Jakarta - Adjust workflow tuning via environment variable

[gsmet]

Hard to find a quoting that actually works on GitHub with the weird
stuff they are doing to set environment variables via a file.

@famod after this is in, I will need your help. I have the GIB build failing for the jakarta-rewrite branch. I tried to remove -q here but the build says nothing except build success but apparently it exits with a 1 status anyway and I have no idea what's going on:
https://github.com/quarkusio/quarkus/runs/7247993104?check_suite_focus=true
Note: this particular build already includes this fix. I also tried to remove $EXCLUDE_JAKARTA_INCOMPATIBLE_MODULES from the GIB command line but to no avail.

The jakarta-rewrite branch will be pushed tonight and after that, you can push whatever you want to jakarta-rewrite to test the workflow, just know that it's force pushed every night with the new state.

[quarkus-bot]

Failing Jobs - Building 24b810c

Status	Name	Step	Failures	Logs	Raw logs
✔️	JVM Tests - JDK 11
⌛	JVM Tests - JDK 11 Windows	Build	⚠️ Check →	Logs	Raw logs
✔️	JVM Tests - JDK 17
✖	JVM Tests - JDK 18	Build	Failures	Logs	Raw logs
⌛	MicroProfile TCKs Tests	Verify	⚠️ Check →	Logs	Raw logs

Full information is available in the Build summary check run.

Failures

⚙️ JVM Tests - JDK 18 #

- Failing: extensions/smallrye-reactive-messaging-kafka/deployment 
! Skipped: integration-tests/kafka-oauth-keycloak integration-tests/kafka-sasl-elytron integration-tests/kubernetes/quarkus-standard-way-kafka and 3 more

📦 extensions/smallrye-reactive-messaging-kafka/deployment

✖ io.quarkus.smallrye.reactivemessaging.kafka.deployment.testing.KafkaDevServicesContinuousTestingWorkingAppPropsTestCase.testContinuousTestingScenario3 line 50 - More details - Source on GitHub

org.awaitility.core.ConditionTimeoutException: Failed to wait for test run 3 State{lastRun=2, running=true, inProgress=false, run=1, passed=0, failed=1, skipped=0, isBrokenOnly=false, isTestOutput=false, isInstrumentationBasedReload=false, isLiveReload=true}
	at io.quarkus.test.ContinuousTestingTestUtils.waitForNextCompletion(ContinuousTestingTestUtils.java:44)
	at io.quarkus.smallrye.reactivemessaging.kafka.deployment.testing.KafkaDevServicesContinuousTestingWorkingAppPropsTestCase.testContinuousTestingScenario3(KafkaDevServicesContinuousTestingWorkingAppPropsTestCase.java:50)

[famod]

IIRC, over the last weeks I had a few BUILD SUCCESS runs with return code > 1 too, but that was locally and without GIB.
At first glance, I don't see how GIB can turn a successful build into exit code > 0.

I suppose it's reproducible?

[famod]

Hum, there is a chunk of shell code after the mvnw call and I think the problem is there somewhere:

          ./mvnw -q -T1C $COMMON_MAVEN_ARGS -Dtcks -Dquickly-ci ${{ steps.get-gib-args.outputs.gib_args }} -Dgib.logImpactedTo=gib-impacted.log validate
          if [ -f gib-impacted.log ]
          then
            # TODO: for now, we don't run TCKs and for the jakarta-rewrite branch
            # we filter them here so that it cascades to all the builds
            if [ "${GITHUB_REF_NAME}" == "jakarta-rewrite" ]
            then
              GIB_IMPACTED=$(cat gib-impacted.log | grep -Pv '^(integration-tests/infinispan-client|integration-tests/kafka-avro)')
            else
              GIB_IMPACTED=$(cat gib-impacted.log)
            fi
          else
            GIB_IMPACTED=''
          fi
          echo "GIB_IMPACTED: ${GIB_IMPACTED}"
          echo "::set-output name=impacted_modules::${GIB_IMPACTED//$'\n'/'%0A'}"

== caught my attention. Shouldn't this be just = instead?

Anyhow, adding set -x could help here.