spec_exec: fix can_be_ignored function

[muzarski]

ref: #1083
This is a first iteration of fixing #1083 - one we can merge before 0.15. I think this is crucial for speculative execution's performance - current implementation does not ignore any of the DbErrors. If any DbError is returned from database, we return prematurely without further investigation.

Previous implementation did not make much sense. We would prematurely return from speculative execution for some errors which could be ignored.

I adjusted the function, so we use #[deny(clippy::wildcard_enum_match_arm)] lint. It will force the developer to think before adjusting this function if new QueryError variant is introduced.

I categorized the errors (see the comments in the code) and justified why each of them needs to be ignored or not.

Note: In the comments, I mention that there are some error variants we don't expect there. It's true, because not all of the QueryError variants could be constructed at this stage. I'm currently working on narrowing the error type passed to this function (as well as errors passed to RetrySession::decide_should_retry and LBP::on_query_failure).

Pre-review checklist

• I have split my patch into logically separate commits.
• All commit messages clearly explain what they change and why.
• I added relevant tests for new features and bug fixes.
• All commits compile, pass static checks and pass test.
• PR description sums up the changes and reasons why they should be introduced.
• I have provided docstrings for the public items that I want to introduce.
• I have adjusted the documentation in ./docs/source/.
• I added appropriate Fixes: annotations to PR description.

[github-actions]

cargo semver-checks found no API-breaking changes in this PR! 🎉🥳
Checked commit: 5f1e007