Allow I/O reliability checks using 16-bit checksums

Checksums are set immediately prior to flush out of shared buffers
and checked when pages are read in again. Hint bit setting will
require full page write when block is dirtied, which causes various
infrastructure changes. Extensive comments, docs and README.

WARNING message thrown if checksum fails on non-all zeroes page;
ERROR thrown but can be disabled with ignore_checksum_failure = on.

Feature enabled by an initdb option, since transition from option off
to option on is long and complex and has not yet been implemented.
Default is not to use checksums.

Checksum used is WAL CRC-32 truncated to 16-bits.

Simon Riggs, Jeff Davis, Greg Smith
Wide input and assistance from many community members. Thank you.

3 files changed, 20 insertions, 3 deletions

diff --git a/src/bin/initdb/initdb.c b/src/bin/initdb/initdb.c
index e16f3e3c80a..bb38796eb28 100644
--- a/src/bin/initdb/initdb.c
+++ b/src/bin/initdb/initdb.c
@@ -119,6 +119,7 @@ static bool noclean = false;
 static bool do_sync = true;
 static bool sync_only = false;
 static bool show_setting = false;
+static bool data_checksums = false;
 static char *xlog_dir = "";
 
 
@@ -1441,8 +1442,10 @@ bootstrap_template1(void)
 	unsetenv("PGCLIENTENCODING");
 
 	snprintf(cmd, sizeof(cmd),
-			 "\"%s\" --boot -x1 %s %s",
-			 backend_exec, boot_options, talkargs);
+			 "\"%s\" --boot -x1 %s %s %s",
+			 backend_exec,
+			 data_checksums ? "-k" : "",
+			 boot_options, talkargs);
 
 	PG_CMD_OPEN;
 
@@ -2748,6 +2751,7 @@ usage(const char *progname)
 	printf(_("  -X, --xlogdir=XLOGDIR     location for the transaction log directory\n"));
 	printf(_("\nLess commonly used options:\n"));
 	printf(_("  -d, --debug               generate lots of debugging output\n"));
+	printf(_("  -k, --data-checksums      data page checksums\n"));
 	printf(_("  -L DIRECTORY              where to find the input files\n"));
 	printf(_("  -n, --noclean             do not clean up after errors\n"));
 	printf(_("  -N, --nosync              do not wait for changes to be written safely to disk\n"));
@@ -3424,6 +3428,7 @@ main(int argc, char *argv[])
 		{"nosync", no_argument, NULL, 'N'},
 		{"sync-only", no_argument, NULL, 'S'},
 		{"xlogdir", required_argument, NULL, 'X'},
+		{"data-checksums", no_argument, NULL, 'k'},
 		{NULL, 0, NULL, 0}
 	};
 
@@ -3455,7 +3460,7 @@ main(int argc, char *argv[])
 
 	/* process command-line options */
 
-	while ((c = getopt_long(argc, argv, "dD:E:L:nNU:WA:sST:X:", long_options, &option_index)) != -1)
+	while ((c = getopt_long(argc, argv, "dD:E:kL:nNU:WA:sST:X:", long_options, &option_index)) != -1)
 	{
 		switch (c)
 		{
@@ -3504,6 +3509,9 @@ main(int argc, char *argv[])
 			case 'S':
 				sync_only = true;
 				break;
+			case 'k':
+				data_checksums = true;
+				break;
 			case 'L':
 				share_path = pg_strdup(optarg);
 				break;
@@ -3615,6 +3623,11 @@ main(int argc, char *argv[])
 	setup_locale_encoding();
 
 	setup_text_search();
+
+	if (data_checksums)
+		printf(_("Data page checksums are enabled.\n"));
+	else
+		printf(_("Data page checksums are disabled.\n"));
 	
 	printf("\n");
 
diff --git a/src/bin/pg_controldata/pg_controldata.c b/src/bin/pg_controldata/pg_controldata.c
index 369f2fb0759..ceb412505bf 100644
--- a/src/bin/pg_controldata/pg_controldata.c
+++ b/src/bin/pg_controldata/pg_controldata.c
@@ -287,5 +287,7 @@ main(int argc, char *argv[])
 		   (ControlFile.float4ByVal ? _("by value") : _("by reference")));
 	printf(_("Float8 argument passing:              %s\n"),
 		   (ControlFile.float8ByVal ? _("by value") : _("by reference")));
+	printf(_("Data page checksums:                  %s\n"),
+		   (ControlFile.data_checksums ? _("enabled") : _("disabled")));
 	return 0;
 }
diff --git a/src/bin/pg_resetxlog/pg_resetxlog.c b/src/bin/pg_resetxlog/pg_resetxlog.c
index acf9f8dd3e3..124dcfb7772 100644
--- a/src/bin/pg_resetxlog/pg_resetxlog.c
+++ b/src/bin/pg_resetxlog/pg_resetxlog.c
@@ -624,6 +624,8 @@ PrintControlValues(bool guessed)
 		   (ControlFile.float4ByVal ? _("by value") : _("by reference")));
 	printf(_("Float8 argument passing:              %s\n"),
 		   (ControlFile.float8ByVal ? _("by value") : _("by reference")));
+	printf(_("Data page checksums:                  %s\n"),
+		   (ControlFile.data_checksums ? _("enabled") : _("disabled")));
 }