[#25329] YSQL: Address YB_TODOs in pg_upgrade.c::main

Summary:
Addressed a few YB_TODOs in `main` in `pg_upgrade.c`:

=== `copy_xact_xlog_xid` ===

copy_xact_xlog_xid and start_postmaster were both wrapped in an `#ifdef YB_TODO` block.

`copy_xact_xlog_xid` copies Postgres commit and WAL data around. Any data that is actually required
by YB for data integrity should be in DocDB, which isn't impacted by the upgrade. Therefore it
should be safe to fully disable this function.

=== `start_postmaster` / `stop_postmaster` ===
Previous work on `pg_upgrade` had wrapped calls to `start_postmaster` or `stop_postmaster` with a
`!is_yugabyte_enabled()` check, as we don't need to stop / start the postmaster in the YB upgrade.

=== Cleaning up the old cluster ===
Because Yugabyte data resides on DocDB and is reused from pg11 to pg15, we don't need to remove it. The Upgrade RPCs will handle removing old masters during the upgrade finalize step, so there is nothing required to clean up the old cluster.
Jira: DB-14539

Test Plan:
Jenkins

```
./yb_build.sh release --cxx-test pg15_upgrade-test
```

Reviewers: hsunder

Reviewed By: hsunder

Subscribers: yql

Differential Revision: https://phorge.dev.yugabyte.com/D40720

src/postgres/src/bin/pg_upgrade/pg_upgrade.c

@@ -78,10 +78,7 @@ char	   *output_files[] = {
 int
 main(int argc, char **argv)
 {
-#ifdef YB_TODO
-	/* Unused variable due to later YB_TODO create_script_for_old_cluster_deletion */
 	char	   *deletion_script_file_name = NULL;
-#endif
 	bool		live_check = false;
 
 	pg_logging_init(argv[0]);
@@ -182,75 +179,84 @@ main(int argc, char **argv)
 
 	/*
 	 * Destructive Changes to New Cluster
+	 * YB: We don't need to start / stop the postmaster, or copy around commit
+	 * logs, as those are handled by the YB master.
 	 */
-#ifdef YB_TODO
-	/* Investigate whether this should be permanently disabled */
-	copy_xact_xlog_xid();
+	if (!is_yugabyte_enabled())
+	{
+		copy_xact_xlog_xid();
 
-	/* New now using xids of the old system */
+		/* New now using xids of the old system */
 
-	/* -- NEW -- */
-	start_postmaster(&new_cluster, true);
-#endif
+		/* -- NEW -- */
+		start_postmaster(&new_cluster, true);
+	}
 
 	prepare_new_globals();
 
 	create_new_objects();
 
+	/*
+	 * YB: The YB major version upgrade flow does not require disabling the old
+	 * cluster, as the upgrade is online and in place. We don't need to delete
+	 * the old cluster data files, because the relevant data lives on DocDB.
+	 */
 	if (!is_yugabyte_enabled())
+	{
 		stop_postmaster(false);
 
-#ifdef YB_TODO
-	/* Investigate these functions being executed after pg_restore is done */
-	/*
-	 * Most failures happen in create_new_objects(), which has completed at
-	 * this point.  We do this here because it is just before linking, which
-	 * will link the old and new cluster data files, preventing the old
-	 * cluster from being safely started once the new cluster is started.
-	 */
-	if (user_opts.transfer_mode == TRANSFER_MODE_LINK)
-		disable_old_cluster();
+		/*
+		 * Most failures happen in create_new_objects(), which has completed at
+		 * this point.  We do this here because it is just before linking, which
+		 * will link the old and new cluster data files, preventing the old
+		 * cluster from being safely started once the new cluster is started.
+		 */
+		if (user_opts.transfer_mode == TRANSFER_MODE_LINK)
+			disable_old_cluster();
 
-	transfer_all_new_tablespaces(&old_cluster.dbarr, &new_cluster.dbarr,
-								 old_cluster.pgdata, new_cluster.pgdata);
+		transfer_all_new_tablespaces(&old_cluster.dbarr, &new_cluster.dbarr,
+									 old_cluster.pgdata, new_cluster.pgdata);
 
-	/*
-	 * Assuming OIDs are only used in system tables, there is no need to
-	 * restore the OID counter because we have not transferred any OIDs from
-	 * the old system, but we do it anyway just in case.  We do it late here
-	 * because there is no need to have the schema load use new oids.
-	 */
-	prep_status("Setting next OID for new cluster");
-	exec_prog(UTILITY_LOG_FILE, NULL, true, true,
-			  "\"%s/pg_resetwal\" -o %u \"%s\"",
-			  new_cluster.bindir, old_cluster.controldata.chkpnt_nxtoid,
-			  new_cluster.pgdata);
-	check_ok();
-
-	if (user_opts.do_sync)
-	{
-		prep_status("Sync data directory to disk");
+		/*
+		 * Assuming OIDs are only used in system tables, there is no need to
+		 * restore the OID counter because we have not transferred any OIDs from
+		 * the old system, but we do it anyway just in case.  We do it late here
+		 * because there is no need to have the schema load use new oids.
+		 */
+		prep_status("Setting next OID for new cluster");
 		exec_prog(UTILITY_LOG_FILE, NULL, true, true,
-				  "\"%s/initdb\" --sync-only \"%s\"", new_cluster.bindir,
+				  "\"%s/pg_resetwal\" -o %u \"%s\"",
+				  new_cluster.bindir, old_cluster.controldata.chkpnt_nxtoid,
 				  new_cluster.pgdata);
 		check_ok();
-	}
 
-	create_script_for_old_cluster_deletion(&deletion_script_file_name);
+		if (user_opts.do_sync)
+		{
+			prep_status("Sync data directory to disk");
+			exec_prog(UTILITY_LOG_FILE, NULL, true, true,
+					  "\"%s/initdb\" --sync-only \"%s\"", new_cluster.bindir,
+					  new_cluster.pgdata);
+			check_ok();
+		}
 
-	issue_warnings_and_set_wal_level();
+		create_script_for_old_cluster_deletion(&deletion_script_file_name);
+
+		issue_warnings_and_set_wal_level();
+	}
 
 	pg_log(PG_REPORT,
 		   "\n"
 		   "Upgrade Complete\n"
 		   "----------------\n");
 
-	output_completion_banner(deletion_script_file_name);
+	if (!is_yugabyte_enabled())
+	{
+		output_completion_banner(deletion_script_file_name);
 
-	pg_free(deletion_script_file_name);
+		pg_free(deletion_script_file_name);
+	}
 
 	cleanup_output_dirs();
-#endif
 	return 0;
 }