libata: fallback to the other IDENTIFY on device error, take#2

It seems the world isn't as frank as we thought and some devices lie
about who they are.  Fallback to the other IDENTIFY if IDENTIFY is
aborted by the device.  As this is the strategy used by IDE for a long
time, it shouldn't cause too much problem.
Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: William Thompson <wt@electro-mechanical.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

drivers/ata/libata-core.c

@@ -1660,7 +1660,7 @@ int ata_dev_read_id(struct ata_device *dev, unsigned int *p_class,
 	struct ata_taskfile tf;
 	unsigned int err_mask = 0;
 	const char *reason;
-	int tried_spinup = 0;
+	int may_fallback = 1, tried_spinup = 0;
 	int rc;
 
 	if (ata_msg_ctl(ap))
@@ -1704,11 +1704,31 @@ int ata_dev_read_id(struct ata_device *dev, unsigned int *p_class,
 			return -ENOENT;
 		}
 
+		/* Device or controller might have reported the wrong
+		 * device class.  Give a shot at the other IDENTIFY if
+		 * the current one is aborted by the device.
+		 */
+		if (may_fallback &&
+		    (err_mask == AC_ERR_DEV) && (tf.feature & ATA_ABORTED)) {
+			may_fallback = 0;
+
+			if (class == ATA_DEV_ATA)
+				class = ATA_DEV_ATAPI;
+			else
+				class = ATA_DEV_ATA;
+			goto retry;
+		}
+
 		rc = -EIO;
 		reason = "I/O error";
 		goto err_out;
 	}
 
+	/* Falling back doesn't make sense if ID data was read
+	 * successfully at least once.
+	 */
+	may_fallback = 0;
+
 	swap_buf_le16(id, ATA_ID_WORDS);
 
 	/* sanity check */