RFR: 8370947: Mitigate Neoverse-N1 erratum 1542419 negative impact on GCs and JIT performance [v28]

[Evgeny Astigeevich]

On Fri, 20 Feb 2026 15:03:27 GMT, Erik Österlund <eosterlund at openjdk.org> wrote:

> Thank you for the numbers. Is this reporting the speed of running GC cycles?

These are reductions in GC cycle times. 

> It would be encouraging to know that at least some macro benchmarks don't regress, and not just that some micro benchmarks show improvements on GC cycle lengths (assuming that's what we are measuring).

On Graviton 2 I ran SpecJVM, dacapo and renaissance benchmarks. I did not see regressions. There are renaissance and dacapo benchmarks improved. In the description you can find improvements in SpecJVM. I need to update the description.

Another team was assessing performance impact of this change. They got ~1.2% lift in the SpecJBB maxjops held across Graviton generations (2, 3, 4).

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

PR Comment: https://git.openjdk.org/jdk/pull/28328#issuecomment-3936883492