Fix for Bug #16032

[Rob]

Hello,

I stumbled on a windows-only bug in pg_basebackup which I've reported as 
#16032 
(https://www.postgresql.org/message-id/16032-4ba56823a2b2805f%40postgresql.org).

I'm pretty sure I've fixed it in the attached patch.

Many Thanks,
RobFrom 148d525c0d30af35abba1b8c5bbe07e4e72ecfec Mon Sep 17 00:00:00 2001
From: Rob Emery <mints...@users.noreply.github.com>
Date: Tue, 1 Oct 2019 23:51:52 +0100
Subject: [PATCH] Fix bug16032; under windows when the backup is aborted or
 fails, pg_basebackup is unable to cleanup the backup as the filehandles
 haven't been released

---
 src/bin/pg_basebackup/pg_basebackup.c | 22 ++++++++++++++++++++++
 1 file changed, 22 insertions(+)

diff --git a/src/bin/pg_basebackup/pg_basebackup.c b/src/bin/pg_basebackup/pg_basebackup.c
index aa68f59965..b6d68e0256 100644
--- a/src/bin/pg_basebackup/pg_basebackup.c
+++ b/src/bin/pg_basebackup/pg_basebackup.c
@@ -1098,6 +1098,17 @@ ReceiveTarFile(PGconn *conn, PGresult *res, int rownum)
 	{
 		fprintf(stderr, _("%s: could not get COPY data stream: %s"),
 				progname, PQerrorMessage(conn));
+		if (strcmp(basedir, "-") != 0)
+		{
+			/*
+			 * File handles could already be closed so we don't care about the result
+			 */
+#ifdef HAVE_LIBZ
+			gzclose(ztarfile);
+#else
+			fclose(tarfile);
+#endif
+		}
 		disconnect_and_exit(1);
 	}
 
@@ -1179,6 +1190,17 @@ ReceiveTarFile(PGconn *conn, PGresult *res, int rownum)
 		{
 			fprintf(stderr, _("%s: could not read COPY data: %s"),
 					progname, PQerrorMessage(conn));
+			if (strcmp(basedir, "-") != 0)
+			{
+				/*
+				 * File handles could already be closed so we don't care about the result
+				 */
+#ifdef HAVE_LIBZ
+				gzclose(ztarfile);
+#else
+				fclose(tarfile);
+#endif
+			}
 			disconnect_and_exit(1);
 		}
 
-- 
2.11.0