On Mon, 25 Mar 2002, Richard Gooch wrote:
> Christoph Hellwig writes:
> > In article <[email protected]> you wrote:
> > > Hi, Al. What is the intended behaviour of MNT_DETACH
> > > wrt. /proc/mounts? It appears that detaching a FS leaves behind an
> > > apparently stale entry in /proc/mounts. Is this intentional?
> >
> > mount(8) code for MNT_DETACH is mine, not viro's. And yes, leaving
> > stale entry there is a bug - I will submit a patch to clear it.
>
> This is not a problem with mount(8). It's a kernel problem. When
> sys_umount(..., MNT_DETACH) is called from kernel space, it doesn't
> clean up /proc/mounts.
Which kernel version it is?
It works here, so I'd love to see testcase.