Received-SPF: pass (google.com: domain of linux-bluetooth+bounces-4248-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.199.223 as permitted sender) client-ip=147.75.199.223;
From: Sungwoo Kim <iam@sung-woo.kim>
To: 
Cc: daveti@purdue.edu,
	benquike@gmail.com,
	Sungwoo Kim <iam@sung-woo.kim>,
	Marcel Holtmann <marcel@holtmann.org>,
	Johan Hedberg <johan.hedberg@gmail.com>,
	Luiz Augusto von Dentz <luiz.dentz@gmail.com>,
	linux-bluetooth@vger.kernel.org,
	linux-kernel@vger.kernel.org
Subject: [PATCH] Bluetooth: HCI: Fix potential null-ptr-deref
Date: Wed,  1 May 2024 22:17:09 -0400
Message-Id: <20240502021709.141220-1-iam@sung-woo.kim>
Precedence: bulk
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit

Dear all,
hci_le_big_sync_established_evt() has a potential null-ptr-deref bug.

hci_le_big_sync_established_evt()
  bis = hci_conn_hash_lookup_handle(hdev, handle);
  if (!bis)
    bis = hci_conn_add() <- could fail
  /* ... */
  bis = hci_conn_hash_lookup_handle(hdev, handle);
  set_bit(HCI_CONN_BIG_SYNC_FAILED, &bis->flags); <- null-ptr-deref

There is a missing check after hci_conn_hash_lookup_handle(), which can
return NULL. Especially, if a prior hci_conn_add() fails than
hci_conn_hash_lookup_handle() will return NULL.

This patch fixes this by adding a check.

Thanks,
Sungwoo.

Signed-off-by: Sungwoo Kim <iam@sung-woo.kim>
---
 net/bluetooth/hci_event.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c
index 4a27e4a17..d72d238c1 100644
--- a/net/bluetooth/hci_event.c
+++ b/net/bluetooth/hci_event.c
@@ -7037,6 +7037,8 @@ static void hci_le_big_sync_established_evt(struct hci_dev *hdev, void *data,
 			u16 handle = le16_to_cpu(ev->bis[i]);
 
 			bis = hci_conn_hash_lookup_handle(hdev, handle);
+			if (!bis)
+				continue;
 
 			set_bit(HCI_CONN_BIG_SYNC_FAILED, &bis->flags);
 			hci_connect_cfm(bis, ev->status);
-- 
2.34.1