Update jboss-logmanager to 3.0.4.Final

[dmlloyd]

Fixes #36919. Fixes #22844.

[gsmet]

@dmlloyd the errors are probably related as I haven't see these failures before.

[dmlloyd]

The failures do indeed look like they may be logging-related, though I don't see a change that might cause it.

[dmlloyd]

java.util.logging.ErrorManager: 1: Handler publication threw an exception
java.lang.invoke.WrongMethodTypeException: expected (ExtLogRecord,long)LogRecord but found (ExtLogRecord,long)void
	at java.base/java.lang.invoke.Invokers.newWrongMethodTypeException(Invokers.java:523)
	at java.base/java.lang.invoke.Invokers.checkExactType(Invokers.java:532)
	at org.jboss.logmanager.ExtLogRecord.superSetLongThreadID(ExtLogRecord.java:692)
	at org.jboss.logmanager.ExtLogRecord.setLongThreadID(ExtLogRecord.java:711)
	at org.jboss.logmanager.ExtLogRecord.<init>(ExtLogRecord.java:137)

[dmlloyd]

It's a regression: https://issues.redhat.com/browse/LOGMGR-344

I'll update this PR with 3.0.4.Final once the release is done (and the requisite time has elapsed).

[dmlloyd]

Should be OK now.

[dmlloyd]

I'll double check these failures.

[dmlloyd]

Some (or maybe all) of these failures are due to usage of InMemoryHandler, which implements the JUL Handler API. Because of this, and the fix for #22844, the log manager auto-formats printf-style messages, causing tests which rely on the structure of log records to fail.

The simple workaround which I am testing now is for these handlers to extend ExtHandler instead, indicating that they can process printf-formatted records.

[quarkus-bot]

✔️ The latest workflow run for the pull request has completed successfully.

It should be safe to merge provided you have a look at the other checks in the summary.