Bug 42442

Summary: Atheros Bluetooth 0cf3:3002 is not working
Product: Drivers Reporter: AceLan Kao (acelan)
Component: BluetoothAssignee: drivers_bluetooth (drivers_bluetooth)
Status: NEW ---    
Severity: normal CC: a.cuadrado, alex.d.hall, gjakob, james, jm.leddy, jrnieder, kallisti5, rjw, sakhnik, scottsshort
Priority: P1    
Hardware: All   
OS: Linux   
Kernel Version: 3.1-rc4 Subsystem:
Regression: No Bisected commit-id:
Attachments: dmesg log
verbose lsusb output

Description AceLan Kao 2011-09-06 07:57:24 UTC
The Dell V130 comes with Atheros BT(ath3k) and Wifi(ath9k) chip, but the BT device is not working.
The USB ID is
   0cf3:3002 Atheros Communications, Inc.

The issue is related with bug 27402[1] I think, it's closed, but the problem still exists.

Is there any info that I could provide?

1. https://bugzilla.kernel.org/show_bug.cgi?id=27402
Comment 1 AceLan Kao 2011-09-06 07:58:21 UTC
Created attachment 71742 [details]
dmesg log
Comment 2 Jonathan Nieder 2012-01-27 21:47:10 UTC
Regression from 2.6.36, duplicate of bug 27402. Could some kind person with the relevant privileges please close this one and reopen that one?
Comment 3 James M. Leddy 2012-03-06 14:57:20 UTC
This turned out to be a problem with the firmware. It should be fixed by this modification to the firmware:

http://comments.gmane.org/gmane.linux.kernel/1262485
Comment 4 Alexander Hall 2012-04-04 20:56:32 UTC
Issue is still present on one machine....no issue on another...
Comment 5 Jonathan Nieder 2012-04-04 21:02:12 UTC
Does one machine have the firmware fix and another not?  Does one machine have this bluetooth chip and the other not?
Comment 6 Alexander Hall 2012-04-05 00:46:10 UTC
Both have the latest firmware available (Lbuntu and Ubuntu 12.04). The one that doesn't work (on boot) is recognized, but firmware does not load:

Bus 004 Device 002: ID 0cf3:3000 Atheros Communications, Inc. AR3011 Bluetooth (no firmware)

Relevant lines from dmesg:

[   25.965815] Bluetooth: Can't change to loading configuration err
[   25.965868] ath3k: probe of 4-2:1.0 failed with error -110
[   25.965934] usbcore: registered new interface driver ath3k
Comment 7 Alberto Cuadrado 2013-08-22 17:27:13 UTC
Finally this firmware works perfectly in my Dell Vostro V130:

https://bugs.launchpad.net/ubuntu/+source/linux-firmware/+bug/714862/+attachment/2832934/+files/ath3k-1.fw

I think that's the one in comment 3. It has the following md5sum (just in case the previous link breaks):

24cfee59a24f336dfae838a0de6445a0

But despite the fact it has been a long time since that fix was available, it seems it is not in the linux-firmware git repository (and that's why it is not in Debian either). Am I right? Could anybody please take a look at this?

By the way, the problem in comment 6 looks like solved and it had nothing to do with the firmware:

https://bugs.launchpad.net/libmtp/+bug/975222


And of course, for all the people involved in the solution of this problem, thank you very much.
Comment 8 Günther Jakob 2013-12-24 15:25:08 UTC
(In reply to Alberto Cuadrado from comment #7)
> 
> By the way, the problem in comment 6 looks like solved and it had nothing to
> do with the firmware:
>
I'm sure that it has also to do with the firmware bug, since I got exactly the same kernel error message (as reported in comment 6) as long as I was still using the ath3k file from the official kernel git that does not contain this patch even up to now! This firmware file with the md5sum value of 1211fa34c09e10ba48381586b7c3883d is still shipped with the kernels of several Fedora distribution versions (including the latest 20 version) and also provided as part of a corresponding package from ELRepo for RHEL and clones thereof.
>
> 
> https://bugs.launchpad.net/libmtp/+bug/975222
>
Where did you get the idea from that the version of this "media player library" should have an impact on the interaction between the low-level kernel driver and the firmware of that bluetooth adapter!?
After I had replaced the firmware file from the kernel git tree with the one containing the above-mentioned patch, my Atheros Bluetooth adapter worked properly - even with the very old version 1.0.1 of libmtp, which is part of  the Scientific Linux 6.x distribution that I am using...
Comment 9 Alberto Cuadrado 2013-12-30 16:11:06 UTC
(In reply to Günther Jakob from comment #8)
> (In reply to Alberto Cuadrado from comment #7)
> >
> > 
> > https://bugs.launchpad.net/libmtp/+bug/975222
> >
> Where did you get the idea from that the version of this "media player
> library" should have an impact on the interaction between the low-level
> kernel driver and the firmware of that bluetooth adapter!?

The author of comment 6 was later the reporter of that Ubuntu bug, with the same error messages, and there they found out that libmtp was to blame for those errors (I don't know how), not the firmware.

I'm just asking why the last available firmware (md5sum: 24cfee59a24f336dfae838a0de6445a0) is not in the linux-firmware git repo yet, if it hasn't any actual issue reports, while there are issues with the one currently in git (that's why this bug was opened).

> After I had replaced the firmware file from the kernel git tree with the one
> containing the above-mentioned patch, my Atheros Bluetooth adapter worked
> properly - even with the very old version 1.0.1 of libmtp, which is part of 
> the Scientific Linux 6.x distribution that I am using...

If you mean that 24cfee59a24f336dfae838a0de6445a0 works for you, then we agree, don't we? Maybe I didn't explain myself clearly in my previous comment.
Comment 10 Alexander von Gluck 2014-01-10 22:40:38 UTC
I can confirm seeing this bug in SteamOS as well (debian based)

After swapping to the firmware md5 "24cfee59a24f336dfae838a0de6445a0" the -110 issue disappeared for me as well.


Bus 007 Device 003: ID 0cf3:3005 Atheros Communications, Inc. AR3011 Bluetooth


Attaching detailed lsusb of device.
Comment 11 Alexander von Gluck 2014-01-10 22:41:24 UTC
Created attachment 121611 [details]
verbose lsusb output

Linux steamos 3.10-3-amd64 #1 SMP Debian 3.10.11-1st1 (2013-12-20) x86_64 GNU/Linux