Prevent bcrypt from raising a ValueError and log

[anoadragon453]

Fixes #19063

bcrypt 5.0.0 raises a ValueError when a password's length is >72 bytes: pyca/bcrypt#1000 Whereas before it would silently truncate instead.

This truncation is fine, as a password of length 72 is sufficiently long that it will still take millions of years to brute force. The pepper being cut off as well is also fine - its designed to add bytes to short passwords.

We may want to think about transitioning away from bcrypt in future (to scrypt or argon2id), but for now this returns to existing behaviour without needing to downgrade bcrypt.

Question: should we backport this to a 1.140.1?

Pull Request Checklist

• Pull request is based on the develop branch
• Pull request includes a changelog file. The entry should:
  • Be a short description of your change which makes sense to users. "Fixed a bug that prevented receiving messages from other servers." instead of "Moved X method from EventStore to EventWorkerStore.".
  • Use markdown where necessary, mostly for code blocks.
  • End with either a period (.) or an exclamation mark (!).
  • Start with a capital letter.
  • Feel free to credit yourself, by adding a sentence "Contributed by @github_username." or "Contributed by [Your Name]." to the end of the entry.
• Code style is correct (run the linters)