Remove read_page callback from XLogReader.

Previously, the XLogReader module would fetch new input data using a
callback function.  Redesign the interface so that it tells the caller
to insert more data with a special return value instead.  This API suits
later patches for prefetching, encryption and maybe other future
projects that would otherwise require continually extending the callback
interface.

As incidental cleanup work, move global variables readOff, readLen and
readSegNo inside XlogReaderState.

Author: Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp>
Author: Heikki Linnakangas <hlinnaka@iki.fi> (parts of earlier version)
Reviewed-by: Antonin Houska <ah@cybertec.at>
Reviewed-by: Alvaro Herrera <alvherre@2ndquadrant.com>
Reviewed-by: Takashi Menjo <takashi.menjo@gmail.com>
Reviewed-by: Andres Freund <andres@anarazel.de>
Reviewed-by: Thomas Munro <thomas.munro@gmail.com>
Discussion: https://postgr.es/m/20190418.210257.43726183.horiguchi.kyotaro%40lab.ntt.co.jp

src/backend/access/transam/twophase.c

@@ -1330,11 +1330,8 @@ XlogReadTwoPhaseData(XLogRecPtr lsn, char **buf, int *len)
 	char	   *errormsg;
 	TimeLineID	save_currtli = ThisTimeLineID;
 
-	xlogreader = XLogReaderAllocate(wal_segment_size, NULL,
+	xlogreader = XLogReaderAllocate(wal_segment_size, NULL, wal_segment_close);
-									XL_ROUTINE(.page_read = &read_local_xlog_page,
+
-											   .segment_open = &wal_segment_open,
-											   .segment_close = &wal_segment_close),
-									NULL);
 	if (!xlogreader)
 		ereport(ERROR,
 				(errcode(ERRCODE_OUT_OF_MEMORY),
@@ -1342,7 +1339,12 @@ XlogReadTwoPhaseData(XLogRecPtr lsn, char **buf, int *len)
 				 errdetail("Failed while allocating a WAL reading processor.")));
 
 	XLogBeginRead(xlogreader, lsn);
-	record = XLogReadRecord(xlogreader, &errormsg);
+	while (XLogReadRecord(xlogreader, &record, &errormsg) ==
+		   XLREAD_NEED_DATA)
+	{
+		if (!read_local_xlog_page(xlogreader))
+			break;
+	}
 
 	/*
 	 * Restore immediately the timeline where it was previously, as

src/backend/access/transam/xlog.c

@@ -811,17 +811,13 @@ static XLogSegNo openLogSegNo = 0;
  * These variables are used similarly to the ones above, but for reading
  * the XLOG.  Note, however, that readOff generally represents the offset
  * of the page just read, not the seek position of the FD itself, which
- * will be just past that page. readLen indicates how much of the current
+ * will be just past that page. readSource indicates where we got the
- * page has been read into readBuf, and readSource indicates where we got
+ * currently open file from.
- * the currently open file from.
  * Note: we could use Reserve/ReleaseExternalFD to track consumption of
  * this FD too; but it doesn't currently seem worthwhile, since the XLOG is
  * not read by general-purpose sessions.
  */
 static int	readFile = -1;
-static XLogSegNo readSegNo = 0;
-static uint32 readOff = 0;
-static uint32 readLen = 0;
 static XLogSource readSource = XLOG_FROM_ANY;
 
 /*
@@ -838,13 +834,6 @@ static XLogSource currentSource = XLOG_FROM_ANY;
 static bool lastSourceFailed = false;
 static bool pendingWalRcvRestart = false;
 
-typedef struct XLogPageReadPrivate
-{
-	int			emode;
-	bool		fetching_ckpt;	/* are we fetching a checkpoint record? */
-	bool		randAccess;
-} XLogPageReadPrivate;
-
 /*
  * These variables track when we last obtained some WAL data to process,
  * and where we got it from.  (XLogReceiptSource is initially the same as
@@ -920,10 +909,12 @@ static bool InstallXLogFileSegment(XLogSegNo *segno, char *tmppath,
 static int	XLogFileRead(XLogSegNo segno, int emode, TimeLineID tli,
 						 XLogSource source, bool notfoundOk);
 static int	XLogFileReadAnyTLI(XLogSegNo segno, int emode, XLogSource source);
-static int	XLogPageRead(XLogReaderState *xlogreader, XLogRecPtr targetPagePtr,
+static bool XLogPageRead(XLogReaderState *state,
-						 int reqLen, XLogRecPtr targetRecPtr, char *readBuf);
+						 bool fetching_ckpt, int emode, bool randAccess);
 static bool WaitForWALToBecomeAvailable(XLogRecPtr RecPtr, bool randAccess,
-										bool fetching_ckpt, XLogRecPtr tliRecPtr);
+										bool fetching_ckpt,
+										XLogRecPtr tliRecPtr,
+										XLogSegNo readSegNo);
 static int	emode_for_corrupt_record(int emode, XLogRecPtr RecPtr);
 static void XLogFileClose(void);
 static void PreallocXlogFiles(XLogRecPtr endptr);
@@ -1234,8 +1225,7 @@ XLogInsertRecord(XLogRecData *rdata,
 			appendBinaryStringInfo(&recordBuf, rdata->data, rdata->len);
 
 		if (!debug_reader)
-			debug_reader = XLogReaderAllocate(wal_segment_size, NULL,
+			debug_reader = XLogReaderAllocate(wal_segment_size, NULL, NULL);
-											  XL_ROUTINE(), NULL);
 
 		if (!debug_reader)
 		{
@@ -4373,12 +4363,7 @@ ReadRecord(XLogReaderState *xlogreader, int emode,
 		   bool fetching_ckpt)
 {
 	XLogRecord *record;
-	XLogPageReadPrivate *private = (XLogPageReadPrivate *) xlogreader->private_data;
+	bool		randAccess = (xlogreader->ReadRecPtr == InvalidXLogRecPtr);
-
-	/* Pass through parameters to XLogPageRead */
-	private->fetching_ckpt = fetching_ckpt;
-	private->emode = emode;
-	private->randAccess = (xlogreader->ReadRecPtr == InvalidXLogRecPtr);
 
 	/* This is the first attempt to read this page. */
 	lastSourceFailed = false;
@@ -4386,8 +4371,16 @@ ReadRecord(XLogReaderState *xlogreader, int emode,
 	for (;;)
 	{
 		char	   *errormsg;
+		XLogReadRecordResult result;
+
+		while ((result = XLogReadRecord(xlogreader, &record, &errormsg))
+			   == XLREAD_NEED_DATA)
+		{
+			if (!XLogPageRead(xlogreader, fetching_ckpt, emode, randAccess))
+				break;
+
+		}
 
-		record = XLogReadRecord(xlogreader, &errormsg);
 		ReadRecPtr = xlogreader->ReadRecPtr;
 		EndRecPtr = xlogreader->EndRecPtr;
 		if (record == NULL)
@@ -6457,7 +6450,6 @@ StartupXLOG(void)
 	bool		backupFromStandby = false;
 	DBState		dbstate_at_startup;
 	XLogReaderState *xlogreader;
-	XLogPageReadPrivate private;
 	bool		promoted = false;
 	struct stat st;
 
@@ -6616,13 +6608,9 @@ StartupXLOG(void)
 		OwnLatch(&XLogCtl->recoveryWakeupLatch);
 
 	/* Set up XLOG reader facility */
-	MemSet(&private, 0, sizeof(XLogPageReadPrivate));
 	xlogreader =
-		XLogReaderAllocate(wal_segment_size, NULL,
+		XLogReaderAllocate(wal_segment_size, NULL, wal_segment_close);
-						   XL_ROUTINE(.page_read = &XLogPageRead,
+
-									  .segment_open = NULL,
-									  .segment_close = wal_segment_close),
-						   &private);
 	if (!xlogreader)
 		ereport(ERROR,
 				(errcode(ERRCODE_OUT_OF_MEMORY),
@@ -7819,7 +7807,8 @@ StartupXLOG(void)
 		XLogRecPtr	pageBeginPtr;
 
 		pageBeginPtr = EndOfLog - (EndOfLog % XLOG_BLCKSZ);
-		Assert(readOff == XLogSegmentOffset(pageBeginPtr, wal_segment_size));
+		Assert(XLogSegmentOffset(xlogreader->readPagePtr, wal_segment_size) ==
+			   XLogSegmentOffset(pageBeginPtr, wal_segment_size));
 
 		firstIdx = XLogRecPtrToBufIdx(EndOfLog);
 
@@ -12107,13 +12096,15 @@ CancelBackup(void)
  * XLogPageRead() to try fetching the record from another source, or to
  * sleep and retry.
  */
-static int
+static bool
-XLogPageRead(XLogReaderState *xlogreader, XLogRecPtr targetPagePtr, int reqLen,
+XLogPageRead(XLogReaderState *state,
-			 XLogRecPtr targetRecPtr, char *readBuf)
+			 bool fetching_ckpt, int emode, bool randAccess)
 {
-	XLogPageReadPrivate *private =
+	char *readBuf				= state->readBuf;
-	(XLogPageReadPrivate *) xlogreader->private_data;
+	XLogRecPtr	targetPagePtr	= state->readPagePtr;
-	int			emode = private->emode;
+	int			reqLen			= state->reqLen;
+	int			readLen			= 0;
+	XLogRecPtr	targetRecPtr	= state->ReadRecPtr;
 	uint32		targetPageOff;
 	XLogSegNo	targetSegNo PG_USED_FOR_ASSERTS_ONLY;
 	int			r;
@@ -12126,7 +12117,7 @@ XLogPageRead(XLogReaderState *xlogreader, XLogRecPtr targetPagePtr, int reqLen,
 	 * is not in the currently open one.
 	 */
 	if (readFile >= 0 &&
-		!XLByteInSeg(targetPagePtr, readSegNo, wal_segment_size))
+		!XLByteInSeg(targetPagePtr, state->seg.ws_segno, wal_segment_size))
 	{
 		/*
 		 * Request a restartpoint if we've replayed too much xlog since the
@@ -12134,10 +12125,10 @@ XLogPageRead(XLogReaderState *xlogreader, XLogRecPtr targetPagePtr, int reqLen,
 		 */
 		if (bgwriterLaunched)
 		{
-			if (XLogCheckpointNeeded(readSegNo))
+			if (XLogCheckpointNeeded(state->seg.ws_segno))
 			{
 				(void) GetRedoRecPtr();
-				if (XLogCheckpointNeeded(readSegNo))
+				if (XLogCheckpointNeeded(state->seg.ws_segno))
 					RequestCheckpoint(CHECKPOINT_CAUSE_XLOG);
 			}
 		}
@@ -12147,7 +12138,7 @@ XLogPageRead(XLogReaderState *xlogreader, XLogRecPtr targetPagePtr, int reqLen,
 		readSource = XLOG_FROM_ANY;
 	}
 
-	XLByteToSeg(targetPagePtr, readSegNo, wal_segment_size);
+	XLByteToSeg(targetPagePtr, state->seg.ws_segno, wal_segment_size);
 
 retry:
 	/* See if we need to retrieve more data */
@@ -12156,17 +12147,15 @@ retry:
 		 flushedUpto < targetPagePtr + reqLen))
 	{
 		if (!WaitForWALToBecomeAvailable(targetPagePtr + reqLen,
-										 private->randAccess,
+										 randAccess, fetching_ckpt,
-										 private->fetching_ckpt,
+										 targetRecPtr, state->seg.ws_segno))
-										 targetRecPtr))
 		{
 			if (readFile >= 0)
 				close(readFile);
 			readFile = -1;
-			readLen = 0;
 			readSource = XLOG_FROM_ANY;
-
+			XLogReaderSetInputData(state, -1);
-			return -1;
+			return false;
 		}
 	}
 
@@ -12193,40 +12182,36 @@ retry:
 	else
 		readLen = XLOG_BLCKSZ;
 
-	/* Read the requested page */
-	readOff = targetPageOff;
-
 	pgstat_report_wait_start(WAIT_EVENT_WAL_READ);
-	r = pg_pread(readFile, readBuf, XLOG_BLCKSZ, (off_t) readOff);
+	r = pg_pread(readFile, readBuf, XLOG_BLCKSZ, (off_t) targetPageOff);
 	if (r != XLOG_BLCKSZ)
 	{
 		char		fname[MAXFNAMELEN];
 		int			save_errno = errno;
 
 		pgstat_report_wait_end();
-		XLogFileName(fname, curFileTLI, readSegNo, wal_segment_size);
+		XLogFileName(fname, curFileTLI, state->seg.ws_segno, wal_segment_size);
 		if (r < 0)
 		{
 			errno = save_errno;
 			ereport(emode_for_corrupt_record(emode, targetPagePtr + reqLen),
 					(errcode_for_file_access(),
 					 errmsg("could not read from log segment %s, offset %u: %m",
-							fname, readOff)));
+							fname, targetPageOff)));
 		}
 		else
 			ereport(emode_for_corrupt_record(emode, targetPagePtr + reqLen),
 					(errcode(ERRCODE_DATA_CORRUPTED),
 					 errmsg("could not read from log segment %s, offset %u: read %d of %zu",
-							fname, readOff, r, (Size) XLOG_BLCKSZ)));
+							fname, targetPageOff, r, (Size) XLOG_BLCKSZ)));
 		goto next_record_is_invalid;
 	}
 	pgstat_report_wait_end();
 
-	Assert(targetSegNo == readSegNo);
+	Assert(targetSegNo == state->seg.ws_segno);
-	Assert(targetPageOff == readOff);
+	Assert(readLen >= reqLen);
-	Assert(reqLen <= readLen);
 
-	xlogreader->seg.ws_tli = curFileTLI;
+	state->seg.ws_tli = curFileTLI;
 
 	/*
 	 * Check the page header immediately, so that we can retry immediately if
@@ -12254,14 +12239,16 @@ retry:
 	 * Validating the page header is cheap enough that doing it twice
 	 * shouldn't be a big deal from a performance point of view.
 	 */
-	if (!XLogReaderValidatePageHeader(xlogreader, targetPagePtr, readBuf))
+	if (!XLogReaderValidatePageHeader(state, targetPagePtr, readBuf))
 	{
-		/* reset any error XLogReaderValidatePageHeader() might have set */
+		/* reset any error StateValidatePageHeader() might have set */
-		xlogreader->errormsg_buf[0] = '\0';
+		state->errormsg_buf[0] = '\0';
 		goto next_record_is_invalid;
 	}
 
-	return readLen;
+	Assert(state->readPagePtr == targetPagePtr);
+	XLogReaderSetInputData(state, readLen);
+	return true;
 
 next_record_is_invalid:
 	lastSourceFailed = true;
@@ -12269,14 +12256,14 @@ next_record_is_invalid:
 	if (readFile >= 0)
 		close(readFile);
 	readFile = -1;
-	readLen = 0;
 	readSource = XLOG_FROM_ANY;
 
 	/* In standby-mode, keep trying */
 	if (StandbyMode)
 		goto retry;
-	else
+
-		return -1;
+	XLogReaderSetInputData(state, -1);
+	return false;
 }
 
 /*
@@ -12307,7 +12294,8 @@ next_record_is_invalid:
  */
 static bool
 WaitForWALToBecomeAvailable(XLogRecPtr RecPtr, bool randAccess,
-							bool fetching_ckpt, XLogRecPtr tliRecPtr)
+							bool fetching_ckpt, XLogRecPtr tliRecPtr,
+							XLogSegNo readSegNo)
 {
 	static TimestampTz last_fail_time = 0;
 	TimestampTz now;

src/backend/access/transam/xlogutils.c

@@ -686,8 +686,7 @@ XLogTruncateRelation(RelFileNode rnode, ForkNumber forkNum,
 void
 XLogReadDetermineTimeline(XLogReaderState *state, XLogRecPtr wantPage, uint32 wantLength)
 {
-	const XLogRecPtr lastReadPage = (state->seg.ws_segno *
+	const XLogRecPtr lastReadPage = state->readPagePtr;
-									 state->segcxt.ws_segsize + state->segoff);
 
 	Assert(wantPage != InvalidXLogRecPtr && wantPage % XLOG_BLCKSZ == 0);
 	Assert(wantLength <= XLOG_BLCKSZ);
@@ -702,7 +701,7 @@ XLogReadDetermineTimeline(XLogReaderState *state, XLogRecPtr wantPage, uint32 wa
 	 * current TLI has since become historical.
 	 */
 	if (lastReadPage == wantPage &&
-		state->readLen != 0 &&
+		state->page_verified &&
 		lastReadPage + state->readLen >= wantPage + Min(wantLength, XLOG_BLCKSZ - 1))
 		return;
 
@@ -824,10 +823,12 @@ wal_segment_close(XLogReaderState *state)
  * exists for normal backends, so we have to do a check/sleep/repeat style of
  * loop for now.
  */
-int
+bool
-read_local_xlog_page(XLogReaderState *state, XLogRecPtr targetPagePtr,
+read_local_xlog_page(XLogReaderState *state)
-					 int reqLen, XLogRecPtr targetRecPtr, char *cur_page)
 {
+	XLogRecPtr	targetPagePtr = state->readPagePtr;
+	int			reqLen		  = state->reqLen;
+	char	   *cur_page	  = state->readBuf;
 	XLogRecPtr	read_upto,
 				loc;
 	TimeLineID	tli;
@@ -926,7 +927,8 @@ read_local_xlog_page(XLogReaderState *state, XLogRecPtr targetPagePtr,
 	else if (targetPagePtr + reqLen > read_upto)
 	{
 		/* not enough data there */
-		return -1;
+		XLogReaderSetInputData(state,  -1);
+		return false;
 	}
 	else
 	{
@@ -939,12 +941,14 @@ read_local_xlog_page(XLogReaderState *state, XLogRecPtr targetPagePtr,
 	 * as 'count', read the whole page anyway. It's guaranteed to be
 	 * zero-padded up to the page boundary if it's incomplete.
 	 */
-	if (!WALRead(state, cur_page, targetPagePtr, XLOG_BLCKSZ, tli,
+	if (!WALRead(state, wal_segment_open, wal_segment_close,
-				 &errinfo))
+				 cur_page, targetPagePtr, XLOG_BLCKSZ, tli, &errinfo))
 		WALReadRaiseError(&errinfo);
 
 	/* number of valid bytes in the buffer */
-	return count;
+	state->readPagePtr = targetPagePtr;
+	XLogReaderSetInputData(state, count);
+	return true;
 }
 
 /*

src/backend/replication/logical/logical.c

@@ -148,7 +148,8 @@ StartupDecodingContext(List *output_plugin_options,
 					   TransactionId xmin_horizon,
 					   bool need_full_snapshot,
 					   bool fast_forward,
-					   XLogReaderRoutine *xl_routine,
+					   LogicalDecodingXLogPageReadCB page_read,
+					   WALSegmentCleanupCB cleanup_cb,
 					   LogicalOutputPluginWriterPrepareWrite prepare_write,
 					   LogicalOutputPluginWriterWrite do_write,
 					   LogicalOutputPluginWriterUpdateProgress update_progress)
@@ -198,11 +199,12 @@ StartupDecodingContext(List *output_plugin_options,
 
 	ctx->slot = slot;
 
-	ctx->reader = XLogReaderAllocate(wal_segment_size, NULL, xl_routine, ctx);
+	ctx->reader = XLogReaderAllocate(wal_segment_size, NULL, cleanup_cb);
 	if (!ctx->reader)
 		ereport(ERROR,
 				(errcode(ERRCODE_OUT_OF_MEMORY),
 				 errmsg("out of memory")));
+	ctx->page_read = page_read;
 
 	ctx->reorder = ReorderBufferAllocate();
 	ctx->snapshot_builder =
@@ -319,7 +321,8 @@ CreateInitDecodingContext(const char *plugin,
 						  List *output_plugin_options,
 						  bool need_full_snapshot,
 						  XLogRecPtr restart_lsn,
-						  XLogReaderRoutine *xl_routine,
+						  LogicalDecodingXLogPageReadCB page_read,
+						  WALSegmentCleanupCB cleanup_cb,
 						  LogicalOutputPluginWriterPrepareWrite prepare_write,
 						  LogicalOutputPluginWriterWrite do_write,
 						  LogicalOutputPluginWriterUpdateProgress update_progress)
@@ -422,7 +425,7 @@ CreateInitDecodingContext(const char *plugin,
 
 	ctx = StartupDecodingContext(NIL, restart_lsn, xmin_horizon,
 								 need_full_snapshot, false,
-								 xl_routine, prepare_write, do_write,
+								 page_read, cleanup_cb, prepare_write, do_write,
 								 update_progress);
 
 	/* call output plugin initialization callback */
@@ -476,7 +479,8 @@ LogicalDecodingContext *
 CreateDecodingContext(XLogRecPtr start_lsn,
 					  List *output_plugin_options,
 					  bool fast_forward,
-					  XLogReaderRoutine *xl_routine,
+					  LogicalDecodingXLogPageReadCB page_read,
+					  WALSegmentCleanupCB cleanup_cb,
 					  LogicalOutputPluginWriterPrepareWrite prepare_write,
 					  LogicalOutputPluginWriterWrite do_write,
 					  LogicalOutputPluginWriterUpdateProgress update_progress)
@@ -528,8 +532,8 @@ CreateDecodingContext(XLogRecPtr start_lsn,
 
 	ctx = StartupDecodingContext(output_plugin_options,
 								 start_lsn, InvalidTransactionId, false,
-								 fast_forward, xl_routine, prepare_write,
+								 fast_forward, page_read, cleanup_cb,
-								 do_write, update_progress);
+								 prepare_write, do_write, update_progress);
 
 	/* call output plugin initialization callback */
 	old_context = MemoryContextSwitchTo(ctx->context);
@@ -585,7 +589,13 @@ DecodingContextFindStartpoint(LogicalDecodingContext *ctx)
 		char	   *err = NULL;
 
 		/* the read_page callback waits for new WAL */
-		record = XLogReadRecord(ctx->reader, &err);
+		while (XLogReadRecord(ctx->reader, &record, &err) ==
+			   XLREAD_NEED_DATA)
+		{
+			if (!ctx->page_read(ctx->reader))
+				break;
+		}
+
 		if (err)
 			elog(ERROR, "%s", err);
 		if (!record)

src/backend/replication/logical/logicalfuncs.c

@@ -233,9 +233,8 @@ pg_logical_slot_get_changes_guts(FunctionCallInfo fcinfo, bool confirm, bool bin
 		ctx = CreateDecodingContext(InvalidXLogRecPtr,
 									options,
 									false,
-									XL_ROUTINE(.page_read = read_local_xlog_page,
+									read_local_xlog_page,
-											   .segment_open = wal_segment_open,
+									wal_segment_close,
-											   .segment_close = wal_segment_close),
 									LogicalOutputPrepareWrite,
 									LogicalOutputWrite, NULL);
 
@@ -284,7 +283,13 @@ pg_logical_slot_get_changes_guts(FunctionCallInfo fcinfo, bool confirm, bool bin
 			XLogRecord *record;
 			char	   *errm = NULL;
 
-			record = XLogReadRecord(ctx->reader, &errm);
+			while (XLogReadRecord(ctx->reader, &record, &errm) ==
+				   XLREAD_NEED_DATA)
+			{
+				if (!ctx->page_read(ctx->reader))
+					break;
+			}
+
 			if (errm)
 				elog(ERROR, "%s", errm);
 

src/backend/replication/slotfuncs.c

@@ -153,9 +153,8 @@ create_logical_replication_slot(char *name, char *plugin,
 	ctx = CreateInitDecodingContext(plugin, NIL,
 									false,	/* just catalogs is OK */
 									restart_lsn,
-									XL_ROUTINE(.page_read = read_local_xlog_page,
+									read_local_xlog_page,
-											   .segment_open = wal_segment_open,
+									wal_segment_close,
-											   .segment_close = wal_segment_close),
 									NULL, NULL, NULL);
 
 	/*
@@ -512,9 +511,8 @@ pg_logical_replication_slot_advance(XLogRecPtr moveto)
 		ctx = CreateDecodingContext(InvalidXLogRecPtr,
 									NIL,
 									true,	/* fast_forward */
-									XL_ROUTINE(.page_read = read_local_xlog_page,
+									read_local_xlog_page,
-											   .segment_open = wal_segment_open,
+									wal_segment_close,
-											   .segment_close = wal_segment_close),
 									NULL, NULL, NULL);
 
 		/*
@@ -536,7 +534,13 @@ pg_logical_replication_slot_advance(XLogRecPtr moveto)
 			 * Read records.  No changes are generated in fast_forward mode,
 			 * but snapbuilder/slot statuses are updated properly.
 			 */
-			record = XLogReadRecord(ctx->reader, &errm);
+			while (XLogReadRecord(ctx->reader, &record, &errm) ==
+				   XLREAD_NEED_DATA)
+			{
+				if (!ctx->page_read(ctx->reader))
+					break;
+			}
+
 			if (errm)
 				elog(ERROR, "%s", errm);
 

src/backend/replication/walsender.c

@@ -580,10 +580,7 @@ StartReplication(StartReplicationCmd *cmd)
 
 	/* create xlogreader for physical replication */
 	xlogreader =
-		XLogReaderAllocate(wal_segment_size, NULL,
+		XLogReaderAllocate(wal_segment_size, NULL, wal_segment_close);
-						   XL_ROUTINE(.segment_open = WalSndSegmentOpen,
-									  .segment_close = wal_segment_close),
-						   NULL);
 
 	if (!xlogreader)
 		ereport(ERROR,
@@ -806,10 +803,12 @@ StartReplication(StartReplicationCmd *cmd)
  * which has to do a plain sleep/busy loop, because the walsender's latch gets
  * set every time WAL is flushed.
  */
-static int
+static bool
-logical_read_xlog_page(XLogReaderState *state, XLogRecPtr targetPagePtr, int reqLen,
+logical_read_xlog_page(XLogReaderState *state)
-					   XLogRecPtr targetRecPtr, char *cur_page)
 {
+	XLogRecPtr		targetPagePtr = state->readPagePtr;
+	int				reqLen		  = state->reqLen;
+	char		   *cur_page	  = state->readBuf;
 	XLogRecPtr	flushptr;
 	int			count;
 	WALReadError errinfo;
@@ -826,7 +825,10 @@ logical_read_xlog_page(XLogReaderState *state, XLogRecPtr targetPagePtr, int req
 
 	/* fail if not (implies we are going to shut down) */
 	if (flushptr < targetPagePtr + reqLen)
-		return -1;
+	{
+		XLogReaderSetInputData(state, -1);
+		return false;
+	}
 
 	if (targetPagePtr + XLOG_BLCKSZ <= flushptr)
 		count = XLOG_BLCKSZ;	/* more than one block available */
@@ -834,7 +836,7 @@ logical_read_xlog_page(XLogReaderState *state, XLogRecPtr targetPagePtr, int req
 		count = flushptr - targetPagePtr;	/* part of the page available */
 
 	/* now actually read the data, we know it's there */
-	if (!WALRead(state,
+	if (!WALRead(state, WalSndSegmentOpen, wal_segment_close,
 				 cur_page,
 				 targetPagePtr,
 				 XLOG_BLCKSZ,
@@ -854,7 +856,8 @@ logical_read_xlog_page(XLogReaderState *state, XLogRecPtr targetPagePtr, int req
 	XLByteToSeg(targetPagePtr, segno, state->segcxt.ws_segsize);
 	CheckXLogRemoved(segno, state->seg.ws_tli);
 
-	return count;
+	XLogReaderSetInputData(state, count);
+	return true;
 }
 
 /*
@@ -1007,9 +1010,8 @@ CreateReplicationSlot(CreateReplicationSlotCmd *cmd)
 
 		ctx = CreateInitDecodingContext(cmd->plugin, NIL, need_full_snapshot,
 										InvalidXLogRecPtr,
-										XL_ROUTINE(.page_read = logical_read_xlog_page,
+										logical_read_xlog_page,
-												   .segment_open = WalSndSegmentOpen,
+										wal_segment_close,
-												   .segment_close = wal_segment_close),
 										WalSndPrepareWrite, WalSndWriteData,
 										WalSndUpdateProgress);
 
@@ -1167,9 +1169,8 @@ StartLogicalReplication(StartReplicationCmd *cmd)
 	 */
 	logical_decoding_ctx =
 		CreateDecodingContext(cmd->startpoint, cmd->options, false,
-							  XL_ROUTINE(.page_read = logical_read_xlog_page,
+							  logical_read_xlog_page,
-										 .segment_open = WalSndSegmentOpen,
+							  wal_segment_close,
-										 .segment_close = wal_segment_close),
 							  WalSndPrepareWrite, WalSndWriteData,
 							  WalSndUpdateProgress);
 	xlogreader = logical_decoding_ctx->reader;
@@ -2745,7 +2746,7 @@ XLogSendPhysical(void)
 	enlargeStringInfo(&output_message, nbytes);
 
 retry:
-	if (!WALRead(xlogreader,
+	if (!WALRead(xlogreader, WalSndSegmentOpen, wal_segment_close,
 				 &output_message.data[output_message.len],
 				 startptr,
 				 nbytes,
@@ -2843,7 +2844,12 @@ XLogSendLogical(void)
 	 */
 	WalSndCaughtUp = false;
 
-	record = XLogReadRecord(logical_decoding_ctx->reader, &errm);
+	while (XLogReadRecord(logical_decoding_ctx->reader, &record, &errm) ==
+		   XLREAD_NEED_DATA)
+	{
+		if (!logical_decoding_ctx->page_read(logical_decoding_ctx->reader))
+			break;
+	}
 
 	/* xlog record was invalid */
 	if (errm != NULL)

src/bin/pg_rewind/parsexlog.c

@@ -41,15 +41,9 @@ static int	xlogreadfd = -1;
 static XLogSegNo xlogreadsegno = -1;
 static char xlogfpath[MAXPGPATH];
 
-typedef struct XLogPageReadPrivate
+static bool SimpleXLogPageRead(XLogReaderState *xlogreader,
-{
+							   const char *datadir, int *tliIndex,
-	const char *restoreCommand;
+							   const char *restoreCommand);
-	int			tliIndex;
-} XLogPageReadPrivate;
-
-static int	SimpleXLogPageRead(XLogReaderState *xlogreader,
-							   XLogRecPtr targetPagePtr,
-							   int reqLen, XLogRecPtr targetRecPtr, char *readBuf);
 
 /*
  * Read WAL from the datadir/pg_wal, starting from 'startpoint' on timeline
@@ -66,20 +60,22 @@ extractPageMap(const char *datadir, XLogRecPtr startpoint, int tliIndex,
 	XLogRecord *record;
 	XLogReaderState *xlogreader;
 	char	   *errormsg;
-	XLogPageReadPrivate private;
 
-	private.tliIndex = tliIndex;
+	xlogreader = XLogReaderAllocate(WalSegSz, datadir, NULL);
-	private.restoreCommand = restoreCommand;
+
-	xlogreader = XLogReaderAllocate(WalSegSz, datadir,
-									XL_ROUTINE(.page_read = &SimpleXLogPageRead),
-									&private);
 	if (xlogreader == NULL)
 		pg_fatal("out of memory");
 
 	XLogBeginRead(xlogreader, startpoint);
 	do
 	{
-		record = XLogReadRecord(xlogreader, &errormsg);
+		while (XLogReadRecord(xlogreader, &record, &errormsg) ==
+			   XLREAD_NEED_DATA)
+		{
+			if (!SimpleXLogPageRead(xlogreader, datadir,
+									&tliIndex, restoreCommand))
+				break;
+		}
 
 		if (record == NULL)
 		{
@@ -123,19 +119,19 @@ readOneRecord(const char *datadir, XLogRecPtr ptr, int tliIndex,
 	XLogRecord *record;
 	XLogReaderState *xlogreader;
 	char	   *errormsg;
-	XLogPageReadPrivate private;
 	XLogRecPtr	endptr;
 
-	private.tliIndex = tliIndex;
+	xlogreader = XLogReaderAllocate(WalSegSz, datadir, NULL);
-	private.restoreCommand = restoreCommand;
-	xlogreader = XLogReaderAllocate(WalSegSz, datadir,
-									XL_ROUTINE(.page_read = &SimpleXLogPageRead),
-									&private);
 	if (xlogreader == NULL)
 		pg_fatal("out of memory");
 
 	XLogBeginRead(xlogreader, ptr);
-	record = XLogReadRecord(xlogreader, &errormsg);
+	while (XLogReadRecord(xlogreader, &record, &errormsg) ==
+		   XLREAD_NEED_DATA)
+	{
+		if (!SimpleXLogPageRead(xlogreader, datadir, &tliIndex, restoreCommand))
+			break;
+	}
 	if (record == NULL)
 	{
 		if (errormsg)
@@ -170,7 +166,6 @@ findLastCheckpoint(const char *datadir, XLogRecPtr forkptr, int tliIndex,
 	XLogRecPtr	searchptr;
 	XLogReaderState *xlogreader;
 	char	   *errormsg;
-	XLogPageReadPrivate private;
 
 	/*
 	 * The given fork pointer points to the end of the last common record,
@@ -186,11 +181,7 @@ findLastCheckpoint(const char *datadir, XLogRecPtr forkptr, int tliIndex,
 			forkptr += SizeOfXLogShortPHD;
 	}
 
-	private.tliIndex = tliIndex;
+	xlogreader = XLogReaderAllocate(WalSegSz, datadir, NULL);
-	private.restoreCommand = restoreCommand;
-	xlogreader = XLogReaderAllocate(WalSegSz, datadir,
-									XL_ROUTINE(.page_read = &SimpleXLogPageRead),
-									&private);
 	if (xlogreader == NULL)
 		pg_fatal("out of memory");
 
@@ -200,7 +191,13 @@ findLastCheckpoint(const char *datadir, XLogRecPtr forkptr, int tliIndex,
 		uint8		info;
 
 		XLogBeginRead(xlogreader, searchptr);
-		record = XLogReadRecord(xlogreader, &errormsg);
+		while (XLogReadRecord(xlogreader, &record, &errormsg) ==
+			   XLREAD_NEED_DATA)
+		{
+			if (!SimpleXLogPageRead(xlogreader, datadir,
+									&tliIndex, restoreCommand))
+				break;
+		}
 
 		if (record == NULL)
 		{
@@ -246,16 +243,19 @@ findLastCheckpoint(const char *datadir, XLogRecPtr forkptr, int tliIndex,
 }
 
 /* XLogReader callback function, to read a WAL page */
-static int
+static bool
-SimpleXLogPageRead(XLogReaderState *xlogreader, XLogRecPtr targetPagePtr,
+SimpleXLogPageRead(XLogReaderState *xlogreader, const char *datadir,
-				   int reqLen, XLogRecPtr targetRecPtr, char *readBuf)
+				   int *tliIndex, const char *restoreCommand)
 {
-	XLogPageReadPrivate *private = (XLogPageReadPrivate *) xlogreader->private_data;
+	XLogRecPtr	targetPagePtr = xlogreader->readPagePtr;
+	char	   *readBuf		  = xlogreader->readBuf;
 	uint32		targetPageOff;
 	XLogRecPtr	targetSegEnd;
 	XLogSegNo	targetSegNo;
 	int			r;
 
+	Assert(xlogreader->reqLen <= XLOG_BLCKSZ);
+
 	XLByteToSeg(targetPagePtr, targetSegNo, WalSegSz);
 	XLogSegNoOffsetToRecPtr(targetSegNo + 1, 0, WalSegSz, targetSegEnd);
 	targetPageOff = XLogSegmentOffset(targetPagePtr, WalSegSz);
@@ -283,14 +283,14 @@ SimpleXLogPageRead(XLogReaderState *xlogreader, XLogRecPtr targetPagePtr,
 		 * be done both forward and backward, consider also switching timeline
 		 * accordingly.
 		 */
-		while (private->tliIndex < targetNentries - 1 &&
+		while (*tliIndex < targetNentries - 1 &&
-			   targetHistory[private->tliIndex].end < targetSegEnd)
+			   targetHistory[*tliIndex].end < targetSegEnd)
-			private->tliIndex++;
+			(*tliIndex)++;
-		while (private->tliIndex > 0 &&
+		while (*tliIndex > 0 &&
-			   targetHistory[private->tliIndex].begin >= targetSegEnd)
+			   targetHistory[*tliIndex].begin >= targetSegEnd)
-			private->tliIndex--;
+			(*tliIndex)--;
 
-		XLogFileName(xlogfname, targetHistory[private->tliIndex].tli,
+		XLogFileName(xlogfname, targetHistory[*tliIndex].tli,
 					 xlogreadsegno, WalSegSz);
 
 		snprintf(xlogfpath, MAXPGPATH, "%s/" XLOGDIR "/%s",
@@ -303,10 +303,11 @@ SimpleXLogPageRead(XLogReaderState *xlogreader, XLogRecPtr targetPagePtr,
 			/*
 			 * If we have no restore_command to execute, then exit.
 			 */
-			if (private->restoreCommand == NULL)
+			if (restoreCommand == NULL)
 			{
 				pg_log_error("could not open file \"%s\": %m", xlogfpath);
-				return -1;
+				XLogReaderSetInputData(xlogreader, -1);
+				return false;
 			}
 
 			/*
@@ -316,10 +317,13 @@ SimpleXLogPageRead(XLogReaderState *xlogreader, XLogRecPtr targetPagePtr,
 			xlogreadfd = RestoreArchivedFile(xlogreader->segcxt.ws_dir,
 											 xlogfname,
 											 WalSegSz,
-											 private->restoreCommand);
+											 restoreCommand);
 
 			if (xlogreadfd < 0)
-				return -1;
+			{
+				XLogReaderSetInputData(xlogreader, -1);
+				return false;
+			}
 			else
 				pg_log_debug("using file \"%s\" restored from archive",
 							 xlogfpath);
@@ -335,7 +339,8 @@ SimpleXLogPageRead(XLogReaderState *xlogreader, XLogRecPtr targetPagePtr,
 	if (lseek(xlogreadfd, (off_t) targetPageOff, SEEK_SET) < 0)
 	{
 		pg_log_error("could not seek in file \"%s\": %m", xlogfpath);
-		return -1;
+		XLogReaderSetInputData(xlogreader, -1);
+		return false;
 	}
 
 
@@ -348,13 +353,15 @@ SimpleXLogPageRead(XLogReaderState *xlogreader, XLogRecPtr targetPagePtr,
 			pg_log_error("could not read file \"%s\": read %d of %zu",
 						 xlogfpath, r, (Size) XLOG_BLCKSZ);
 
-		return -1;
+		XLogReaderSetInputData(xlogreader, -1);
+		return false;
 	}
 
 	Assert(targetSegNo == xlogreadsegno);
 
-	xlogreader->seg.ws_tli = targetHistory[private->tliIndex].tli;
+	xlogreader->seg.ws_tli = targetHistory[*tliIndex].tli;
-	return XLOG_BLCKSZ;
+	XLogReaderSetInputData(xlogreader, XLOG_BLCKSZ);
+	return true;
 }
 
 /*

src/bin/pg_waldump/pg_waldump.c

@@ -29,14 +29,6 @@ static const char *progname;
 
 static int	WalSegSz;
 
-typedef struct XLogDumpPrivate
-{
-	TimeLineID	timeline;
-	XLogRecPtr	startptr;
-	XLogRecPtr	endptr;
-	bool		endptr_reached;
-} XLogDumpPrivate;
-
 typedef struct XLogDumpConfig
 {
 	/* display options */
@@ -330,30 +322,41 @@ WALDumpCloseSegment(XLogReaderState *state)
 	state->seg.ws_file = -1;
 }
 
-/* pg_waldump's XLogReaderRoutine->page_read callback */
+/*
-static int
+ * pg_waldump's WAL page rader
-WALDumpReadPage(XLogReaderState *state, XLogRecPtr targetPagePtr, int reqLen,
+ *
-				XLogRecPtr targetPtr, char *readBuff)
+ * timeline and startptr specifies the LSN, and reads up to endptr.
+ */
+static bool
+WALDumpReadPage(XLogReaderState *state, TimeLineID timeline,
+				XLogRecPtr startptr, XLogRecPtr endptr)
 {
-	XLogDumpPrivate *private = state->private_data;
+	XLogRecPtr	targetPagePtr = state->readPagePtr;
+	int			reqLen		  = state->reqLen;
+	char	   *readBuff	  = state->readBuf;
 	int			count = XLOG_BLCKSZ;
 	WALReadError errinfo;
 
-	if (private->endptr != InvalidXLogRecPtr)
+	/* determine the number of bytes to read on the page */
+	if (endptr != InvalidXLogRecPtr)
 	{
-		if (targetPagePtr + XLOG_BLCKSZ <= private->endptr)
+		if (targetPagePtr + XLOG_BLCKSZ <= endptr)
 			count = XLOG_BLCKSZ;
-		else if (targetPagePtr + reqLen <= private->endptr)
+		else if (targetPagePtr + reqLen <= endptr)
-			count = private->endptr - targetPagePtr;
+			count = endptr - targetPagePtr;
 		else
 		{
-			private->endptr_reached = true;
+			/* Notify xlogreader that we didn't read at all */
-			return -1;
+			XLogReaderSetInputData(state,  -1);
+			return false;
 		}
 	}
 
-	if (!WALRead(state, readBuff, targetPagePtr, count, private->timeline,
+	/* We should read more than requested by xlogreader */
-				 &errinfo))
+	Assert(count >= state->readLen);
+
+	if (!WALRead(state, WALDumpOpenSegment, WALDumpCloseSegment,
+				 readBuff, targetPagePtr, count, timeline, &errinfo))
 	{
 		WALOpenSegment *seg = &errinfo.wre_seg;
 		char		fname[MAXPGPATH];
@@ -373,7 +376,9 @@ WALDumpReadPage(XLogReaderState *state, XLogRecPtr targetPagePtr, int reqLen,
 						(Size) errinfo.wre_req);
 	}
 
-	return count;
+	/* Notify xlogreader of how many bytes we have read */
+	XLogReaderSetInputData(state, count);
+	return true;
 }
 
 /*
@@ -754,7 +759,10 @@ main(int argc, char **argv)
 	uint32		xlogid;
 	uint32		xrecoff;
 	XLogReaderState *xlogreader_state;
-	XLogDumpPrivate private;
+	XLogFindNextRecordState *findnext_state;
+	TimeLineID	timeline;
+	XLogRecPtr	startptr;
+	XLogRecPtr	endptr;
 	XLogDumpConfig config;
 	XLogDumpStats stats;
 	XLogRecord *record;
@@ -800,14 +808,12 @@ main(int argc, char **argv)
 		}
 	}
 
-	memset(&private, 0, sizeof(XLogDumpPrivate));
 	memset(&config, 0, sizeof(XLogDumpConfig));
 	memset(&stats, 0, sizeof(XLogDumpStats));
 
-	private.timeline = 1;
+	timeline = 1;
-	private.startptr = InvalidXLogRecPtr;
+	startptr = InvalidXLogRecPtr;
-	private.endptr = InvalidXLogRecPtr;
+	endptr = InvalidXLogRecPtr;
-	private.endptr_reached = false;
 
 	config.quiet = false;
 	config.bkp_details = false;
@@ -841,7 +847,7 @@ main(int argc, char **argv)
 								 optarg);
 					goto bad_argument;
 				}
-				private.endptr = (uint64) xlogid << 32 | xrecoff;
+				endptr = (uint64) xlogid << 32 | xrecoff;
 				break;
 			case 'f':
 				config.follow = true;
@@ -894,10 +900,10 @@ main(int argc, char **argv)
 					goto bad_argument;
 				}
 				else
-					private.startptr = (uint64) xlogid << 32 | xrecoff;
+					startptr = (uint64) xlogid << 32 | xrecoff;
 				break;
 			case 't':
-				if (sscanf(optarg, "%d", &private.timeline) != 1)
+				if (sscanf(optarg, "%d", &timeline) != 1)
 				{
 					pg_log_error("could not parse timeline \"%s\"", optarg);
 					goto bad_argument;
@@ -974,21 +980,21 @@ main(int argc, char **argv)
 		close(fd);
 
 		/* parse position from file */
-		XLogFromFileName(fname, &private.timeline, &segno, WalSegSz);
+		XLogFromFileName(fname, &timeline, &segno, WalSegSz);
 
-		if (XLogRecPtrIsInvalid(private.startptr))
+		if (XLogRecPtrIsInvalid(startptr))
-			XLogSegNoOffsetToRecPtr(segno, 0, WalSegSz, private.startptr);
+			XLogSegNoOffsetToRecPtr(segno, 0, WalSegSz, startptr);
-		else if (!XLByteInSeg(private.startptr, segno, WalSegSz))
+		else if (!XLByteInSeg(startptr, segno, WalSegSz))
 		{
 			pg_log_error("start WAL location %X/%X is not inside file \"%s\"",
-						 LSN_FORMAT_ARGS(private.startptr),
+						 LSN_FORMAT_ARGS(startptr),
 						 fname);
 			goto bad_argument;
 		}
 
 		/* no second file specified, set end position */
-		if (!(optind + 1 < argc) && XLogRecPtrIsInvalid(private.endptr))
+		if (!(optind + 1 < argc) && XLogRecPtrIsInvalid(endptr))
-			XLogSegNoOffsetToRecPtr(segno + 1, 0, WalSegSz, private.endptr);
+			XLogSegNoOffsetToRecPtr(segno + 1, 0, WalSegSz, endptr);
 
 		/* parse ENDSEG if passed */
 		if (optind + 1 < argc)
@@ -1004,26 +1010,26 @@ main(int argc, char **argv)
 			close(fd);
 
 			/* parse position from file */
-			XLogFromFileName(fname, &private.timeline, &endsegno, WalSegSz);
+			XLogFromFileName(fname, &timeline, &endsegno, WalSegSz);
 
 			if (endsegno < segno)
 				fatal_error("ENDSEG %s is before STARTSEG %s",
 							argv[optind + 1], argv[optind]);
 
-			if (XLogRecPtrIsInvalid(private.endptr))
+			if (XLogRecPtrIsInvalid(endptr))
 				XLogSegNoOffsetToRecPtr(endsegno + 1, 0, WalSegSz,
-										private.endptr);
+										endptr);
 
 			/* set segno to endsegno for check of --end */
 			segno = endsegno;
 		}
 
 
-		if (!XLByteInSeg(private.endptr, segno, WalSegSz) &&
+		if (!XLByteInSeg(endptr, segno, WalSegSz) &&
-			private.endptr != (segno + 1) * WalSegSz)
+			endptr != (segno + 1) * WalSegSz)
 		{
 			pg_log_error("end WAL location %X/%X is not inside file \"%s\"",
-						 LSN_FORMAT_ARGS(private.endptr),
+						 LSN_FORMAT_ARGS(endptr),
 						 argv[argc - 1]);
 			goto bad_argument;
 		}
@@ -1032,7 +1038,7 @@ main(int argc, char **argv)
 		waldir = identify_target_directory(waldir, NULL);
 
 	/* we don't know what to print */
-	if (XLogRecPtrIsInvalid(private.startptr))
+	if (XLogRecPtrIsInvalid(startptr))
 	{
 		pg_log_error("no start WAL location given");
 		goto bad_argument;
@@ -1042,42 +1048,56 @@ main(int argc, char **argv)
 
 	/* we have everything we need, start reading */
 	xlogreader_state =
-		XLogReaderAllocate(WalSegSz, waldir,
+		XLogReaderAllocate(WalSegSz, waldir, WALDumpCloseSegment);
-						   XL_ROUTINE(.page_read = WALDumpReadPage,
+
-									  .segment_open = WALDumpOpenSegment,
-									  .segment_close = WALDumpCloseSegment),
-						   &private);
 	if (!xlogreader_state)
 		fatal_error("out of memory");
 
-	/* first find a valid recptr to start from */
+	findnext_state =
-	first_record = XLogFindNextRecord(xlogreader_state, private.startptr);
+		InitXLogFindNextRecord(xlogreader_state, startptr);
 
+	if (!findnext_state)
+		fatal_error("out of memory");
+
+	/* first find a valid recptr to start from */
+	while (XLogFindNextRecord(findnext_state))
+	{
+		if (!WALDumpReadPage(xlogreader_state, timeline, startptr, endptr))
+			break;
+	}
+
+	first_record = findnext_state->currRecPtr;
 	if (first_record == InvalidXLogRecPtr)
 		fatal_error("could not find a valid record after %X/%X",
-					LSN_FORMAT_ARGS(private.startptr));
+					LSN_FORMAT_ARGS(startptr));
 
 	/*
 	 * Display a message that we're skipping data if `from` wasn't a pointer
 	 * to the start of a record and also wasn't a pointer to the beginning of
 	 * a segment (e.g. we were used in file mode).
 	 */
-	if (first_record != private.startptr &&
+	if (first_record != startptr &&
-		XLogSegmentOffset(private.startptr, WalSegSz) != 0)
+		XLogSegmentOffset(startptr, WalSegSz) != 0)
 		printf(ngettext("first record is after %X/%X, at %X/%X, skipping over %u byte\n",
 						"first record is after %X/%X, at %X/%X, skipping over %u bytes\n",
-						(first_record - private.startptr)),
+						(first_record - startptr)),
-			   LSN_FORMAT_ARGS(private.startptr),
+			   LSN_FORMAT_ARGS(startptr),
 			   LSN_FORMAT_ARGS(first_record),
-			   (uint32) (first_record - private.startptr));
+			   (uint32) (first_record - startptr));
 
 	for (;;)
 	{
 		/* try to read the next record */
-		record = XLogReadRecord(xlogreader_state, &errormsg);
+		while (XLogReadRecord(xlogreader_state, &record, &errormsg) ==
+			   XLREAD_NEED_DATA)
+		{
+			if (!WALDumpReadPage(xlogreader_state, timeline, startptr, endptr))
+				break;
+		}
+
 		if (!record)
 		{
-			if (!config.follow || private.endptr_reached)
+			if (!config.follow)
 				break;
 			else
 			{

src/include/access/xlogreader.h

@@ -56,65 +56,17 @@ typedef struct WALSegmentContext
 } WALSegmentContext;
 
 typedef struct XLogReaderState XLogReaderState;
+typedef struct XLogFindNextRecordState XLogFindNextRecordState;
 
-/* Function type definitions for various xlogreader interactions */
+/* Function type definition for the segment cleanup callback */
-typedef int (*XLogPageReadCB) (XLogReaderState *xlogreader,
+typedef void (*WALSegmentCleanupCB) (XLogReaderState *xlogreader);
-							   XLogRecPtr targetPagePtr,
+
-							   int reqLen,
+/* Function type definition for the open/close callbacks for WALRead() */
-							   XLogRecPtr targetRecPtr,
-							   char *readBuf);
 typedef void (*WALSegmentOpenCB) (XLogReaderState *xlogreader,
 								  XLogSegNo nextSegNo,
 								  TimeLineID *tli_p);
 typedef void (*WALSegmentCloseCB) (XLogReaderState *xlogreader);
 
-typedef struct XLogReaderRoutine
-{
-	/*
-	 * Data input callback
-	 *
-	 * This callback shall read at least reqLen valid bytes of the xlog page
-	 * starting at targetPagePtr, and store them in readBuf.  The callback
-	 * shall return the number of bytes read (never more than XLOG_BLCKSZ), or
-	 * -1 on failure.  The callback shall sleep, if necessary, to wait for the
-	 * requested bytes to become available.  The callback will not be invoked
-	 * again for the same page unless more than the returned number of bytes
-	 * are needed.
-	 *
-	 * targetRecPtr is the position of the WAL record we're reading.  Usually
-	 * it is equal to targetPagePtr + reqLen, but sometimes xlogreader needs
-	 * to read and verify the page or segment header, before it reads the
-	 * actual WAL record it's interested in.  In that case, targetRecPtr can
-	 * be used to determine which timeline to read the page from.
-	 *
-	 * The callback shall set ->seg.ws_tli to the TLI of the file the page was
-	 * read from.
-	 */
-	XLogPageReadCB page_read;
-
-	/*
-	 * Callback to open the specified WAL segment for reading.  ->seg.ws_file
-	 * shall be set to the file descriptor of the opened segment.  In case of
-	 * failure, an error shall be raised by the callback and it shall not
-	 * return.
-	 *
-	 * "nextSegNo" is the number of the segment to be opened.
-	 *
-	 * "tli_p" is an input/output argument. WALRead() uses it to pass the
-	 * timeline in which the new segment should be found, but the callback can
-	 * use it to return the TLI that it actually opened.
-	 */
-	WALSegmentOpenCB segment_open;
-
-	/*
-	 * WAL segment close callback.  ->seg.ws_file shall be set to a negative
-	 * number.
-	 */
-	WALSegmentCloseCB segment_close;
-} XLogReaderRoutine;
-
-#define XL_ROUTINE(...) &(XLogReaderRoutine){__VA_ARGS__}
-
 typedef struct
 {
 	/* Is this block ref in use? */
@@ -144,12 +96,36 @@ typedef struct
 	uint16		data_bufsz;
 } DecodedBkpBlock;
 
+/* Return code from XLogReadRecord */
+typedef enum XLogReadRecordResult
+{
+	XLREAD_SUCCESS,				/* record is successfully read */
+	XLREAD_NEED_DATA,			/* need more data. see XLogReadRecord. */
+	XLREAD_FAIL					/* failed during reading a record */
+}			XLogReadRecordResult;
+
+/*
+ * internal state of XLogReadRecord
+ *
+ * XLogReadState runs a state machine while reading a record. Theses states
+ * are not seen outside the function. Each state may repeat several times
+ * exiting requesting caller for new data. See the comment of XLogReadRecrod
+ * for details.
+ */
+typedef enum XLogReadRecordState
+{
+	XLREAD_NEXT_RECORD,
+	XLREAD_TOT_LEN,
+	XLREAD_FIRST_FRAGMENT,
+	XLREAD_CONTINUATION
+}			XLogReadRecordState;
+
 struct XLogReaderState
 {
 	/*
 	 * Operational callbacks
 	 */
-	XLogReaderRoutine routine;
+	WALSegmentCleanupCB cleanup_cb;
 
 	/* ----------------------------------------
 	 * Public parameters
@@ -162,19 +138,31 @@ struct XLogReaderState
 	 */
 	uint64		system_identifier;
 
-	/*
-	 * Opaque data for callbacks to use.  Not used by XLogReader.
-	 */
-	void	   *private_data;
-
 	/*
 	 * Start and end point of last record read.  EndRecPtr is also used as the
 	 * position to read next.  Calling XLogBeginRead() sets EndRecPtr to the
 	 * starting position and ReadRecPtr to invalid.
 	 */
-	XLogRecPtr	ReadRecPtr;		/* start of last record read */
+	XLogRecPtr	ReadRecPtr;		/* start of last record read or being read */
 	XLogRecPtr	EndRecPtr;		/* end+1 of last record read */
+	XLogRecPtr	PrevRecPtr;		/* start of previous record read */
 
+	/* ----------------------------------------
+	 * Communication with page reader
+	 * readBuf is XLOG_BLCKSZ bytes, valid up to at least reqLen bytes.
+	 * ----------------------------------------
+	 */
+	/* variables the clients of xlogreader can examine */
+	XLogRecPtr	readPagePtr;	/* page pointer to read */
+	int32		reqLen;			/* bytes requested to the caller */
+	char	   *readBuf;		/* buffer to store data */
+	bool		page_verified;	/* is the page header on the buffer verified? */
+	bool		record_verified;/* is the current record header verified? */
+
+	/* variables set by the client of xlogreader */
+	int32		readLen;		/* actual bytes copied into readBuf by client,
+								 * which should be >= reqLen.  Client should
+								 * use XLogReaderSetInputData() to set. */
 
 	/* ----------------------------------------
 	 * Decoded representation of current record
@@ -203,13 +191,6 @@ struct XLogReaderState
 	 * ----------------------------------------
 	 */
 
-	/*
-	 * Buffer for currently read page (XLOG_BLCKSZ bytes, valid up to at least
-	 * readLen bytes)
-	 */
-	char	   *readBuf;
-	uint32		readLen;
-
 	/* last read XLOG position for data currently in readBuf */
 	WALSegmentContext segcxt;
 	WALOpenSegment seg;
@@ -222,8 +203,6 @@ struct XLogReaderState
 	XLogRecPtr	latestPagePtr;
 	TimeLineID	latestPageTLI;
 
-	/* beginning of the WAL record being read. */
-	XLogRecPtr	currRecPtr;
 	/* timeline to read it from, 0 if a lookup is required */
 	TimeLineID	currTLI;
 
@@ -250,16 +229,37 @@ struct XLogReaderState
 	char	   *readRecordBuf;
 	uint32		readRecordBufSize;
 
+	/*
+	 * XLogReadRecord() state
+	 */
+	XLogReadRecordState readRecordState;	/* state machine state */
+	int			recordGotLen;	/* amount of current record that has already
+								 * been read */
+	int			recordRemainLen;	/* length of current record that remains */
+	XLogRecPtr	recordContRecPtr;	/* where the current record continues */
+
 	/* Buffer to hold error message */
 	char	   *errormsg_buf;
 };
 
+struct XLogFindNextRecordState
+{
+	XLogReaderState *reader_state;
+	XLogRecPtr		targetRecPtr;
+	XLogRecPtr		currRecPtr;
+};
+
+/* Report that data is available for decoding. */
+static inline void
+XLogReaderSetInputData(XLogReaderState *state, int32 len)
+{
+	state->readLen = len;
+}
+
 /* Get a new XLogReader */
 extern XLogReaderState *XLogReaderAllocate(int wal_segment_size,
 										   const char *waldir,
-										   XLogReaderRoutine *routine,
+										   WALSegmentCleanupCB cleanup_cb);
-										   void *private_data);
-extern XLogReaderRoutine *LocalXLogReaderRoutine(void);
 
 /* Free an XLogReader */
 extern void XLogReaderFree(XLogReaderState *state);
@@ -267,12 +267,14 @@ extern void XLogReaderFree(XLogReaderState *state);
 /* Position the XLogReader to given record */
 extern void XLogBeginRead(XLogReaderState *state, XLogRecPtr RecPtr);
 #ifdef FRONTEND
-extern XLogRecPtr XLogFindNextRecord(XLogReaderState *state, XLogRecPtr RecPtr);
+extern XLogFindNextRecordState *InitXLogFindNextRecord(XLogReaderState *reader_state, XLogRecPtr start_ptr);
+extern bool XLogFindNextRecord(XLogFindNextRecordState *state);
 #endif							/* FRONTEND */
 
 /* Read the next XLog record. Returns NULL on end-of-WAL or failure */
-extern struct XLogRecord *XLogReadRecord(XLogReaderState *state,
+extern XLogReadRecordResult XLogReadRecord(XLogReaderState *state,
-										 char **errormsg);
+										   XLogRecord **record,
+										   char **errormsg);
 
 /* Validate a page */
 extern bool XLogReaderValidatePageHeader(XLogReaderState *state,
@@ -292,6 +294,7 @@ typedef struct WALReadError
 } WALReadError;
 
 extern bool WALRead(XLogReaderState *state,
+					WALSegmentOpenCB segopenfn, WALSegmentCloseCB sgclosefn,
 					char *buf, XLogRecPtr startptr, Size count,
 					TimeLineID tli, WALReadError *errinfo);
 

src/include/access/xlogutils.h

@@ -47,9 +47,7 @@ extern Buffer XLogReadBufferExtended(RelFileNode rnode, ForkNumber forknum,
 extern Relation CreateFakeRelcacheEntry(RelFileNode rnode);
 extern void FreeFakeRelcacheEntry(Relation fakerel);
 
-extern int	read_local_xlog_page(XLogReaderState *state,
+extern bool read_local_xlog_page(XLogReaderState *state);
-								 XLogRecPtr targetPagePtr, int reqLen,
-								 XLogRecPtr targetRecPtr, char *cur_page);
 extern void wal_segment_open(XLogReaderState *state,
 							 XLogSegNo nextSegNo,
 							 TimeLineID *tli_p);

src/include/replication/logical.h

@@ -29,6 +29,10 @@ typedef void (*LogicalOutputPluginWriterUpdateProgress) (struct LogicalDecodingC
 														 TransactionId xid
 );
 
+typedef struct LogicalDecodingContext LogicalDecodingContext;
+
+typedef bool (*LogicalDecodingXLogPageReadCB)(XLogReaderState *ctx);
+
 typedef struct LogicalDecodingContext
 {
 	/* memory context this is all allocated in */
@@ -39,6 +43,7 @@ typedef struct LogicalDecodingContext
 
 	/* infrastructure pieces for decoding */
 	XLogReaderState *reader;
+	LogicalDecodingXLogPageReadCB page_read;
 	struct ReorderBuffer *reorder;
 	struct SnapBuild *snapshot_builder;
 
@@ -105,14 +110,16 @@ extern LogicalDecodingContext *CreateInitDecodingContext(const char *plugin,
 														 List *output_plugin_options,
 														 bool need_full_snapshot,
 														 XLogRecPtr restart_lsn,
-														 XLogReaderRoutine *xl_routine,
+														 LogicalDecodingXLogPageReadCB page_read,
+														 WALSegmentCleanupCB cleanup_cb,
 														 LogicalOutputPluginWriterPrepareWrite prepare_write,
 														 LogicalOutputPluginWriterWrite do_write,
 														 LogicalOutputPluginWriterUpdateProgress update_progress);
 extern LogicalDecodingContext *CreateDecodingContext(XLogRecPtr start_lsn,
 													 List *output_plugin_options,
 													 bool fast_forward,
-													 XLogReaderRoutine *xl_routine,
+													 LogicalDecodingXLogPageReadCB page_read,
+													 WALSegmentCleanupCB cleanup_cb,
 													 LogicalOutputPluginWriterPrepareWrite prepare_write,
 													 LogicalOutputPluginWriterWrite do_write,
 													 LogicalOutputPluginWriterUpdateProgress update_progress);