TEST: After tablesyn-early-exit patch code change to avoid tablesync processing...

...

2020-12-08 14:25:12.820 AEDT [8895] LOG:  !!>> apply worker: LogicalRepApplyLoop
2020-12-08 14:25:12.820 AEDT [8895] LOG:  !!>> apply worker: called process_syncing_tables
2020-12-08 14:25:13.821 AEDT [8895] LOG:  !!>> apply worker: LogicalRepApplyLoop
2020-12-08 14:25:13.821 AEDT [8895] LOG:  !!>> apply worker: called process_syncing_tables
2020-12-08 14:25:14.125 AEDT [10125] LOG:  logical decoding found consistent point at 0/162A760
2020-12-08 14:25:14.125 AEDT [10125] DETAIL:  There are no running transactions.
2020-12-08 14:25:14.125 AEDT [10125] STATEMENT:  CREATE_REPLICATION_SLOT "tap_sub_16438_sync_16385" TEMPORARY LOGICAL pgoutput USE_SNAPSHOT
2020-12-08 14:25:14.130 AEDT [8895] LOG:  !!>> apply worker: LogicalRepApplyLoop
2020-12-08 14:25:14.130 AEDT [8895] LOG:  !!>> apply worker: called process_syncing_tables

2020-12-08 14:25:14.144 AEDT [8904] LOG:  !!>> tablesync worker: wait for CATCHUP state notification
2020-12-08 14:25:14.144 AEDT [8895] LOG:  !!>> apply worker: LogicalRepApplyLoop
2020-12-08 14:25:14.144 AEDT [8895] LOG:  !!>> apply worker: called process_syncing_tables

## Attached and paused in debugger where the "tablesync" is told to CATCHUP.

## At this point we might execute any number of SQL commands
## e.g 10 inserts

psql -d test_pub -c "INSERT INTO test_tab VALUES(1, 'foo');"
psql -d test_pub -c "INSERT INTO test_tab VALUES(2, 'foo');"
psql -d test_pub -c "INSERT INTO test_tab VALUES(3, 'foo');"
psql -d test_pub -c "INSERT INTO test_tab VALUES(4, 'foo');"
psql -d test_pub -c "INSERT INTO test_tab VALUES(5, 'foo');"
psql -d test_pub -c "INSERT INTO test_tab VALUES(6, 'foo');"
psql -d test_pub -c "INSERT INTO test_tab VALUES(7, 'foo');"
psql -d test_pub -c "INSERT INTO test_tab VALUES(8, 'foo');"
psql -d test_pub -c "INSERT INTO test_tab VALUES(9, 'foo');"
psql -d test_pub -c "INSERT INTO test_tab VALUES(10, 'foo');"

## Now continue in the "tablesync" worker
## But because of this patch the "tablesync" will exit before doing anything in the LogicalRepApplyLoop

2020-12-08 14:26:43.339 AEDT [8904] LOG:  !!>> tablesync worker: called process_syncing_tables
2020-12-08 14:26:43.339 AEDT [8904] LOG:  logical replication table synchronization worker for subscription "tap_sub", table "test_tab" has finished

## So the "apply" worker now will process all 10x INSERTs

2020-12-08 14:26:43.340 AEDT [8895] LOG:  !!>> apply worker: LogicalRepApplyLoop
2020-12-08 14:26:43.340 AEDT [8895] LOG:  !!>> apply worker: apply_dispatch for message kind 'B'
2020-12-08 14:26:43.340 AEDT [8895] LOG:  !!>> apply worker: apply_dispatch for message kind 'R'
2020-12-08 14:26:43.340 AEDT [8895] LOG:  !!>> apply worker: apply_dispatch for message kind 'I'
2020-12-08 14:26:43.341 AEDT [8895] LOG:  !!>> apply worker: should_apply_changes_for_rel: true
	rel->state == READY: false
	rel->state == SYNCDONE: true
	rel->statelsn <= remote_final_lsn: true
2020-12-08 14:26:43.342 AEDT [8895] LOG:  !!>> apply worker: apply_dispatch for message kind 'C'
2020-12-08 14:26:43.342 AEDT [8895] LOG:  !!>> apply worker: called process_syncing_tables

X 10

...

## Then apply worker just waiting for next incoming message like normal

2020-12-08 14:26:43.344 AEDT [8895] LOG:  !!>> apply worker: called process_syncing_tables
2020-12-08 14:26:43.344 AEDT [8895] LOG:  !!>> apply worker: LogicalRepApplyLoop
2020-12-08 14:26:43.344 AEDT [8895] LOG:  !!>> apply worker: called process_syncing_tables
2020-12-08 14:26:43.545 AEDT [8895] LOG:  !!>> apply worker: LogicalRepApplyLoop
2020-12-08 14:26:43.545 AEDT [8895] LOG:  !!>> apply worker: called process_syncing_tables
2020-12-08 14:26:43.748 AEDT [8895] LOG:  !!>> apply worker: LogicalRepApplyLoop
...