Pathname resolution starts from the user's root or working directory. The LFS looks up the first component of the pathname in that directory. This often yields another directory, and the LFS looks up the second component of the name in this new directory. The LFS looks up each successive component of the name in the directory that was returned from the previous lookup, until the end of the pathname is reached.

When the LFS encounters a directory that is a mount point, it switches to the root directory of the file system that was mounted there. The next lookup is done in the mounted file system's root directory, rather than in the directory that was returned from the previous lookup. This is called crossing mount points; it is because of these mount points that pathname resolution has to be done one component at a time.

Resolving pathnames and identifying mount points is a function of the LFS. Except for the individual vn_lookup operations that are invoked, the PFS is not involved.