Wireless Networking disapears after update

[Paul Boyle]

Hi,

I recently updated my CENTOS 7 system. After the update, the wireless support was gone. I can not even search for wireless networks. Some relevant information:
Kernel Version:
3.10.0-1062.1.1.el7.x86_64 #1 SMP Fri Sep 13 22:55:44 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux

I can boot an older kernel, but still problem persists.

/usr/sbin/lspci shows the Atheros wireless network adapter:
03:00.0 Network controller: Qualcomm Atheros QCA9377 802.11ac Wireless Network Adapter (rev 31)

/usr/sbin/lsmod | grep ath10k shows that the kernel modules for the network driver are loaded (I think):
ath10k_pci 47418 0
ath10k_core 325711 1 ath10k_pci
ath 33542 1 ath10k_core
mac80211 718956 1 ath10k_core
cfg80211 632335 3 ath,mac80211,ath10k_core

/usr/sbin/ip -a address doesn't show network adapter in the list of devices.

I am suspecting that NetworkManager may be the culprit, but I'm not sure. I would appreciate some help.

Thanks,

Paul

[gostal]

Assuming your hardware is configured and working. Is your NetworkManager enabled? Code:

Code: Select all

systemctl status NetworkManager

if not enable it (as root):

Code: Select all

systemctl enable NetworkManager

Then do (as root)

Code: Select all

systemctl start network.service

If all of the above goes well your connections should show up if you do:

Code: Select all

nmcli connection show --active

Cheers,
gostal

[Paul Boyle]

Hi,

Sorry for the delay in getting back to this issue.

Thanks for the feedback. When I do:
[root@kuden ~]# systemctl status NetworkManager
● NetworkManager.service - Network Manager
Loaded: loaded (/usr/lib/systemd/system/NetworkManager.service; enabled; vendor preset: enabled)
Active: active (running) since Thu 2019-10-17 17:10:51 EDT; 35min ago
Docs: man:NetworkManager(8)
Main PID: 1192 (NetworkManager)

So, it looks like the NetworkManager is working OK. When I run the nmcli command,
[root@kuden ~]# nmcli connection show --active
NAME UUID TYPE DEVICE
enp4s0f1 42e06683-f127-4cba-85d9-3821aabb17f7 ethernet enp4s0f1
lxdbr0 1abcc363-323d-4957-bfd2-65f9dc03a9aa bridge lxdbr0
virbr0 501a1297-74b2-420b-9f31-3b13e11adbea bridge virbr0

There is no wifi device in this list.

In investigating more, I see this in /var/log/messages after bootup:
Oct 17 17:10:33 kuden kernel: ath10k_pci 0000:03:00.0: qca9377 hw1.1 target 0x05020001 chip_id 0x003821ff sub 105b:e0a1
Oct 17 17:10:33 kuden kernel: ath10k_pci 0000:03:00.0: kconfig debug 0 debugfs 1 tracing 0 dfs 0 testmode 0
Oct 17 17:10:33 kuden kernel: ath10k_pci 0000:03:00.0: firmware ver WLAN.TF.2.1-00021-QCARMSWP-1 api 6 features wowlan,ignore-otp crc32 42e41877
Oct 17 17:10:36 kuden kernel: ath10k_pci 0000:03:00.0: unable to read from the device
Oct 17 17:10:36 kuden kernel: ath10k_pci 0000:03:00.0: could not execute otp for board id check: -110
Oct 17 17:10:36 kuden kernel: ath10k_pci 0000:03:00.0: failed to get board id from otp: -110
Oct 17 17:10:36 kuden kernel: ath10k_pci 0000:03:00.0: could not probe fw (-110)

I don't know what this means, but I assume not being able to read from the device is a bad thing. Nevertheless, 'lsmod' indicates that the drivers for the wifi adapter have loaded.

Thanks for any continued guidance.

[TrevorH]

What do the similar lines in dmesg report on the kernel that does work? Compare the two and see what the differences are.

[Paul Boyle]

I tried booting the oldest kernel I have on my computer (which previously had usable wifi). I get exactly the same error message from dmesg. This would indicate to me, two possibilities:
1) The problem is not in the kernel
2) The wifi adapter suffered a hardware malfunction which was fortuitously showed up after the upgrade.

Paul

[Paul Boyle]

There was a large update last night which included a new kernel. I get the same error message:

[root@kuden ~]# dmesg | grep ath10k
[ 13.245320] ath10k_pci 0000:03:00.0: irq 133 for MSI/MSI-X
[ 13.245408] ath10k_pci 0000:03:00.0: pci irq msi oper_irq_mode 2 irq_mode 0 reset_mode 0
[ 13.837067] ath10k_pci 0000:03:00.0: qca9377 hw1.1 target 0x05020001 chip_id 0x003821ff sub 105b:e0a1
[ 13.837081] ath10k_pci 0000:03:00.0: kconfig debug 0 debugfs 1 tracing 0 dfs 0 testmode 0
[ 13.839069] ath10k_pci 0000:03:00.0: firmware ver WLAN.TF.2.1-00021-QCARMSWP-1 api 6 features wowlan,ignore-otp crc32 42e41877
[ 16.901930] ath10k_pci 0000:03:00.0: unable to read from the device
[ 16.901934] ath10k_pci 0000:03:00.0: could not execute otp for board id check: -110
[ 16.901936] ath10k_pci 0000:03:00.0: failed to get board id from otp: -110
[ 16.901975] ath10k_pci 0000:03:00.0: could not probe fw (-110)

[root@kuden ~]# uname -a
Linux kuden 3.10.0-1062.4.1.el7.x86_64 #1 SMP Fri Oct 18 17:15:30 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux
[root@kuden ~]#

I can use the 'nmcli' command to see that the computer believes the wifi is enabled:
[root@kuden ~]# nmcli general status
STATE CONNECTIVITY WIFI-HW WIFI WWAN-HW WWAN
connected full enabled enabled enabled enabled
[root@kuden ~]#

However, if I look for active connections, the wifi has no device associated with it.
[root@kuden ~]# nmcli connection show --active
NAME UUID TYPE DEVICE
enp4s0f1 42e06683-f127-4cba-85d9-3821aabb17f7 ethernet enp4s0f1
lxdbr0 7687394b-ad8d-46e3-a9f4-98120ae21f71 bridge lxdbr0
virbr0 4d05adce-fbb5-4cc2-9ed1-688d5ed0f9e7 bridge virbr0

[TrevorH]

So what was the last kernel version where your device did function correctly?

[Paul Boyle]

Before the fateful update, I was running 3.10.0-957.27.2.el7

I just booted that kernel and no wifi. Here is the output from dmesg, uname, and a couple if nmcli commands:

[root@kuden ~]# dmesg | grep -i ath10k
[ 14.823266] ath10k_pci 0000:03:00.0: irq 133 for MSI/MSI-X
[ 14.823339] ath10k_pci 0000:03:00.0: pci irq msi oper_irq_mode 2 irq_mode 0 reset_mode 0
[ 15.422624] ath10k_pci 0000:03:00.0: qca9377 hw1.1 target 0x05020001 chip_id 0x003821ff sub 105b:e0a1
[ 15.422641] ath10k_pci 0000:03:00.0: kconfig debug 0 debugfs 1 tracing 0 dfs 0 testmode 0
[ 15.424418] ath10k_pci 0000:03:00.0: firmware ver WLAN.TF.2.1-00021-QCARMSWP-1 api 6 features wowlan,ignore-otp crc32 42e41877
[ 18.490666] ath10k_pci 0000:03:00.0: unable to read from the device
[ 18.490669] ath10k_pci 0000:03:00.0: could not execute otp for board id check: -110
[ 18.491359] ath10k_pci 0000:03:00.0: failed to get board id from otp: -110
[ 18.492086] ath10k_pci 0000:03:00.0: could not probe fw (-110)
[root@kuden ~]# uname -a
Linux kuden 3.10.0-957.27.2.el7.x86_64 #1 SMP Mon Jul 29 17:46:05 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux
[root@kuden ~]# nmcli general status
STATE CONNECTIVITY WIFI-HW WIFI WWAN-HW WWAN
connected full enabled enabled enabled enabled
[root@kuden ~]# nmcli device
DEVICE TYPE STATE CONNECTION
enp4s0f1 ethernet connected enp4s0f1
lxdbr0 bridge connected lxdbr0
virbr0 bridge connected virbr0
lo loopback unmanaged --
virbr0-nic tun unmanaged --

[TrevorH]

The only other thing I can think of is that the firmware for your device has changed. Running locate -i qca9377 on my up to date CentOS 7 system yields a list of files in /usr/lib/firmware/ath10k/QCA9377. I have no idea which of those files relate to your model of card. An rpm -qf /usr/lib/firmware/ath10k/QCA9377/hw1.0/firmware-6.bin (one picked at random) tells me that it belongs to the linux-firmware package. You could look in /var/log/yum.log and see when this was last updated and what the previous one was and you could try backleveling to that.

[Paul Boyle]

Thanks, I look into that. I appreciate all your help so far.