postmaster.c

PostgreSQL7.4.6 for Linux

		return CAC_SHUTDOWN;
	if (StartupPID)
		return CAC_STARTUP;
	if (FatalError)
		return CAC_RECOVERY;

	/*
	 * Don't start too many children.
	 *
	 * We allow more connections than we can have backends here because some
	 * might still be authenticating; they might fail auth, or some
	 * existing backend might exit before the auth cycle is completed. The
	 * exact MaxBackends limit is enforced when a new backend tries to
	 * join the shared-inval backend array.
	 */
	if (CountChildren() >= 2 * MaxBackends)
		return CAC_TOOMANY;

	return CAC_OK;
}


/*
 * ConnCreate -- create a local connection data structure
 */
static Port *
ConnCreate(int serverFd)
{
	Port	   *port;

	if (!(port = (Port *) calloc(1, sizeof(Port))))
	{
		ereport(LOG,
				(errcode(ERRCODE_OUT_OF_MEMORY),
				 errmsg("out of memory")));
		ExitPostmaster(1);
	}

	if (StreamConnection(serverFd, port) != STATUS_OK)
	{
		StreamClose(port->sock);
		ConnFree(port);
		port = NULL;
	}
	else
	{
		/*
		 * Precompute password salt values to use for this connection.
		 * It's slightly annoying to do this long in advance of knowing
		 * whether we'll need 'em or not, but we must do the random()
		 * calls before we fork, not after.  Else the postmaster's random
		 * sequence won't get advanced, and all backends would end up
		 * using the same salt...
		 */
		RandomSalt(port->cryptSalt, port->md5Salt);
	}

	return port;
}


/*
 * ConnFree -- free a local connection data structure
 */
static void
ConnFree(Port *conn)
{
#ifdef USE_SSL
	secure_close(conn);
#endif
	free(conn);
}


/*
 * ClosePostmasterPorts -- close all the postmaster's open sockets
 *
 * This is called during child process startup to release file descriptors
 * that are not needed by that child process.  The postmaster still has
 * them open, of course.
 */
void
ClosePostmasterPorts(bool pgstat_too)
{
	int			i;

	/* Close the listen sockets */
	for (i = 0; i < MAXLISTEN; i++)
	{
		if (ListenSocket[i] != -1)
		{
			StreamClose(ListenSocket[i]);
			ListenSocket[i] = -1;
		}
	}

	/* Close pgstat control sockets, unless we're starting pgstat itself */
	if (pgstat_too)
		pgstat_close_sockets();
}


/*
 * reset_shared -- reset shared memory and semaphores
 */
static void
reset_shared(unsigned short port)
{
	/*
	 * Create or re-create shared memory and semaphores.
	 *
	 * Note: in each "cycle of life" we will normally assign the same IPC
	 * keys (if using SysV shmem and/or semas), since the port number is
	 * used to determine IPC keys.	This helps ensure that we will clean
	 * up dead IPC objects if the postmaster crashes and is restarted.
	 */
	CreateSharedMemoryAndSemaphores(false, MaxBackends, port);
}


/*
 * SIGHUP -- reread config files, and tell children to do same
 */
static void
SIGHUP_handler(SIGNAL_ARGS)
{
	int			save_errno = errno;

	PG_SETMASK(&BlockSig);

	if (Shutdown <= SmartShutdown)
	{
		ereport(LOG,
			 (errmsg("received SIGHUP, reloading configuration files")));
		ProcessConfigFile(PGC_SIGHUP);
#ifdef EXEC_BACKEND
		write_nondefault_variables(PGC_SIGHUP);
#endif
		SignalChildren(SIGHUP);
		load_hba();
		load_ident();
	}

	PG_SETMASK(&UnBlockSig);

	errno = save_errno;
}



/*
 * pmdie -- signal handler for processing various postmaster signals.
 */
static void
pmdie(SIGNAL_ARGS)
{
	int			save_errno = errno;

	PG_SETMASK(&BlockSig);

	ereport(DEBUG2,
			(errmsg_internal("postmaster received signal %d",
							 postgres_signal_arg)));

	switch (postgres_signal_arg)
	{
		case SIGTERM:

			/*
			 * Smart Shutdown:
			 *
			 * Wait for children to end their work and ShutdownDataBase.
			 */
			if (Shutdown >= SmartShutdown)
				break;
			Shutdown = SmartShutdown;
			ereport(LOG,
					(errmsg("received smart shutdown request")));
			if (DLGetHead(BackendList)) /* let reaper() handle this */
				break;

			/*
			 * No children left. Shutdown data base system.
			 */
			if (StartupPID > 0 || FatalError)	/* let reaper() handle
												 * this */
				break;
			if (ShutdownPID > 0)
			{
				elog(PANIC, "shutdown process %d already running",
					 (int) ShutdownPID);
				abort();
			}

			ShutdownPID = ShutdownDataBase();
			break;

		case SIGINT:

			/*
			 * Fast Shutdown:
			 *
			 * abort all children with SIGTERM (rollback active transactions
			 * and exit) and ShutdownDataBase when they are gone.
			 */
			if (Shutdown >= FastShutdown)
				break;
			ereport(LOG,
					(errmsg("received fast shutdown request")));
			if (DLGetHead(BackendList)) /* let reaper() handle this */
			{
				Shutdown = FastShutdown;
				if (!FatalError)
				{
					ereport(LOG,
							(errmsg("aborting any active transactions")));
					SignalChildren(SIGTERM);
				}
				break;
			}
			if (Shutdown > NoShutdown)
			{
				Shutdown = FastShutdown;
				break;
			}
			Shutdown = FastShutdown;

			/*
			 * No children left. Shutdown data base system.
			 */
			if (StartupPID > 0 || FatalError)	/* let reaper() handle
												 * this */
				break;
			if (ShutdownPID > 0)
			{
				elog(PANIC, "shutdown process %d already running",
					 (int) ShutdownPID);
				abort();
			}

			ShutdownPID = ShutdownDataBase();
			break;

		case SIGQUIT:

			/*
			 * Immediate Shutdown:
			 *
			 * abort all children with SIGQUIT and exit without attempt to
			 * properly shutdown data base system.
			 */
			ereport(LOG,
					(errmsg("received immediate shutdown request")));
			if (ShutdownPID > 0)
				kill(ShutdownPID, SIGQUIT);
			if (StartupPID > 0)
				kill(StartupPID, SIGQUIT);
			if (DLGetHead(BackendList))
				SignalChildren(SIGQUIT);
			ExitPostmaster(0);
			break;
	}

	PG_SETMASK(&UnBlockSig);

	errno = save_errno;
}

/*
 * Reaper -- signal handler to cleanup after a backend (child) dies.
 */
static void
reaper(SIGNAL_ARGS)
{
	int			save_errno = errno;

#ifdef WIN32
#warning fix waidpid for Win32
#else
#ifdef HAVE_WAITPID
	int			status;			/* backend exit status */

#else
	union wait	status;			/* backend exit status */
#endif
	int			exitstatus;
	int			pid;			/* process id of dead backend */

	PG_SETMASK(&BlockSig);

	ereport(DEBUG4,
			(errmsg_internal("reaping dead processes")));
#ifdef HAVE_WAITPID
	while ((pid = waitpid(-1, &status, WNOHANG)) > 0)
	{
		exitstatus = status;
#else
	while ((pid = wait3(&status, WNOHANG, NULL)) > 0)
	{
		exitstatus = status.w_status;
#endif

		/*
		 * Check if this child was the statistics collector. If so, try to
		 * start a new one.  (If fail, we'll try again in future cycles of
		 * the main loop.)
		 */
		if (pgstat_ispgstat(pid))
		{
			LogChildExit(LOG, gettext("statistics collector process"),
						 pid, exitstatus);
			pgstat_start();
			continue;
		}

		/*
		 * Check if this child was a shutdown or startup process.
		 */
		if (ShutdownPID > 0 && pid == ShutdownPID)
		{
			if (exitstatus != 0)
			{
				LogChildExit(LOG, gettext("shutdown process"),
							 pid, exitstatus);
				ExitPostmaster(1);
			}
			/* Normal postmaster exit is here */
			ExitPostmaster(0);
		}

		if (StartupPID > 0 && pid == StartupPID)
		{
			if (exitstatus != 0)
			{
				LogChildExit(LOG, gettext("startup process"),
							 pid, exitstatus);
				ereport(LOG,
						(errmsg("aborting startup due to startup process failure")));
				ExitPostmaster(1);
			}
			StartupPID = 0;

			/*
			 * Startup succeeded - remember its ID and RedoRecPtr.
			 *
			 * NB: this MUST happen before we fork a checkpoint or shutdown
			 * subprocess, else they will have wrong local ThisStartUpId.
			 */
			SetThisStartUpID();

			FatalError = false; /* done with recovery */

			/*
			 * Arrange for first checkpoint to occur after standard delay.
			 */
			CheckPointPID = 0;
			checkpointed = time(NULL);

			/*
			 * Go to shutdown mode if a shutdown request was pending.
			 */
			if (Shutdown > NoShutdown)
			{
				if (ShutdownPID > 0)
				{
					elog(PANIC, "startup process %d died while shutdown process %d already running",
						 pid, (int) ShutdownPID);
					abort();
				}
				ShutdownPID = ShutdownDataBase();
			}

			goto reaper_done;
		}

		/*
		 * Else do standard child cleanup.
		 */
		CleanupProc(pid, exitstatus);

	}							/* loop over pending child-death reports */
#endif

	if (FatalError)
	{
		/*
		 * Wait for all children exit, then reset shmem and
		 * StartupDataBase.
		 */
		if (DLGetHead(BackendList) || StartupPID > 0 || ShutdownPID > 0)
			goto reaper_done;
		ereport(LOG,
			(errmsg("all server processes terminated; reinitializing")));

		shmem_exit(0);
		reset_shared(PostPortNumber);

		StartupPID = StartupDataBase();

		goto reaper_done;
	}

	if (Shutdown > NoShutdown)
	{
		if (DLGetHead(BackendList))
			goto reaper_done;
		if (StartupPID > 0 || ShutdownPID > 0)
			goto reaper_done;
		ShutdownPID = ShutdownDataBase();
	}

reaper_done:
	PG_SETMASK(&UnBlockSig);

	errno = save_errno;
}

/*
 * CleanupProc -- cleanup after terminated backend.
 *
 * Remove all local state associated with backend.
 */
static void
CleanupProc(int pid,
			int exitstatus)		/* child's exit status. */
{
	Dlelem	   *curr,
			   *next;
	Backend    *bp;

	LogChildExit(DEBUG2, gettext("child process"), pid, exitstatus);

	/*
	 * If a backend dies in an ugly way (i.e. exit status not 0) then we
	 * must signal all other backends to quickdie.	If exit status is zero
	 * we assume everything is hunky dory and simply remove the backend
	 * from the active backend list.
	 */
	if (exitstatus == 0)
	{
		curr = DLGetHead(BackendList);
		while (curr)
		{
			bp = (Backend *) DLE_VAL(curr);
			if (bp->pid == pid)
			{
				DLRemove(curr);
				free(bp);
				DLFreeElem(curr);
				break;
			}
			curr = DLGetSucc(curr);
		}

		if (pid == CheckPointPID)
		{
			CheckPointPID = 0;
			if (!FatalError)
			{
				checkpointed = time(NULL);
				/* Update RedoRecPtr for future child backends */
				GetSavedRedoRecPtr();
			}
		}
		else
			pgstat_beterm(pid);

		return;
	}

	/* below here we're dealing with a non-normal exit */

	/* Make log entry unless we did so already */
	if (!FatalError)
	{
		LogChildExit(LOG,
				 (pid == CheckPointPID) ? gettext("checkpoint process") :
					 gettext("server process"),
					 pid, exitstatus);
		ereport(LOG,
			  (errmsg("terminating any other active server processes")));
	}

	curr = DLGetHead(BackendList);
	while (curr)
	{
		next = DLGetSucc(curr);
		bp = (Backend *) DLE_VAL(curr);
		if (bp->pid != pid)
		{
			/*
			 * This backend is still alive.  Unless we did so already,
			 * tell it to commit hara-kiri.
			 *
			 * SIGQUIT is the special signal that says exit without proc_exit
			 * and let the user know what's going on. But if SendStop is
			 * set (-s on command line), then we send SIGSTOP instead, so
			 * that we can get core dumps from all backends by hand.
			 */
			if (!FatalError)
			{
				ereport(DEBUG2,
						(errmsg_internal("sending %s to process %d",
										 (SendStop ? "SIGSTOP" : "SIGQUIT"),
										 (int) bp->pid)));
				kill(bp->pid, (SendStop ? SIGSTOP : SIGQUIT));
			}
		}
		else