User namespaces: Only put the userns when we unhash the uid
authorSerge E. Hallyn <serue@us.ibm.com>
Fri, 13 Feb 2009 14:04:21 +0000 (14:04 +0000)
committerLinus Torvalds <torvalds@linux-foundation.org>
Fri, 13 Feb 2009 16:07:40 +0000 (08:07 -0800)
commitfb5ae64fdde29236e1a15e0366946df7060f41f2
tree86be58998529fa84282ec1b98040d707e4b61686
parent37bed90094fdb1eea6e4afec6a200d4e60143e55
User namespaces: Only put the userns when we unhash the uid

uids in namespaces other than init don't get a sysfs entry.

For those in the init namespace, while we're waiting to remove
the sysfs entry for the uid the uid is still hashed, and
alloc_uid() may re-grab that uid without getting a new
reference to the user_ns, which we've already put in free_user
before scheduling remove_user_sysfs_dir().

Reported-and-tested-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: Serge E. Hallyn <serue@us.ibm.com>
Acked-by: David Howells <dhowells@redhat.com>
Tested-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
kernel/user.c