Rework activation of commit timestamps during recovery

The activation and deactivation of commit timestamp tracking has not
been handled consistently for a primary or standbys at recovery.  The
facility can be activated at three different moments of recovery:
- The beginning, where a primary would use the GUC value for the
decision-making, and where a standby relies on the contents of the
control file.
- When replaying a XLOG_PARAMETER_CHANGE record at redo.
- The end, where both primary and standby rely on the GUC value.

Using the GUC value for a primary at the beginning of recovery causes
problems with commit timestamp access when doing crash recovery.
Particularly, when replaying transaction commits, it could be possible
that an attempt to read commit timestamps is done for a transaction
which committed at a moment when track_commit_timestamp was disabled.

A test case is added to reproduce the failure.  The test works down to
v11 as it takes advantage of transaction commits within procedures.

Reported-by: Hailong Li
Author: Masahiko Sawasa, Michael Paquier
Reviewed-by: Kyotaro Horiguchi
Discussion: https://postgr.es/m/11224478-a782-203b-1f17-e4797b39bdf0@qunar.com
Backpatch-through: 9.5, where commit timestamps have been introduced.

1 files changed, 31 insertions, 5 deletions

diff --git a/src/test/modules/commit_ts/t/004_restart.pl b/src/test/modules/commit_ts/t/004_restart.pl
index daf42d3a029..241b0b08dcf 100644
--- a/src/test/modules/commit_ts/t/004_restart.pl
+++ b/src/test/modules/commit_ts/t/004_restart.pl
@@ -1,4 +1,4 @@
-# Testing of commit timestamps preservation across clean restarts
+# Testing of commit timestamps preservation across restarts
 use strict;
 use warnings;
 use PostgresNode;
@@ -71,12 +71,36 @@ is($after_restart_ts, $before_restart_ts,
 	'timestamps before and after restart are equal');
 
 # Now disable commit timestamps
-
 $node_master->append_conf('postgresql.conf', 'track_commit_timestamp = off');
-
 $node_master->stop('fast');
+
+# Start the server, which generates a XLOG_PARAMETER_CHANGE record where
+# the parameter change is registered.
 $node_master->start;
 
+# Now restart again the server so as no XLOG_PARAMETER_CHANGE record are
+# replayed with the follow-up immediate shutdown.
+$node_master->restart;
+
+# Move commit timestamps across page boundaries.  Things should still
+# be able to work across restarts with those transactions committed while
+# track_commit_timestamp is disabled.
+$node_master->safe_psql('postgres',
+qq(CREATE PROCEDURE consume_xid(cnt int)
+AS \$\$
+DECLARE
+    i int;
+    BEGIN
+        FOR i in 1..cnt LOOP
+            EXECUTE 'SELECT txid_current()';
+            COMMIT;
+        END LOOP;
+    END;
+\$\$
+LANGUAGE plpgsql;
+));
+$node_master->safe_psql('postgres', 'CALL consume_xid(2000)');
+
 ($ret, $stdout, $stderr) = $node_master->psql('postgres',
 	qq[SELECT pg_xact_commit_timestamp('$xid');]);
 is($ret, 3, 'no commit timestamp from enable tx when cts disabled');
@@ -106,10 +130,12 @@ like(
 # Re-enable, restart and ensure we can still get the old timestamps
 $node_master->append_conf('postgresql.conf', 'track_commit_timestamp = on');
 
-$node_master->stop('fast');
+# An immediate shutdown is used here.  At next startup recovery will
+# replay transactions which committed when track_commit_timestamp was
+# disabled, and the facility should be able to work properly.
+$node_master->stop('immediate');
 $node_master->start;
 
-
 my $after_enable_ts = $node_master->safe_psql('postgres',
 	qq[SELECT pg_xact_commit_timestamp('$xid');]);
 is($after_enable_ts, '', 'timestamp of enabled tx null after re-enable');