The RBD image size field in protobuf is 32 bits which is not big enough

[gbregman]

When we used an image size of 8GB we got an error in grpc.c:

Jan 03 14:01:46 ceph-nvme-vm1 bash[103930]: INFO:control.grpc:list_namespaces: [{'nqn': 'nqn.2016-06.io.spdk:cnode1', 'subtype': 'NVMe', 'listen_addresses': [], 'allow_any_host': True, 'hosts': [], 'serial_number': 'SPDK24287840519620', 'model_number': 'SPDK bdev Controller', 'max_namespaces': 256, 'min_cntlid': 1, 'max_cntlid': 65519, 'namespaces': [{'nsid': 1, 'bdev_name': 'bdev_2ad8834b-52a7-4563-b7c2-c35fec4a542d', 'name': 'bdev_2ad8834b-52a7-4563-b7c2-c35fec4a542d', 'nguid': '2AD8834B52A74563B7C2C35FEC4A542D', 'uuid': '2ad8834b-52a7-4563-b7c2-c35fec4a542d'}]}]
Jan 03 14:01:46 ceph-nvme-vm1 bash[103930]: ERROR:control.grpc:ns_bdev={'name': 'bdev_2ad8834b-52a7-4563-b7c2-c35fec4a542d', 'aliases': ['2ad8834b-52a7-4563-b7c2-c35fec4a542d'], 'product_name': 'Ceph Rbd Disk', 'block_size': 512, 'num_blocks': 16777216, 'uuid': '2ad8834b-52a7-4563-b7c2-c35fec4a542d', 'assigned_rate_limits': {'rw_ios_per_sec': 0, 'rw_mbytes_per_sec': 0, 'r_mbytes_per_sec': 0, 'w_mbytes_per_sec': 0}, 'claimed': True, 'zoned': False, 'supported_io_types': {'read': True, 'write': True, 'unmap': True, 'write_zeroes': True, 'flush': True, 'reset': True, 'compare': False, 'compare_and_write': False, 'abort': False, 'nvme_admin': False, 'nvme_io': False}, 'driver_specific': {'rbd': {'pool_name': 'mypool', 'rbd_name': 'myimage', 'user_id': 'nvmeof.mypool.ceph-nvme-vm1.kbdjnt'}}} parse error:
Jan 03 14:01:46 ceph-nvme-vm1 bash[103930]: Traceback (most recent call last):
Jan 03 14:01:46 ceph-nvme-vm1 bash[103930]:   File "/src/control/grpc.py", line 875, in list_namespaces
Jan 03 14:01:46 ceph-nvme-vm1 bash[103930]:     one_ns.rbd_image_size = ns_bdev["block_size"] * ns_bdev["num_blocks"]
Jan 03 14:01:46 ceph-nvme-vm1 bash[103930]: ValueError: Value out of range: 8589934592

Looking at the definition of rbd_image_size I see it's uint32 which is not big enough for 8GB. We need to change it to uint64.