You are not logged in.
Hello everyone,
My Bluetooth mouse (Microsoft 3600) randomly stops working after a while when using Arch. First I thought it might be some kind of power-saving issues, so I blacklisted its usb id in /etc/default/tlp looking like this:
USB_BLACKLIST="8087:0a2b"
But it didn't work, it still stops working randomly, and what's more, the mouse is still shown as "Connected" even though it is not working propperly. In fact, I cannot find the moment when it stops working. Here is my dmesg:
[dollo@Arch-T470 mnt]$ dmesg | grep -i Bluetooth |tail
[ 3565.742904] Bluetooth: hci0: Minimum firmware build 1 week 10 2014
[ 3565.747713] Bluetooth: hci0: Found device firmware: intel/ibt-12-16.sfi
[ 3567.506142] Bluetooth: hci0: Waiting for firmware download to complete
[ 3567.506769] Bluetooth: hci0: Firmware loaded in 1728201 usecs
[ 3567.506917] Bluetooth: hci0: Waiting for device to boot
[ 3567.519209] Bluetooth: hci0: Device booted in 12030 usecs
[ 3567.527320] Bluetooth: hci0: Found Intel DDC parameters: intel/ibt-12-16.ddc
[ 3567.528908] Bluetooth: hci0: Applying Intel DDC parameters completed
[ 3568.191702] input: BluetoothMouse3600 as /devices/virtual/misc/uhid/0005:045E:0916.0002/input/input20
[ 3568.192342] hid-generic 0005:045E:0916.0002: input,hidraw0: BLUETOOTH HID v1.00 Mouse [BluetoothMouse3600] on BC:A8:A6:E7:6F:9D
Any ideas? The Bluetooth device is an Intel 8265 Dual-band Wireless Bluetooth.
Last edited by dollo (2017-11-27 19:59:40)
Offline
Hey,
I have a similar issue, with a Thinkpad 470p but with my Bose bluetooth headset. After a few seconds it disconnects.
I suppose is the same problem and has to do with one of the latest updates.
I checked and bluez and related programs have not been updated, so I think is a driver issue. I just need to figure out which driver it is using.
Let me gather some additional info and come back at you with some logs.
Offline
Hello,
Wanted to chime in and say I'm having a similar issue. I have a Dell XPS 13 and Logitech M535. Duration before disconnect is not consistent for me, usually under a minute. The mouse works fine on other devices.
Downgrading bluez to a previous version that I know worked did not help. I don't know enough to diagnose the problem further.
Offline
I think it is definitely a problem with "iwlwifi".
The strange thing is that it doesn't seem that "linux-firmware" has been updated recently.
If I disable the Wifi, the problem gets solved.
Maybe this works for you:
https://wiki.archlinux.org/index.php/Wi … on#iwlwifi
EDIT: downgrading "linux-firmware" does not help:
pacman -U /var/cache/pacman/pkg/linux-firmware-20170907.a61ac5c-1-any.pkg.tar.xz
EDIT 2: disabling WiFi coexistence *seems* to work, I get sutter now and then tough!
Last edited by R3DKn16h7 (2017-11-25 08:14:56)
Offline
I've had a similar issue with the Intel 8265 bluetooth recently on my Thinkpad. It seemed to be a firmware issue, because it would work if I rebooted from Windows, but not after a cold start.
After a cold boot, the relevant dmesg output was:
[jkohler@thorondor ~]$ dmesg | grep -i blue
[ 8.672746] Bluetooth: Core ver 2.22
[ 8.672760] Bluetooth: HCI device and connection manager initialized
[ 8.672762] Bluetooth: HCI socket layer initialized
[ 8.672764] Bluetooth: L2CAP socket layer initialized
[ 8.672767] Bluetooth: SCO socket layer initialized
[ 8.678706] Bluetooth: HCI UART driver ver 2.3
[ 8.678708] Bluetooth: HCI UART protocol H4 registered
[ 8.678709] Bluetooth: HCI UART protocol BCSP registered
[ 8.678726] Bluetooth: HCI UART protocol LL registered
[ 8.678727] Bluetooth: HCI UART protocol ATH3K registered
[ 8.678728] Bluetooth: HCI UART protocol Three-wire (H5) registered
[ 8.678748] Bluetooth: HCI UART protocol Intel registered
[ 8.678757] Bluetooth: HCI UART protocol Broadcom registered
[ 8.678758] Bluetooth: HCI UART protocol QCA registered
[ 8.678758] Bluetooth: HCI UART protocol AG6XX registered
[ 8.678759] Bluetooth: HCI UART protocol Marvell registered
[ 8.688843] thinkpad_acpi: rfkill switch tpacpi_bluetooth_sw: radio is unblocked
[ 9.944161] Bluetooth: hci0: Bootloader revision 0.0 build 26 week 38 2015
[ 9.945164] Bluetooth: hci0: Device revision is 16
[ 9.945164] Bluetooth: hci0: Secure boot is enabled
[ 9.945165] Bluetooth: hci0: OTP lock is enabled
[ 9.945165] Bluetooth: hci0: API lock is enabled
[ 9.945165] Bluetooth: hci0: Debug lock is disabled
[ 9.945166] Bluetooth: hci0: Minimum firmware build 1 week 10 2014
[ 9.946792] Bluetooth: hci0: Found device firmware: intel/ibt-12-16.sfi
[ 11.374469] Bluetooth: hci0: Waiting for firmware download to complete
[ 11.375152] Bluetooth: hci0: Firmware loaded in 1397752 usecs
[ 11.375202] Bluetooth: hci0: Waiting for device to boot
[ 11.387148] Bluetooth: hci0: Device booted in 11680 usecs
[ 11.387310] Bluetooth: hci0: Found Intel DDC parameters: intel/ibt-12-16.ddc
[ 11.389183] Bluetooth: hci0: Applying Intel DDC parameters completed
[ 14.197538] Bluetooth: BNEP (Ethernet Emulation) ver 1.3
[ 14.197540] Bluetooth: BNEP filters: protocol multicast
[ 14.197547] Bluetooth: BNEP socket layer initialized
[ 20.033552] Bluetooth: RFCOMM TTY layer initialized
[ 20.033568] Bluetooth: RFCOMM socket layer initialized
[ 20.033578] Bluetooth: RFCOMM ver 1.11
And after a reboot from windows:
[jkohler@thorondor ~]$ dmesg | grep -i blue
[ 9.181298] Bluetooth: Core ver 2.22
[ 9.181312] Bluetooth: HCI device and connection manager initialized
[ 9.181315] Bluetooth: HCI socket layer initialized
[ 9.181317] Bluetooth: L2CAP socket layer initialized
[ 9.181320] Bluetooth: SCO socket layer initialized
[ 9.184559] Bluetooth: HCI UART driver ver 2.3
[ 9.184561] Bluetooth: HCI UART protocol H4 registered
[ 9.184561] Bluetooth: HCI UART protocol BCSP registered
[ 9.184571] Bluetooth: HCI UART protocol LL registered
[ 9.184571] Bluetooth: HCI UART protocol ATH3K registered
[ 9.184572] Bluetooth: HCI UART protocol Three-wire (H5) registered
[ 9.184594] Bluetooth: HCI UART protocol Intel registered
[ 9.184601] Bluetooth: HCI UART protocol Broadcom registered
[ 9.184602] Bluetooth: HCI UART protocol QCA registered
[ 9.184602] Bluetooth: HCI UART protocol AG6XX registered
[ 9.184603] Bluetooth: HCI UART protocol Marvell registered
[ 9.260224] thinkpad_acpi: rfkill switch tpacpi_bluetooth_sw: radio is unblocked
[ 10.467348] Bluetooth: hci0: Firmware revision 0.1 build 151 week 24 2017
[ 14.726752] Bluetooth: BNEP (Ethernet Emulation) ver 1.3
[ 14.726753] Bluetooth: BNEP filters: protocol multicast
[ 14.726756] Bluetooth: BNEP socket layer initialized
[ 20.847658] Bluetooth: RFCOMM TTY layer initialized
[ 20.847674] Bluetooth: RFCOMM socket layer initialized
[ 20.847686] Bluetooth: RFCOMM ver 1.11
It seems the bluetooth controller maintains an updated firmware loaded by Windows, and the linux kernel doesn't overwrite it on restart. I identified the corresponding *.sfi file from the windows driver (either in the driver installer, or finding it through the device driver files), and copied it over intel/ibt-12-16.sfi (after backing up), and now Bluetooth works like a charm in linux after a cold boot too. A warm reboot also reports the newer firmware (Firmware revision 0.1 build 151 week 24 2017). It seems this firmware needs to be updated in linux-firmware upstream, but I don't know how to find the appropriate license documentation.
*** EDIT ***
See also https://bbs.archlinux.org/viewtopic.php?id=231119 for a likely related issue
Last edited by jkohler (2017-11-26 04:10:32)
Offline
Wow, I didn't noticed that difference between a cold boot and rebooting from Windows.
However, I have to say that I have been using my bluetooth mouse with kernel linux-lts for the whole day without any issues at all! But that's a temporary solution until linux-lts catches linux... So, taking into account @jkohler's solution, there might be a regression in how the latest version of the kernel handles the firmware because linux and linux-lts share the same firmware within the package linux-firmware, isn't it?
Besides, my dmesg matches your "cold boot" output, so Arch has loaded the firmware into the controller
Last edited by dollo (2017-11-26 16:47:52)
Offline
Yes, seems it must be an issue between the newer kernel and the old firmware.
I opened a bug report on the upstream kernel (https://bugzilla.kernel.org/show_bug.cgi?id=197997), and the folks at Intel provided a pre-release firmware update there which also resolves the problem for me.
Offline
Great! I think we can mark this thread as solved
Offline