Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 15 Nov 2005 10:22:53 -0600
From:      Eric Anderson <anderson@centtech.com>
To:        Maksim Yevmenkin <maksim.yevmenkin@savvis.net>
Cc:        freebsd-bluetooth@freebsd.org
Subject:   Re: Bluetooth mouse breaks on -current
Message-ID:  <437A0B5D.1000205@centtech.com>
In-Reply-To: <4342F41A.1050706@savvis.net>
References:  <431F388E.2010204@centtech.com>	<43217D56.3050508@centtech.com>	<4321D6C4.6050300@savvis.net>	<432AC7E9.5060206@centtech.com> <433F4B50.1050108@centtech.com> <433F62F7.3060607@centtech.com> <43427296.6090404@centtech.com> <4342BC20.8000001@savvis.net> <4342CEEB.5010501@centtech.com> <4342F41A.1050706@savvis.net>

next in thread | previous in thread | raw e-mail | index | archive | help
Maksim Yevmenkin wrote:
> Eric,
> 
> [...]
> 
>>> this looks like the mouse is loosing the connection and trying to 
>>> re-connect.
>>
>>
>> Correct - the mouse would just stop working while I was using it, and 
>> this particular mouse needs a button click to tell it to re-connect 
>> (as if it was in power save mode).  The reconnects you see, are me 
>> clicking buttons to get it to come back alive.
> 
> 
> fine
> 
>>>>>> [died completely at that vkbd message]
>>>>>>
>>>>>> Oct  1 21:01:16 neutrino su: anderson to root on /dev/ttyp2
>>>>>> Oct  1 21:06:22 neutrino kernel: ng_hci_process_command_timeout: 
>>>>>> ubt0hci - unable to complete HCI command OGF=0x1, OCF=0x1. Timeout
>>>>>> Oct  1 21:06:22 neutrino kernel: ubt_request_complete2: ubt0 - 
>>>>>> Control request failed. TIMEOUT (15)
>>>>>> Oct  1 21:07:39 neutrino kernel: ng_hci_process_command_timeout: 
>>>>>> ubt0hci - unable to complete HCI command OGF=0x3, OCF=0x3. Timeout
>>>>>> Oct  1 21:07:39 neutrino kernel: ubt_request_complete2: ubt0 - 
>>>>>> Control request failed. TIMEOUT (15)
>>>
>>>
>>> this is bad. something screwed up at usb level. basically we are 
>>> trying to send commands to the local device (not mouse) and it does 
>>> not respond.
>>
>>
>> I agree - almost seems as though it's a faulty usb bluetooth adapter.
> 
> 
> well, may be. may be not. since the load is high during buildworld i 
> wonder if this has something to do with interrupt latency.
> 
>>> what is the model/make of the local bluetooth device (not mouse)?
>>
>>
>> It's a built-in bluetooth adapter in my Sony laptop:
>>
>> ubt0: ALPS UGX, rev 1.10/11.68, addr 3
>> ubt0: ALPS UGX, rev 1.10/11.68, addr 3
>> ubt0: Interface 0 endpoints: interrupt=0x81, bulk-in=0x82, bulk-out=0x2
>> ubt0: Interface 1 (alt.config 5) endpoints: isoc-in=0x83, 
>> isoc-out=0x3; wMaxPacketSize=49; nframes=6, buffer size=294
> 
> 
> thanks
> 
>>> was buildworld still running?
>>
>>
>> Yes
>>
>>> if buildworld was running can you tell what the load was?
>>
>>
>> I can't recall what the load was at the time, but the CPU was close to 
>> 100% utilized.
> 
> 
> hmm...
> 
>>> what scheduler are you using BSD or ULE?
>>
>>
>> ULE
> 
> 
> can you please try BSD? just to see if it makes any difference. also try 
> to disable PREEMPTION (with both ULE and BSD).
> 
>>> are you running with WITNESS enabled?
>>
>>
>> No.
> 
> 
> good
> 
>>> also how fast is the cpu on your machine?
>>
>>
>> 1500MHZ.
> 
> 
> hmm... 1.5ghz, i wonder why your cpu is close to 100% during buildworld. 
> it would be nice to know how much time system spends in "sys" etc. you 
> have pretty fast cpu, so its kinda strange.
> 
>> Full dmesg, and other info here:
>>
>> http://www.googlebit.com/freebsd/
>>
>> Look for the most recent date.
> 
> 
> thanks
> 
>>>>>> Oct  1 21:09:33 neutrino kernel: ubt0: detached
>>>
>>>
>>> ok, this is really bad. usb device should not just detach by itself 
>>> (unless you pull it out).
>>
>>
>> Actually, that was me flipping the switch (enable/disable wireless, 
>> including bluetooth) on the laptop.
> 
> 
> ah, i see.

For what it's worth, I've been using a new Logitech Bluetooth mouse now 
for a month, with absolutely no problems.  I think it was the mouse 
causing all my troubles.

Thanks again for the great work (and also the recent rc.d commits!)

Eric




-- 
------------------------------------------------------------------------
Eric Anderson        Sr. Systems Administrator        Centaur Technology
Anything that works is better than anything that doesn't.
------------------------------------------------------------------------



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?437A0B5D.1000205>