media: i2c: ov8858: Don't set fwnode in the driver

This makes the driver work with the new check in
v4l2_async_register_subdev() that was introduced recently in 6.6-rc1.
Without this change, probe fails with:

ov8858 1-0036: Detected OV8858 sensor, revision 0xb2
ov8858 1-0036: sub-device fwnode is an endpoint!
ov8858 1-0036: v4l2 async register subdev failed
ov8858: probe of 1-0036 failed with error -22

This also simplifies the driver a bit.

Signed-off-by: Ondrej Jirman <[e-mail]>
Reviewed-by: Jacopo Mondi <[e-mail]>
Signed-off-by: Sakari Ailus <[e-mail]>
Signed-off-by: Mauro Carvalho Chehab <[e-mail]>
Signed-off-by: Hans Verkuil <[e-mail]>

drivers/media/i2c/ov8858.c

···
 	}
 
 	ret = v4l2_fwnode_endpoint_parse(endpoint, &vep);
+	fwnode_handle_put(endpoint);
 	if (ret) {
 		dev_err(dev, "Failed to parse endpoint: %d\n", ret);
-		fwnode_handle_put(endpoint);
 		return ret;
 	}
 
···
 	default:
 		dev_err(dev, "Unsupported number of data lanes %u\n",
 			ov8858->num_lanes);
-		fwnode_handle_put(endpoint);
 		return -EINVAL;
 	}
-
-	ov8858->subdev.fwnode = endpoint;
 
 	return 0;
 }
···
 
 	ret = ov8858_init_ctrls(ov8858);
 	if (ret)
-		goto err_put_fwnode;
+		return ret;
 
 	sd = &ov8858->subdev;
 	sd->flags |= V4L2_SUBDEV_FL_HAS_DEVNODE | V4L2_SUBDEV_FL_HAS_EVENTS;
···
 	media_entity_cleanup(&sd->entity);
 err_free_handler:
 	v4l2_ctrl_handler_free(&ov8858->ctrl_handler);
-err_put_fwnode:
-	fwnode_handle_put(ov8858->subdev.fwnode);
 
 	return ret;
 }
···
 	v4l2_async_unregister_subdev(sd);
 	media_entity_cleanup(&sd->entity);
 	v4l2_ctrl_handler_free(&ov8858->ctrl_handler);
-	fwnode_handle_put(ov8858->subdev.fwnode);
 
 	pm_runtime_disable(&client->dev);
 	if (!pm_runtime_status_suspended(&client->dev))