Limit disk usage of TCK reports

We don't need to keep a copy of the whole JDK around in the reports
of every single build...

Note this is only necessary in the 3.1 TCK, as the 3.2 TCK doesn't seem
to do that anymore.

Author: yrodiere

ci/jpa-3.1-tck.Jenkinsfile

@@ -1,4 +1,4 @@
-@Library('hibernate-jenkins-pipeline-helpers@1.13') _
+@Library('hibernate-jenkins-pipeline-helpers') _
 
 // Avoid running the pipeline on branch indexing
 if (currentBuild.getBuildCauses().toString().contains('BranchIndexingCause')) {
@@ -80,6 +80,7 @@ pipeline {
                             docker volume create tck-vol
                             docker run -v ~/.m2/repository/org/hibernate:/root/.m2/repository/org/hibernate:z -v tck-vol:/tck/persistence-tck/tmp/:z -e NO_SLEEP=${params.NO_SLEEP} -e HIBERNATE_VERSION=$HIBERNATE_VERSION --name tck jakarta-tck-runner
                             docker cp tck:/tck/persistence-tck/tmp/ ./results
+                            rm -Rf ./results/jdk-bundles
                         """
                         archiveArtifacts artifacts: 'results/**'
                         script {