kernel/linux.git/net, branch v4.3.1

Bluetooth: Fix missing hdev locking for LE scan cleanup

2015-12-09T19:34:28+00:00

commit 8ce783dc5ea3af3a213ac9b4d9d2ccfeeb9c9058 upstream. The hci_conn objects don't have a dedicated lock themselves but rely on the caller to hold the hci_dev lock for most types of access. The hci_conn_timeout() function has so far sent certain HCI commands based on the hci_conn state which has been possible without holding the hci_dev lock. The recent changes to do LE scanning before connect attempts added even more operations to hci_conn and hci_dev from hci_conn_timeout, thereby exposing potential race conditions with the hci_dev and hci_conn states. As an example of such a race, here there's a timeout but an l2cap_sock_connect() call manages to race with the cleanup routine: [Oct21 08:14] l2cap_chan_timeout: chan ee4b12c0 state BT_CONNECT [ +0.000004] l2cap_chan_close: chan ee4b12c0 state BT_CONNECT [ +0.000002] l2cap_chan_del: chan ee4b12c0, conn f3141580, err 111, state BT_CONNECT [ +0.000002] l2cap_sock_teardown_cb: chan ee4b12c0 state BT_CONNECT [ +0.000005] l2cap_chan_put: chan ee4b12c0 orig refcnt 4 [ +0.000010] hci_conn_drop: hcon f53d56e0 orig refcnt 1 [ +0.000013] l2cap_chan_put: chan ee4b12c0 orig refcnt 3 [ +0.000063] hci_conn_timeout: hcon f53d56e0 state BT_CONNECT [ +0.000049] hci_conn_params_del: addr ee:0d:30:09:53:1f (type 1) [ +0.000002] hci_chan_list_flush: hcon f53d56e0 [ +0.000001] hci_chan_del: hci0 hcon f53d56e0 chan f4e7ccc0 [ +0.004528] l2cap_sock_create: sock e708fc00 [ +0.000023] l2cap_chan_create: chan ee4b1770 [ +0.000001] l2cap_chan_hold: chan ee4b1770 orig refcnt 1 [ +0.000002] l2cap_sock_init: sk ee4b3390 [ +0.000029] l2cap_sock_bind: sk ee4b3390 [ +0.000010] l2cap_sock_setsockopt: sk ee4b3390 [ +0.000037] l2cap_sock_connect: sk ee4b3390 [ +0.000002] l2cap_chan_connect: 00:02:72:d9:e5:8b -> ee:0d:30:09:53:1f (type 2) psm 0x00 [ +0.000002] hci_get_route: 00:02:72:d9:e5:8b -> ee:0d:30:09:53:1f [ +0.000001] hci_dev_hold: hci0 orig refcnt 8 [ +0.000003] hci_conn_hold: hcon f53d56e0 orig refcnt 0 Above the l2cap_chan_connect() shouldn't have been able to reach the hci_conn f53d56e0 anymore but since hci_conn_timeout didn't do proper locking that's not the case. The end result is a reference to hci_conn that's not in the conn_hash list, resulting in list corruption when trying to remove it later: [Oct21 08:15] l2cap_chan_timeout: chan ee4b1770 state BT_CONNECT [ +0.000004] l2cap_chan_close: chan ee4b1770 state BT_CONNECT [ +0.000003] l2cap_chan_del: chan ee4b1770, conn f3141580, err 111, state BT_CONNECT [ +0.000001] l2cap_sock_teardown_cb: chan ee4b1770 state BT_CONNECT [ +0.000005] l2cap_chan_put: chan ee4b1770 orig refcnt 4 [ +0.000002] hci_conn_drop: hcon f53d56e0 orig refcnt 1 [ +0.000015] l2cap_chan_put: chan ee4b1770 orig refcnt 3 [ +0.000038] hci_conn_timeout: hcon f53d56e0 state BT_CONNECT [ +0.000003] hci_chan_list_flush: hcon f53d56e0 [ +0.000002] hci_conn_hash_del: hci0 hcon f53d56e0 [ +0.000001] ------------[ cut here ]------------ [ +0.000461] WARNING: CPU: 0 PID: 1782 at lib/list_debug.c:56 __list_del_entry+0x3f/0x71() [ +0.000839] list_del corruption, f53d56e0->prev is LIST_POISON2 (00000200) The necessary fix is unfortunately more complicated than just adding hci_dev_lock/unlock calls to the hci_conn_timeout() call path. Particularly, the hci_conn_del() API, which expects the hci_dev lock to be held, performs a cancel_delayed_work_sync(&hcon->disc_work) which would lead to a deadlock if the hci_conn_timeout() call path tries to acquire the same lock. This patch solves the problem by deferring the cleanup work to a separate work callback. To protect against the hci_dev or hci_conn going away meanwhile temporary references are taken with the help of hci_dev_hold() and hci_conn_get(). Signed-off-by: Johan Hedberg Signed-off-by: Marcel Holtmann Signed-off-by: Greg Kroah-Hartman

Bluetooth: Fix removing connection parameters when unpairing

2015-12-09T19:34:28+00:00

commit a6ad2a6b9cc1d9d791aee5462cfb8528f366f1d4 upstream. The commit 89cbb0638e9b7 introduced support for deferred connection parameter removal when unpairing by removing them only once an existing connection gets disconnected. However, it failed to address the scenario when we're *not* connected and do an unpair operation. What makes things worse is that most user space BlueZ versions will first issue a disconnect request and only then unpair, meaning the buggy code will be triggered every time. This effectively causes the kernel to resume scanning and reconnect to a device for which we've removed all keys and GATT database information. This patch fixes the issue by adding the missing call to the hci_conn_params_del() function to a branch which handles the case of no existing connection. Signed-off-by: Johan Hedberg Signed-off-by: Marcel Holtmann Signed-off-by: Greg Kroah-Hartman

Bluetooth: hidp: fix device disconnect on idle timeout

2015-12-09T19:34:26+00:00

commit 660f0fc07d21114549c1862e67e78b1cf0c90c29 upstream. The HIDP specs define an idle-timeout which automatically disconnects a device. This has always been implemented in the HIDP layer and forced a synchronous shutdown of the hidp-scheduler. This works just fine, but lacks a forced disconnect on the underlying l2cap channels. This has been broken since: commit 5205185d461d5902325e457ca80bd421127b7308 Author: David Herrmann Date: Sat Apr 6 20:28:47 2013 +0200 Bluetooth: hidp: remove old session-management The old session-management always forced an l2cap error on the ctrl/intr channels when shutting down. The new session-management skips this, as we don't want to enforce channel policy on the caller. In other words, if user-space removes an HIDP device, the underlying channels (which are *owned* and *referenced* by user-space) are still left active. User-space needs to call shutdown(2) or close(2) to release them. Unfortunately, this does not work with idle-timeouts. There is no way to signal user-space that the HIDP layer has been stopped. The API simply does not support any event-passing except for poll(2). Hence, we restore old behavior and force EUNATCH on the sockets if the HIDP layer is disconnected due to idle-timeouts (behavior of explicit disconnects remains unmodified). User-space can still call getsockopt(..., SO_ERROR, ...) ..to retrieve the EUNATCH error and clear sk_err. Hence, the channels can still be re-used (which nobody does so far, though). Therefore, the API still supports the new behavior, but with this patch it's also compatible to the old implicit channel shutdown. Reported-by: Mark Haun Reported-by: Luiz Augusto von Dentz Signed-off-by: David Herrmann Signed-off-by: Marcel Holtmann Signed-off-by: Greg Kroah-Hartman

NFC: nci: extract pipe value using NCI_HCP_MSG_GET_PIPE

2015-12-09T19:34:23+00:00

commit e65917b6d54f8b47d8293ea96adfa604fd46cf0d upstream. When receiving data in nci_hci_msg_rx_work, extract pipe value using NCI_HCP_MSG_GET_PIPE macro. Signed-off-by: Christophe Ricard Signed-off-by: Samuel Ortiz Signed-off-by: Greg Kroah-Hartman

NFC: nci: Fix improper management of HCI return code

2015-12-09T19:34:23+00:00

commit d8cd37ed2fc871c66b4c79c59f651dc2cdf7091c upstream. When sending HCI data over NCI, HCI return code is part of the NCI data. In order to get correctly the HCI return code, we assume the NCI communication is successful and extract the return code for the nci_hci functions return code. This is done because nci_to_errno does not match hci return code value. Signed-off-by: Christophe Ricard Signed-off-by: Samuel Ortiz Signed-off-by: Greg Kroah-Hartman

NFC: nci: Fix incorrect data chaining when sending data

2015-12-09T19:34:23+00:00

commit 500c4ef02277eaadbfe20537f963b6221f6ac007 upstream. When sending HCI data over NCI, cmd information should be present only on the first packet. Each packet shall be specifically allocated and sent to the NCI layer. Signed-off-by: Christophe Ricard Signed-off-by: Samuel Ortiz Signed-off-by: Greg Kroah-Hartman

nl80211: Fix potential memory leak from parse_acl_data

2015-12-09T19:34:22+00:00

commit 4baf6bea37247e59f1971e8009d13aeda95edba2 upstream. If parse_acl_data succeeds but the subsequent parsing of smps attributes fails, there will be a memory leak due to early returns. Fix that by moving the ACL parsing later. Fixes: 18998c381b19b ("cfg80211: allow requesting SMPS mode on ap start") Signed-off-by: Ola Olsson Signed-off-by: Johannes Berg Signed-off-by: Greg Kroah-Hartman

mac80211: fix divide by zero when NOA update

2015-12-09T19:34:21+00:00

commit 519ee6918b91abdc4bc9720deae17599a109eb40 upstream. In case of one shot NOA the interval can be 0, catch that instead of potentially (depending on the driver) crashing like this: divide error: 0000 [#1] SMP [...] Call Trace: [] ieee80211_extend_absent_time+0x6c/0xb0 [mac80211] [] ieee80211_update_p2p_noa+0xb7/0xe0 [mac80211] [] ath9k_p2p_ps_timer+0x170/0x190 [ath9k] [] ath_gen_timer_isr+0xc8/0xf0 [ath9k_hw] [] ath9k_tasklet+0x296/0x2f0 [ath9k] [] tasklet_action+0xe5/0xf0 [...] Signed-off-by: Janusz Dziedzic Signed-off-by: Johannes Berg Signed-off-by: Greg Kroah-Hartman

mac80211: allow null chandef in tracing

2015-12-09T19:34:21+00:00

commit 254d3dfe445f94a764e399ca12e04365ac9413ed upstream. In TDLS channel-switch operations the chandef can sometimes be NULL. Avoid an oops in the trace code for these cases and just print a chandef full of zeros. Fixes: a7a6bdd0670fe ("mac80211: introduce TDLS channel switch ops") Signed-off-by: Arik Nemtsov Signed-off-by: Emmanuel Grumbach Signed-off-by: Johannes Berg Signed-off-by: Greg Kroah-Hartman

mac80211: fix driver RSSI event calculations

2015-12-09T19:34:20+00:00

commit 8ec6d97871f37e4743678ea4a455bd59580aa0f4 upstream. The ifmgd->ave_beacon_signal value cannot be taken as is for comparisons, it must be divided by since it's represented like that for better accuracy of the EWMA calculations. This would lead to invalid driver RSSI events. Fix the used value. Fixes: 615f7b9bb1f8 ("mac80211: add driver RSSI threshold events") Signed-off-by: Johannes Berg Signed-off-by: Greg Kroah-Hartman