Test Rerun holds a machine unnecessarily

[AdamBrousseau]

Related #3431 runtimes/backlog/issues/1022
Observed recently on a build that did a rerun. The rerun timed out waiting for a machine, because it was deadlocked. You can see in the last line, the node block isn't exited until after the rerun comes back.

06:19:27.801  Test_openjdk17_j9_extended.openjdk_aarch64_mac_rerun jobIsRunnable: true
[Pipeline] parallel
06:19:27.805  No branches to run
[Pipeline] // parallel
[Pipeline] build (Building Test_openjdk17_j9_extended.openjdk_aarch64_mac_rerun)
06:19:28.050  Scheduling project: [Test_openjdk17_j9_extended.openjdk_aarch64_mac_rerun](https://hyc-runtimes-jenkins.swg-devops.com/job/Test_openjdk17_j9_extended.openjdk_aarch64_mac_rerun/)
06:19:34.781  Starting building: [Test_openjdk17_j9_extended.openjdk_aarch64_mac_rerun #2](https://hyc-runtimes-jenkins.swg-devops.com/job/Test_openjdk17_j9_extended.openjdk_aarch64_mac_rerun/2/)
25:21:03.409  Cancelling nested steps due to timeout
[Pipeline] }
[Pipeline] // stage
[Pipeline] }
[Pipeline] // stage
[Pipeline] echo
25:21:06.871  PROCESSCATCH: Terminating any hung/left over test processes:
[Pipeline] sh
25:21:09.344  + aqa-tests/terminateTestProcesses.sh jenkins
25:21:09.345  Unix type machine..
25:21:09.345  Woohoo - no rogue processes detected!
[Pipeline] cleanWs
25:21:10.067  [WS-CLEANUP] Deleting project workspace...
25:21:10.068  [WS-CLEANUP] Deferred wipeout is disabled by the job configuration...
25:21:24.107  [WS-CLEANUP] done
[Pipeline] sh
25:21:24.744  + find /cores -name '*core*' -print -exec rm -f '{}' ';'
25:21:24.745  /cores
[Pipeline] }
[Pipeline] // timeout
[Pipeline] echo
25:21:24.902  Exception: org.jenkinsci.plugins.workflow.steps.FlowInterruptedException
[Pipeline] timeout
25:21:24.909  Timeout set to expire in 5 min 0 sec
[Pipeline] {
[Pipeline] }
[Pipeline] // timeout
[Pipeline] }
[Pipeline] // node

[llxia]

just for the record, rerun is turned off (RERUN_ITERATIONS is set to 0).