Bug 62301 - Bluetooth userspace is gravely out of sync with kernelspace
Summary: Bluetooth userspace is gravely out of sync with kernelspace
Status: RESOLVED INVALID
Alias: None
Product: Drivers
Classification: Unclassified
Component: Bluetooth (show other bugs)
Hardware: All Linux
: P1 high
Assignee: linux-bluetooth@vger.kernel.org
URL: https://bugs.launchpad.net/ubuntu/+so...
Keywords:
Depends on:
Blocks:
 
Reported: 2013-09-28 22:22 UTC by Sergey Kondakov
Modified: 2013-12-11 11:52 UTC (History)
1 user (show)

See Also:
Kernel Version: 3.11.1
Subsystem:
Regression: No
Bisected commit-id:


Attachments

Description Sergey Kondakov 2013-09-28 22:22:04 UTC
The problem is: 
1) there is no place for tracking bugs in Linux Bluetooth userspace
As were said in bug #43841: "It's [this bug tracker] only used for tracking kernel bugs so it's not the right place [to be poking in bluez userspace]. I suspect you need to rattle cages and repost patches on the bluex mailing lists."
2) in only official place for Linux Bluetooth userspace discussion the interest is so lax, that willing contributors and distribution maintainers gave up long ago
As said in https://iwilcox.me.uk/2012/sixaxis-ubuntu, pertaining to Ubuntu's bug 124743 from the main link: "the linux-bluetooth list’s indifference and unresponsiveness is disheartening and demotivating enough that I can’t even see much point in raising the general bug"
See history of http://thread.gmane.org/gmane.linux.bluez.kernel/15308/ on how nice efforts go to waste.
3) which led to userspace not utilising kernel's abilities
so much so in fact, that in my example of Sony's PS3 controllers, proper, ready, mainline driver gets just plainly ignored and not loaded, rendering a device useless.

And all that in time, when Valve decided to reach out for Linux gaming and focus on on-couch-with-gamepad-in-hands PC gaming experience, where main input tools are Xbox360 and PS3 gamepads, with Valve's own creation just been announced. Not the best time for ignoring half of viable hardware of that purpose, just because not giving a damn became a habit at this point.

Personally, i can't utilise my DualShock 3 via Bluetooth on Fedora 19 either with the same problem described on those year old links:
[51505.042488] usb 4-4: new full-speed USB device number 2 using ohci-pci
[51505.243364] usb 4-4: New USB device found, idVendor=054c, idProduct=0268
[51505.243373] usb 4-4: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[51505.243378] usb 4-4: Product: PLAYSTATION(R)3 Controller
[51505.243382] usb 4-4: Manufacturer: Sony
[51505.877888] sony 0003:054C:0268.0004: Fixing up Sony Sixaxis report descriptor
[51505.925055] input: Sony PLAYSTATION(R)3 Controller as /devices/pci0000:00/0000:00:12.0/usb4/4-4/4-4:1.0/input/input15
[51505.925638] sony 0003:054C:0268.0004: input,hiddev0,hidraw3: USB HID v1.11 Joystick [Sony PLAYSTATION(R)3 Controller] on usb-0000:00:12.0-4/input0
"The cause, though not the root cause, was that the hid-core module only loads hid-blah modules when devices with corresponding registered vendor/product IDs appear. BlueZ passes 0000:0000 when the SIXAXIS connects over Bluetooth meaning that hid-sony never gets called."
[51564.526624] hid-generic 0005:0000:0000.0005: unknown main item tag 0x0
[51564.526730] input: PLAYSTATION(R)3 Controller as /devices/pci0000:00/0000:00:13.0/usb5/5-1/5-1:1.0/bluetooth/hci0/hci0:1/input16
[51564.526918] hid-generic 0005:0000:0000.0005: input,hidraw3: BLUETOOTH HID v0.00 Joystick [PLAYSTATION(R)3 Controller] on 00:11:67:d6:16:90

Maybe this example is too one-sided, being the exception, and Linux Bluetooth crowd just hates gaming, but that's the worst time for it to be this way.
And don't let me get started on BT headsets' LED screens and AVRCP not being utilised in Linux software. Even Android people had to rely on third-party patches for years for that to work.

Now, i'm just one disgruntled user, capable of at least comprehending the issue and writing here, which can only pity the poor bastards who still bear intentions to tackle these problems.
But the point is: kernel's BT stack is nothing without userspace properly utilising it and vice versa. Please, don't fail Linux gaming now. Wireless input is important.
Comment 1 Alan 2013-11-13 14:26:05 UTC
If its that bad then perhaps ubuntu needs to simply fork the bluetooth userspace. They've got the power to do that either themselves or with others, as has everyone else.

Still not a kernel problem for this bugzilla, and pointing it out here won't help.
Comment 2 Sergey Kondakov 2013-12-11 08:27:28 UTC
I don't see how Ubuntu can help anything, let alone this. But having official bugzilla for kernel's userspace such as bluez already would be nice.

Interestingly, with recently started major distributions' adoption, i found out that there is bluez5-branch that addresses _all of those issues_ (except the bugtracker). It even has AVRCP and Sony gamepad fixes !

Looks like time to break out champagne comes early this New Year's Eve.
Comment 3 Alan 2013-12-11 11:52:35 UTC
Good to hear but please note "bluez" is not "the kernel's bluetooth userspace", it's just one of several. In particular Android historically uses its own. In the kernel's eyes all are equals, none are "the kernels".

Enjoy the champagne 8)

Note You need to log in before you can comment on or make changes to this bug.