kvserver: add TestFlowControlSendQueueRangeSplitMerge test

[kvoli]

First commit from #136330.

---

Add a new rac2 flow control integration test,
TestFlowControlSendQueueRangeSplitMerge.

This test takes the following steps:

-- We will exhaust the tokens across all streams while admission is blocked on
-- n3, using a single 4 MiB (deduction, the write itself is small) write. Then,
-- we will write a 1 MiB put to the range, split it, write a 1 MiB put to the
-- RHS range, merge the ranges, and write a 1 MiB put to the merged range. We
-- expect that at each stage where a send queue develops n1->s3, the send queue
-- will be flushed by the range merge and range split range operations.

Part of: #132614
Release note: None

[blathers-crl]

Your pull request contains more than 1000 changes. It is strongly encouraged to split big PRs into smaller chunks.

_{🦉 Hoot! I am a Blathers, a bot for CockroachDB. My owner is dev-inf.}

[cockroach-teamcity]

This change is 

[kvoli]

Should be good for review, ignoring the first commit (I won't merge this until after we merge the split-out fixes from #136077).

ptal @sumeerbhola.

[sumeerbhola]

Reviewed 2 of 2 files at r2.
Reviewable status: complete! 1 of 0 LGTMs obtained (waiting on @kvoli)

---

pkg/kv/kvserver/flow_control_integration_test.go line 5606 at r2 (raw file):

	h.comment(`(Sending 1 MiB put request to post-split RHS range)`)
	h.put(ctx, k, 1, admissionpb.NormalPri)

should this be writing to right.StartKey since that is the start of the RHS. k seems to be the LHS.

Perhaps we should write to both LHS and RHS, to make sure everything is working as expected. The SubsumeRequest force-flush will flush on the RHS, and the txn commit (whose txn record is on the LHS) should cause a force-flush on the LHS.