[BUG] Regression in v2.6.28 introduced by: 'USB: skip Set-Interface(0) if already in altsetting 0'

[BUG] Regression in v2.6.28 introduced by: 'USB: skip Set-Interface(0) if already in altsetting 0'

[Jan Scholz]

Hi,

In v2.6.28 my usb bluetooth device disappears after suspend to ram:

$ hcitool dev
Devices:
	hci0	00:0D:93:15:D6:E8
$ echo mem > /sys/power/state 
$ hcitool dev
Devices:
$

The problem does not occur with the (older) hci_usb.ko, but only with
btusb.ko.
 
This is the output of lsusb:
Bus 001 Device 003: ID 05ac:8203 Apple Computer, Inc. Bluetooth HCI

Bisecting finds commit 24c0996 (USB: skip Set-Interface(0) if already in
altsetting 0) as the first bad commit, and indeed if I use 2.6.28 with
24c0996 reverted, the bluetooth device survives suspend to ram.

If you need more information/test/etc please let me know.

Regards,
Jan

> commit 24c0996a6b73e2554104961afcc8659534503e0d
> Author: Alan Stern <stern@rowland.harvard.edu>
> Date:   Mon Dec 1 10:24:41 2008 -0500
>
>     USB: skip Set-Interface(0) if already in altsetting 0
>
>     When a driver unbinds from an interface, usbcore always sends a
>     Set-Interface request to reinstall altsetting 0.  Unforunately, quite
>     a few devices have buggy firmware that crashes when it receives this
>     request.
>
>     To avoid such problems, this patch (as1180) arranges to send the
>     Set-Interface request only when the interface is not already in
>     altsetting 0.
>
>     Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
>     Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

--
 Jan Scholz                                ____  ____    __    ___ 
                                          ( ___)(_  _)  /__\  / __)
 Frankfurt Institute for Advanced Studies  )__)  _)(_  /(__)\ \__ \
                                          (__)  (____)(__)(__)(___/
 Goethe Universitaet Frankfurt
 Ruth-Moufang-Str. 1                             Tel. 069-798-47534
 60438 Frankfurt am Main

Re: [BUG] Regression in v2.6.28 introduced by: 'USB: skip Set-Interface(0) if already in altsetting 0'

[Ingo Molnar]

* Jan Scholz <scholz@fias.uni-frankfurt.de> wrote:

> Hi,
> 
> In v2.6.28 my usb bluetooth device disappears after suspend to ram:
> 
> $ hcitool dev
> Devices:
> 	hci0	00:0D:93:15:D6:E8
> $ echo mem > /sys/power/state 
> $ hcitool dev
> Devices:
> $
> 
> The problem does not occur with the (older) hci_usb.ko, but only with
> btusb.ko.
>  
> This is the output of lsusb:
> Bus 001 Device 003: ID 05ac:8203 Apple Computer, Inc. Bluetooth HCI
> 
> Bisecting finds commit 24c0996 (USB: skip Set-Interface(0) if already in
> altsetting 0) as the first bad commit, and indeed if I use 2.6.28 with
> 24c0996 reverted, the bluetooth device survives suspend to ram.
> 
> If you need more information/test/etc please let me know.
> 
> Regards,
> Jan
> 
> > commit 24c0996a6b73e2554104961afcc8659534503e0d
> > Author: Alan Stern <stern@rowland.harvard.edu>
> > Date:   Mon Dec 1 10:24:41 2008 -0500
> >
> >     USB: skip Set-Interface(0) if already in altsetting 0
> >
> >     When a driver unbinds from an interface, usbcore always sends a
> >     Set-Interface request to reinstall altsetting 0.  Unforunately, quite
> >     a few devices have buggy firmware that crashes when it receives this
> >     request.
> >
> >     To avoid such problems, this patch (as1180) arranges to send the
> >     Set-Interface request only when the interface is not already in
> >     altsetting 0.
> >
> >     Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
> >     Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

here's a fix patch that appears to be dealing with aspects of this commit:

  http://lkml.org/lkml/2009/1/2/47

you might want to give it a try.

	Ingo

Re: [BUG] Regression in v2.6.28 introduced by: 'USB: skip Set-Interface(0) if already in altsetting 0'

[Jan Scholz]

Ingo Molnar <mingo@elte.hu> writes:

> * Jan Scholz <scholz@fias.uni-frankfurt.de> wrote:
>
>> Hi,
>> 
>> In v2.6.28 my usb bluetooth device disappears after suspend to ram:
>> 
>> $ hcitool dev
>> Devices:
>> 	hci0	00:0D:93:15:D6:E8
>> $ echo mem > /sys/power/state 
>> $ hcitool dev
>> Devices:
>> $
>> 
>> The problem does not occur with the (older) hci_usb.ko, but only with
>> btusb.ko.
>>  
>> This is the output of lsusb:
>> Bus 001 Device 003: ID 05ac:8203 Apple Computer, Inc. Bluetooth HCI
>> 
>> Bisecting finds commit 24c0996 (USB: skip Set-Interface(0) if already in
>> altsetting 0) as the first bad commit, and indeed if I use 2.6.28 with
>> 24c0996 reverted, the bluetooth device survives suspend to ram.
>> 
>> If you need more information/test/etc please let me know.
>> 
>> Regards,
>> Jan
>> 
>> > commit 24c0996a6b73e2554104961afcc8659534503e0d
>> > Author: Alan Stern <stern@rowland.harvard.edu>
>> > Date:   Mon Dec 1 10:24:41 2008 -0500
>> >
>> >     USB: skip Set-Interface(0) if already in altsetting 0
>> >
>> >     When a driver unbinds from an interface, usbcore always sends a
>> >     Set-Interface request to reinstall altsetting 0.  Unforunately, quite
>> >     a few devices have buggy firmware that crashes when it receives this
>> >     request.
>> >
>> >     To avoid such problems, this patch (as1180) arranges to send the
>> >     Set-Interface request only when the interface is not already in
>> >     altsetting 0.
>> >
>> >     Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
>> >     Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
>
> here's a fix patch that appears to be dealing with aspects of this commit:
>
>   http://lkml.org/lkml/2009/1/2/47
>
> you might want to give it a try.
>
> 	Ingo
>

Hi Ingo,

That patch is actually just 24c099 reverted, and as I wrote in the
previous post it indeed fixes the problem.

Thanks anyway,
   Jan