Add copy_file_range strategy to create database

Author: jodiew

src/backend/commands/dbcommands.c

@@ -83,6 +83,7 @@ typedef enum CreateDBStrategy
 {
 	CREATEDB_WAL_LOG,
 	CREATEDB_FILE_COPY,
+	CREATEDB_COPY_FILE_RANGE,
 } CreateDBStrategy;
 
 typedef struct
@@ -136,7 +137,8 @@ static CreateDBRelInfo *ScanSourceDatabasePgClassTuple(HeapTupleData *tuple,
 static void CreateDirAndVersionFile(char *dbpath, Oid dbid, Oid tsid,
 									bool isRedo);
 static void CreateDatabaseUsingFileCopy(Oid src_dboid, Oid dst_dboid,
-										Oid src_tsid, Oid dst_tsid);
+										Oid src_tsid, Oid dst_tsid,
+										bool range_copy);
 static void recovery_create_dbdir(char *path, bool only_tblspc);
 
 /*
@@ -548,7 +550,7 @@ CreateDirAndVersionFile(char *dbpath, Oid dbid, Oid tsid, bool isRedo)
  */
 static void
 CreateDatabaseUsingFileCopy(Oid src_dboid, Oid dst_dboid, Oid src_tsid,
-							Oid dst_tsid)
+							Oid dst_tsid, bool range_copy)
 {
 	TableScanDesc scan;
 	Relation	rel;
@@ -608,7 +610,7 @@ CreateDatabaseUsingFileCopy(Oid src_dboid, Oid dst_dboid, Oid src_tsid,
 		 *
 		 * We don't need to copy subdirectories
 		 */
-		copydir(srcpath, dstpath, false);
+		copydir(srcpath, dstpath, false, range_copy);
 
 		/* Record the filesystem change in XLOG */
 		{
@@ -1022,6 +1024,8 @@ createdb(ParseState *pstate, const CreatedbStmt *stmt)
 			dbstrategy = CREATEDB_WAL_LOG;
 		else if (pg_strcasecmp(strategy, "file_copy") == 0)
 			dbstrategy = CREATEDB_FILE_COPY;
+		else if (pg_strcasecmp(strategy, "copy_file_range") == 0)
+			dbstrategy = CREATEDB_COPY_FILE_RANGE;
 		else
 			ereport(ERROR,
 					(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
@@ -1508,9 +1512,12 @@ createdb(ParseState *pstate, const CreatedbStmt *stmt)
 		if (dbstrategy == CREATEDB_WAL_LOG)
 			CreateDatabaseUsingWalLog(src_dboid, dboid, src_deftablespace,
 									  dst_deftablespace);
+		else if (dbstrategy == CREATEDB_COPY_FILE_RANGE)
+			CreateDatabaseUsingFileCopy(src_dboid, dboid, src_deftablespace,
+										dst_deftablespace, true);
 		else
 			CreateDatabaseUsingFileCopy(src_dboid, dboid, src_deftablespace,
-										dst_deftablespace);
+										dst_deftablespace, false);
 
 		/*
 		 * Close pg_database, but keep lock till commit.
@@ -2156,7 +2163,7 @@ movedb(const char *dbname, const char *tblspcname)
 		/*
 		 * Copy files from the old tablespace to the new one
 		 */
-		copydir(src_dbpath, dst_dbpath, false);
+		copydir(src_dbpath, dst_dbpath, false, false);
 
 		/*
 		 * Record the filesystem change in XLOG
@@ -3341,7 +3348,7 @@ dbase_redo(XLogReaderState *record)
 		 *
 		 * We don't need to copy subdirectories
 		 */
-		copydir(src_path, dst_path, false);
+		copydir(src_path, dst_path, false, false);
 
 		pfree(src_path);
 		pfree(dst_path);

src/backend/storage/file/copydir.c

@@ -26,6 +26,7 @@
 #include "pgstat.h"
 #include "storage/copydir.h"
 #include "storage/fd.h"
+#include "pg_config.h"
 
 /*
  * copydir: copy a directory
@@ -34,7 +35,7 @@
  * a directory or a regular file is ignored.
  */
 void
-copydir(const char *fromdir, const char *todir, bool recurse)
+copydir(const char *fromdir, const char *todir, bool recurse, bool range_copy)
 {
 	DIR		   *xldir;
 	struct dirent *xlde;
@@ -68,10 +69,15 @@ copydir(const char *fromdir, const char *todir, bool recurse)
 		{
 			/* recurse to handle subdirectories */
 			if (recurse)
-				copydir(fromfile, tofile, true);
+				copydir(fromfile, tofile, true, range_copy);
 		}
 		else if (xlde_type == PGFILETYPE_REG)
-			copy_file(fromfile, tofile);
+		{
+			if (range_copy)
+				copy_file_reflink(fromfile, tofile);
+			else
+				copy_file(fromfile, tofile);
+		}
 	}
 	FreeDir(xldir);
 
@@ -214,3 +220,63 @@ copy_file(const char *fromfile, const char *tofile)
 
 	pfree(buffer);
 }
+
+/*
+  * copy one file using copy_file_range to give filesystem a chance to do COW optimization
+  */
+void
+copy_file_reflink(const char *fromfile, const char *tofile)
+{
+#if  defined(HAVE_COPY_FILE_RANGE)
+	int			srcfd;
+	int			dstfd;
+	ssize_t		nbytes;
+
+	/*
+		* Open the files
+		*/
+	srcfd = OpenTransientFile(fromfile, O_RDONLY | PG_BINARY);
+
+	if (srcfd < 0)
+		ereport(ERROR,
+				errcode_for_file_access(),
+				errmsg("could not open file \"%s\": %m", fromfile));
+
+	dstfd = OpenTransientFile(tofile, O_RDWR | O_CREAT | O_EXCL | PG_BINARY);
+
+	if (dstfd < 0)
+		ereport(ERROR,
+				errcode_for_file_access(),
+				errmsg("could not create file \"%s\": %m", tofile));
+
+	/*
+		* Do the data copying.
+		*/
+	do
+	{
+		/* If we got a cancel signal during the copy of the file, quit */
+		CHECK_FOR_INTERRUPTS();
+
+		pgstat_report_wait_start(WAIT_EVENT_COPY_FILE_READ);
+		nbytes = copy_file_range(srcfd, NULL, dstfd, NULL, SSIZE_MAX, 0);
+		pgstat_report_wait_end();
+		if (nbytes < 0)
+			ereport(ERROR,
+					errcode_for_file_access(),
+					errmsg("could not copy file \"%s\": %m", fromfile));
+	}
+	while (nbytes > 0);
+
+	if (CloseTransientFile(dstfd) != 0)
+		ereport(ERROR,
+				errcode_for_file_access(),
+				errmsg("could not close file \"%s\": %m", tofile));
+
+	if (CloseTransientFile(srcfd) != 0)
+		ereport(ERROR,
+				errcode_for_file_access(),
+				errmsg("could not close file \"%s\": %m", fromfile));
+#else
+	pg_fatal("copy_file_range not available on this platform");
+#endif
+}

src/include/storage/copydir.h

@@ -13,7 +13,8 @@
 #ifndef COPYDIR_H
 #define COPYDIR_H
 
-extern void copydir(const char *fromdir, const char *todir, bool recurse);
+extern void copydir(const char *fromdir, const char *todir, bool recurse, bool range_copy);
 extern void copy_file(const char *fromfile, const char *tofile);
+extern void copy_file_reflink(const char *fromfile, const char *tofile);
 
 #endif							/* COPYDIR_H */