search

Frogging-Family / linux-tkg

linux-tkg custom kernels
GNU General Public License v2.0
1.34k stars 168 forks source link

Bluetooth is broken on 6.4.1 #766

Open gardotd426 opened 1 year ago

gardotd426 commented 1 year ago

So I went to try and connect my Switch Pro Controller that I use at least 3 times a week, and I noticed my bluetooth was off. I went into the Plasma settings and it said BT was disabled. Clicking enable did nothing. No errors in systemctl status bluetooth. HCI0 showed up in /sys/class/bluetooth and wasn't blocked by rfkill, but bluetoothctl showed that there was no controller available.

I rebooted into my 6.3 tkg-bore kernel and BT worked. So then I rebooted into 6.4.1-arch to see if it was an upstream kernel bug, and nope, BT worked fine again (that's the kernel I'm currently booted into).

I tried booting into 6.4.1-tkg-bore like 4 times, BT wouldn't work. I'm almost certain it worked in 6.4.0 though.

I can try to get some logs later if needed.

What's weird, is that my journalctl logs from that boot 6.4.0-273-tkg-bore. No idea why that is. It's fetching 6.4.1 from upstream though, so that part doesn't really matter.

journalctl -b -2 | grep bluetooth doesn't show any errors, everything shows success (-2 is the last boot I had in that kernel).

gardotd426 commented 1 year ago

Like I said, it works in 6.4.1-arch. That wouldn't happen with a bad conf.

On Tue, Jul 4, 2023 at 11:27 PM Ryan Musante @.***> wrote:

Check your pacnew files, Bluetooth had a conf update

— Reply to this email directly, view it on GitHub https://github.com/Frogging-Family/linux-tkg/issues/766#issuecomment-1620962025, or unsubscribe https://github.com/notifications/unsubscribe-auth/AM5Y335UTDY3OI7KKQBSTETXOTNILANCNFSM6AAAAAAZ6KV2FE . You are receiving this because you authored the thread.Message ID: @.***>

ryanmusante commented 1 year ago

Oh, sorry. I see now, I skimmed past that.

ryanmusante commented 1 year ago

I found information on this that might be helpful:

Bug in bluez new version https://bugs.archlinux.org/task/78937

https://bbs.archlinux.org/viewtopic.php?id=286864

gardotd426 commented 1 year ago

I'm not getting a coredump, either.

On Sat, Jul 8, 2023 at 4:31 AM Ryan Musante @.***> wrote:

I found information on this that might be helpful:

https://bugs.archlinux.org/task/78937

— Reply to this email directly, view it on GitHub https://github.com/Frogging-Family/linux-tkg/issues/766#issuecomment-1626912466, or unsubscribe https://github.com/notifications/unsubscribe-auth/AM5Y33YX3TVM2CY2D3GPE4LXPELH5ANCNFSM6AAAAAAZ6KV2FE . You are receiving this because you authored the thread.Message ID: @.***>

ryanmusante commented 1 year ago

Bluetooth it seems a mess recently. Reminds me of how wifi was many years ago

gardotd426 commented 1 year ago

I found something:

Jul 09 23:54:44 matt-archlinux systemd[1]: Started Bluetooth service.
Jul 09 23:54:44 matt-archlinux kernel: Bluetooth: hci0: command 0xfc05 tx timeout
Jul 09 23:54:44 matt-archlinux kernel: Bluetooth: hci0: Reading Intel version command failed (-110)

There are a ton of other lines in the journal that contain btusb, bluetooth, or hci0, but that's the only error and it doesn't show up in 6.4.1-arch or 6.4.1-zen.

gardotd426 commented 1 year ago

It's a reintroduction of an old bug.

Current bugzilla page:

https://bugzilla.kernel.org/show_bug.cgi?id=215167

I don't see any actual fixes mentioned yet, but it's a known issue. Right now the only fix is either a very hacky workaround patch (that extends the timeout on boot) or modprobe -r btusb; modprobe btusb.

ryanmusante commented 1 year ago

Hopefully it's patched soon

AdelKS commented 9 months ago

Any news on this front ?