always sync flush fatal log messages

[davidmarkclements]

it occurred to me today that if you ever write a fatal log, it's because the server is crashing or because you need to crash the server.

e.g.

if (thisIsAnImpossibleSituation) {
  pino.fatal(`there's no where to go, we have to crash`)
  process.exit(1)
}

In this scenario, we always want to sync flush, and there's no legit scenarios where fatal would be used other than in a death spiral.

[mcollina]

LGTM

[jsumners]

Noting here that this usage of fatal is basically the standard assumption -- https://logging.apache.org/log4j/1.2/apidocs/org/apache/log4j/Level.html#FATAL

The FATAL level designates very severe error events that will presumably lead the application to abort.

That being said, I would consider this change a semver major change since it alters the behavior of the method.

[github-actions]

This pull request has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.