The typical use of pivot_root() is during system startup, when the system mounts a temporary root file system (e.g. an initrd), then mounts the real root file system, and eventually turns the latter into the current root of all relevant processes or threads.
pivot_root() may or may not change the current root and the current working directory (cwd) of any processes or threads which use the old root directory. The caller of pivot_root() must ensure that processes with root or cwd at the old root operate correctly in either case. An easy way to ensure this is to change their root and cwd to new_root before invoking pivot_root().
The paragraph above is intentionally vague because the implementation of pivot_root() may change in the future. At the time of writing, pivot_root() changes root and cwd of each process or thread to new_root if they point to the old root directory. This is necessary in order to prevent kernel threads from keeping the old root directory busy with their root and cwd, even if they never access the file system in any way. In the future, there may be a mechanism for kernel threads to explicitly relinquish any access to the file system, such that this fairly intrusive mechanism can be removed from pivot_root().
Note that this also applies to the current process: pivot_root() may or may not affect its cwd. It is therefore recommended to call chdir("/") immediately after pivot_root().
The following restrictions apply to new_root and put_old:
See also pivot_root(8) for additional usage examples.
If the current root is not a mount point (e.g. after chroot(2) or pivot_root(), see also below), not the old root directory, but the mount point of that file system is mounted on put_old.
new_root does not have to be a mount point. In this case, /proc/mounts will show the mount point of the file system containing new_root as root (/).
Some of the more obscure uses of pivot_root() may quickly lead to insanity.