WARN-ON in eee80211_vif_release_channel in 4.20.2+, ath10k

WARN-ON in eee80211_vif_release_channel in 4.20.2+, ath10k

[Ben Greear]

Hello,

I was attempting 6 vap devices on one radio using my hacked ath10k driver and firmware...

I'm using ch 100 to make it even more buggy.

It is having issues, and one thing I saw is this WARN_ON hit:

void ieee80211_vif_release_channel(struct ieee80211_sub_if_data *sdata)
{
	WARN_ON(sdata->dev && netif_carrier_ok(sdata->dev));

	lockdep_assert_held(&sdata->local->mtx);

	mutex_lock(&sdata->local->chanctx_mtx);
	__ieee80211_vif_release_channel(sdata);
	mutex_unlock(&sdata->local->chanctx_mtx);
}

There are no obvious errors in the logs immediately preceeding this, but several
more similar splats hit lower.

[ 3225.690458] WARNING: CPU: 0 PID: 30791 at net/mac80211/chan.c:1760 ieee80211_vif_release_channel+0x18/0x50 [mac80211]
[ 3225.690460] Modules linked in: ath10k_pci ath10k_core nf_conntrack_netlink nfnetlink iptable_raw xt_CT nf_conntrack bluetooth nf_defrag_ipv6 nf_defrag_ipv4 
libcrc32c bpfilter vrf 8021q garp mrp stp llc fuse macvlan wanlink(O) pktgen nfsv3 nfs fscache snd_hda_codec_hdmi coretemp snd_hda_intel intel_rapl 
x86_pkg_temp_thermal snd_hda_codec intel_powerclamp snd_hda_core kvm_intel snd_hwdep iTCO_wdt iTCO_vendor_support snd_seq ath kvm snd_seq_device snd_pcm 
irqbypass mac80211 joydev snd_timer snd cfg80211 i2c_i801 soundcore intel_pch_thermal mei_wdt acpi_pad pcc_cpufreq nfsd auth_rpcgss nfs_acl lockd grace sunrpc 
sch_fq_codel igb hwmon serio_raw e1000e dca i915 i2c_algo_bit drm_kms_helper drm i2c_core video ipv6 crc_ccitt [last unloaded: ath10k_core]
[ 3225.690492] CPU: 0 PID: 30791 Comm: kworker/u8:4 Tainted: G        W  O      4.20.2+ #3
[ 3225.690493] Hardware name: _ _/, BIOS 5.11 08/26/2016
[ 3225.690506] Workqueue: phy11 ieee80211_dfs_cac_timer_work [mac80211]
[ 3225.690517] RIP: 0010:ieee80211_vif_release_channel+0x18/0x50 [mac80211]
[ 3225.690518] Code: bd ea ff ff ff eb a9 66 66 2e 0f 1f 84 00 00 00 00 00 53 48 8b 87 18 0c 00 00 48 89 fb 48 85 c0 74 0a 48 8b 40 48 a8 04 75 02 <0f> 0b 48 8b 
83 20 0c 00 00 48 8d b8 b8 20 00 00 e8 13 bd 31 e1 48
[ 3225.690519] RSP: 0018:ffffc900031e3e60 EFLAGS: 00010246
[ 3225.690520] RAX: 0000000000000003 RBX: ffff88813be0c940 RCX: ffff88816d824020
[ 3225.690520] RDX: ffff888135641e40 RSI: ffff88813d0990b0 RDI: ffff88813be0c940
[ 3225.690521] RBP: ffff88816d824000 R08: 0000000000000000 R09: 0000003131796870
[ 3225.690521] R10: 8080808080808080 R11: 0000000000000000 R12: ffff88813d03cd00
[ 3225.690522] R13: 0000000000000000 R14: ffff888169d8fc00 R15: 0ffff88813d03cd2
[ 3225.690523] FS:  0000000000000000(0000) GS:ffff888172c00000(0000) knlGS:0000000000000000
[ 3225.690523] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 3225.690524] CR2: 000055714bd5eaf0 CR3: 000000000220a005 CR4: 00000000003606f0
[ 3225.690525] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 3225.690525] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[ 3225.690526] Call Trace:
[ 3225.690539]  ieee80211_dfs_cac_timer_work+0x8c/0xb0 [mac80211]
[ 3225.690542]  process_one_work+0x1f3/0x410
[ 3225.690544]  worker_thread+0x28/0x3c0
[ 3225.690546]  ? wq_calc_node_cpumask+0x70/0x70
[ 3225.690547]  kthread+0x10b/0x130
[ 3225.690548]  ? kthread_create_on_node+0x60/0x60
[ 3225.690550]  ret_from_fork+0x1f/0x30
[ 3225.690552] ---[ end trace b4302d8e8360d501 ]---

Eventually, log is filled with this:

[ 3227.469556] ath10k_pci 0000:04:00.0: could not get mac80211 beacon
[ 3227.531081] ath10k_pci 0000:04:00.0: could not get mac80211 beacon
[ 3227.592435] ath10k_pci 0000:04:00.0: could not get mac80211 beacon
[ 3227.653867] ath10k_pci 0000:04:00.0: could not get mac80211 beacon
[ 3227.715322] ath10k_pci 0000:04:00.0: could not get mac80211 beacon
[ 3227.776834] ath10k_pci 0000:04:00.0: could not get mac80211 beacon
[ 3227.838186] ath10k_pci 0000:04:00.0: could not get mac80211 beacon
[ 3227.899703] ath10k_pci 0000:04:00.0: could not get mac80211 beacon
[ 3227.961149] ath10k_pci 0000:04:00.0: could not get mac80211 beacon

Any ideas for places to look for bugs are welcome.

Thanks,
Ben

-- 
Ben Greear <greearb@candelatech.com>
Candela Technologies Inc  http://www.candelatech.com