WLAN is deactivated after every 2nd boot!? RT3090 Wireless driver rt2800pci v: 2.3.0 on lenovo ideapad s205

Hello,
i have an odd bug in 17.0.5 - 64 bit stable. Every 2nd time i boot, my wlan is de-activated.
i tried it for e.g. ten times, the wlan goes on... and after a reboot it stays off?
When it is off, i cannot select it to re-acitvate it, in the network section. (see png)

It happens on kernel
3.16.49-1 and
4.9 and
4.13.10-1-MANJARO #1 SMP PREEMPT Fri Oct 27 11:13:27 UTC 2017 x86_64 GNU/Linux
4.14-manjaro rc version

thanx in advance
LF

EDIT:
booting on a manjaro 17.0.5 live usb/stick in principle wlan is working.

on ubuntu live boot usb stick there is no error

ubuntu 4.10.0-28-generic

ubuntu@ubuntu:~$ uname -a
Linux ubuntu 4.10.0-28-generic #32~16.04.2-Ubuntu SMP Thu Jul 20 10:19:48 UTC 2017 x86_64 x86_64 x86_64 GNU/Linux

dmesg | grep wlp3s0
[ 42.013951] rt2800pci 0000:03:00.0 wlp3s0: renamed from wlan0
[ 42.235807] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[ 42.410133] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[ 42.488865] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[ 71.094322] wlp3s0: authenticate with 00:00:00:00:00:00
[ 71.139730] wlp3s0: send auth to 00:00:00:00:00:00 (try 1/3)
[ 71.143513] wlp3s0: authenticated
[ 71.147221] wlp3s0: associate with 00:00:00:00:00:00 (try 1/3)
[ 71.152275] wlp3s0: RX AssocResp from 00:00:00:00:00:00 (capab=0x431 status=0 aid=1)
[ 71.152415] wlp3s0: associated
[ 71.152497] IPv6: ADDRCONF(NETDEV_CHANGE): wlp3s0: link becomes ready
[ 99.378324] wlp3s0: deauthenticating from 00:00:00:00:00:00 by local choice (Reason: 3=DEAUTH_LEAVING)
[ 101.004253] wlp3s0: authenticate with 00:00:00:00:00:00
[ 101.045325] wlp3s0: send auth to 00:00:00:00:00:00 (try 1/3)
[ 101.063698] wlp3s0: authenticated
[ 101.065074] wlp3s0: associate with 00:00:00:00:00:00 (try 1/3)
[ 101.085948] wlp3s0: RX AssocResp from 00:00:00:00:00:00 (capab=0x431 status=0 aid=1)
[ 101.086118] wlp3s0: associated

Also no problems on manjaro developer version: live stick 17.1 pre, 4.13.3-2-MANJARO #1

System:    Host: user Kernel: 4.13.5-1-MANJARO x86_64 bits: 64 gcc: 7.2.0 Console: tty 0 dm: lightdm
           Distro: Manjaro Linux
Machine:   Device: laptop System: LENOVO product: 1038D9G v: Ideapad S205 serial: <filter>
           Mobo: LENOVO model: Inagua v: 109-B78210-00A serial: <filter>
           UEFI: LENOVO v: 4BCN24WW date: 08/10/2011
           Chassis: type: 10 serial: N/A
CPU:       Dual core AMD E-450 APU with Radeon HD Graphics (-MCP-) arch: Bobcat rev.0 cache: 512 KB
           flags: (lm nx sse sse2 sse3 sse4a ssse3 svm) bmips: 6588
           clock speeds: min/max: 825/1650 MHz 1: 1646 MHz 2: 1646 MHz
Network:   Card-1: **Realtek RTL8101/2/6E** PCI Express Fast/Gigabit Ethernet controller
           driver: **r8169 v: 2.3**LK-NAPI port: 1000 bus-ID: 02:00.0 chip-ID: 10ec:8136
           IF: enp2s0 state: up speed: 100 Mbps duplex: full mac: <filter>
           Card-2: Ralink RT3090 Wireless 802.11n 1T/1R PCIe
           driver: rt2800pci v: 2.3.0 bus-ID: 03:00.0 chip-ID: 1814:3090
           IF: wlp3s0 state: down mac: <filter>
Info:      Processes: 128 Uptime: 27 min Memory: 1157.2/3555.1MB Init: systemd v: 234 Gcc sys: 7.2.0
           Client: Shell (sudo running in bash) inxi: 2.3.40 

(networkwlan)

On the bad boots, if you execute sudo modprobe r8169 does it start working?

Ehrm.. That's the realtek ethernet nic...

Ralink wireless adapters are instead supported in mainline kernel (in rt2x00) since quite a long time
So, I dunno, I suspect a regression.

Though stuff happening in-between boots could be efi or acpi related, and that's a crazy big PITA to solve.
Could you post dmesg?

i cannot upload txt files? why is that? should i post all of dmesg, this would make the post ultra-long..?

Just paste it to any text storage site ( like pastebin.com) and link it to here.

Well... this is weird? now i have to ask. :slight_smile:
Why using this forum, and must paste files to another forum? I've seen now, that pastebin is recomended by some users, but why not simple paste in this forum? It could be google drive as well.. (-; horrible

PS: its because i normally get made at this..
NOTE: Your guest paste has been posted. If you sign up for a free account, you can edit and delete your pastes!
i must use another account... (-;
ok just kidding..

https://pastebin.com/PP9vStBf

Mhh.. Seems totally normal til there.

Maybe post dmesg of the "broken" boot?
Also with journalctl -b -1 so we can see if at any time during the previous shut down the system played any joke.

You can paste everything here, and if you think it's too long, just use the Hide Details option

Summary

You can find it if you click on the little cogwheel on the right side next to the emoji icon when you are writing a post. With this you can hide the selected text and make it shown only if someone clicks on the Summary. You can also change the "Summary" to anything else you find more fitting.

You could also check journalctl | grep NetworkManager too see if that gives you any useful info.

Hi,
here it is,
https://pastebin.com/cN7nCVWV

thanx

wlan related things...

journalctl -b -1 | grep wlan
Okt 17 09:57:38 user NetworkManager[391]: [1508227058.7174] wifi-nl80211: (wlan0): using nl80211 for WiFi device control
Okt 17 09:57:38 user NetworkManager[391]: [1508227058.7181] device (wlan0): driver supports Access Point (AP) mode
Okt 17 09:57:38 user NetworkManager[391]: [1508227058.7287] manager: (wlan0): new 802.11 WiFi device (/org/freedesktop/NetworkManager/Devices/3)
Okt 17 09:57:41 user kernel: rt2800pci 0000:03:00.0 wlp3s0: renamed from wlan0
Okt 17 09:57:41 user NetworkManager[391]: [1508227061.5562] device (wlan0): interface index 3 renamed iface from 'wlan0' to 'wlp3s0'

journalctl -b -1 | grep killswitch
Okt 17 09:57:32 user NetworkManager[391]: [1508227052.8045] rfkill1: found WiFi radio killswitch (at /sys/devices/platform/acer-wmi/rfkill/rfkill1) (platform driver acer-wmi)
Okt 17 09:57:33 user NetworkManager[391]: [1508227053.8505] manager: rfkill: WiFi enabled by radio killswitch; enabled by state file
Okt 17 09:57:33 user NetworkManager[391]: [1508227053.8506] manager: rfkill: WWAN enabled by radio killswitch; enabled by state file
Okt 17 09:57:38 user NetworkManager[391]: [1508227058.7525] rfkill2: found WiFi radio killswitch (at /sys/devices/pci0000:00/0000:00:15.2/0000:03:00.0/ieee80211/phy0/rfkill2) (driver rt2800pci)

So.. That has nothing to do with the idea I had in mind
Nonetheless there I noticed WiFi radio killswitch is handled by acer-wmi.

Which in the dmesg had reportedly a problem.
...
https://patchwork.kernel.org/patch/9303381/
http://www.spinics.net/lists/platform-driver-x86/msg09314.html
https://www.mail-archive.com/platform-driver-x86@vger.kernel.org/msg02880.html


Bingo


I guess this isn't going to be an official patch, so for starters, could you try to compile a kernel with this lines added down there?

	{
		.callback = dmi_matched,
		.ident = "Lenovo Ideapad S205-1038D9G",
		.matches = {
			DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
			DMI_MATCH(DMI_PRODUCT_NAME, "1038D9G"),
		},
		.driver_data = &quirk_lenovo_ideapad_s205,
},
`

Hi,
wow, thank you. I'll try that.
I could not find a manjaro kernel 4.9 or 4.13 on github/manjaro.
Where do i find it, or should i take the last 4.9 from kernel.org?

regards LF

edit:
here it is:

use linux-mainline from aur

With yaourt, tell him to edit pkgbuild
Then somewhere around the start of package/build block put patch -i /path/to/file/with/lines/above
When asked provide it /whatever/drivers/platform/x86/acer-wmi.c

ok, so i need a git clone with: https://aur.archlinux.org/linux-mainline.git

Zusammenfassung

sudo git clone https://aur.archlinux.org/linux-mainline.git manjaro_kernel
Klone nach 'manjaro_kernel' ...
remote: Counting objects: 680, done.
remote: Compressing objects: 100% (415/415), done.
remote: Total 680 (delta 472), reused 422 (delta 265)
Empfange Objekte: 100% (680/680), 346.30 KiB | 1.45 MiB/s, Fertig.
Löse Unterschiede auf: 100% (472/472), Fertig.

where do i set yaourt to the new pkbuild?
[manjaro_kernel]$ ls
90-linux.hook config.x86_64 linux.preset PKGBUILD
config.i686 linux.install merge-with-testing

yaourt

You *install* an aur package, you don't git clone it.
and yaourt is what you use in this process.

ok thanx,
i've loaded kernel 4.13 but yaourt says:
linux-mainline 4.14rc6-1 (2017-10-23 13:12)
(not suporrted pacet: potentially dangerous!)

There was an error: fatal: workingdirectory could not been checked out
do i need to run yaourt with sudo?

Checke Dateien aus: 100% (61287/61287), Fertig.
fatal: Arbeitsverzeichnis konnte nicht ausgecheckt werden
warning: Klonen erfolgreich, Auschecken ist aber fehlgeschlagen.
Sie können mit 'git status' prüfen, was ausgecheckt worden ist
und das Auschecken mit 'git checkout -f HEAD' erneut versuchen.

==> FEHLER: Fehler beim Erstellen der Arbeitskopie des linux git Repos
    Breche ab...
==> FEHLER:Makepkg konnte linux-mainline nicht erstellen.
==> Erstellen von linux-mainline neu starten?[j/N]

Mhh.. I'm too lazy to translate german, but I'll assume that's about the standard error everybody that has been trying to compile something seriously big from AUR always get.

sudo gedit /etc/yaourtrc
Then, have TMPDIR point to some random folder to fill in during compilation.

1 Like

Maybe it would be a useful tip to use LANG=C before typing a command and posting the output? :slightly_smiling_face:

1 Like

Forum kindly sponsored by