Invalidate non-local summation candidates across calls

[a7ehuo]

Prevent summation reduction optimization when non-local symbols (static, shadow, or method metadata) may be used or killed by call nodes in the loop. This ensures intermediate values are not incorrectly optimized away when they may be observed by other calls.

Fixes: eclipse-openj9/openj9#23277

[a7ehuo]

@hzongaro May I ask you to review this change? Thank you!

[hzongaro]

I think this looks correct. Thanks!

[hzongaro]

Jenkins build all

[hzongaro]

Jenkins build win

[a7ehuo]

The current three failures look to be either infrastructure related or happen in other PR build test too:

zos_390

Did you forget the `def` keyword? WorkflowScript seems to be setting a field named defaultCompile (to a value of type String) which could lead to memory leaks or other issues.
...
14:51:38  Still waiting to schedule task
14:51:38  All nodes of label ‘compile:zos’ are offline
22:51:22  Cancelling nested steps due to timeout

linux_riscv64. It also fails in https://ci.eclipse.org/omr/job/PullRequest-linux_riscv64/570/

First time build. Skipping changelog.
[Pipeline] Start of Pipeline
Did you forget the `def` keyword? WorkflowScript seems to be setting a field named defaultCompile (to a value of type String) which could lead to memory leaks or other issues.
...
Caused: java.io.IOException: Remote call on riscv-build2 failed
	at hudson.remoting.Channel.call(Channel.java:1112)
	at hudson.FilePath.act(FilePath.java:1204)
	at hudson.FilePath.act(FilePath.java:1193)
	at hudson.FilePath.mkdirs(FilePath.java:1384)
	at PluginClassLoader for workflow-basic-steps//org.jenkinsci.plugins.workflow.steps.CoreStep$Execution.run(CoreStep.java:97)
	at PluginClassLoader for workflow-basic-steps//org.jenkinsci.plugins.workflow.steps.CoreStep$Execution.run(CoreStep.java:71)
	at PluginClassLoader for workflow-step-api//org.jenkinsci.plugins.workflow.steps.SynchronousNonBlockingStepExecution.lambda$start$0(SynchronousNonBlockingStepExecution.java:49)
	at java.base/java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:539)
	at java.base/java.util.concurrent.FutureTask.run(FutureTask.java:264)
	at java.base/java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1136)
	at java.base/java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:635)
	at java.base/java.lang.Thread.run(Thread.java:853)
Setting status of d8dcf4754c69e2e1d233f5c396cd1fe90a31de37 to FAILURE with url https://ci.eclipse.org/omr/job/PullRequest-linux_riscv64/569/ and message: 'Build finished. '
Using context: continuous-integration/eclipse-omr/pr/linux_riscv64
Finished: FAILURE

ppc-64_le. It also fails in https://ci.eclipse.org/omr/job/PullRequest-linux_ppc-64_le_gcc/5025/

15:02:29  30: free(): invalid next size (normal)
15:02:29  30/30 Test #30: compunittest ......................Child aborted***Exception:   0.76 sec
15:02:29  
15:02:29  97% tests passed, 1 tests failed out of 30
15:02:29  
15:02:29  Total Test time (real) = 488.34 sec
15:02:29  
15:02:29  The following tests FAILED:
15:02:29  	 30 - compunittest (Child aborted)
15:02:29  Errors while running CTest

[hzongaro]

@a7ehuo, thank you for checking the test results. All failures are due to infrastructure issues or known problems. Merging.