lkml.org 
[lkml]   [2003]   [Oct]   [1]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectCLONE_NEWNS & chroot(2) behavior
From
Date
Hello,

I have problems to understand the CLONE_NEWNS semantic because I see
some strange behavior in combination with chroot(2).

E.g. look at the program at

http://www.tu-chemnitz.de/~ensc/nst.c



There, the umount(2) in 'child()' fails when the chroot is outside
/dev/root. E.g. when /var/tmp is on a separate partition, I get


# mkdir -p /var/tmp/foo/bar && strace -f ./nst /var/tmp/foo bar
| mount("none", "/var/tmp/foo/bar", "proc", , 0) = 0
| chroot("/var/tmp/foo") = 0
| chdir("/") = 0
| clone(Process 28706 attached
| child_stack=0xbffff7bc, flags=CLONE_NEWNS|SIGCHLD) = 28706
| [pid 28706] --- SIGSTOP (Stopped (signal)) @ 0 (0) ---
| [pid 28706] oldumount("bar") = -1 EINVAL (Invalid argument)


The umount(2) can be replaced with anything which accesses files in
"bar"---it would fail too, since "bar" is not mounted in the new
namespace.

But when doing the same in /, things are ok:

# mkdir -p /foo/bar && strace -f ./nst /foo bar
| mount("none", "/foo/bar", "proc", , 0) = 0
| chroot("/foo") = 0
| chdir("/") = 0
| clone(Process 28733 attached
| child_stack=0xbffff7cc, flags=CLONE_NEWNS|SIGCHLD) = 28733
| [pid 28733] --- SIGSTOP (Stopped (signal)) @ 0 (0) ---
| [pid 28733] oldumount("bar") = 0



When this behavior is wanted, what is the formal description of the
CLONE_NEWNS behavior?




Enrico
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:49    [W:0.095 / U:0.008 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site