[PATCH 325/379][SRU][OEM-5.6] ath11k: mac: remove unused conf_mutex to solve a deadlock
You-Sheng Yang
vicamo.yang at canonical.com
Wed Dec 23 08:50:58 UTC 2020
From: Wen Gong <wgong at codeaurora.org>
BugLink: https://bugs.launchpad.net/bugs/1879633
The conf_mutex is not use and lead below deadlock, remove it to solve
the deadlock issue.
[ 44.967496] NET: Registered protocol family 42
[ 45.119629] ath11k_pci 0000:06:00.0: WARNING: ath11k PCI support is experimental!
[ 45.120087] ath11k_pci 0000:06:00.0: BAR 0: assigned [mem 0xdc000000-0xdc0fffff 64bit]
[ 45.120108] ath11k_pci 0000:06:00.0: enabling device (0000 -> 0002)
[ 45.206525] ath11k_pci 0000:06:00.0: aspm 0x42 changed to 0x40
[ 45.207430] mhi 0000:06:00.0: Requested to power ON
[ 45.208609] mhi 0000:06:00.0: Power on setup success
[ 46.190711] ath11k_pci 0000:06:00.0: chip_id 0x0 chip_family 0xb board_id 0x101 soc_id 0xffffffff
[ 46.190729] ath11k_pci 0000:06:00.0: fw_version 0x306a70f fw_build_timestamp 2000-01-01 00:00 fw_build_id
1]: Starting Load/Save RF Kill Switch Status...
[ 46.385118] ath11k_pci 0000:06:00.0 wlp6s0: renamed from wlan0
1]: Started Load/Save RF Kill Switch Status.
[ 53.566669] wlp6s0: authenticate with 00:03:7f:48:dd:bf
[ 53.809092] wlp6s0: send auth to 00:03:7f:48:dd:bf (try 1/3)
[ 53.816490] wlp6s0: authenticated
[ 53.818618] wlp6s0: associate with 00:03:7f:48:dd:bf (try 1/3)
[ 53.820839] wlp6s0: RX AssocResp from 00:03:7f:48:dd:bf (capab=0x1 status=0 aid=2)
[ 53.834859]
[ 53.834861] ======================================================
[ 53.834862] WARNING: possible circular locking dependency detected
[ 53.834863] 5.9.0-rc5-wt-ath+ #198 Not tainted
[ 53.834864] ------------------------------------------------------
[ 53.834865] kworker/u16:3/166 is trying to acquire lock:
[ 53.834866] ffff8c4b37184f78 (&ar->conf_mutex){+.+.}-{3:3}, at: ath11k_mac_op_config+0x16/0x30 [ath11k]
[ 53.834875]
[ 53.834875] but task is already holding lock:
[ 53.834876] ffff8c4b37182808 (&local->iflist_mtx){+.+.}-{3:3}, at: ieee80211_set_associated+0x167/0x360
[ 53.834879]
[ 53.834879] which lock already depends on the new lock.
[ 53.834879]
[ 53.834880]
[ 53.834880] the existing dependency chain (in reverse order) is:
[ 53.834881]
[ 53.834881] -> #1 (&local->iflist_mtx){+.+.}-{3:3}:
[ 53.834884] __lock_acquire+0x3bf/0x6e0
[ 53.834886] lock_acquire+0xb6/0x270
[ 53.834887] __mutex_lock+0x88/0x8e0
[ 53.834890] ieee80211_set_hw_80211_encap+0x3e/0x1f0
[ 53.834895] ath11k_mac_op_add_interface+0x348/0x7f0 [ath11k]
[ 53.834897] drv_add_interface+0x7c/0x190
[ 53.834899] ieee80211_do_open+0x552/0x9a0
[ 53.834901] __dev_open+0xe5/0x190
[ 53.834902] __dev_change_flags+0x1c6/0x230
[ 53.834903] dev_change_flags+0x1c/0x50
[ 53.834905] do_setlink+0x246/0xc60
[ 53.834906] __rtnl_newlink+0x607/0x990
[ 53.834907] rtnl_newlink+0x3f/0x60
[ 53.834908] rtnetlink_rcv_msg+0x174/0x490
[ 53.834910] netlink_rcv_skb+0x42/0x100
[ 53.834911] netlink_unicast+0x18c/0x250
[ 53.834912] netlink_sendmsg+0x227/0x460
[ 53.834914] sock_sendmsg+0x59/0x60
[ 53.834915] ____sys_sendmsg+0x1f5/0x230
[ 53.834916] ___sys_sendmsg+0x70/0xb0
[ 53.834917] __sys_sendmsg+0x54/0xa0
[ 53.834919] do_syscall_64+0x33/0x40
[ 53.834920] entry_SYSCALL_64_after_hwframe+0x44/0xa9
[ 53.834921]
[ 53.834921] -> #0 (&ar->conf_mutex){+.+.}-{3:3}:
[ 53.834923] check_prev_add+0x98/0x9f0
[ 53.834925] validate_chain+0x404/0x6c0
[ 53.834926] __lock_acquire+0x3bf/0x6e0
[ 53.834927] lock_acquire+0xb6/0x270
[ 53.834929] __mutex_lock+0x88/0x8e0
[ 53.834934] ath11k_mac_op_config+0x16/0x30 [ath11k]
[ 53.834935] ieee80211_hw_config+0xb3/0x270
[ 53.834937] ieee80211_set_associated+0x17c/0x360
[ 53.834938] ieee80211_assoc_success.constprop.0+0x5a2/0xc80
[ 53.834940] ieee80211_rx_mgmt_assoc_resp+0x16a/0x350
[ 53.834941] ieee80211_sta_rx_queued_mgmt+0xca/0x410
[ 53.834943] ieee80211_iface_work+0x1f3/0x350
[ 53.834945] process_one_work+0x265/0x5d0
[ 53.834946] worker_thread+0x49/0x300
[ 53.834948] kthread+0x135/0x150
[ 53.834949] ret_from_fork+0x22/0x30
[ 53.834950]
[ 53.834950] other info that might help us debug this:
[ 53.834950]
[ 53.834951] Possible unsafe locking scenario:
[ 53.834951]
[ 53.834952] CPU0 CPU1
[ 53.834952] ---- ----
[ 53.834953] lock(&local->iflist_mtx);
[ 53.834954] lock(&ar->conf_mutex);
[ 53.834955] lock(&local->iflist_mtx);
[ 53.834956] lock(&ar->conf_mutex);
[ 53.834957]
[ 53.834957] *** DEADLOCK ***
[ 53.834957]
[ 53.834958] 4 locks held by kworker/u16:3/166:
[ 53.834959] #0: ffff8c4b37c22948 ((wq_completion)phy0){+.+.}-{0:0}, at: process_one_work+0x1d3/0x5d0
[ 53.834961] #1: ffffa98300abfe70 ((work_completion)(&sdata->work)){+.+.}-{0:0}, at: process_one_work+0x1d3/0x5d0
[ 53.834963] #2: ffff8c4b371e4cd0 (&wdev->mtx){+.+.}-{3:3}, at: ieee80211_sta_rx_queued_mgmt+0x4b/0x410
[ 53.834965] #3: ffff8c4b37182808 (&local->iflist_mtx){+.+.}-{3:3}, at: ieee80211_set_associated+0x167/0x360
[ 53.834968]
[ 53.834968] stack backtrace:
[ 53.834969] CPU: 1 PID: 166 Comm: kworker/u16:3 Not tainted 5.9.0-rc5-wt-ath+ #198
[ 53.834970] Hardware name: Intel(R) Client Systems NUC8i7HVK/NUC8i7HVB, BIOS HNKBLi70.86A.0049.2018.0801.1601 08/01/2018
[ 53.834972] Workqueue: phy0 ieee80211_iface_work
[ 53.834974] Call Trace:
[ 53.834976] dump_stack+0x77/0xa0
[ 53.834978] check_noncircular+0x15d/0x180
[ 53.834980] check_prev_add+0x98/0x9f0
[ 53.834982] ? add_chain_cache+0x143/0x440
[ 53.834984] validate_chain+0x404/0x6c0
[ 53.834986] __lock_acquire+0x3bf/0x6e0
[ 53.834988] lock_acquire+0xb6/0x270
[ 53.834993] ? ath11k_mac_op_config+0x16/0x30 [ath11k]
[ 53.834999] ? ath11k_mac_op_config+0x16/0x30 [ath11k]
[ 53.835001] __mutex_lock+0x88/0x8e0
[ 53.835006] ? ath11k_mac_op_config+0x16/0x30 [ath11k]
[ 53.835007] ? sched_clock_cpu+0xc/0xb0
[ 53.835009] ? __lock_release+0x179/0x2c0
[ 53.835014] ath11k_mac_op_config+0x16/0x30 [ath11k]
[ 53.835016] ieee80211_hw_config+0xb3/0x270
[ 53.835018] ieee80211_set_associated+0x17c/0x360
[ 53.835019] ieee80211_assoc_success.constprop.0+0x5a2/0xc80
[ 53.835021] ? lockdep_hardirqs_on_prepare.part.0+0x9f/0x140
[ 53.835023] ? cmpxchg_double_slab.constprop.0+0x185/0x1a0
[ 53.835025] ? trace_hardirqs_on+0x1c/0x100
[ 53.835027] ? __slab_free+0x8f/0x330
[ 53.835029] ? slab_free_freelist_hook+0xf8/0x150
[ 53.835031] ? ieee802_11_parse_elems_crc+0x147/0x1d0
[ 53.835032] ? kfree+0x2b0/0x2d0
[ 53.835034] ? ieee802_11_parse_elems_crc+0x147/0x1d0
[ 53.835036] ieee80211_rx_mgmt_assoc_resp+0x16a/0x350
[ 53.835041] ieee80211_sta_rx_queued_mgmt+0xca/0x410
[ 53.835043] ? __lock_acquire+0x3bf/0x6e0
[ 53.835045] ? lock_acquire+0xb6/0x270
[ 53.835046] ? skb_dequeue+0x13/0x70
[ 53.835048] ? find_held_lock+0x32/0x90
[ 53.835049] ? sched_clock_cpu+0xc/0xb0
[ 53.835051] ? mark_held_locks+0x50/0x80
[ 53.835053] ? lockdep_hardirqs_on_prepare.part.0+0x9f/0x140
[ 53.835054] ? _raw_spin_unlock_irqrestore+0x34/0x40
[ 53.835056] ? trace_hardirqs_on+0x1c/0x100
[ 53.835058] ieee80211_iface_work+0x1f3/0x350
[ 53.835060] process_one_work+0x265/0x5d0
[ 53.835062] worker_thread+0x49/0x300
[ 53.835063] ? process_one_work+0x5d0/0x5d0
[ 53.835065] kthread+0x135/0x150
[ 53.835066] ? kthread_create_worker_on_cpu+0x60/0x60
[ 53.835068] ret_from_fork+0x22/0x30
[ 53.835075] wlp6s0: associated
[ 53.835132] IPv6: ADDRCONF(NETDEV_CHANGE): wlp6s0: link becomes ready
Tested-on: QCA6390 hw2.0 PCI WLAN.HST.1.0.1-01740-QCAHSTSWPLZ_V2_TO_X86-1
Signed-off-by: Wen Gong <wgong at codeaurora.org>
Signed-off-by: Kalle Valo <kvalo at codeaurora.org>
Link: https://lore.kernel.org/r/1601463073-12106-4-git-send-email-kvalo@codeaurora.org
(cherry picked from commit 7fd85517c359a56a09711f4f2f12f9ef18bfe38c)
Signed-off-by: You-Sheng Yang <vicamo.yang at canonical.com>
---
drivers/net/wireless/ath/ath11k/mac.c | 11 +----------
1 file changed, 1 insertion(+), 10 deletions(-)
diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c
index 1625bd7e1ae1..38a8c419ef21 100644
--- a/drivers/net/wireless/ath/ath11k/mac.c
+++ b/drivers/net/wireless/ath/ath11k/mac.c
@@ -758,21 +758,12 @@ static int ath11k_monitor_vdev_up(struct ath11k *ar, int vdev_id)
static int ath11k_mac_op_config(struct ieee80211_hw *hw, u32 changed)
{
- struct ath11k *ar = hw->priv;
- int ret = 0;
-
/* mac80211 requires this op to be present and that's why
* there's an empty function, this can be extended when
* required.
*/
- mutex_lock(&ar->conf_mutex);
-
- /* TODO: Handle configuration changes as appropriate */
-
- mutex_unlock(&ar->conf_mutex);
-
- return ret;
+ return 0;
}
static int ath11k_mac_setup_bcn_tmpl(struct ath11k_vif *arvif)
--
2.29.2
More information about the kernel-team
mailing list