release-22.1: ccl/sqlproxyccl: fix inaccurate CurConnCount metric due to goroutine leak

[jaylim-crl]

Backport 1/1 commits from #82652.

/cc @cockroachdb/release

---

Previously, there was a possibility where a processor can return from
resuming because the client's connection was closed before waitResumed even
has the chance to wake up to check on the resumed field. When that happens,
the connection goroutine will be blocked forever, and the CurConnCount metric
will never be decremented, even if the connection has already been terminated.

When the client's connection was closed, the forwarder's context will be
cancelled as well. The ideal behavior would be to terminate all waiters when
that happens, but the current code does not do that. This commit fixes that
issue by adding a new closed state to the processors, and ensuring that the
processor is closed whenever resume returns with an error. waitResumed can
then check on this state before going back to wait.

Release note: None

Release justification: sqlproxy bug fix. Internal only.

[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 

[jaylim-crl]

Note to self to only merge this once #82652 has been merged to master.

[jeffswenson]

LGTM

[jaylim-crl]

TFTR!

Merging as the Bazel CI failure is unrelated to this PR (i.e. docker image test failures).