Am Do., 6. Juni 2019 um 13:20 Uhr schrieb Marcel Holtmann <mar...@holtmann.org>: > > Hi Joerg, > > >>> In 5.2.0-rcx I see a new error message on startup probably after > >>> loading the Bluetooth firmware: > >>> [ 1.609460] Bluetooth: hci0: unexpected event for opcode 0xfc2f > >>> > >>>> dmesg | grep Bluetooth > >>> [ 0.130969] Bluetooth: Core ver 2.22 > >>> [ 0.130973] Bluetooth: HCI device and connection manager initialized > >>> [ 0.130974] Bluetooth: HCI socket layer initialized > >>> [ 0.130975] Bluetooth: L2CAP socket layer initialized > >>> [ 0.130976] Bluetooth: SCO socket layer initialized > >>> [ 0.374716] Bluetooth: RFCOMM TTY layer initialized > >>> [ 0.374718] Bluetooth: RFCOMM socket layer initialized > >>> [ 0.374718] Bluetooth: RFCOMM ver 1.11 > >>> [ 0.374719] Bluetooth: BNEP (Ethernet Emulation) ver 1.3 > >>> [ 0.374720] Bluetooth: BNEP socket layer initialized > >>> [ 0.374721] Bluetooth: HIDP (Human Interface Emulation) ver 1.2 > >>> [ 0.374722] Bluetooth: HIDP socket layer initialized > >>> [ 1.422530] Bluetooth: hci0: read Intel version: 370710018002030d00 > >>> [ 1.422533] Bluetooth: hci0: Intel Bluetooth firmware file: > >>> intel/ibt-hw-37.7.10-fw-1.80.2.3.d.bseq > >>> [ 1.609460] Bluetooth: hci0: unexpected event for opcode 0xfc2f > >>> [ 1.625557] Bluetooth: hci0: Intel firmware patch completed and > >>> activated > >>> [ 21.986125] input: BluetoothMouse3600 Mouse as > >>> /devices/virtual/misc/uhid/0005:045E:0916.0004/input/input15 > >>> [ 21.986329] input: BluetoothMouse3600 Consumer Control as > >>> /devices/virtual/misc/uhid/0005:045E:0916.0004/input/input16 > >>> [ 21.986408] hid-generic 0005:045E:0916.0004: input,hidraw3: > >>> BLUETOOTH HID v1.10 Mouse [BluetoothMouse3600] on 80:19:34:4D:31:44 > >>> > >>> > >>> The error message goes away if I revert following patch: > >>> f80c5dad7b64 Bluetooth: Ignore CC events not matching the last HCI command > >> > >> if you can send btmon trace (or better btmon -w trace.log) for this event > >> triggering it, then we can look if this is a hardware issue. > > > > The problem is that it happens only once during startup, especially at > > the very first startup after power-on only. So I can't issue any > > command. > > try to blacklist btusb.ko module. Create /etc/modprobe.d/blacklist-btusb.conf > with the content of "blacklist vc4”.
I hhink you mean "blacklist btusb" >Then once booted, start “btmon -w trace.log” and then “modprobe btusb”. This >should give you the initial firmware loading trace. > > I am just assuming that the module is connected via USB, if not then you need > to let me know. > > >> We have only seen this with Atheros hardware so far, but it might happen > >> for others as well. It indicates that this is an unexpected event. > >> Normally you can ignore this warning since it just indicates an existing > >> issue that we just papered over before. So if everything works as before, > >> just ignore it, > > > > Yes for me BT works as usual so ignoring it would be no problem (but > > it looks ugly because the error message is painted right on the > > boot-screen). > > The 0xfc2f command is never issued by btusb.c or btintel.c actually. It is a > command to apply the BDDATA information used only by Intel AG6xx devices > which are UART only. So I am almost certain that this is a bug in the > hardware / firmware and the patch above just started to highlight it. The > trace will show if that is the case. Done!. Here comes trace.log. Thank's Jörg
btsnoop Ñ $ $ÿÿ âh]ÈvÏ?Linux version 5.2.0-rc3-BT (x86_64) ! !ÿÿ âh]ÈvÏCBluetooth subsystem version 2.22 ÿÿ âh]ÈvÏj btmon âh]Éؼ hci0 âh]ÉØ âh]ÉØ âh]ÉØ | âh]ÉØ9¯ âh]ÉØ9Óü âh]ÉØ= ü 7 âh]ÉØ=ü âh]ÉØAwü âh]ÉØAü W âh]ÉØEhü ý ý âh]ÉØEyüú ! ô øÂÿá! Ê a ÑÀà~èH ÌH 0ZÐ ¢ÁV S Ø®=`0¢Ààx°H ô©pÈÿÔH Ð Ù (I ö Òà pXÓ`xàÀ¸àxH ¸%xäâIö! ñ ¸¡àx öàøàx H ã!ðàxê % ¯ !P àxH àxÒc exÂÐ ð Ø9À%H XÓ âh]ÉØM$ü ý ý âh]ÉØM9üúô ! ô¹,H Èà~àx4H áÇü´ è^¬» £,E# # þ¦t <S ]5 V 0H ÉpÁÇÁÆàÁÅ8H »^Ø ¼ ìØ H áÇü´üH´üð6Øb ÑdH ØØîñhH Àþè ÞJ#@ J&