AArch64: Enable SymbolValidationManager

[knn-k]

This commit enables SymbolValidationManager on AArch64.

[knn-k]

Jenkins test sanity,sanity.system amac,alinux64 jdk21

[knn-k]

Jenkins test sanity,sanity.system amac,alinux64 jdk21

[knn-k]

The test failure (timeout in java/lang/Thread/virtual/Collectable.java) in Test_openjdk21_j9_sanity.openjdk_aarch64_linux_Personal seems to be the same as a known issue: #18463

[knn-k]

Jenkins test sanity.openjdk alinux64 jdk21

[knn-k]

@0xdaryl Would you review this PR when you have time, please? Thanks.

[0xdaryl]

Looks OK to me, but I'd like to wait for @dsouzai 's opinion on whether anything else needs to be done to enable this.

[dsouzai]

Looks fine to me.

That said, this really just enables the infra; I don't know if all of the places that need support have been addressed. #3028 (comment) (note, this was from 2019) lists out places that have be addressed to support SVM AOT. Assuming all of that is addressed, it should be fine to enable SVM on the platform.

Now, the SVM is disabled during startup - by that I mean when the JVM Phase is not in J9VM_PHASE_NOT_STARTUP. We also don't tend to compile AOT after startup. So, it may be worth adding -Xaot:enableSVMDuringStartup as an additional mode to some tests. But at the same time, unless we're running server workloads like Liberty, we get out of startup pretty fast so maybe we just need to run with forceAOT. So (assuming this hasn't already been done) kicking off tests with +aot would be a good way of ensuring the SVM infra gets at least partially tested.

[0xdaryl]

Thanks for the comments @dsouzai . @knn-k, before we enable this, can you go through that checklist that Irwin linked to make sure all of those are complete for AArch64?