network problem after kernel update

[ftriscari]

My current kernel is
3.12.25-desktop-3.mga4
and everything works ok.

When I upgrade to a newer kernel I'm unable to use the wifi. the system is unable to connect to the router.

the wifi driver is

Code: Select all

lspci | grep -i wireless
05:00.0 Network controller: Qualcomm Atheros AR928X Wireless Network Adapter (PCI-Express) (rev 01)

when I boot with a recent kernel I can see

Code: Select all

 iwconfig
wlp5s0    IEEE 802.11bgn  ESSID:off/any 
          Mode:Managed  Access Point: Not-Associated   Tx-Power=16 dBm   
          Retry short limit:7   RTS thr:off   Fragment thr:off
          Encryption key:off
          Power Management:off
         
enp3s0    no wireless extensions.

lo        no wireless extensions.


from what I can read in the wiki this is a problem but I'm unable to find a way forward.
could you help my with solving this problem?
thanks
FT

[doktor5000]

Could you please show the output of

Code: Select all

lspcidrake -v | grep -i wireless
ps -ef | grep -v grep | grep -ie net -e wpa


and, as root:

Code: Select all

journalctl -b -u network --no-pager -a
journalctl -ab | grep -iE "fw|firmware|iwl|wifi|wire|80211"

[doktor5000]

Hi doktor5000
the outputs are:

Code: Select all

lspcidrake -v | grep -i wireless
ath9k           : Qualcomm Atheros|AR928X Wireless Network Adapter (PCI-Express) [NETWORK_OTHER] (vendor:168c device:002a subv:105b subd:e01f) (rev: 01)

Code: Select all

ps -ef | grep -v grep | grep -ie net -e wpa
root        24     2  0 10:28 ?        00:00:00 [netns]
root      2986     1  0 10:28 ?        00:00:00 /usr/sbin/wpa_supplicant -B -i wlp5s0 -c /etc/wpa_supplicant.conf -D wext
500       4574  3316  1 10:29 ?        00:00:01 /usr/bin/perl /usr/bin/net_applet

Code: Select all

journalctl -b -u network --no-pager -a
-- Logs begin at Tue 2014-04-29 21:21:41 BST, end at Sat 2014-11-22 10:30:52 GMT. --
Nov 22 10:28:52 localhost.localdomain ifplugd(enp3s0)[2878]: ifplugd 0.28 initializing.
Nov 22 10:28:52 localhost.localdomain ifplugd(enp3s0)[2878]: Using interface enp3s0/00:26:2D:93:B9:AC with driver <tg3> (version: 3.136)
Nov 22 10:28:52 localhost.localdomain ifplugd(enp3s0)[2878]: Using detection mode: SIOCETHTOOL
Nov 22 10:28:52 localhost.localdomain ifplugd(enp3s0)[2878]: Initialization complete, link beat not detected.
Nov 22 10:28:52 localhost.localdomain network[2614]: Bringing up loopback interface:  [  OK  ]
Nov 22 10:28:52 localhost.localdomain network[2614]: Configuring wireless regulatory domain  [  OK  ]
Nov 22 10:28:52 localhost.localdomain network[2614]: Bringing up interface enp0s29u1u3c4i2:  ERROR    : [/etc/sysconfig/network-scripts/ifup-eth] Device enp0s29u1u3c4i2 does not seem to be present, delaying initialization.
Nov 22 10:28:52 localhost.localdomain network[2614]: [FAILED]
Nov 22 10:28:52 localhost.localdomain network[2614]: Bringing up interface enp3s0:  [  OK  ]
Nov 22 10:28:52 localhost.localdomain network[2614]: Bringing up interface macvtap0:  ERROR    : [/etc/sysconfig/network-scripts/ifup-eth] Device macvtap0 does not seem to be present, delaying initialization.
Nov 22 10:28:52 localhost.localdomain network[2614]: [FAILED]
Nov 22 10:28:52 localhost.localdomain network[2614]: Bringing up interface wlp5s0:  Error for wireless request "Set Encode" (8B2A) :
Nov 22 10:28:52 localhost.localdomain network[2614]: SET failed on device wlp5s0 ; Invalid argument.
Nov 22 10:28:52 localhost.localdomain network[2614]: ioctl[SIOCSIWENCODEEXT]: Invalid argument
Nov 22 10:28:52 localhost.localdomain network[2614]: ioctl[SIOCSIWENCODEEXT]: Invalid argument
Nov 22 10:28:52 localhost.localdomain ifplugd(wlp5s0)[2988]: ifplugd 0.28 initializing.
Nov 22 10:28:52 localhost.localdomain ifplugd(wlp5s0)[2988]: Using interface wlp5s0/F0:7B:CB:18:2B:53 with driver <ath9k> (version: 3.14.24-desktop-1.mga4)
Nov 22 10:28:52 localhost.localdomain ifplugd(wlp5s0)[2988]: Using detection mode: SIOCETHTOOL
Nov 22 10:28:52 localhost.localdomain ifplugd(wlp5s0)[2988]: Initialization complete, link beat not detected.
Nov 22 10:28:52 localhost.localdomain network[2614]: [  OK  ]
Nov 22 10:28:52 localhost.localdomain systemd[1]: network.service: control process exited, code=exited status=1
Nov 22 10:28:52 localhost.localdomain systemd[1]: Failed to start LSB: Bring up/down networking.
Nov 22 10:28:52 localhost.localdomain systemd[1]: Unit network.service entered failed state.

Code: Select all

journalctl -ab | grep -iE "fw|firmware|iwl|wifi|wire|80211"
Nov 22 10:28:21 localhost.localdomain kernel: [Firmware Bug]: ACPI: BIOS _OSI(Linux) query ignored
Nov 22 10:28:21 localhost.localdomain kernel: psmouse serio1: synaptics: Touchpad model: 1, fw: 7.2, id: 0x1c0b1, caps: 0xd04733/0xa44000/0xa0000, board id: 3655, fw id: 529512
Nov 22 10:28:23 localhost.localdomain kernel: cfg80211: Calling CRDA to update world regulatory domain
Nov 22 10:28:23 localhost.localdomain kernel: cfg80211: World regulatory domain updated:
Nov 22 10:28:23 localhost.localdomain kernel: cfg80211:  DFS Master region: unset
Nov 22 10:28:23 localhost.localdomain kernel: cfg80211:   (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
Nov 22 10:28:23 localhost.localdomain kernel: cfg80211:   (2402000 KHz - 2472000 KHz @ 40000 KHz), (N/A, 2000 mBm)
Nov 22 10:28:23 localhost.localdomain kernel: cfg80211:   (2457000 KHz - 2482000 KHz @ 40000 KHz), (N/A, 2000 mBm)
Nov 22 10:28:23 localhost.localdomain kernel: cfg80211:   (2474000 KHz - 2494000 KHz @ 20000 KHz), (N/A, 2000 mBm)
Nov 22 10:28:23 localhost.localdomain kernel: cfg80211:   (5170000 KHz - 5250000 KHz @ 80000 KHz), (N/A, 2000 mBm)
Nov 22 10:28:23 localhost.localdomain kernel: cfg80211:   (5735000 KHz - 5835000 KHz @ 80000 KHz), (N/A, 2000 mBm)
Nov 22 10:28:23 localhost.localdomain kernel: cfg80211:   (57240000 KHz - 63720000 KHz @ 2160000 KHz), (N/A, 0 mBm)
Nov 22 10:28:26 localhost.localdomain kernel: ieee80211 phy0: Selected rate control algorithm 'minstrel_ht'
Nov 22 10:28:26 localhost.localdomain kernel: ieee80211 phy0: Atheros AR9280 Rev:2 mem=0xffffc900113e0000, irq=17
Nov 22 10:28:26 localhost.localdomain kernel: cfg80211: Calling CRDA for country: US
Nov 22 10:28:26 localhost.localdomain kernel: cfg80211: Regulatory domain changed to country: US
Nov 22 10:28:26 localhost.localdomain kernel: cfg80211:  DFS Master region: unset
Nov 22 10:28:26 localhost.localdomain kernel: cfg80211:   (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
Nov 22 10:28:26 localhost.localdomain kernel: cfg80211:   (2402000 KHz - 2472000 KHz @ 40000 KHz), (N/A, 3000 mBm)
Nov 22 10:28:26 localhost.localdomain kernel: cfg80211:   (5170000 KHz - 5250000 KHz @ 80000 KHz), (N/A, 1700 mBm)
Nov 22 10:28:26 localhost.localdomain kernel: cfg80211:   (5250000 KHz - 5330000 KHz @ 80000 KHz), (N/A, 2300 mBm)
Nov 22 10:28:26 localhost.localdomain kernel: cfg80211:   (5735000 KHz - 5835000 KHz @ 80000 KHz), (N/A, 3000 mBm)
Nov 22 10:28:26 localhost.localdomain kernel: cfg80211:   (57240000 KHz - 63720000 KHz @ 2160000 KHz), (N/A, 4000 mBm)
Nov 22 10:28:52 localhost.localdomain network[2614]: Configuring wireless regulatory domain  [  OK  ]
Nov 22 10:28:52 localhost.localdomain network[2614]: Bringing up interface wlp5s0:  Error for wireless request "Set Encode" (8B2A) :
Nov 22 10:29:04 localhost.localdomain kernel: IFWLOG: register target

thanks
FT

[ftriscari]

I'm still having the same problem even with the latest kernel update.
can anyone help with this?
thanks
FT

[oj]

The update today (Jan 7 2015) prevents my system from booting entirely. In 'failsafe' I get to a desktop with no wifi. I had the exact same thing happen today with an update to fedora 20. Is there some unseen impetus behind today's "update?"

I had to revert to kernel ..12.23 and still had to use 'failsafe' to get to a desktop. It'd be nice to have a windows-like systemrestore in times like this.

[doktor5000]

Dou you have the same wifi adapter and driver as the OP?

For the issue with the system not booting, maybe see viewtopic.php?f=23&t=8863

[oj]

No, my adapter is broadcom:

Code: Select all

wl              : Broadcom Corporation|BCM4312 802.11b/g LP-PHY

I'll give the nvidia driver fix a try. Meantime, it seems the broadcom module is not being rebuilt. As I mentioned I have the same problem with a fedora update, also with broadcom hardware.

EDIT: got it fixed, thanks. The nvidia rebuild did the trick. For the wireless all I did was remove the connection and run the setup again. Fortunately the transmitter was on and it found the device. I haven't been so lucky on that fedora machine, it can't see the hardware. (transmitter is off, no switch)

[ftriscari]

no resolution for me... I'm still unable to use a newer kernel as this will kill my wifi...
do you have any ideas on how to solve this?
thanks
FT

[oj]

ftriscari,

Have you tried what I did, removing the connection then setting it up again with the tool in mcc?

If it doesn't find the device it's most likely that the transmitter is off. Too many laptops do not have a hardware switch to toggle the transmitter, they rely on a "software switch" meaning a proprietary (aka 'secret') windows kernel "switch." The trick for most Linux distros is to get the transmitter on so normal setup can ensue.

I've used a number of tricks. If the machine has dual boot, I've booted into windows first, which turned the xmitter on, the did a reboot into Linux. (that is not a complete shutdown/startup) This worked on about 90% of such machines. Some laptops shut the xmitter off when reading BIOS no matter what. If you have one of them you may never get wireless working. At least not with the distro you want to use. As an example, I had one laptop that fedora would start the transmitter but Mageia, PCLinuxOS and others I like to use did not. I just bit the bullet and used fedora on that one.

Any way you can think of to do the reboot trick would work, it doesn't have to be windows. Maybe you could boot a live cd/dvd of a distro that does handle the device, then restart back into Mageia. (hoping the transmitter stays on)

The only other advice is to poke around system setup (BIOS) and see if there's any mention of wifi, and make changes. (when testing BIOS changes only change one thing at a time, and if it doesn't help, set it back to where it was) I had one system that had a wifi-related option like "allow soft switch." It seemed counter intuitive, but turning that off actually resulted in the transmitter being on full time. Alas, I only saw that work on that one computer.

Nevertheless, it can't hurt to change some settings.

And of course I must mention that if you do have a hardware switch, make sure it's on, or try turning it off and on to see if the system awakes to the arrival of 'new hardware.'