From: Christian Brauner <christian@brauner.io>
To: Daniel Colascione <dancol@google.com>
Cc: Andy Lutomirski <luto@amacapital.net>,
Aleksa Sarai <cyphar@cyphar.com>,
Andy Lutomirski <luto@kernel.org>,
Randy Dunlap <rdunlap@infradead.org>,
"Eric W. Biederman" <ebiederm@xmission.com>,
LKML <linux-kernel@vger.kernel.org>,
"Serge E. Hallyn" <serge@hallyn.com>,
Jann Horn <jannh@google.com>,
Andrew Morton <akpm@linux-foundation.org>,
Oleg Nesterov <oleg@redhat.com>,
Al Viro <viro@zeniv.linux.org.uk>,
Linux FS Devel <linux-fsdevel@vger.kernel.org>,
Linux API <linux-api@vger.kernel.org>,
Tim Murray <timmurray@google.com>,
Kees Cook <keescook@chromium.org>,
Jan Engelhardt <jengelh@inai.de>
Subject: Re: [PATCH] proc: allow killing processes via file descriptors
Date: Mon, 19 Nov 2018 01:40:39 +0100 [thread overview]
Message-ID: <20181119004037.d4avmjyiooa7ujyf@brauner.io> (raw)
In-Reply-To: <CAKOZuetT8mzZJR-K_W0VM7Dg1c1SnH8tW-HXxMT77yS0DE453w@mail.gmail.com>
On Sun, Nov 18, 2018 at 04:31:22PM -0800, Daniel Colascione wrote:
> On Sun, Nov 18, 2018 at 1:30 PM, Christian Brauner <christian@brauner.io> wrote:
> > On Sun, Nov 18, 2018 at 10:23:36PM +0100, Christian Brauner wrote:
> >> On Sun, Nov 18, 2018 at 12:54:10PM -0800, Daniel Colascione wrote:
> >> > On Sun, Nov 18, 2018 at 12:43 PM, Christian Brauner
> >> > <christian@brauner.io> wrote:
> >> > > On Sun, Nov 18, 2018 at 01:28:41PM -0700, Andy Lutomirski wrote:
> >> > >>
> >> > >>
> >> > >> > On Nov 18, 2018, at 12:44 PM, Daniel Colascione <dancol@google.com> wrote:
> >> > >> >
> >> > >>
> >> > >> >
> >> > >> > That is, I'm proposing an API that looks like this:
> >> > >> >
> >> > >> > int process_kill(int procfs_dfd, int signo, const union sigval value)
> >> > >> >
> >> > >> > If, later, process_kill were to *also* accept process-capability FDs,
> >> > >> > nothing would break.
> >> > >>
> >> > >> Except that this makes it ambiguous to the caller as to whether their current creds are considered. So it would need to be a different syscall or at least a flag. Otherwise a lot of those nice theoretical properties go away.
> >> > >
> >> > > I can add a flag argument
> >> > > int process_signal(int procfs_dfd, int signo, siginfo_t *info, int flags)
> >> > > The way I see it process_signal() should be equivalent to kill(pid, signal) for now.
> >> > > That is siginfo_t is cleared and set to:
> >> > >
> >> > > info.si_signo = sig;
> >> > > info.si_errno = 0;
> >> > > info.si_code = SI_USER;
> >> > > info.si_pid = task_tgid_vnr(current);
> >> > > info.si_uid = from_kuid_munged(current_user_ns(), current_uid());
> >> >
> >> > That makes sense. I just don't want to get into a situation where
> >> > callers feel that they *have* to use the PID-based APIs to send a
> >> > signal because process_kill doesn't offer some bit of functionality.
> >>
> >> Yeah.
> >>
> >> >
> >> > Are you imagining something like requiring info t be NULL unless flags
> >> > contains some "I have a siginfo_t" value?
> >>
> >> Well, I was actually thinking about something like:
> >>
> >> /**
> >> * sys_process_signal - send a signal to a process trough a process file descriptor
> >> * @fd: the file descriptor of the process
> >> * @sig: signal to be sent
> >> * @info: the signal info
> >> * @flags: future flags to be passed
> >> */
> >> SYSCALL_DEFINE4(process_signal, int, fd, int, sig, siginfo_t __user *, info,
> >> int, flags)
> >> {
> >> struct pid *pid;
> >> struct fd *f;
> >> kernel_siginfo_t kinfo;
> >>
> >> /* Do not allow users to pass garbage. */
> >> if (flags)
> >> return -EINVAL;
> >>
> >> int ret = __copy_siginfo_from_user(sig, &kinfo, info);
> >> if (unlikely(ret))
> >> return ret;
> >>
> >> /* For now, enforce that caller's creds are used. */
> >> kinfo.si_pid = task_tgid_vnr(current);
> >> kinfo.si_uid = from_kuid_munged(current_user_ns(), current_uid());
>
> How about doing it this way? If info is NULL, act like kill(2);
> otherwise, act like rt_sigqueueinfo(2).
>
> (Not actual working or compiled code.)
>
> SYSCALL_DEFINE4(process_signal, int, fd, int, sig, siginfo_t __user *, info,
> int, flags)
> {
> struct fd f = { 0 };
> kernel_siginfo_t kinfo;
> int ret;
>
> /* Make API extension possible. */
> ret = -EINVAL;
> if (flags)
> goto out;
>
> ret = -EBADF;
> f = fdget(fd);
> if (!f.file)
> goto out;
>
> ret = mumble_mumble_check_real_proc_file(f.file);
> if (ret)
> goto out;
>
> /* Act like kill(2) or rt_sigqueueinfo(2) depending on whether
> * the user gave us a siginfo structure.
> */
> if (info) {
> ret = __copy_siginfo_from_user(sig, &kinfo, info);
> if (ret)
> goto out;
> /* Combine this logic with rt_sigqueueinfo(2) */
> ret = -EPERM;
> if ((info->si_code >= 0 || info->si_code == SI_TKILL) &&
> (task_pid_vnr(current) != pid))
> goto out;
>
> } else {
> /* Combine this logic with kill(2) */
> clear_siginfo(&kinfo);
> kinfo.si_signo = sig;
> kinfo.si_errno = 0;
> kinfo.si_code = SI_USER;
> kinfo.si_pid = task_tgid_vnr(current);
> kinfo.si_uid = from_kuid_munged(current_user_ns(),
> current_uid());
> }
>
> ret = kill_pid_info(sig, &kinfo, proc_pid(file_inode(f.file)));
>
> out:
> if (f.file)
> fput(f);
> return ret;
> }
Right, allowing to ass NULL might make sense. I had:
/**
* sys_process_signal - send a signal to a process trough a process file descriptor
* @fd: the file descriptor of the process
* @sig: signal to be sent
* @info: the signal info
* @flags: future flags to be passed
*/
SYSCALL_DEFINE4(procfd_kill, int, fd, int, sig, siginfo_t __user *, info, int, flags)
{
int ret;
struct pid *pid;
kernel_siginfo_t kinfo;
struct fd f;
/* Do not allow users to pass garbage. */
if (flags)
return -EINVAL;
ret = __copy_siginfo_from_user(sig, &kinfo, info);
if (unlikely(ret))
return ret;
/* For now, enforce that caller's creds are used. */
kinfo.si_pid = task_tgid_vnr(current);
kinfo.si_uid = from_kuid_munged(current_user_ns(), current_uid());
f = fdget_raw(fd);
if (!f.file)
return -EBADF;
ret = -EINVAL;
/* Is this a process file descriptor? */
if (!proc_is_procfd(f.file) || !d_is_dir(f.file->f_path.dentry))
goto err;
pid = f.file->private_data;
if (!pid)
goto err;
ret = -EPERM;
/*
* Not even root can pretend to send signals from the kernel.
* Nor can they impersonate a kill()/tgkill(), which adds source info.
*/
if ((kinfo.si_code >= 0 || kinfo.si_code == SI_TKILL) &&
(task_pid(current) != pid))
goto err;
ret = kill_pid_info(sig, &kinfo, pid);
err:
fdput(f);
return ret;
}
next prev parent reply other threads:[~2018-11-19 0:40 UTC|newest]
Thread overview: 53+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-11-18 11:17 [PATCH] proc: allow killing processes via file descriptors Christian Brauner
2018-11-18 13:59 ` Daniel Colascione
2018-11-18 15:38 ` Andy Lutomirski
2018-11-18 15:53 ` Daniel Colascione
2018-11-18 16:17 ` Andy Lutomirski
2018-11-18 16:29 ` Daniel Colascione
2018-11-18 17:13 ` Andy Lutomirski
2018-11-18 17:17 ` Daniel Colascione
2018-11-18 17:43 ` Eric W. Biederman
2018-11-18 17:45 ` Andy Lutomirski
2018-11-18 17:56 ` Daniel Colascione
2018-11-18 16:33 ` Randy Dunlap
2018-11-18 16:48 ` Daniel Colascione
2018-11-18 17:09 ` Andy Lutomirski
2018-11-18 17:24 ` Daniel Colascione
2018-11-18 17:42 ` Andy Lutomirski
2018-11-18 17:51 ` Daniel Colascione
2018-11-18 18:28 ` Andy Lutomirski
2018-11-18 18:43 ` Daniel Colascione
2018-11-18 19:05 ` Aleksa Sarai
2018-11-18 19:44 ` Daniel Colascione
2018-11-18 20:15 ` Christian Brauner
2018-11-18 20:21 ` Daniel Colascione
2018-11-18 20:28 ` Andy Lutomirski
2018-11-18 20:32 ` Daniel Colascione
2018-11-19 1:43 ` Andy Lutomirski
2018-11-18 20:43 ` Christian Brauner
2018-11-18 20:54 ` Daniel Colascione
2018-11-18 21:23 ` Christian Brauner
2018-11-18 21:30 ` Christian Brauner
2018-11-19 0:31 ` Daniel Colascione
2018-11-19 0:40 ` Christian Brauner [this message]
2018-11-19 0:09 ` Aleksa Sarai
2018-11-19 0:53 ` Daniel Colascione
2018-11-19 1:16 ` Daniel Colascione
2018-11-19 16:13 ` Dmitry Safonov
2018-11-19 16:26 ` [PATCH] proc: allow killing processes via file descriptors (Larger pids) Eric W. Biederman
2018-11-19 16:27 ` [PATCH] proc: allow killing processes via file descriptors Daniel Colascione
2018-11-19 20:21 ` Aleksa Sarai
2018-11-19 2:47 ` Al Viro
2018-11-19 3:01 ` Andy Lutomirski
2018-11-18 17:41 ` Christian Brauner
2018-11-18 17:44 ` Andy Lutomirski
2018-11-18 18:07 ` Daniel Colascione
2018-11-18 18:15 ` Andy Lutomirski
2018-11-18 18:31 ` Daniel Colascione
2018-11-18 19:24 ` Christian Brauner
2018-11-19 0:08 ` Aleksa Sarai
2018-11-19 1:14 ` Daniel Colascione
2018-11-18 16:03 ` Daniel Colascione
2018-11-19 10:56 ` kbuild test robot
2018-11-19 14:15 ` David Laight
2018-11-19 15:49 ` Dave Martin
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=20181119004037.d4avmjyiooa7ujyf@brauner.io \
--to=christian@brauner.io \
--cc=akpm@linux-foundation.org \
--cc=cyphar@cyphar.com \
--cc=dancol@google.com \
--cc=ebiederm@xmission.com \
--cc=jannh@google.com \
--cc=jengelh@inai.de \
--cc=keescook@chromium.org \
--cc=linux-api@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=luto@amacapital.net \
--cc=luto@kernel.org \
--cc=oleg@redhat.com \
--cc=rdunlap@infradead.org \
--cc=serge@hallyn.com \
--cc=timmurray@google.com \
--cc=viro@zeniv.linux.org.uk \
/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;
as well as URLs for NNTP newsgroup(s).