Bug 206513
Summary: | [RTLWIFI/RTL8188EE] Bad Wifi and frequent connection drops with rtlwifi/rtl8188ee | ||
---|---|---|---|
Product: | Networking | Reporter: | Hélder (helder.armando) |
Component: | Wireless | Assignee: | networking_wireless (networking_wireless) |
Status: | NEW --- | ||
Severity: | high | CC: | bidokok587, merpuh, nixgod, pmenzel+bugzilla.kernel.org, ptf, sidpranjale127 |
Priority: | P1 | ||
Hardware: | x86-64 | ||
OS: | Linux | ||
Kernel Version: | 5.0.0-25-generic | Subsystem: | |
Regression: | No | Bisected commit-id: | |
Attachments: |
uname -a
module params modinfo lspci -k lshw network dmesg 0 dmesg 1 dmesg 2 Snippet of systemd journal from when the disconnect occurs |
Description
Hélder
2020-02-13 11:52:14 UTC
Created attachment 287335 [details]
module params
for f in /sys/module/rtl8188ee/parameters/*; do echo $f; cat $f; done
Created attachment 287337 [details] modinfo >> modinfo rtl8188ee Created attachment 287339 [details] lspci -k >> lspci -k Created attachment 287341 [details] lshw network >> lshw -class network Created attachment 287343 [details] dmesg 0 >> dmesg | grep rtlwifi Created attachment 287345 [details] dmesg 1 >> dmesg | grep wlp Created attachment 287347 [details] dmesg 2 >> dmesg | grep -i firm I have the same issue with my HP 15-ba018wm. It has an RTL8818EE. Every time the kernel updates, I need to assure that I have a copy of this code: https://github.com/lwfinger/rtlwifi_new When I follow the instructions provided then my WiFi works fine. Maintainers, please incorporate the code into the mainstream kernel. Thank you! I have the same problem with the wifi card rtl8188ee and 5.4.0-56-generic kernel. Upload is dropping to 100KiB/s (instead of 5 Mb/s with other devices), Download is 500KiB/s (instead of 5 Mb/s). At all connection is slow and unstable. Couldn't find a fix as the repo https://github.com/lwfinger/rtlwifi_new is down. Suggestions to fix this annoying problem? > Maintainers, please incorporate the code into the mainstream kernel. I believe the code was integrated [1]. > If you are looking for a driver for chips such as RTL8188EE, RTL8192CE, > RTL8192CU, RTL8192DE, RTL8192EE, RTL8192SE, RTL8723AE, RTL8723BE, or > RTL8821AE, these should be provided by your kernel. If not, then you should > go to the Backports Project > (https://backports.wiki.kernel.org/index.php/Main_Page) to obtain the > necessary code. So please test the latest Linux kernel version, currently 6.0.7 and 6.1-rc4, and report back. If the problem persists, please send an email to the contacts for rtlwifi listed in the file `MAINTAINERS`. [1]: https://github.com/lwfinger/rtw88 I've tested with 6.0.10, and can confirm this is still an issue, at least on RTL8821AE. The network disconnects and asks me for the wifi password over and over, no matter what secret service I use (kdewallet, gnome-keyring, all the same). I'm not sure if I should be using a driver from backports for this. Created attachment 303362 [details]
Snippet of systemd journal from when the disconnect occurs
After digging around a little, I isolated the exact logs from when this happens on rtl8821ae. It appears that the supplicant service disconnects from the network because it is unable to re-authenticate when the band changes. I will also e-mail the maintainer about this.
Can confirm this is still a problem. RTL8723BE (same family) Linux dora 6.2.8-200.fc37.x86_64 #1 SMP PREEMPT_DYNAMIC Wed Mar 22 19:11:02 UTC 2023 x86_64 x86_64 x86_64 GNU/Linux Mar 30 22:12:03 dora kernel: wlp7s0: deauthenticated from 74:4d:28:4b:b6:eb (Reason: 6=CLASS2_FRAME_FROM_NONAUTH_STA) Mar 30 22:12:03 dora wpa_supplicant[922]: wlp7s0: CTRL-EVENT-DISCONNECTED bssid=74:4d:28:4b:b6:eb reason=6 Mar 30 22:26:58 dora wpa_supplicant[922]: wlp7s0: WPA: Group rekeying completed with 74:4d:28:4b:26:ee [GTK=CCMP] Mar 30 22:31:58 dora wpa_supplicant[922]: wlp7s0: WPA: Group rekeying completed with 74:4d:28:4b:26:ee [GTK=CCMP] it's incredibly frustrating, you can't simply change the WiFI card in this old laptop lenovo hardware locks them so you can only put their part number in it. Please always attach your full log (`dmesg`), and also list what driver is used. It looks like the Linux kernel bug tracker is not used by the subsystem maintainers, so I recommend to contact the maintainer and list, listed in `MAINTAINERS` [1]. @Josh, what old Lenovo laptop is that? Maybe you can use coreboot, or patch the vendor firmware. [1]: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/MAINTAINERS The plan was to include dmesg but it wouldn't format right (need markdown or something). I'll look up the maintainer. however this is a bug that has existed sine 2017 even now there are several bugs in this bug tracker about it.. not to mention every linux distro tracker i've checked.. it's either an elusiive issue that only hapens on some dies or defective dies. I've noticed that, it starts happening when a bluetooth device is connected to, if i leave bluetooth off its fine.. but once i connect to a bt speaker etc. it starts happening at the very least much much more frequentatly after that only way to fix it is modprobe -r and reload it will work better until i connect to another BT device.. sad thing is it has horrible speakers so a BT speaker is almost a requirement.. Anyway, it's a Lenovo B50-45 it's an older APU machine, and i don't know if coreboot is supported as for patching the firmware, i've read about it it seems to require a jtag/clip and directly writing to the EPROM.. If i'm being honest, the potato isn't worth that much to me, soon as i can aquire a replacement it will get thrown in my server room as a small webserver or something Thank you for the update. (I would have filed a separate issue, if using Bluetooth causes the issue.) Please do not paste the Linux messages into the comment, but attach it as a text file. If the maintainers do not use the bug tracker, then they won’t know about the problem. So, contacting the mailing list is best. Fingers crossed, that they have the resources to look into it. The issues resolve for me when I prevent the firmware from being loaded, either by deleting it or moving it. |