I goofed when I made unshare(CLONE_NEWPID) only work in a single-threaded process. There is no need for that requirement and in fact I analyzied things right for setns. The hard requirement is for tasks that share a VM to all be in the pid namespace and we properly prevent that in do_fork. Just to be certain I took a look through do_wait and forget_original_parent and there are no cases that make it any harder for children to be in the multiple pid namespaces than it is for children to be in the same pid namespace. I also performed a check to see if there were in uses of task->nsproxy_pid_ns I was not familiar with, but it is only used when allocating a new pid for a new task, and in checks to prevent craziness from happening. Signed-off-by: "Eric W. Biederman" <ebiederm@xxxxxxxxxxxx> --- kernel/fork.c | 5 ----- 1 files changed, 0 insertions(+), 5 deletions(-) diff --git a/kernel/fork.c b/kernel/fork.c index 66635c8..eb45f1d 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -1818,11 +1818,6 @@ SYSCALL_DEFINE1(unshare, unsigned long, unshare_flags) if (unshare_flags & CLONE_NEWUSER) unshare_flags |= CLONE_THREAD | CLONE_FS; /* - * If unsharing a pid namespace must also unshare the thread. - */ - if (unshare_flags & CLONE_NEWPID) - unshare_flags |= CLONE_THREAD; - /* * If unsharing a thread from a thread group, must also unshare vm. */ if (unshare_flags & CLONE_THREAD) -- 1.7.5.4 _______________________________________________ Containers mailing list Containers@xxxxxxxxxxxxxxxxxxxxxxxxxx https://lists.linuxfoundation.org/mailman/listinfo/containers
