence if possible, assuming that made the cut for this release.

[zl-wang]

• Inlining strategy adjustment to improve throughput performance

@mpirvu it might be good for you to have one sentence on -Xtune:throughput enhancement and fix.

[vijaysun-omr]

I think we may have to generalize it as "Inlining strategy adjustment to improve throughput performance"

[mpirvu]

@mpirvu it might be good for you to have one sentence on -Xtune:throughput enhancement and fix.

-Xtune:throughput will not be documented at this time as we plan to do more tuning on it.

[0xdaryl]

should we mention some of the improvement in method handle performance that Nazim/Devin have (I believe) merged ?

None of the recent performance changes have been merged yet unfortunately. I'm expecting these to land for 0.33 (branching approx. May 23).

[doveye]

Updated draft to add jitserver on zLinux and z16 exploitation (hope I got the z terminology right!) Sounds like there's nothing else to add, judging by the other comments.