ceph_statfs currently stuffs the cluster fsid into the f_fsid field. This was fine when we only had a single filesystem per cluster, but now that we have multiples we need to use something that will vary between them. Change ceph_statfs to mix the current 64 bit hashed fsid down to 32 bits using a trivial xor hash. Shift that to be the upper 32 bits in the 64 bit field and then fold in the fs_cluster_id from the mon client. That should give us a value that is guaranteed to be unique between filesystems within a cluster, and should minimize the chance of collisions between mounts of different clusters. URL: https://tracker.ceph.com/issues/52812 Reported-by: Sachin Prabhu <sprabhu@xxxxxxxxxx> Signed-off-by: Jeff Layton <jlayton@xxxxxxxxxx> --- fs/ceph/super.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/fs/ceph/super.c b/fs/ceph/super.c index 42c502eee80a..b6d2e20e857b 100644 --- a/fs/ceph/super.c +++ b/fs/ceph/super.c @@ -104,6 +104,12 @@ static int ceph_statfs(struct dentry *dentry, struct kstatfs *buf) le64_to_cpu(*((__le64 *)&monc->monmap->fsid + 1)); mutex_unlock(&monc->mutex); + /* mix the fsid down to 32 bits */ + fsid = *(u32 *)&fsid ^ *((u32 *)&fsid + 1); + + /* fold the fs_cluster_id into the upper bits */ + fsid |= ((u64)monc->fs_cluster_id << 32); + buf->f_fsid = u64_to_fsid(fsid); return 0; -- 2.31.1
