oldestActiveXid_fixed.v2.patch

diff --git a/src/backend/access/transam/clog.c b/src/backend/access/transam/clog.c
index 3a6c391..363db0e 100644
--- a/src/backend/access/transam/clog.c
+++ b/src/backend/access/transam/clog.c
@@ -479,6 +479,16 @@ ZeroCLOGPage(int pageno, bool writeXlog)
 void
 StartupCLOG(void)
 {
+	/* Currently this is a no-op */
+	return;
+}
+
+/*
+ * This must be called ONCE at the end of startup/recovery.
+ */
+void
+TrimCLOG(void)
+{
 	TransactionId xid = ShmemVariableCache->nextXid;
 	int			pageno = TransactionIdToPage(xid);
 
diff --git a/src/backend/access/transam/xlog.c b/src/backend/access/transam/xlog.c
index befb507..1535896 100644
--- a/src/backend/access/transam/xlog.c
+++ b/src/backend/access/transam/xlog.c
@@ -6917,6 +6917,11 @@ StartupXLOG(void)
 		StartupMultiXact();
 	}
 
+	/*
+	 * Trim commit log to the nextxid, so that any later xid states are reset.
+	 */
+	TrimCLOG();
+
 	/* Reload shared-memory state for prepared transactions */
 	RecoverPreparedTransactions();
 
@@ -7623,6 +7628,16 @@ CreateCheckPoint(int flags)
 	checkPoint.time = (pg_time_t) time(NULL);
 
 	/*
+	 * For Hot Standby, derive the oldestActiveXid before we fix the redo pointer.
+	 * This allows us to begin accumulating changes to assemble our starting
+	 * snapshot of locks and transactions.
+	 */
+	if (!shutdown && XLogStandbyInfoActive())
+		checkPoint.oldestActiveXid = GetOldestActiveTransactionId();
+	else
+		checkPoint.oldestActiveXid = InvalidTransactionId;
+
+	/*
 	 * We must hold WALInsertLock while examining insert state to determine
 	 * the checkpoint REDO pointer.
 	 */
@@ -7808,9 +7823,7 @@ CreateCheckPoint(int flags)
 	 * Update checkPoint.nextXid since we have a later value
 	 */
 	if (!shutdown && XLogStandbyInfoActive())
-		LogStandbySnapshot(&checkPoint.oldestActiveXid, &checkPoint.nextXid);
-	else
-		checkPoint.oldestActiveXid = InvalidTransactionId;
+		LogStandbySnapshot(&checkPoint.nextXid);
 
 	START_CRIT_SECTION();
 
diff --git a/src/backend/storage/ipc/procarray.c b/src/backend/storage/ipc/procarray.c
index 9489012..bbc5f44 100644
--- a/src/backend/storage/ipc/procarray.c
+++ b/src/backend/storage/ipc/procarray.c
@@ -1504,6 +1504,63 @@ GetRunningTransactionData(void)
 }
 
 /*
+ * GetOldestActiveTransactionId()
+ *
+ * Similar to GetSnapshotData but returns just oldestActiveXid. We include
+ * all PGPROCs with an assigned TransactionId, even VACUUM processes.
+ * We look at all databases, though there is no need to include WALSender
+ * since this has no effect on hot standby conflicts.
+ *
+ * This is never executed during recovery so there is no need to look at
+ * KnownAssignedXids.
+ *
+ * We don't worry about updating other counters, we want to keep this as
+ * simple as possible and leave GetSnapshotData() as the primary code for
+ * that bookkeeping.
+ */
+TransactionId
+GetOldestActiveTransactionId(void)
+{
+	ProcArrayStruct *arrayP = procArray;
+	TransactionId oldestRunningXid;
+	int			index;
+
+	Assert(!RecoveryInProgress());
+
+	LWLockAcquire(ProcArrayLock, LW_SHARED);
+
+	oldestRunningXid = ShmemVariableCache->nextXid;
+
+	/*
+	 * Spin over procArray collecting all xids and subxids.
+	 */
+	for (index = 0; index < arrayP->numProcs; index++)
+	{
+		volatile PGPROC *proc = arrayP->procs[index];
+		TransactionId xid;
+
+		/* Fetch xid just once - see GetNewTransactionId */
+		xid = proc->xid;
+
+		if (!TransactionIdIsNormal(xid))
+			continue;
+
+		if (TransactionIdPrecedes(xid, oldestRunningXid))
+			oldestRunningXid = xid;
+
+		/*
+		 * Top-level XID of a transaction is always less than any of its
+		 * subxids, so we don't need to check if any of the subxids are
+		 * smaller than oldestRunningXid
+		 */
+	}
+
+	LWLockRelease(ProcArrayLock);
+
+	return oldestRunningXid;
+}
+
+/*
  * GetTransactionsInCommit -- Get the XIDs of transactions that are committing
  *
  * Constructs an array of XIDs of transactions that are currently in commit
diff --git a/src/backend/storage/ipc/standby.c b/src/backend/storage/ipc/standby.c
index 72c6b97..3f2fc81 100644
--- a/src/backend/storage/ipc/standby.c
+++ b/src/backend/storage/ipc/standby.c
@@ -815,7 +815,7 @@ standby_desc(StringInfo buf, uint8 xl_info, char *rec)
  * making WAL entries.
  */
 void
-LogStandbySnapshot(TransactionId *oldestActiveXid, TransactionId *nextXid)
+LogStandbySnapshot(TransactionId *nextXid)
 {
 	RunningTransactions running;
 	xl_standby_lock *locks;
@@ -845,7 +845,6 @@ LogStandbySnapshot(TransactionId *oldestActiveXid, TransactionId *nextXid)
 	/* GetRunningTransactionData() acquired XidGenLock, we must release it */
 	LWLockRelease(XidGenLock);
 
-	*oldestActiveXid = running->oldestRunningXid;
 	*nextXid = running->nextXid;
 }
 
diff --git a/src/include/access/clog.h b/src/include/access/clog.h
index 7a8918e..691430c 100644
--- a/src/include/access/clog.h
+++ b/src/include/access/clog.h
@@ -40,6 +40,7 @@ extern Size CLOGShmemSize(void);
 extern void CLOGShmemInit(void);
 extern void BootStrapCLOG(void);
 extern void StartupCLOG(void);
+extern void TrimCLOG(void);
 extern void ShutdownCLOG(void);
 extern void CheckPointCLOG(void);
 extern void ExtendCLOG(TransactionId newestXact);
diff --git a/src/include/storage/procarray.h b/src/include/storage/procarray.h
index a11d438..7eccd5f 100644
--- a/src/include/storage/procarray.h
+++ b/src/include/storage/procarray.h
@@ -44,6 +44,7 @@ extern Snapshot GetSnapshotData(Snapshot snapshot);
 extern bool TransactionIdIsInProgress(TransactionId xid);
 extern bool TransactionIdIsActive(TransactionId xid);
 extern TransactionId GetOldestXmin(bool allDbs, bool ignoreVacuum);
+extern TransactionId GetOldestActiveTransactionId(void);
 
 extern int	GetTransactionsInCommit(TransactionId **xids_p);
 extern bool HaveTransactionsInCommit(TransactionId *xids, int nxids);
diff --git a/src/include/storage/standby.h b/src/include/storage/standby.h
index 6ebac62..e587a5c 100644
--- a/src/include/storage/standby.h
+++ b/src/include/storage/standby.h
@@ -111,6 +111,6 @@ typedef RunningTransactionsData *RunningTransactions;
 extern void LogAccessExclusiveLock(Oid dbOid, Oid relOid);
 extern void LogAccessExclusiveLockPrepare(void);
 
-extern void LogStandbySnapshot(TransactionId *oldestActiveXid, TransactionId *nextXid);
+extern void LogStandbySnapshot(TransactionId *nextXid);
 
 #endif   /* STANDBY_H */