PINE64 dev

31 Oct 2024 13:18

[I] <d​simic> no worries

PINE64 dev

31 Oct 2024 12:57

[I] <d​simic> ah, I see, thanks

PINE64 dev

31 Oct 2024 12:47

[I] <D​aanct12> yeah but there's nothing to that

PINE64 dev

05 Oct 2024 13:11

[I] <m​egi> yes

PINE64 dev

04 Oct 2024 19:33

[M] <p​iggz> megi: good to know the issues are known about, thx

PINE64 dev

04 Oct 2024 18:34

[I] <m​egi> driver author knows of the issues, it's just not clear how to solve them

PINE64 dev

04 Oct 2024 18:09

currently I'm experimenting with disabling power save and lps mode

PINE64 dev

04 Oct 2024 18:07

[I] <d​simic> let's have that reported on the mailing list

PINE64 dev

04 Oct 2024 18:06

[I] <d​simic> but again, I think that providing feedback to the author of the upstream driver can only be helpful

PINE64 dev

04 Oct 2024 18:03

[M] <A​ndrey Skvortsov> and I see poor performance of upstream driver as well on 6.12.

PINE64 dev

04 Oct 2024 18:02

[M] <p​iggz> ah i missed that in his log...

PINE64 dev

04 Oct 2024 17:33

[I] <d​simic> I might be able to have a look into all that, but I'm really not sure when that might happen

PINE64 dev

04 Oct 2024 16:11

[M] <p​iggz> Watching logs I noticed something that might be useful for wlan driver developers.
Wifi is very faulty (I know... I'm redundant...).
I have many of theese before connection occurs:
-
Sep 06 00:19:12 PinePhone wpa_supplicant[2886]: wlan0: SME: Trying to authenticate with xx:xx:xx:xx:xx:xx (SSID='my-SSID' freq=2412 MHz)
Sep 06 00:19:13 PinePhone kernel: wlan0: authenticate with xx:xx:xx:xx:xx:xx (local address=yy:yy:yy:yy:yy:yy)
Sep 06 00:19:13 PinePhone kernel: wlan0: send auth to xx:xx:xx:xx:xx:xx (try 1/3)
Sep 06 00:19:13 PinePhone kernel: wlan0: send auth to xx:xx:xx:xx:xx:xx (try 2/3)
Sep 06 00:19:14 PinePhone kernel: wlan0: send auth to xx:xx:xx:xx:xx:xx (try 3/3)
Sep 06 00:19:15 PinePhone kernel: wlan0: authentication with xx:xx:xx:xx:xx:xx timed out
Sep 06 00:19:15 PinePhone wpa_supplicant[2886]: wlan0: CTRL-EVENT-SSID-TEMP-DISABLED id=0 ssid="my-SSID" auth_failures=1 duration=10 reason=CONN_FAILED
-
then, if connection happens, lots of theese:
-
Sep 06 00:22:39 PinePhone n-selector[5712]: [D] unknown:0 - Selected service "my-SSID" path "/net/connman/service/wifi_zzzzzzzzzzzz_zzzzzzzzzzzzzzzzzzzzzzzz_managed_psk"
Sep 06 00:22:39 PinePhone n-selector[5712]: [D] unknown:0 - Updating default route
-
and then tons of theese:
-
Sep 06 00:22:47 PinePhone wpa_supplicant[2886]: wlan0: CTRL-EVENT-BEACON-LOSS
-
data transfer eg. reading logs on ssh is always stuttering when not stopping.

PINE64 dev

04 Oct 2024 15:47

[M] <p​iggz> id rather the new, future driver was used ... i wonder if there are any params that need set to make it work better

PINE64 dev

04 Oct 2024 15:45

piggz: that's kinda surprising, that old, crappy realtek driver performs better.

PINE64 dev

31 Oct 2024 13:08

sorry for the noise.

PINE64 dev

31 Oct 2024 12:47

[I] <D​aanct12> just them misclicking the voice button on telegram

PINE64 dev

31 Oct 2024 12:39

[I] <d​simic> was there some file or picture? they disapper from messages coming from Telegram while crossing the bridge

PINE64 dev

04 Oct 2024 23:19

by author who knows about the issues do you mean Fiona Klute. right?

PINE64 dev

04 Oct 2024 18:37

[I] <m​egi> there are some patches that are not upstream that may be worth trying for you https://github.com/airtower-luna/linux/commits/rtw88_8723cs-optimize (I tried, there are still issues, there are also some issues with queueing packets for TX, which show up when doing UDP iperf3 test with TX speeds over what network can handle)

PINE64 dev

04 Oct 2024 18:09

[I] <d​simic> you don't need to be subscribed, and you can get the messages to reply to from the public inbox

PINE64 dev

04 Oct 2024 18:08

[M] <p​iggz> i can do it, but im not currently subscribed, so happy for someone who is to take that on

PINE64 dev

04 Oct 2024 18:06

I saw shuttering in ssh session as well

PINE64 dev

04 Oct 2024 18:04

[I] <d​simic> hmm, how can we differentiate between lower speed, stuttering and poor performance in general?

PINE64 dev

04 Oct 2024 18:02

[M] <p​iggz> good to know its not just me

PINE64 dev

04 Oct 2024 17:52

I think it's not something new. megi wrote about limited upstream driver performance https://xnux.eu/log/097.html

PINE64 dev

04 Oct 2024 17:32

[I] <d​simic> @piggz : hmm, maybe you could try sending all this information to the mailing list, in the thread that introduced the upstream WiFi driver?

PINE64 dev

04 Oct 2024 16:11

[M] <p​iggz> this is type of issue using new driver, compared to old:

PINE64 dev

04 Oct 2024 15:46

[M] <p​iggz> yeah, frustrating ... ive told them to use the new driver as much as possible ... but I may end up shippng both, and allowing one to be blacklisted as desired

PINE64 dev

04 Oct 2024 15:43

[M] <p​iggz> My users are really complaiing about the quality of the upstream PP wlan driver .. i hope it gets better, as theyre reverting to using the old staging driver :/