mjhjmtu

I am running Linux.

I have a single process in a mount namespace.
I did in this process a mount -t tmpfs tmpfs /mountpoint.

What happens if the process exits and there are no more processes in that mount namespace?
Will the filesystem be automatically unmounted?
Will the mount namespace be destroyed?
If the namespaces and the mount are still active how do I access it?

What happens to tun/tap/macvtap interfaces if a network namespace has no more processes?

It seems that the mounts remain, but become unaccessible.

I did the following as a test:

1. Entered a new namespace and mounted tmpfs:

root@localhost:~# mkdir tmp
root@localhost:~# unshare -m bash
root@localhost:~# mount -t tmpfs tmpfs tmp

2. Checked memory usage before and after creating a 200 megabyte file on the tmpfs. You can notice the "shared" usage going from 404 megabytes to 604 megabytes:

root@localhost:~# free -m
 total used free shared buff/cache available
Mem: 9885 2966 681 404 6237 6148
Swap: 8191 293 7898
root@localhost:~# dd if=/dev/urandom of=tmp/dummy bs=1M count=200
200+0 records in
200+0 records out
209715200 bytes (210 MB, 200 MiB) copied, 12.0075 s, 17.5 MB/s
root@localhost:~# free -m
 total used free shared buff/cache available
Mem: 9885 2966 481 604 6437 5948
Swap: 8191 293 7898

3. Exited the unshare shell, the memory did not get reclaimed:

root@localhost:~# exit
root@localhost:~# free -m
 total used free shared buff/cache available
Mem: 9885 2963 484 604 6437 5951
Swap: 8191 293 7898

Whereas if I unmount the tmpfs, the shared count goes back to initial value.

Cannot add a comment, because not enough rep, so here it goes:
I did exactly the same test as jpa on a unmodified debian stretch and the memory gets back to the initial value, even though I did NOT unmount. So it seems this behaviour has changed.

http://man7.org/linux/man-pages/man7/namespaces.7.html

implies this at least by saying:

Bind mounting (see mount(2)) one of the files in this directory to
somewhere else in the filesystem keeps the corresponding namespace of
the process specified by pid alive even if all processes currently in
the namespace terminate.

So not doing that destroys the namespace (?).