Package: mount Version: 2.12p-4 Severity: wishlist Problem description:
The 2.4 Linux kernel provides features that can be used by mount to mount a mounted filesystem to an additional mount point. There are two different ways how to do this: either with --bind, or with --rbind. The documentation says that --bind will just bind the mounted volume, excluding sub-mounts, while --rbind will include sub-mounts. This, however, does not specify how the system behaves if one mounts additional volumes below the new mount point. Will they also become available under the old mount point as well? In fact, they do not, and this is highly useful behaviour which one would sometimes want to rely on. But as this is not documented semantics, one cannot. Imagine the following scenario: There are various situations where one wants to provide an image of "mostly the presently running installation" (i.e. /, /usr, /var, etc.) - or a LVM snapshot of it. Examples: ad-hoc creation of a CD-bootable system, NFS-export of a bootable system, etc. One pretty elegant solution is to just use mount --bind to create the desired structure under a new mount point from the available volumes, and then perform some extra modifications e.g. by loopback-monting image files into the --bind-mounted directory tree. This should not only work, but one should also have guarantees (i.e. a documented promise) that it does work. -- regards, [EMAIL PROTECTED] (o_ Thomas Fischbacher - http://www.cip.physik.uni-muenchen.de/~tf //\ (lambda (n) ((lambda (p q r) (p p q r)) (lambda (g x y) V_/_ (if (= x 0) y (g g (- x 1) (* x y)))) n 1)) (Debian GNU) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
