Author:     Noah Misch <noah@leadboat.com>
Commit:     Noah Misch <noah@leadboat.com>

    Fix test race between primary XLOG_RUNNING_XACTS and standby logical slot.
    
    Before the previous commit, the test could hang until
    LOG_SNAPSHOT_INTERVAL_MS (15s), until checkpoint_timeout (300s), or
    indefinitely.  An indefinite hang was awfully improbable.  It entailed
    the test reaching checkpoint_timeout after a CREATE SUBSCRIPTION and
    before its DecodingContextFindStartpoint().  Back-patch to v16, which
    introduced the test.
    
    Reviewed by FIXME.
    
    Discussion: https://postgr.es/m/FIXME

diff --git a/src/test/recovery/t/035_standby_logical_decoding.pl b/src/test/recovery/t/035_standby_logical_decoding.pl
index 61da915..f7b7fc7 100644
--- a/src/test/recovery/t/035_standby_logical_decoding.pl
+++ b/src/test/recovery/t/035_standby_logical_decoding.pl
@@ -466,4 +466,11 @@ $psql_subscriber{subscriber_stdin} .= "\n";
 $psql_subscriber{run}->pump_nb();
 
+# See corresponding create_logical_slot_on_standby() code.
+$node_standby->poll_query_until(
+	'postgres', qq[
+		SELECT restart_lsn IS NOT NULL
+		FROM pg_catalog.pg_replication_slots WHERE slot_name = 'tap_sub'
+	]) or die "timed out waiting for logical slot to calculate its restart_lsn";
+
 # Speed up the subscription creation
 $node_primary->safe_psql('postgres', "SELECT pg_log_standby_snapshot()");