[PATCH 6.1.y 0/2] Bluetooth: backport CVE-2023-53762 to 6.1.y

[PATCH 6.1.y 0/2] Bluetooth: backport CVE-2023-53762 to 6.1.y

[Artem Dinaburg]

Hi Greg, Sasha, and Bluetooth maintainers,

I am backporting CVEss still missing from 6.1.y.
This pair handles CVE-2023-53762 in disconnect-all cleanup. The
second patch is the immediate upstream liveness follow-up to the UAF fix,
so I kept the two together.

Both changes are already inherited by v6.6 and every later mainline
release.

Could you please queue the pair for 6.1.y?

Thanks,
Artem Dinaburg

AI assistance: An LLM helped find the missing CVE fix, adapt
it, and prepare the draft; I reviewed the code and build output.

Luiz Augusto von Dentz (1):
  Bluetooth: hci_sync: Fix UAF in hci_disconnect_all_sync

Pauli Virtanen (1):
  Bluetooth: hci_sync: always check if connection is alive before
    deleting

 net/bluetooth/hci_sync.c | 58 +++++++++++++++++++++++++---------------
 1 file changed, 37 insertions(+), 21 deletions(-)

-- 
2.39.5

[PATCH 6.1.y 1/2] Bluetooth: hci_sync: Fix UAF in hci_disconnect_all_sync

[Artem Dinaburg]

From: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>

[ Upstream commit 94d9ba9f9888b748d4abd2aa1547af56ae85f772 ]

Use-after-free can occur in hci_disconnect_all_sync if a connection is
deleted by concurrent processing of a controller event.

To prevent this the code now tries to iterate over the list backwards
to ensure the links are cleanup before its parents, also it no longer
relies on a cursor, instead it always uses the last element since
hci_abort_conn_sync is guaranteed to call hci_conn_del.

UAF crash log:
==================================================================
BUG: KASAN: slab-use-after-free in hci_set_powered_sync
(net/bluetooth/hci_sync.c:5424) [bluetooth]
Read of size 8 at addr ffff888009d9c000 by task kworker/u9:0/124

CPU: 0 PID: 124 Comm: kworker/u9:0 Tainted: G        W
6.5.0-rc1+ #10
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS
1.16.2-1.fc38 04/01/2014
Workqueue: hci0 hci_cmd_sync_work [bluetooth]
Call Trace:
 <TASK>
 dump_stack_lvl+0x5b/0x90
 print_report+0xcf/0x670
 ? __virt_addr_valid+0xdd/0x160
 ? hci_set_powered_sync+0x2c9/0x4a0 [bluetooth]
 kasan_report+0xa6/0xe0
 ? hci_set_powered_sync+0x2c9/0x4a0 [bluetooth]
 ? __pfx_set_powered_sync+0x10/0x10 [bluetooth]
 hci_set_powered_sync+0x2c9/0x4a0 [bluetooth]
 ? __pfx_hci_set_powered_sync+0x10/0x10 [bluetooth]
 ? __pfx_lock_release+0x10/0x10
 ? __pfx_set_powered_sync+0x10/0x10 [bluetooth]
 hci_cmd_sync_work+0x137/0x220 [bluetooth]
 process_one_work+0x526/0x9d0
 ? __pfx_process_one_work+0x10/0x10
 ? __pfx_do_raw_spin_lock+0x10/0x10
 ? mark_held_locks+0x1a/0x90
 worker_thread+0x92/0x630
 ? __pfx_worker_thread+0x10/0x10
 kthread+0x196/0x1e0
 ? __pfx_kthread+0x10/0x10
 ret_from_fork+0x2c/0x50
 </TASK>

Allocated by task 1782:
 kasan_save_stack+0x33/0x60
 kasan_set_track+0x25/0x30
 __kasan_kmalloc+0x8f/0xa0
 hci_conn_add+0xa5/0xa80 [bluetooth]
 hci_bind_cis+0x881/0x9b0 [bluetooth]
 iso_connect_cis+0x121/0x520 [bluetooth]
 iso_sock_connect+0x3f6/0x790 [bluetooth]
 __sys_connect+0x109/0x130
 __x64_sys_connect+0x40/0x50
 do_syscall_64+0x60/0x90
 entry_SYSCALL_64_after_hwframe+0x6e/0xd8

Freed by task 695:
 kasan_save_stack+0x33/0x60
 kasan_set_track+0x25/0x30
 kasan_save_free_info+0x2b/0x50
 __kasan_slab_free+0x10a/0x180
 __kmem_cache_free+0x14d/0x2e0
 device_release+0x5d/0xf0
 kobject_put+0xdf/0x270
 hci_disconn_complete_evt+0x274/0x3a0 [bluetooth]
 hci_event_packet+0x579/0x7e0 [bluetooth]
 hci_rx_work+0x287/0xaa0 [bluetooth]
 process_one_work+0x526/0x9d0
 worker_thread+0x92/0x630
 kthread+0x196/0x1e0
 ret_from_fork+0x2c/0x50
==================================================================

Fixes: 182ee45da083 ("Bluetooth: hci_sync: Rework hci_suspend_notifier")
Signed-off-by: Pauli Virtanen <pav@iki.fi>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>

[ Backport to 6.1.y: the source change is unchanged; only hunk locations
  in hci_sync.c differ. ]
Assisted-by: LLM
Signed-off-by: Artem Dinaburg <artem@trailofbits.com>
---
Stable submission note: patch 1/2 in the CVE-2023-53762 series for Linux
6.1.y.

CVE: CVE-2023-53762

Build: This patch was included in an x86_64 allmodconfig and
CONFIG_WERROR=y build.
It produced vmlinux and modules with no new warnings or errors.


 net/bluetooth/hci_sync.c | 55 +++++++++++++++++++++++++---------------
 1 file changed, 35 insertions(+), 20 deletions(-)

diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c
index d5856448d46782..da9e6e55bc1d00 100644
--- a/net/bluetooth/hci_sync.c
+++ b/net/bluetooth/hci_sync.c
@@ -5587,6 +5587,7 @@ int hci_abort_conn_sync(struct hci_dev *hdev, struct hci_conn *conn, u8 reason)
 {
 	int err = 0;
 	u16 handle = conn->handle;
+	struct hci_conn *c;
 
 	switch (conn->state) {
 	case BT_CONNECTED:
@@ -5600,43 +5601,57 @@ int hci_abort_conn_sync(struct hci_dev *hdev, struct hci_conn *conn, u8 reason)
 		err = hci_reject_conn_sync(hdev, conn, reason);
 		break;
 	default:
+		hci_dev_lock(hdev);
 		conn->state = BT_CLOSED;
+		hci_disconn_cfm(conn, reason);
+		hci_conn_del(conn);
+		hci_dev_unlock(hdev);
 		return 0;
 	}
 
+	hci_dev_lock(hdev);
+
+	/* Check if the connection hasn't been cleanup while waiting
+	 * commands to complete.
+	 */
+	c = hci_conn_hash_lookup_handle(hdev, handle);
+	if (!c || c != conn) {
+		err = 0;
+		goto unlock;
+	}
+
 	/* Cleanup hci_conn object if it cannot be cancelled as it
 	 * likelly means the controller and host stack are out of sync
 	 * or in case of LE it was still scanning so it can be cleanup
 	 * safely.
 	 */
-	if (err) {
-		struct hci_conn *c;
-
-		/* Check if the connection hasn't been cleanup while waiting
-		 * commands to complete.
-		 */
-		c = hci_conn_hash_lookup_handle(hdev, handle);
-		if (!c || c != conn)
-			return 0;
-
-		hci_dev_lock(hdev);
-		hci_conn_failed(conn, err);
-		hci_dev_unlock(hdev);
-	}
+	hci_conn_failed(conn, reason);
 
+unlock:
+	hci_dev_unlock(hdev);
 	return err;
 }
 
 static int hci_disconnect_all_sync(struct hci_dev *hdev, u8 reason)
 {
-	struct hci_conn *conn, *tmp;
-	int err;
+	struct list_head *head = &hdev->conn_hash.list;
+	struct hci_conn *conn;
 
-	list_for_each_entry_safe(conn, tmp, &hdev->conn_hash.list, list) {
-		err = hci_abort_conn_sync(hdev, conn, reason);
-		if (err)
-			return err;
+	rcu_read_lock();
+	while ((conn = list_first_or_null_rcu(head, struct hci_conn, list))) {
+		/* Make sure the connection is not freed while unlocking */
+		conn = hci_conn_get(conn);
+		rcu_read_unlock();
+		/* Disregard possible errors since hci_conn_del shall have been
+		 * called even in case of errors had occurred since it would
+		 * then cause hci_conn_failed to be called which calls
+		 * hci_conn_del internally.
+		 */
+		hci_abort_conn_sync(hdev, conn, reason);
+		hci_conn_put(conn);
+		rcu_read_lock();
 	}
+	rcu_read_unlock();
 
 	return 0;
 }
-- 
2.39.5

[PATCH 6.1.y 2/2] Bluetooth: hci_sync: always check if connection is alive before deleting

[Artem Dinaburg]

From: Pauli Virtanen <pav@iki.fi>

[ Upstream commit a239110ee8e0b0aafa265f0d54f7a16744855e70 ]

In hci_abort_conn_sync it is possible that conn is deleted concurrently
by something else, also e.g. when waiting for hdev->lock.  This causes
double deletion of the conn, so UAF or conn_hash.list corruption.

Fix by having all code paths check that the connection is still in
conn_hash before deleting it, while holding hdev->lock which prevents
any races.

Log (when powering off while BAP streaming, occurs rarely):
=======================================================================
kernel BUG at lib/list_debug.c:56!
...
 ? __list_del_entry_valid (lib/list_debug.c:56)
 hci_conn_del (net/bluetooth/hci_conn.c:154) bluetooth
 hci_abort_conn_sync (net/bluetooth/hci_sync.c:5415) bluetooth
 ? __pfx_hci_abort_conn_sync+0x10/0x10 [bluetooth]
 ? lock_release+0x1d5/0x3c0
 ? hci_disconnect_all_sync.constprop.0+0xb2/0x230 [bluetooth]
 ? __pfx_lock_release+0x10/0x10
 ? __kmem_cache_free+0x14d/0x2e0
 hci_disconnect_all_sync.constprop.0+0xda/0x230 [bluetooth]
 ? __pfx_hci_disconnect_all_sync.constprop.0+0x10/0x10 [bluetooth]
 ? hci_clear_adv_sync+0x14f/0x170 [bluetooth]
 ? __pfx_set_powered_sync+0x10/0x10 [bluetooth]
 hci_set_powered_sync+0x293/0x450 [bluetooth]
=======================================================================

Fixes: 94d9ba9f9888 ("Bluetooth: hci_sync: Fix UAF in hci_disconnect_all_sync")
Signed-off-by: Pauli Virtanen <pav@iki.fi>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>

[ Backport to 6.1.y follow-up to 94d9ba9f9888: this tree lacks the later
  BT_OPEN and BT_BOUND special cases, so retain its existing state
  switch and add the common liveness check. ]
Assisted-by: LLM
Signed-off-by: Artem Dinaburg <artem@trailofbits.com>
---
Stable submission note: patch 2/2 in the CVE-2023-53762 series for Linux
6.1.y.

CVE: CVE-2023-53762

Build: This patch was included in an x86_64 allmodconfig and
CONFIG_WERROR=y build.
It produced vmlinux and modules with no new warnings or errors.

 net/bluetooth/hci_sync.c | 21 +++++++++++----------
 1 file changed, 11 insertions(+), 10 deletions(-)

diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c
index da9e6e55bc1d00..08e90dab73459b 100644
--- a/net/bluetooth/hci_sync.c
+++ b/net/bluetooth/hci_sync.c
@@ -5587,6 +5587,7 @@ int hci_abort_conn_sync(struct hci_dev *hdev, struct hci_conn *conn, u8 reason)
 {
 	int err = 0;
 	u16 handle = conn->handle;
+	bool disconnect = false;
 	struct hci_conn *c;
 
 	switch (conn->state) {
@@ -5601,19 +5602,13 @@ int hci_abort_conn_sync(struct hci_dev *hdev, struct hci_conn *conn, u8 reason)
 		err = hci_reject_conn_sync(hdev, conn, reason);
 		break;
 	default:
-		hci_dev_lock(hdev);
-		conn->state = BT_CLOSED;
-		hci_disconn_cfm(conn, reason);
-		hci_conn_del(conn);
-		hci_dev_unlock(hdev);
-		return 0;
+		disconnect = true;
+		break;
 	}
 
 	hci_dev_lock(hdev);
 
-	/* Check if the connection hasn't been cleanup while waiting
-	 * commands to complete.
-	 */
+	/* Check if the connection has been cleaned up concurrently */
 	c = hci_conn_hash_lookup_handle(hdev, handle);
 	if (!c || c != conn) {
 		err = 0;
@@ -5625,7 +5620,13 @@ int hci_abort_conn_sync(struct hci_dev *hdev, struct hci_conn *conn, u8 reason)
 	 * or in case of LE it was still scanning so it can be cleanup
 	 * safely.
 	 */
-	hci_conn_failed(conn, reason);
+	if (disconnect) {
+		conn->state = BT_CLOSED;
+		hci_disconn_cfm(conn, reason);
+		hci_conn_del(conn);
+	} else {
+		hci_conn_failed(conn, reason);
+	}
 
 unlock:
 	hci_dev_unlock(hdev);
-- 
2.39.5