Re: [PATCH v2 2/2] pidfd: change pidfd_send_signal() to respect PIDFD_THREAD

[Christian Brauner <brauner@kernel.org>]

[-- Attachment #1: Type: text/plain, Size: 1293 bytes --]

On Wed, Feb 14, 2024 at 01:36:56PM +0100, Oleg Nesterov wrote:
> On 02/10, Oleg Nesterov wrote:
> >
> > On 02/10, Christian Brauner wrote:
> > >
> > > +	if (type == PIDFD_SIGNAL_PROCESS_GROUP)
> > > +		ret = kill_pgrp_info(sig, &kinfo, pid);
> >
> > I guess you meant
> >
> > 	if (type == PIDTYPE_PGID)
> >
> > other than that,
> >
> > Reviewed-by: Oleg Nesterov <oleg@redhat.com>
> 
> Yes, but there is another thing I hadn't thought of...
> 
> sys_pidfd_send_signal() does
> 
> 	/* Only allow sending arbitrary signals to yourself. */
> 	ret = -EPERM;
> 	if ((task_pid(current) != pid) &&
> 	    (kinfo.si_code >= 0 || kinfo.si_code == SI_TKILL))
> 		goto err;
> 
> and I am not sure that task_pid(current) == pid should allow
> the "arbitrary signals" if PIDFD_SIGNAL_PROCESS_GROUP.
> 
> Perhaps
> 
> 	/* Only allow sending arbitrary signals to yourself. */
> 	ret = -EPERM;
> 	if ((task_pid(current) != pid || type == PIDTYPE_PGID) &&
> 	    (kinfo.si_code >= 0 || kinfo.si_code == SI_TKILL)
> 		goto err;

Honestly, we should probably just do:

if (kinfo->si_code != SI_USER)
        goto err

and be done with it. If we get regressions reports about this then it's
easy to fix that up. But I find that unlikely. So why not try to get
away with something much simpler. What do you think?

[-- Attachment #2: 0001-signal-disallow-non-SI_USER-signals-in-pidfd_send_si.patch --]
[-- Type: text/x-diff, Size: 1739 bytes --]

From 82a0d641e6f0bcf1a81731e06462df6911ecdd4e Mon Sep 17 00:00:00 2001
From: Christian Brauner <brauner@kernel.org>
Date: Fri, 16 Feb 2024 13:21:11 +0100
Subject: [PATCH] signal: disallow non-SI_USER signals in pidfd_send_signal()

Oleg pointed out that the following condition:

/* Only allow sending arbitrary signals to yourself. */
ret = -EPERM;
if ((task_pid(current) != pid) &&
    (kinfo.si_code >= 0 || kinfo.si_code == SI_TKILL))
        goto err;

doesn't account for PIDFD_SIGNAL_PROCESS_GROUP. He suggested:

/* Only allow sending arbitrary signals to yourself. */
ret = -EPERM;
if ((task_pid(current) != pid || type == PIDTYPE_PGID) &&
    (kinfo.si_code >= 0 || kinfo.si_code == SI_TKILL)
        goto err;

but I think we should just go all the way and error out if userspace
specifies anything else than SI_USER as si_code. It's probably an unused
feature right now anyway and if someone needs it than it's easy to add
back.

Reported-by: Oleg Nesterov <oleg@redhat.com>
Link: https://lore.kernel.org/r/20240214123655.GB16265@redhat.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
---
 kernel/signal.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

diff --git a/kernel/signal.c b/kernel/signal.c
index cf6539a6b1cb..92a80e8d6b22 100644
--- a/kernel/signal.c
+++ b/kernel/signal.c
@@ -3954,10 +3954,7 @@ SYSCALL_DEFINE4(pidfd_send_signal, int, pidfd, int, sig,
 		if (unlikely(sig != kinfo.si_signo))
 			goto err;
 
-		/* Only allow sending arbitrary signals to yourself. */
-		ret = -EPERM;
-		if ((task_pid(current) != pid) &&
-		    (kinfo.si_code >= 0 || kinfo.si_code == SI_TKILL))
+		if (kinfo.si_code != SI_USER)
 			goto err;
 	} else {
 		prepare_kill_siginfo(sig, &kinfo, type);
-- 
2.43.0