Re: [gentoo-user] Re: Which USB device on which controller?

[Joshua Murphy <poisonbl@gmail.com>]

On Fri, Feb 27, 2009 at 3:21 PM, Grant <emailgrant@gmail.com> wrote:
>>>> I'm looking at the box of one of the webcams and it says "USB 2.0
>>>> compatible".
>>>>
>>>
>>> A USB 1.1 device _is_ "USB 2.0 compatible" because a USB 2.0
>>> will slow down and run at 1.1 speed.  Does the device say it's
>>> "high speed" USB?  "USB 2.0 compatible" generally means it's a
>>> USB 1.1 device.
>>>
>>>
>>>> It's not a cable problem because the cable is built right into
>>>> the webcam.  I'm not trying to get the webcam to go faster
>>>> back and forth to the controller, I just need to make sure I
>>>> don't have both webcams on the same OHCI (1.1) USB controller.
>>>> I would think buying a USB expansion card would work, but I
>>>> have an EHCI (2.0) controller on this system and a second OHCI
>>>> (1.1) controller.
>>>>
>>>> Does anyone have any idea on this.  It really doesn't make sense.
>>>>
>>>
>>> Sounds like it's a 1.1 device to me.
>>>
>>>
>>
>> Yep, that's what it sounds like to me too.
>>
>> Dale
>
> But that's OK isn't it?  I don't need 2.0 speeds between each webcam
> and the controller, I just need the increased overall bandwidth of a
> 2.0 controller so one of the 1.1 webcams doesn't use all of it.  I get
> the feeling I have a misconception somewhere along the line here.
> Could someone straighten me out?
>
> - Grant

The controller itself has the full bandwidth of the 2.0 spec, it's
just only able to allocate the bandwidth of the 1.0/1.1 spec to each
of the cameras because they're being recognized as 1.1 capable only.
The controller, though, I'm pretty sure is able to allocate more than
enough overall bandwidth to the pair. Based on your original remarks
that "one of them isn't functioning", I worry more that there's a
problem with the driver for the camera itself that somehow breaks with
two of the same camera present, broken udev rule somewhere that's not
creating the device for the second camera, or... *something* more
along those lines, rather than an issue with the on-chip
implementation of your usb controller.

-- 
Poison [BLX]
Joshua M. Murphy