Promote BoxedPrimitiveEquality and ReferenceEquality

[iamdanfox]

Before this PR

@Jolyon-S reported in #dev-foundry-infra that a Long == Long snippet of code wasn't flagged by any of our static analysis.

There are existing error-prone checks that can catch this though, but they were just at Warning (which atlasdb didn't opt-in to with -Werror.

• https://errorprone.info/bugpattern/ReferenceEquality
• https://errorprone.info/bugpattern/BoxedPrimitiveEquality

After this PR

==COMMIT_MSG==
Error-prone's BoxedPrimitiveEquality and ReferenceEquality checks will now fail the build, as we consider them errors not just warnings.
==COMMIT_MSG==

Possible downsides?

• Every time we promote checks like this there's the danger of introducing unnecessary friction to developers if there are false positives (or harmless cases). For these two checks though, I don't anticipate there to be lots of false positives, and getting it wrong seems pretty serious.

[changelog-app]

Generate changelog in changelog/@unreleased

Type

• Feature
• Improvement
• Fix
• Break
• Deprecation
• Manual task
• Migration

Description

Error-prone's BoxedPrimitiveEquality and ReferenceEquality checks will now fail the build, as we consider them errors not just warnings.

Check the box to generate changelog(s)

• Generate changelog entry

[ferozco]

👍

[svc-autorelease]

Released 3.43.0