release-22.2: sql: introduce array_cat_agg aggregate builtin

[yuzefovich]

Backport 1/1 commits from #97826.

/cc @cockroachdb/release

---

This commit introduces a new array_cat_agg aggregate builtin function
that takes in an array type as its input, and then unnests each array
and appends all its elements into a single result array. In other
words, it behaves similar to array_agg(unnest(array_column)). This
function doesn't have an analogue in Postgres. However, some of our SQL
observability tools need this functionality, and the current workaround
of using a LATERAL JOIN often results in slow apply joins, so this new
builtin should speed things up significantly. In particular,
crdb_internal.statement_statistics view is now refactored to use the
new builtin which removes an apply join from it. The choice of this
particular name comes from the fact that we have the array_cat builtin
which concatenates two arrays.

Fixes: #97502.

Release note (sql change): New aggregate builtin function
array_cat_agg is introduced. It behaves similar to how
array_agg(unnest(array_column)) would - namely, it takes arrays as its
input, unnests them into the array elements which are then aggregated
into a single result array (i.e. it's similar to concatenating all input
arrays into a single one).

Release justification: performance improvement of observability tooling.

[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]

assuming the oid changes are ok

Reviewed 19 of 19 files at r1, all commit messages.
Reviewable status: complete! 1 of 0 LGTMs obtained (waiting on @DrewKimball and @yuzefovich)

---

pkg/sql/logictest/testdata/logic_test/pg_catalog line 3273 at r1 (raw file):

WHERE t.typname = '_int4'
----
2046  array_in  array_in

Are these changes ok? (this file and pgoidtype)

---

pkg/sql/function_resolver_test.go line 370 at r1 (raw file):

			searchPath:       []string{"sc1", "sc2"},
			expectedFuncBody: "",
			expectedFuncOID:  853,

is this change ok?

[yuzefovich]

TFTR! The oid changes are ok - we didn't have fixed oids until 5ee5d85 which is only present on 23.1 development branch.