Re: [PATCH] driverfs support for USB

[Greg KH <greg@kroah.com>]

On Mon, Jan 28, 2002 at 12:52:04PM +0000, Pavel Machek wrote:
> Hi!
> 
> > .
> >         |-- pci1
> >         |   |-- 01:08.0
> >         |   |   |-- power
> >         |   |   --- status
> >         |   |-- 01:0d.2
> >         |   |   |-- power
> >         |   |   --- status
> >         |   --- usb
> >         |       |-- 002
> >         |       |   |-- power
> >         |       |   --- status
> >         |       --- 003
> >         |           |-- power
> >         |           --- status
> 
> Should not usb bus hang off some pci device (uhci/ohci?)

You are correct, The "usb" device should hang off of the actual pci
device that is the USB host controller.  But due to the current split
between iobus and devices in the device core code, a usb bus has to
attach to the pci bus, which places it in the tree like you see here.

So to be correct the above portion of the tree should look something
like:
        |-- pci1
        |   |-- 01:08.0
        |   |   |-- power
        |   |   --- status
        |   |-- 01:0d.2
        |       |-- power
        |       |-- status
        |       --- usb
        |           |-- 002
        |               |-- power
        |               |-- status
        |               --- 003
        |                   |-- power
        |                   --- status


It's still a work in progress :)

thanks,

greg k-h