linux-kernel - Re: [PATCH v2 1/4] seccomp: add a return code to trap to userspace

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20180517153957.GA3831@cisco>
Date:   Thu, 17 May 2018 09:39:57 -0600
From:   Tycho Andersen <tycho@...ho.ws>
To:     Oleg Nesterov <oleg@...hat.com>
Cc:     linux-kernel@...r.kernel.org,
        containers@...ts.linux-foundation.org,
        Kees Cook <keescook@...omium.org>,
        Andy Lutomirski <luto@...capital.net>,
        "Eric W . Biederman" <ebiederm@...ssion.com>,
        "Serge E . Hallyn" <serge@...lyn.com>,
        Christian Brauner <christian.brauner@...ntu.com>,
        Tyler Hicks <tyhicks@...onical.com>,
        Akihiro Suda <suda.akihiro@....ntt.co.jp>,
        "Tobin C . Harding" <me@...in.cc>
Subject: Re: [PATCH v2 1/4] seccomp: add a return code to trap to userspace

Hi Oleg,

Thanks for taking a look!

On Thu, May 17, 2018 at 05:33:24PM +0200, Oleg Nesterov wrote:
> I didn't read this series yet, and I don't even understand what are you
> trying to do, just one question...
> 
> On 05/17, Tycho Andersen wrote:
> >
> > +static struct file *init_listener(struct task_struct *task,
> > +				  struct seccomp_filter *filter)
> > +{
> > +	struct file *ret = ERR_PTR(-EBUSY);
> > +	struct seccomp_filter *cur;
> > +	bool have_listener = false;
> > +
> > +	for (cur = task->seccomp.filter; cur; cur = cur->prev) {
> > +		mutex_lock(&cur->notify_lock);
> 
> Did you test this patch with CONFIG_LOCKDEP ?

Yes, with,

CONFIG_LOCKDEP=y
CONFIG_DEBUG_LOCKDEP=y
CONFIG_DEBUG_ATOMIC_SLEEP=y

> From lockdep pov this loop tries to take the same lock twice or more, it shoul
> complain.

I didn't, but I guess that's because it's not trying to take the same lock
twice -- the pointer cur is changing in the loop. Unless I'm misunderstanding
what you're saying.

The idea behind this code is to lock the entire chain of filters up to the
parent so that we can ensure none of them have a listener installed. This is
based on a suggestion from Andy last review cycle to not allow two listeners,
since nesting would be confusing.

Tycho