Received: by 2002:a05:6358:3188:b0:123:57c1:9b43 with SMTP id q8csp1222372rwd; Thu, 8 Jun 2023 14:20:16 -0700 (PDT) X-Google-Smtp-Source: ACHHUZ6ttQBqTOJpoE/VSI45laVGKvzQ3pVlzBXAPzRAPpNo9QKm7TkcHkiNAGMWYVBvLzGgm+5j X-Received: by 2002:a17:903:1207:b0:1b0:ec0:7d01 with SMTP id l7-20020a170903120700b001b00ec07d01mr6048219plh.35.1686259216672; Thu, 08 Jun 2023 14:20:16 -0700 (PDT) ARC-Seal: i=2; a=rsa-sha256; t=1686259216; cv=pass; d=google.com; s=arc-20160816; b=um498MCqzmjNLPL8EuT63jJTo3gDAKalq71nRZ54r+u7R1QxdNu73ttwd2PcvnTEvE IpB/Mp3dId40TjVEBhBaBg11MdtvM2F6R2rE3bOf2RZVtWg6Bm/e+1Y82h+VoGRliW5x keth/bOWegL0zF1t09+6tafA24bS83IedpZ7k8t/ZMZgPifwt1ydKP1ZpRWCwR5+U0vE V6EaSYXefPDQzzN3n4LZW0/949iSCqATFvuwS/QnfdiI26tBmUNMkpVWwg4zIjILWu4R Y2vLiAccNum8GPHzXXGCIeydZVHaXuz9QU7/ZSNi76NOACAsXxM+cL/iCJj1mp/SytwF 6lqw== ARC-Message-Signature: i=2; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=list-id:precedence:content-transfer-encoding:mime-version :references:in-reply-to:message-id:date:subject:cc:to:from :dkim-signature; bh=qrs4q5qgVBu4pnZE4HY+ySu9EXbDTYqE7CgsUDti+e4=; b=vjYr40luSYW69jJzdmRN3vyTiy05tbomkGbPalBVVfZr/HXwY1YN6KYFpZv8WE7pl7 y+E3bxkNnID9ypHjxXwV/CGsQKeViYda21J+6xgebJulvbTWY9Y0iy5uvgDVzP4OZJ+T mmFU6L0cw9GwMqocXNyaYtiLMAtu4BLsd+4jOZjMStVLrptb7A75oFJu6n+C9m4fnR9n 0fYx4hrAAgA7AETFf3uLvgax0Zr9xEcyr8uT6SQWnDCj6i7dJZo99sCsujlu2sdgzwUx nm+RRKv9ra/+TazjdDhhWu/Mi4fJm7tqqoWZVAWtlbgGpvTXC2wZs9KkfhtcR/G/eacY 3TOw== ARC-Authentication-Results: i=2; mx.google.com; dkim=pass header.i=@iki.fi header.s=meesny header.b=mRhwnqXU; arc=pass (i=1); spf=pass (google.com: domain of linux-bluetooth-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) smtp.mailfrom=linux-bluetooth-owner@vger.kernel.org Return-Path: Received: from out1.vger.email (out1.vger.email. [2620:137:e000::1:20]) by mx.google.com with ESMTP id la8-20020a170902fa0800b001aafb271d16si1510304plb.461.2023.06.08.14.20.01; Thu, 08 Jun 2023 14:20:16 -0700 (PDT) Received-SPF: pass (google.com: domain of linux-bluetooth-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20; Authentication-Results: mx.google.com; dkim=pass header.i=@iki.fi header.s=meesny header.b=mRhwnqXU; arc=pass (i=1); spf=pass (google.com: domain of linux-bluetooth-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) smtp.mailfrom=linux-bluetooth-owner@vger.kernel.org Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S236968AbjFHVGz (ORCPT + 99 others); Thu, 8 Jun 2023 17:06:55 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:57420 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S236935AbjFHVGy (ORCPT ); Thu, 8 Jun 2023 17:06:54 -0400 Received: from meesny.iki.fi (meesny.iki.fi [IPv6:2001:67c:2b0:1c1::201]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 04F7F30C1 for ; Thu, 8 Jun 2023 14:06:50 -0700 (PDT) Received: from monolith.lan (91-152-120-101.elisa-laajakaista.fi [91.152.120.101]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256) (No client certificate requested) (Authenticated sender: pav) by meesny.iki.fi (Postfix) with ESMTPSA id 4QccGJ3Mh2z10MD; Fri, 9 Jun 2023 00:06:48 +0300 (EEST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=iki.fi; s=meesny; t=1686258408; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=qrs4q5qgVBu4pnZE4HY+ySu9EXbDTYqE7CgsUDti+e4=; b=mRhwnqXUC3Tar6kpPoNOuMpinC4G+zZQy1Q2hm1E2YswBLhC7H/IWu4kJ287G2jzCWsCpf NcpUKZUjmp+Roq2Fl12MHrREyJ6AYgz4dHSuDpwRsagNrAvp6gvNkgcNb4GhBEj4gUxR9q w/mOWAk6Q2aZoZVhViKU3/cE2wQtvvc= ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=iki.fi; s=meesny; t=1686258408; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=qrs4q5qgVBu4pnZE4HY+ySu9EXbDTYqE7CgsUDti+e4=; b=kTCDo8nl4PuoBlZsErC4VikbV7GM26oPOer9vrHOtQQ67rl0/myS2AtsiAwl/MKeq4L0Uw oHSqmFjn7bSMySwBgNRP7FYeH4iwhSKxa2LjxrlcJYs770n+1qp2hK2VB0YdrQWML6GeZ0 64+DvEzidgU3uQYAKGSd26KND2SUZ88= ARC-Authentication-Results: i=1; ORIGINATING; auth=pass smtp.auth=pav smtp.mailfrom=pav@iki.fi ARC-Seal: i=1; s=meesny; d=iki.fi; t=1686258408; a=rsa-sha256; cv=none; b=MLmTqeilg+ZI/8hCpHk0ekZI1yt/sI8rlfRTJ1c1Hp1Cu0rKfPKXboXOO77htsI0Owy90P 6yaaoyi+vC9PHeiXGuLNgG6x6wPRo7W0f7CzfH/2q2fBYWkK1KNpa/tY8K5OO9VOU0spns g8pQlGN7L50jXPrvn/NLB8HdNY4I8Gk= From: Pauli Virtanen To: linux-bluetooth@vger.kernel.org Cc: Pauli Virtanen Subject: [PATCH 3/3] Bluetooth: ISO: fix iso_conn related locking and validity issues Date: Fri, 9 Jun 2023 00:06:44 +0300 Message-Id: X-Mailer: git-send-email 2.40.1 In-Reply-To: References: MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Spam-Status: No, score=-2.8 required=5.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,DKIM_VALID_EF,RCVD_IN_DNSWL_LOW,SPF_HELO_NONE, SPF_PASS,T_SCC_BODY_TEXT_LINE autolearn=ham autolearn_force=no version=3.4.6 X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on lindbergh.monkeyblade.net Precedence: bulk List-ID: X-Mailing-List: linux-bluetooth@vger.kernel.org sk->sk_state indicates whether iso_pi(sk)->conn is valid. Operations that check/update sk_state and access conn should hold lock_sock, otherwise they can race. The order of taking locks is hci_dev_lock > lock_sock > iso_conn_lock, which is how it is in connect/disconnect_cfm -> iso_conn_del -> iso_chan_del. Fix locking in iso_connect_cis/bis and sendmsg/recvmsg to take lock_sock around updating sk_state and conn. iso_conn_del must not occur during iso_connect_cis/bis, as it frees the iso_conn. Hold hdev->lock longer to prevent that. This should not reintroduce the issue fixed in commit 241f51931c35 ("Bluetooth: ISO: Avoid circular locking dependency"), since the we acquire locks in order. We retain the fix in iso_sock_connect to release lock_sock before iso_connect_* acquires hdev->lock. Similarly for commit 6a5ad251b7cd ("Bluetooth: ISO: Fix possible circular locking dependency"). We retain the fix in iso_conn_ready to not acquire iso_conn_lock before lock_sock. iso_conn_add shall return iso_conn with valid hcon. Make it so also when reusing an old CIS connection waiting for disconnect timeout (see __iso_sock_close where conn->hcon is set to NULL). Trace with iso_conn_del after iso_chan_add in iso_connect_cis: =============================================================== iso_sock_create:771: sock 00000000be9b69b7 iso_sock_init:693: sk 000000004dff667e iso_sock_bind:827: sk 000000004dff667e 70:1a:b8:98:ff:a2 type 1 iso_sock_setsockopt:1289: sk 000000004dff667e iso_sock_setsockopt:1289: sk 000000004dff667e iso_sock_setsockopt:1289: sk 000000004dff667e iso_sock_connect:875: sk 000000004dff667e iso_connect_cis:353: 70:1a:b8:98:ff:a2 -> 28:3d:c2:4a:7e:da hci_get_route:1199: 70:1a:b8:98:ff:a2 -> 28:3d:c2:4a:7e:da hci_conn_add:1005: hci0 dst 28:3d:c2:4a:7e:da iso_conn_add:140: hcon 000000007b65d182 conn 00000000daf8625e __iso_chan_add:214: conn 00000000daf8625e iso_connect_cfm:1700: hcon 000000007b65d182 bdaddr 28:3d:c2:4a:7e:da status 12 iso_conn_del:187: hcon 000000007b65d182 conn 00000000daf8625e, err 16 iso_sock_clear_timer:117: sock 000000004dff667e state 3 iso_chan_del:153: sk 000000004dff667e, conn 00000000daf8625e, err 16 hci_conn_del:1151: hci0 hcon 000000007b65d182 handle 65535 hci_conn_unlink:1102: hci0: hcon 000000007b65d182 hci_chan_list_flush:2780: hcon 000000007b65d182 iso_sock_getsockopt:1376: sk 000000004dff667e iso_sock_getname:1070: sock 00000000be9b69b7, sk 000000004dff667e iso_sock_getname:1070: sock 00000000be9b69b7, sk 000000004dff667e iso_sock_getsockopt:1376: sk 000000004dff667e iso_sock_getname:1070: sock 00000000be9b69b7, sk 000000004dff667e iso_sock_getname:1070: sock 00000000be9b69b7, sk 000000004dff667e iso_sock_shutdown:1434: sock 00000000be9b69b7, sk 000000004dff667e, how 1 __iso_sock_close:632: sk 000000004dff667e state 5 socket 00000000be9b69b7 BUG: kernel NULL pointer dereference, address: 0000000000000000 PGD 8000000006467067 P4D 8000000006467067 PUD 3f5f067 PMD 0 Oops: 0000 [#1] PREEMPT SMP PTI Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.2-1.fc38 04/01/2014 RIP: 0010:__iso_sock_close (net/bluetooth/iso.c:664) bluetooth =============================================================== Trace with iso_conn_del before iso_chan_add in iso_connect_cis: =============================================================== iso_connect_cis:356: 70:1a:b8:98:ff:a2 -> 28:3d:c2:4a:7e:da ... iso_conn_add:140: hcon 0000000093bc551f conn 00000000768ae504 hci_dev_put:1487: hci0 orig refcnt 21 hci_event_packet:7607: hci0: event 0x0e hci_cmd_complete_evt:4231: hci0: opcode 0x2062 hci_cc_le_set_cig_params:3846: hci0: status 0x07 hci_sent_cmd_data:3107: hci0 opcode 0x2062 iso_connect_cfm:1703: hcon 0000000093bc551f bdaddr 28:3d:c2:4a:7e:da status 7 iso_conn_del:187: hcon 0000000093bc551f conn 00000000768ae504, err 12 hci_conn_del:1151: hci0 hcon 0000000093bc551f handle 65535 hci_conn_unlink:1102: hci0: hcon 0000000093bc551f hci_chan_list_flush:2780: hcon 0000000093bc551f __iso_chan_add:214: conn 00000000768ae504 iso_sock_clear_timer:117: sock 0000000098323f95 state 3 general protection fault, probably for non-canonical address 0x30b29c630930aec8: 0000 [#1] PREEMPT SMP PTI CPU: 1 PID: 1920 Comm: bluetoothd Tainted: G E 6.3.0-rc7+ #4 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.2-1.fc38 04/01/2014 RIP: 0010:detach_if_pending+0x28/0xd0 Code: 90 90 0f 1f 44 00 00 48 8b 47 08 48 85 c0 0f 84 ad 00 00 00 55 89 d5 53 48 83 3f 00 48 89 fb 74 7d 66 90 48 8b 03 48 8b 53 08 <> RSP: 0018:ffffb90841a67d08 EFLAGS: 00010007 RAX: 0000000000000000 RBX: ffff9141bd5061b8 RCX: 0000000000000000 RDX: 30b29c630930aec8 RSI: ffff9141fdd21e80 RDI: ffff9141bd5061b8 RBP: 0000000000000001 R08: 0000000000000000 R09: ffffb90841a67b88 R10: 0000000000000003 R11: ffffffff8613f558 R12: ffff9141fdd21e80 R13: 0000000000000000 R14: ffff9141b5976010 R15: ffff914185755338 FS: 00007f45768bd840(0000) GS:ffff9141fdd00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000619000424074 CR3: 0000000009f5e005 CR4: 0000000000170ee0 Call Trace: timer_delete+0x48/0x80 try_to_grab_pending+0xdf/0x170 __cancel_work+0x37/0xb0 iso_connect_cis+0x141/0x400 [bluetooth] =============================================================== Trace with NULL conn->hcon in state BT_CONNECT: =============================================================== __iso_sock_close:619: sk 00000000f7c71fc5 state 1 socket 00000000d90c5fe5 ... __iso_sock_close:619: sk 00000000f7c71fc5 state 8 socket 00000000d90c5fe5 iso_chan_del:153: sk 00000000f7c71fc5, conn 0000000022c03a7e, err 104 ... iso_sock_connect:862: sk 00000000129b56c3 iso_connect_cis:348: 70:1a:b8:98:ff:a2 -> 28:3d:c2:4a:7d:2a hci_get_route:1199: 70:1a:b8:98:ff:a2 -> 28:3d:c2:4a:7d:2a hci_dev_hold:1495: hci0 orig refcnt 19 __iso_chan_add:214: conn 0000000022c03a7e iso_sock_clear_timer:117: sock 00000000129b56c3 state 3 ... iso_sock_ready:1485: sk 00000000129b56c3 ... iso_sock_sendmsg:1077: sock 00000000e5013966, sk 00000000129b56c3 BUG: kernel NULL pointer dereference, address: 00000000000006a8 PGD 0 P4D 0 Oops: 0000 [#1] PREEMPT SMP PTI CPU: 1 PID: 1403 Comm: wireplumber Tainted: G E 6.3.0-rc7+ #4 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.2-1.fc38 04/01/2014 RIP: 0010:iso_sock_sendmsg+0x63/0x2a0 [bluetooth] =============================================================== Fixes: 241f51931c35 ("Bluetooth: ISO: Avoid circular locking dependency") Fixes: 6a5ad251b7cd ("Bluetooth: ISO: Fix possible circular locking dependency") Signed-off-by: Pauli Virtanen --- Notes: Don't have a clean reproducer, but these occur sometimes on bluetooth-next/master when sending data over ISO and disconnecting + reconnecting immediately, although the timing window required for these to occur is not so wide. These appear to be somewhat more easily reproduced if hci_le_set_cig_params is changed to return errors if CIG is busy. Enabling printks also seems to help. net/bluetooth/iso.c | 53 ++++++++++++++++++++++++++------------------- 1 file changed, 31 insertions(+), 22 deletions(-) diff --git a/net/bluetooth/iso.c b/net/bluetooth/iso.c index b9a008fd10b1..8f570ab66ae7 100644 --- a/net/bluetooth/iso.c +++ b/net/bluetooth/iso.c @@ -123,8 +123,11 @@ static struct iso_conn *iso_conn_add(struct hci_conn *hcon) { struct iso_conn *conn = hcon->iso_data; - if (conn) + if (conn) { + if (!conn->hcon) + conn->hcon = hcon; return conn; + } conn = kzalloc(sizeof(*conn), GFP_KERNEL); if (!conn) @@ -311,14 +314,13 @@ static int iso_connect_bis(struct sock *sk) goto unlock; } - hci_dev_unlock(hdev); - hci_dev_put(hdev); + lock_sock(sk); err = iso_chan_add(conn, sk, NULL); - if (err) - return err; - - lock_sock(sk); + if (err) { + release_sock(sk); + goto unlock; + } /* Update source addr of the socket */ bacpy(&iso_pi(sk)->src, &hcon->src); @@ -335,7 +337,6 @@ static int iso_connect_bis(struct sock *sk) } release_sock(sk); - return err; unlock: hci_dev_unlock(hdev); @@ -403,14 +404,13 @@ static int iso_connect_cis(struct sock *sk) goto unlock; } - hci_dev_unlock(hdev); - hci_dev_put(hdev); + lock_sock(sk); err = iso_chan_add(conn, sk, NULL); - if (err) - return err; - - lock_sock(sk); + if (err) { + release_sock(sk); + goto unlock; + } /* Update source addr of the socket */ bacpy(&iso_pi(sk)->src, &hcon->src); @@ -427,7 +427,6 @@ static int iso_connect_cis(struct sock *sk) } release_sock(sk); - return err; unlock: hci_dev_unlock(hdev); @@ -1078,8 +1077,8 @@ static int iso_sock_sendmsg(struct socket *sock, struct msghdr *msg, size_t len) { struct sock *sk = sock->sk; - struct iso_conn *conn = iso_pi(sk)->conn; struct sk_buff *skb, **frag; + size_t mtu; int err; BT_DBG("sock %p, sk %p", sock, sk); @@ -1091,11 +1090,18 @@ static int iso_sock_sendmsg(struct socket *sock, struct msghdr *msg, if (msg->msg_flags & MSG_OOB) return -EOPNOTSUPP; - if (sk->sk_state != BT_CONNECTED) + lock_sock(sk); + + if (sk->sk_state != BT_CONNECTED) { + release_sock(sk); return -ENOTCONN; + } + + mtu = iso_pi(sk)->conn->hcon->hdev->iso_mtu; + + release_sock(sk); - skb = bt_skb_sendmsg(sk, msg, len, conn->hcon->hdev->iso_mtu, - HCI_ISO_DATA_HDR_SIZE, 0); + skb = bt_skb_sendmsg(sk, msg, len, mtu, HCI_ISO_DATA_HDR_SIZE, 0); if (IS_ERR(skb)) return PTR_ERR(skb); @@ -1108,8 +1114,7 @@ static int iso_sock_sendmsg(struct socket *sock, struct msghdr *msg, while (len) { struct sk_buff *tmp; - tmp = bt_skb_sendmsg(sk, msg, len, conn->hcon->hdev->iso_mtu, - 0, 0); + tmp = bt_skb_sendmsg(sk, msg, len, mtu, 0, 0); if (IS_ERR(tmp)) { kfree_skb(skb); return PTR_ERR(tmp); @@ -1164,15 +1169,19 @@ static int iso_sock_recvmsg(struct socket *sock, struct msghdr *msg, BT_DBG("sk %p", sk); if (test_and_clear_bit(BT_SK_DEFER_SETUP, &bt_sk(sk)->flags)) { + lock_sock(sk); switch (sk->sk_state) { case BT_CONNECT2: - lock_sock(sk); iso_conn_defer_accept(pi->conn->hcon); sk->sk_state = BT_CONFIG; release_sock(sk); return 0; case BT_CONNECT: + release_sock(sk); return iso_connect_cis(sk); + default: + release_sock(sk); + break; } } -- 2.40.1