linux-kernel - Re: [PATCH v2 1/5] pidfd: verify task is alive when printing fdinfo

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: <20191016163107.5zwt6fmjyd5mkqqw@wittgenstein>
Date:   Wed, 16 Oct 2019 18:31:08 +0200
From:   Christian Brauner <christian.brauner@...ntu.com>
To:     Oleg Nesterov <oleg@...hat.com>
Cc:     linux-kernel@...r.kernel.org, aarcange@...hat.com,
        akpm@...ux-foundation.org, christian@...lner.me, cyphar@...har.com,
        elena.reshetova@...el.com, guro@...com, jannh@...gle.com,
        ldv@...linux.org, linux-api@...r.kernel.org,
        linux-kselftest@...r.kernel.org, mhocko@...e.com, mingo@...nel.org,
        peterz@...radead.org, shuah@...nel.org, tglx@...utronix.de,
        viro@...iv.linux.org.uk
Subject: Re: [PATCH v2 1/5] pidfd: verify task is alive when printing fdinfo

On Wed, Oct 16, 2019 at 06:24:09PM +0200, Oleg Nesterov wrote:
> On 10/16, Christian Brauner wrote:
> >
> > +static inline bool task_alive(struct pid *pid, enum pid_type type)
> > +{
> > +	return !hlist_empty(&pid->tasks[type]);
> > +}
> 
> So you decided to add a helper ;) OK, but note that its name is very
> confusing and misleading. Even more than pid_alive() we already have.

That's why I chose that name. This is the second time I get bitten by
taking inspiration from prior naming examples. :)

> 
> What does "alive" actually mean? Say, task_alive(pid, PIDTYPE_SID) == F
> after fork(). Then it becomes T if this task does setsid().

Yes, that annoyed me to. If you think about pidfd_open() you have a
similar problem. The question we are asking in pidfd_open() is not
task_alive() but rather was-this-pid-used-as.

> 
> And why task_ if it accepts pid+pid_type? May be pid_has_task() or
> something like this...

Given what I said above that might be a decent name.

> 
> OK, since I can't suggest a better name I won't really argue. Feel free
> to add my reviewed-by to this series.

No, naming is important. Thanks for being picky about that too and I'll
happily resend. :)

Thanks!
Christian