libata: Fix use-after-iounmap

Jens Axboe pointed out that the iounmap() call in libata was occurring
too early, and some drivers (ahci, probably others) were using ioremap'd
memory after it had been unmapped.

The patch should address that problem by way of improving the libata
driver API:

* move ->host_stop() call after all ->port_stop() calls have occurred.

* create default helper function ata_host_stop(), and move iounmap()
call there.

* add ->host_stop_prewalk() hook, use it in sata_qstor.c (hi Mark).
sata_qstor appears to require the host-stop-before-port-stop ordering
that existed prior to applying the attached patch.

drivers/scsi/sata_vsc.c

@@ -230,6 +230,7 @@ static struct ata_port_operations vsc_sata_ops = {
 	.scr_write		= vsc_sata_scr_write,
 	.port_start		= ata_port_start,
 	.port_stop		= ata_port_stop,
+	.host_stop		= ata_host_stop,
 };
 
 static void __devinit vsc_sata_setup_port(struct ata_ioports *port, unsigned long base)