RFR: 8343789: Move mutable nmethod data out of CodeCache [v6]
Vladimir Kozlov
kvn at openjdk.org
Fri Jan 31 19:08:52 UTC 2025
On Tue, 28 Jan 2025 12:28:08 GMT, Boris Ulasevich <bulasevich at openjdk.org> wrote:
>> Looks like this also triggers with `compiler/escapeAnalysis/TestFindInstMemRecursion.java` on Linux x64 and `-XX:-TieredCompilation -XX:+StressReflectiveCode -XX:-ReduceInitialCardMarks -XX:-ReduceBulkZeroing -XX:-ReduceFieldZeroing`. This change needs more testing before integration.
>
> @TobiHartmann
>
>> I see this failure in our testing with an internal test
>
> Thanks for catching that issue! I've addressed it.
>
>> This change needs more testing before integration.
>
> Thank you for pointing that out. Here is my test plan for additional validation. Please let me know if you have any suggestions or if adjustments are needed:
>
> Functional Testing:
> - Platforms: x86 and AArch64
> - Build Configurations: Release, Fastdebug, and Slowdebug
> - Java Options: Default and -XX:+UseShenandoahGC
> - Tests: jtreg tests covering Hotspot and JDK, tiers 1–3
>
> Performance Testing:
> - Benchmarks: DaCapo and Renaissance
> - Modes: Default and -XX:+UseShenandoahGC
@bulasevich, please wait my review before integration.
-------------
PR Comment: https://git.openjdk.org/jdk/pull/21276#issuecomment-2628101640
More information about the hotspot-compiler-dev
mailing list