Hello, Open bug in launchpad.net https://bugs.launchpad.net/bugs/1807375 dmesg: Bluetooth: hci0: RTL: rtl: unknown IC info, lmp subver a99e, hci rev 826c, hci ver 0008 Best regards, -- Cristian Aravena Romero (caravena)
Created attachment 280099 [details] dmesg 4.20.0-042000rc7-generic I really do not know if the problem was definitively solved. Because Bluetooth devices now appear in GNOME.
(In reply to Cristian Aravena Romero from comment #1) > Created attachment 280099 [details] > dmesg 4.20.0-042000rc7-generic > > I really do not know if the problem was definitively solved. Because > Bluetooth devices now appear in GNOME. It is still affecting me on vanilla 4.20 and 4.19.12. I was able to modify drivers/bluetooth/btrtl.c to use rtl_bt/rtl8821c_fw.bin firmware for this device.
Created attachment 280449 [details] Add IC_INFO(a99e, 826c) to load rtl8821c firmware With this patch, it has been working fine for a week using Bluetooth mouse. Other things tested: - File transfer - A2DP Source
vasthusya, I also see this issue, but this only happens to reboot. Do you see this on cold boot?
Created attachment 280459 [details] Kernel messages from various boot states. Yes, the coldboot lmp_subver=8221 & hci_rev=c are correct and Bluetooth works on vanilla kernel. Unfortunately, I must restart the computer in order to gain CPU clock boost on my Ryzen 5 2500u. Separate issue, haven't checked around here for that bug yet. https://community.amd.com/thread/229980#2893988
Created attachment 280713 [details] 0001-Bluetooth-btrtl-Let-rtl_load_file-always-return-erro.patch
Created attachment 280715 [details] 0002-Bluetooth-btrtl-Let-btrtl_initialize-always-return-e.patch
Created attachment 280717 [details] 0003-Bluetooth-btrtl-Skip-initialization-if-firmware-is-a.patch
Please try these patches. It should the issue.
Created attachment 280781 [details] 0004-Bluetooth-btrtl-Expose-btrtl_device_info.patch Hi Kai, Your patch #2 had two issues: - btrtl_device_info definition is unavailable in hci_h5.c as it's not in header - a couple of argument type mismatch It would have compiled fine when you don't have CONFIG_BT_HCIUART_RTL=y I have made it compilable and running it now with working Bluetooth. Thanks
Turns out it's a regression, I sent V2 patch to address the issue. https://lkml.org/lkml/2019/1/27/16
The patch has been working for me since kernel 5.1. Thank you.