413c:8140 [Dell Latitude D630] Bluetooth incorrectly removes input devices

[KarlRelton]

Linux kernels after 3.2 introduced battery status reporting for devices such as bluetooth keyboards and mouse.

Unfortunately there is an issue when such input devices are removed:

- a battery status report is requested
- this fails, because the device is being removed (communications with it are now down)
- the failure results in a 5 second wait (the code has a 5 second timeout)

- the bluetooth hci code has a race condition: the 5 second delay means one thread definitely loses the race
- the race condition leads to devices in the sysfs tree being removed in the wrong order
- this leads to udev events being sent out with 'truncated paths'
- this leads to userspace programs like Xorg evdev ignoring the event, but then getting confused about what keyboard/mice are actually attached
- this results in bluetooth keyboard/mice not working after a suspend/resume (because on resume the input devices are first removed, and then re-added)
---
ApportVersion: 2.6.1-0ubuntu9
Architecture: i386
AudioDevicesInUse:
 USER PID ACCESS COMMAND
 /dev/snd/controlC0: karl 2108 F.... pulseaudio
DistroRelease: Ubuntu 12.10
HibernationDevice: RESUME=UUID=a82d4a59-432d-4d4b-9d27-557f669eb637
MachineType: Dell Inc. Latitude D630
MarkForUpload: True
Package: linux (not installed)
PccardctlIdent:
 Socket 0:
   no product info available
PccardctlStatus:
 Socket 0:
   no card
ProcEnviron:
 TERM=xterm
 PATH=(custom, no user)
 XDG_RUNTIME_DIR=<set>
 LANG=en_GB.UTF-8
 SHELL=/usr/bin/tcsh
ProcFB: 0 inteldrmfb
ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-3.8.0-4-generic root=UUID=84c73eda-43e4-429c-8428-0a37de7b6de5 ro quiet splash vt.handoff=7
ProcVersionSignature: Ubuntu 3.8.0-4.8-generic 3.8.0-rc6
RelatedPackageVersions:
 linux-restricted-modules-3.8.0-4-generic N/A
 linux-backports-modules-3.8.0-4-generic N/A
 linux-firmware 1.95
Tags: quantal running-unity
Uname: Linux 3.8.0-4-generic i686
UpgradeStatus: Upgraded to quantal on 2012-10-23 (107 days ago)
UserGroups: adm admin audio cdrom dialout dip fax floppy lpadmin netdev plugdev powerdev scanner tape video
WpaSupplicantLog:

dmi.bios.date: 06/20/2008
dmi.bios.vendor: Dell Inc.
dmi.bios.version: A12
dmi.board.name: 0KU184
dmi.board.vendor: Dell Inc.
dmi.chassis.type: 8
dmi.chassis.vendor: Dell Inc.
dmi.modalias: dmi:bvnDellInc.:bvrA12:bd06/20/2008:svnDellInc.:pnLatitudeD630:pvr:rvnDellInc.:rn0KU184:rvr:cvnDellInc.:ct8:cvr:
dmi.product.name: Latitude D630
dmi.sys.vendor: Dell Inc.

[KarlRelton (karllinuxtest-relton)]

cat /proc/version_signature
Ubuntu 3.8.0-4.8-generic 3.8.0-rc6

[KarlRelton (karllinuxtest-relton)]

Upstream bug is https://bugzilla.kernel.org/show_bug.cgi?id=52471

Note two patches proposed there, which if applied together fix the problem (I have fixed it on my 3.8-rc6 kernel)

[KarlRelton (karllinuxtest-relton)]

I believe this bug is the real root cause of
https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-input-evdev/+bug/1075478

Xorg is not at fault, it is racy code (plus the hideous 5 second timeout) in the kernel

[KarlRelton (karllinuxtest-relton)]

Note also the race conditions in the bluetooth code can also lead to a kernel OOPS, as reported in
https://bugzilla.kernel.org/show_bug.cgi?id=50541

Removing the 5 second delay avoids that race condition however.

[Brad Figg (brad-figg)]

This bug is missing log files that will aid in diagnosing the problem. From a terminal window please run:

apport-collect 1117608

and then change the status of the bug to 'Confirmed'.

If, due to the nature of the issue you have encountered, you are unable to run this command, please add a comment stating that fact and change the bug status to 'Confirmed'.

This change has been made by an automated script, maintained by the Ubuntu Kernel Team.

[KarlRelton (karllinuxtest-relton)]

apport information

[KarlRelton (karllinuxtest-relton)]

apport information

[KarlRelton (karllinuxtest-relton)]

apport information

[KarlRelton (karllinuxtest-relton)]

apport information

[KarlRelton (karllinuxtest-relton)]

apport information

[KarlRelton (karllinuxtest-relton)]

apport information

[KarlRelton (karllinuxtest-relton)]

apport information

[KarlRelton (karllinuxtest-relton)]

apport information

[KarlRelton (karllinuxtest-relton)]

apport information

[KarlRelton (karllinuxtest-relton)]

apport information

[KarlRelton (karllinuxtest-relton)]

apport information

[KarlRelton (karllinuxtest-relton)]

apport information

[KarlRelton (karllinuxtest-relton)]

apport information

[KarlRelton (karllinuxtest-relton)]

apport information

[KarlRelton (karllinuxtest-relton)]

apport information

[KarlRelton (karllinuxtest-relton)]

Note the truncated device paths for the removal of the input/event devices, which comes after the removal of the 'hci' device.

The correct order should work from the bottom up of the tree (so hci device removed last).

[KarlRelton (karllinuxtest-relton)]

Here is a snippet of kern.log after a resume, with printk added to hidp_session() and hidp_get_raw_report().

When the connection is taken down, hidp_session() exists its main loop (prints 'out of loop' message).
It then does a hid_destroy_device() ... which invokes sysfs removal of the input device. BUT that invokes a battery lookup (in power_supply code), which calls back hidp_get_raw_report ... which stalls and waits for its 5 second timeout. Only then does hidp_session finally complete. Notice the 5 second time jump in this log snippet.

[penalvch (penalvch)]

KarlRelton, as per http://www.dell.com/support/troubleshooting/us/en/19/Product/latitude-d630 an update is available for your BIOS (A19). If you update to this, does it change anything?

If not, could you please both specify what happened, and provide the output of the following terminal command:
sudo dmidecode -s bios-version && sudo dmidecode -s bios-release-date

Thank you for your understanding.

[KarlRelton (karllinuxtest-relton)]

I would be very surprised if the Dell BIOS update changed anything.

Actually this bug can now be closed. Good bluetooth work was done in the kernel, which made it into the 3.10 tree. So Ubuntu 13.10 installations won't have this bug.