From: ebiederm@xmission.com (Eric W. Biederman)
To: Oleg Nesterov <oleg@redhat.com>
Cc: Linux Containers <containers@lists.linux-foundation.org>,
linux-kernel@vger.kernel.org, Serge Hallyn <serge@hallyn.com>,
Gao feng <gaofeng@cn.fujitsu.com>,
Andrew Morton <akpm@linux-foundation.org>
Subject: Re: [PATCH 07/11] pidns: Wait in zap_pid_ns_processes until pid_ns->nr_hashed == 1
Date: Fri, 21 Dec 2012 10:42:57 -0800 [thread overview]
Message-ID: <8738yzl00e.fsf@xmission.com> (raw)
In-Reply-To: <20121221153152.GA17250@redhat.com> (Oleg Nesterov's message of "Fri, 21 Dec 2012 16:31:52 +0100")
Oleg Nesterov <oleg@redhat.com> writes:
> On 12/21, Oleg Nesterov wrote:
>>
>> Once again, the parent namespace injects the task T after ns->reaper
>> sees nr_hashed == 1 and returns. Suppose that reaper's parent does
>> do_wait() and free_pidmap() clears the bit == 1.
>>
>> Now, what if T doesn't exit but forks? We must not re-create the
>> task with pid_nr == 1 in the dead namespace. Normally this can't
>> happen, RESERVED_PIDS logic in alloc_pidmap() saves us. But it
>> seems that we need
>>
>> - .extra1 = &zero,
>> + .extra1 = &one,
>>
>> in pid_ns_ctl_table.
>
> Oh, and another problem, or I am totally confused.
>
> T forks and creates the child C1. C1 creates C2. What if C1 exits?
> It will try to reparent C2 to the dead/freed ns->child_reaper.
>
> In short. We shouldn't allow alloc_pid() if ns->child_reaper is dying,
> I think. nr_hashed == -1 doesn't really work.
Certainly nr_hashed == -1 is insufficient.
Injecting a processes when nr_hashed == 1 seems to be the magic poison.
I wonder if we could just say.
if (ns->nr_hashed == -1)
goto out_unlock;
if ((ns->nr_hashed >= 1) && (ns->child_reaper->flags & PF_EXITING))
goto out_unlock;
I don't know if the locking is sufficient at that point.
Eric
next prev parent reply other threads:[~2012-12-21 18:43 UTC|newest]
Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-11-16 16:32 [REVIEW][PATCH 0/11] pid namespace cleanups and enhancements Eric W. Biederman
2012-11-16 16:35 ` [PATCH 01/11] procfs: Use the proc generic infrastructure for proc/self Eric W. Biederman
2012-11-16 16:35 ` [PATCH 02/11] procfs: Don't cache a pid in the root inode Eric W. Biederman
2012-11-21 1:07 ` Gao feng
2012-11-16 16:35 ` [PATCH 03/11] pidns: Capture the user namespace and filter ns_last_pid Eric W. Biederman
2012-11-21 1:26 ` Gao feng
2012-11-16 16:35 ` [PATCH 04/11] pidns: Use task_active_pid_ns where appropriate Eric W. Biederman
2012-11-21 2:02 ` Gao feng
2012-11-16 16:35 ` [PATCH 05/11] pidns: Make the pidns proc mount/umount logic obvious Eric W. Biederman
2012-11-19 11:02 ` Gao feng
2012-11-16 16:35 ` [PATCH 06/11] pidns: Don't allow new processes in a dead pid namespace Eric W. Biederman
2012-11-21 2:17 ` Gao feng
2012-11-16 16:35 ` [PATCH 07/11] pidns: Wait in zap_pid_ns_processes until pid_ns->nr_hashed == 1 Eric W. Biederman
2012-11-21 2:24 ` Gao feng
2012-12-19 18:47 ` Oleg Nesterov
2012-12-21 1:19 ` Eric W. Biederman
2012-12-21 14:11 ` Oleg Nesterov
2012-12-21 15:02 ` Oleg Nesterov
2012-12-21 15:31 ` Oleg Nesterov
2012-12-21 18:42 ` Eric W. Biederman [this message]
2012-12-21 18:33 ` Eric W. Biederman
2012-11-16 16:35 ` [PATCH 08/11] pidns: Deny strange cases when creating pid namespaces Eric W. Biederman
2012-11-21 2:25 ` Gao feng
2012-11-16 16:35 ` [PATCH 09/11] pidns: Add setns support Eric W. Biederman
2012-11-19 9:11 ` Gao feng
2012-11-19 9:27 ` Eric W. Biederman
2012-11-21 2:36 ` Gao feng
2012-11-16 16:35 ` [PATCH 10/11] pidns: Consolidate initialzation of special init task state Eric W. Biederman
2012-11-21 2:56 ` Gao feng
2012-11-16 16:35 ` [PATCH 11/11] pidns: Support unsharing the pid namespace Eric W. Biederman
2012-11-21 2:55 ` Gao feng
2012-12-19 18:14 ` Oleg Nesterov
2012-12-21 1:43 ` Eric W. Biederman
2012-12-21 15:49 ` Oleg Nesterov
2012-12-21 17:51 ` Eric W. Biederman
2012-12-21 19:24 ` Rob Landley
2012-12-21 22:58 ` namespace documentation Eric W. Biederman
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=8738yzl00e.fsf@xmission.com \
--to=ebiederm@xmission.com \
--cc=akpm@linux-foundation.org \
--cc=containers@lists.linux-foundation.org \
--cc=gaofeng@cn.fujitsu.com \
--cc=linux-kernel@vger.kernel.org \
--cc=oleg@redhat.com \
--cc=serge@hallyn.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
Powered by JetHome