Fix timeout overflow in WAIT FOR LSN

Commit 447aae13b0 accepted TIMEOUT values up to INT64_MAX
milliseconds, but computing the deadline multiplies by 1000, which
caused an int64 overflow. And the command reported a timeout at once
instead of waiting.

Parse the value as an int instead, which simplifies the code and is
consistent with other timeout values such as the statement_timeout GUC.

The specified values are now rounded to the nearest millisecond, so a
positive timeout below half a millisecond becomes zero and waits
indefinitely.

Backpatch to v19, where the TIMEOUT option was introduced.

Reported-by: ChangAo Chen <cca5507@qq.com>
Author: ChangAo Chen <cca5507@qq.com>
Reviewed-by: Xuneng Zhou <xunengzhou@gmail.com>
Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com>
Discussion: https://postgr.es/m/tencent_86B83240785807077600A1778566B5C12908@qq.com
Backpatch-through: 19

doc/src/sgml/ref/wait_for.sgml

@@ -145,13 +145,21 @@ WAIT FOR LSN '<replaceable class="parameter">lsn</replaceable>'
          <para>
           When specified and <parameter>timeout</parameter> is greater than zero,
           the command waits until <parameter>lsn</parameter> is reached or
-          the specified <parameter>timeout</parameter> has elapsed.
+          the specified <parameter>timeout</parameter> has elapsed.  A value
+          of zero (the default) means the command waits indefinitely.
          </para>
          <para>
-          The <parameter>timeout</parameter> might be given as integer number of
-          milliseconds.  Also it might be given as string literal with
-          integer number of milliseconds or a number with unit
-          (see <xref linkend="config-setting-names-values"/>).
+          The <parameter>timeout</parameter> is an amount of time in
+          milliseconds.  It may also be specified as a string containing the
+          numerical value followed by a time unit
+          (see <xref linkend="config-setting-names-values"/>).  The maximum
+          value is <literal>2147483647 ms</literal>.
+         </para>
+         <para>
+          Fractional values are rounded to the nearest millisecond.  Note
+          that a <parameter>timeout</parameter> of half a millisecond or
+          less therefore rounds down to zero, which means waiting
+          indefinitely.
          </para>
         </listitem>
        </varlistentry>

src/backend/access/transam/xlogwait.c

@@ -440,7 +440,7 @@ WaitLSNTypeRequiresRecovery(WaitLSNType t)
  * or replica got promoted before the target LSN reached.
  */
 WaitLSNResult
-WaitForLSN(WaitLSNType lsnType, XLogRecPtr targetLSN, int64 timeout)
+WaitForLSN(WaitLSNType lsnType, XLogRecPtr targetLSN, int timeout)
 {
 	XLogRecPtr	currentLSN;
 	WaitLSNProcInfo *procInfo;

src/backend/commands/repack_worker.c

@@ -453,7 +453,7 @@ decode_concurrent_changes(LogicalDecodingContext *ctx,
 
 		if (record == NULL)
 		{
-			int64		timeout = 0;
+			int			timeout = 0;
 			WaitLSNResult res;
 
 			/*
@@ -472,7 +472,7 @@ decode_concurrent_changes(LogicalDecodingContext *ctx,
 			 * should already have been flushed to disk.
 			 */
 			if (!XLogRecPtrIsValid(lsn_upto))
-				timeout = 100L;
+				timeout = 100;
 			res = WaitForLSN(WAIT_LSN_TYPE_PRIMARY_FLUSH,
 							 ctx->reader->EndRecPtr + 1,
 							 timeout);

src/backend/commands/wait.c

@@ -13,8 +13,6 @@
  */
 #include "postgres.h"
 
-#include <math.h>
-
 #include "access/xlog.h"
 #include "access/xlogrecovery.h"
 #include "access/xlogwait.h"
@@ -37,7 +35,7 @@ ExecWaitStmt(ParseState *pstate, WaitStmt *stmt, bool isTopLevel,
 			 DestReceiver *dest)
 {
 	XLogRecPtr	lsn;
-	int64		timeout = 0;
+	int			timeout = 0;
 	WaitLSNResult waitLSNResult;
 	WaitLSNType lsnType = WAIT_LSN_TYPE_STANDBY_REPLAY; /* default */
 	bool		throw = true;
@@ -94,7 +92,6 @@ ExecWaitStmt(ParseState *pstate, WaitStmt *stmt, bool isTopLevel,
 		{
 			char	   *timeout_str;
 			const char *hintmsg;
-			double		dval;
 
 			if (timeout_specified)
 				errorConflictingDefElem(defel, pstate);
@@ -102,33 +99,18 @@ ExecWaitStmt(ParseState *pstate, WaitStmt *stmt, bool isTopLevel,
 
 			timeout_str = defGetString(defel);
 
-			if (!parse_real(timeout_str, &dval, GUC_UNIT_MS, &hintmsg))
-			{
+			if (!parse_int(timeout_str, &timeout, GUC_UNIT_MS, &hintmsg))
 				ereport(ERROR,
 						errcode(ERRCODE_INVALID_PARAMETER_VALUE),
 						errmsg("invalid timeout value: \"%s\"", timeout_str),
-						hintmsg ? errhint("%s", _(hintmsg)) : 0);
-			}
+						hintmsg ? errhint("%s", _(hintmsg)) : 0,
+						parser_errposition(pstate, defel->location));
 
-			/*
-			 * Get rid of any fractional part in the input. This is so we
-			 * don't fail on just-out-of-range values that would round into
-			 * range.
-			 */
-			dval = rint(dval);
-
-			/* Range check */
-			if (unlikely(isnan(dval) || !FLOAT8_FITS_IN_INT64(dval)))
-				ereport(ERROR,
-						errcode(ERRCODE_NUMERIC_VALUE_OUT_OF_RANGE),
-						errmsg("timeout value is out of range"));
-
-			if (dval < 0)
+			if (timeout < 0)
 				ereport(ERROR,
 						errcode(ERRCODE_INVALID_PARAMETER_VALUE),
-						errmsg("timeout cannot be negative"));
-
-			timeout = (int64) dval;
+						errmsg("timeout cannot be negative"),
+						parser_errposition(pstate, defel->location));
 		}
 		else if (strcmp(defel->defname, "no_throw") == 0)
 		{

src/include/access/xlogwait.h

@@ -104,6 +104,6 @@ extern XLogRecPtr GetCurrentLSNForWaitType(WaitLSNType lsnType);
 extern void WaitLSNWakeup(WaitLSNType lsnType, XLogRecPtr currentLSN);
 extern void WaitLSNCleanup(void);
 extern WaitLSNResult WaitForLSN(WaitLSNType lsnType, XLogRecPtr targetLSN,
-								int64 timeout);
+								int timeout);
 
 #endif							/* XLOG_WAIT_H */

src/test/recovery/t/049_wait_for_lsn.pl

@@ -343,6 +343,14 @@ $node_standby->psql(
 	stderr => \$stderr);
 ok($stderr =~ /timeout cannot be negative/, "get error for negative timeout");
 
+# Test out of range timeout
+$node_standby->psql(
+	'postgres',
+	"WAIT FOR LSN '${test_lsn}' WITH (timeout '2147483648ms');",
+	stderr => \$stderr);
+ok($stderr =~ /invalid timeout value: "2147483648ms"/,
+	"get error for out of range timeout");
+
 # Test unknown parameter with WITH clause
 $node_standby->psql(
 	'postgres',
@@ -407,6 +415,12 @@ $output = $node_standby->safe_psql(
 ok($output eq "timeout",
 	"WAIT FOR WITH clause returns correct timeout status");
 
+# Test maximum timeout
+$output = $node_standby->safe_psql(
+	'postgres', qq[
+	WAIT FOR LSN '${lsn2}' WITH (timeout '2147483647ms', no_throw);]);
+ok($output eq "success", "maximum timeout value is accepted");
+
 # Test WITH clause error case - invalid option
 $node_standby->psql(
 	'postgres',