Received-SPF: softfail (google.com: domain of transitioning linux-kernel-owner@vger.kernel.org does not designate 23.128.96.19 as permitted sender) client-ip=23.128.96.19;
From:   Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To:     linux-kernel@vger.kernel.org
Cc:     Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        stable@vger.kernel.org,
        "Eric W. Biederman" <ebiederm@xmission.com>,
        Linus Torvalds <torvalds@linuxfoundation.org>,
        =?UTF-8?q?Michal=20Koutn=C3=BD?= <mkoutny@suse.com>,
        Tejun Heo <tj@kernel.org>,
        Ovidiu Panait <ovidiu.panait@windriver.com>
Subject: [PATCH 5.10 162/171] cgroup: Use open-time credentials for process migraton perm checks
Date:   Tue, 12 Apr 2022 08:30:53 +0200
Message-Id: <20220412062932.586417436@linuxfoundation.org>
In-Reply-To: <20220412062927.870347203@linuxfoundation.org>
References: <20220412062927.870347203@linuxfoundation.org>
User-Agent: quilt/0.66
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Precedence: bulk

From: Tejun Heo <tj@kernel.org>

commit 1756d7994ad85c2479af6ae5a9750b92324685af upstream.

cgroup process migration permission checks are performed at write time as
whether a given operation is allowed or not is dependent on the content of
the write - the PID. This currently uses current's credentials which is a
potential security weakness as it may allow scenarios where a less
privileged process tricks a more privileged one into writing into a fd that
it created.

This patch makes both cgroup2 and cgroup1 process migration interfaces to
use the credentials saved at the time of open (file->f_cred) instead of
current's.

Reported-by: "Eric W. Biederman" <ebiederm@xmission.com>
Suggested-by: Linus Torvalds <torvalds@linuxfoundation.org>
Fixes: 187fe84067bd ("cgroup: require write perm on common ancestor when moving processes on the default hierarchy")
Reviewed-by: Michal Koutný <mkoutny@suse.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
[OP: apply original __cgroup_procs_write() changes to cgroup_threads_write()
and cgroup_procs_write(), as the refactoring commit da70862efe006 ("cgroup:
cgroup.{procs,threads} factor out common parts") is not present in 5.10-stable]
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 kernel/cgroup/cgroup-v1.c |    7 ++++---
 kernel/cgroup/cgroup.c    |   17 ++++++++++++++++-
 2 files changed, 20 insertions(+), 4 deletions(-)

--- a/kernel/cgroup/cgroup-v1.c
+++ b/kernel/cgroup/cgroup-v1.c
@@ -505,10 +505,11 @@ static ssize_t __cgroup1_procs_write(str
 		goto out_unlock;
 
 	/*
-	 * Even if we're attaching all tasks in the thread group, we only
-	 * need to check permissions on one of them.
+	 * Even if we're attaching all tasks in the thread group, we only need
+	 * to check permissions on one of them. Check permissions using the
+	 * credentials from file open to protect against inherited fd attacks.
 	 */
-	cred = current_cred();
+	cred = of->file->f_cred;
 	tcred = get_task_cred(task);
 	if (!uid_eq(cred->euid, GLOBAL_ROOT_UID) &&
 	    !uid_eq(cred->euid, tcred->uid) &&
--- a/kernel/cgroup/cgroup.c
+++ b/kernel/cgroup/cgroup.c
@@ -4788,6 +4788,7 @@ static ssize_t cgroup_procs_write(struct
 	struct cgroup_file_ctx *ctx = of->priv;
 	struct cgroup *src_cgrp, *dst_cgrp;
 	struct task_struct *task;
+	const struct cred *saved_cred;
 	ssize_t ret;
 	bool locked;
 
@@ -4805,9 +4806,16 @@ static ssize_t cgroup_procs_write(struct
 	src_cgrp = task_cgroup_from_root(task, &cgrp_dfl_root);
 	spin_unlock_irq(&css_set_lock);
 
+	/*
+	 * Process and thread migrations follow same delegation rule. Check
+	 * permissions using the credentials from file open to protect against
+	 * inherited fd attacks.
+	 */
+	saved_cred = override_creds(of->file->f_cred);
 	ret = cgroup_attach_permissions(src_cgrp, dst_cgrp,
 					of->file->f_path.dentry->d_sb, true,
 					ctx->ns);
+	revert_creds(saved_cred);
 	if (ret)
 		goto out_finish;
 
@@ -4832,6 +4840,7 @@ static ssize_t cgroup_threads_write(stru
 	struct cgroup_file_ctx *ctx = of->priv;
 	struct cgroup *src_cgrp, *dst_cgrp;
 	struct task_struct *task;
+	const struct cred *saved_cred;
 	ssize_t ret;
 	bool locked;
 
@@ -4851,10 +4860,16 @@ static ssize_t cgroup_threads_write(stru
 	src_cgrp = task_cgroup_from_root(task, &cgrp_dfl_root);
 	spin_unlock_irq(&css_set_lock);
 
-	/* thread migrations follow the cgroup.procs delegation rule */
+	/*
+	 * Process and thread migrations follow same delegation rule. Check
+	 * permissions using the credentials from file open to protect against
+	 * inherited fd attacks.
+	 */
+	saved_cred = override_creds(of->file->f_cred);
 	ret = cgroup_attach_permissions(src_cgrp, dst_cgrp,
 					of->file->f_path.dentry->d_sb, false,
 					ctx->ns);
+	revert_creds(saved_cred);
 	if (ret)
 		goto out_finish;