release-22.2: opt: fix normalization of comparisons with constants

[mgartner]

Backport 1/2 commits from #90266.

/cc @cockroachdb/release

---

opt: fix normalization of comparisons with constants

A prior commit in #88199 attempted to fix a bug in the
NormalizeCmpPlusConst, NormalizeCmpMinusConst, and
NormalizeCmpConstMinus rules by checking for overflow/underflow in the
addition/subtraction of constants in a comparison expression. This was
insufficient to completely fix the bug because the transformation is
invalid if the non-normalized expression would have overflowed. Consider
an expression:

t::TIME + '-11 hrs'::INTERVAL > '01:00'::TIME

NormalizeCmpPlusConst would successively normalize it to this:

t::TIME > '01:00'::TIME - '-11 hrs'::INTERVAL
=> t::TIME > '12:00'::TIME

This expression is not semantically equivalent to the original
expression. It yields different results when t is a value that would
underflow when eleven hours is subtracted from it. For example, consider
t = '03:00'::TIME:

Original expression:
t::TIME + '-11 hrs'::INTERVAL > '01:00'::TIME
=> '03:00'::TIME + '-11 hrs'::INTERVAL > '01:00'::TIME
=> '16:00'::TIME > '01:00'::TIME
=> true

Normalized expression:
t::TIME > '12:00'::TIME
=> '03:00'::TIME > '12:00'::TIME
=> false

These normalization rules are only valid with types where overflow or
underflow during addition and subtraction results in an error.

This commit restricts these normalization rules to only operate on
integers, floats, and decimals, which will error if there is underflow
or overflow.

Fixes #90053

Release note (bug fix): A bug has been fixed that caused incorrect
evaluation of comparison expressions involving time and interval types,
like col::TIME + '10 hrs'::INTERVAL' > '01:00'::TIME.

---

Release justification: Fixes a long-standing correctness bug in the optimizer.

[blathers-crl]

Thanks for opening a backport.

Please check the backport criteria before merging:

• Patches should only be created for serious issues or test-only changes.
• Patches should not break backwards-compatibility.
• Patches should change as little code as possible.
• Patches should not change on-disk formats or node communication protocols.
• Patches should not add new functionality.
• Patches must not add, edit, or otherwise modify cluster versions; or add version gates.

If some of the basic criteria cannot be satisfied, ensure that the exceptional criteria are satisfied within.

• There is a high priority need for the functionality that cannot wait until the next release and is difficult to address in another way.
• The new functionality is additive-only and only runs for clusters which have specifically “opted in” to it (e.g. by a cluster setting).
• New code is protected by a conditional check that is trivial to verify and ensures that it only runs for opt-in clusters.
• The PM and TL on the team that owns the changed code have signed off that the change obeys the above rules.

Add a brief release justification to the body of your PR to justify this backport.

Some other things to consider:

• What did we do to ensure that a user that doesn’t know & care about this backport, has no idea that it happened?
• Will this work in a cluster of mixed patch versions? Did we test that?
• If a user upgrades a patch version, uses this feature, and then downgrades, what happens?

[cockroach-teamcity]

This change is 

[rytaft]

Reviewed 5 of 5 files at r1, all commit messages.
Reviewable status: complete! 1 of 0 LGTMs obtained (waiting on @msirek)