On Mon, Apr 27, 2009 at 01:16:17PM -0400, Alan Stern wrote: > On Mon, 27 Apr 2009, Greg KH wrote: > > > > The host controller device (which is the parent of the root hub device) > > > is the best choice. > > > > But isn't that the same thing as the usb_host device is today? Or am I > > totally confused? > > You're partially confused. :-) > > Here's an example. On my system, there's an EHCI controller at PCI > address 1d.7: > > /sys/devices/pci0000:00/0000:00:1d.7/ > > Inside that directory is the root hub: > > /sys/devices/pci0000:00/0000:00:1d.7/usb1/ > > But also inside it is the class device directory: > > /sys/devices/pci0000:00/0000:00:1d.7/usb_host/usb_host1/ > > The "companion" attribute is currently there, inside usb_host1. But it > could easily be moved up to 0000:00:1d.7. Ok, I've now created a patch to move this. > So IIUC, Kay would like to remove the usb_host and usb_host1 > subdirectories entirely. The symlinks under /sys/class can remain, but > their targets would be different. Right now I've got: > > $ ls -l /sys/class/usb_host/usb_host1 > lrwxrwxrwx 1 root root 0 2009-04-27 11:54 /sys/class/usb_host/usb_host1 -> ../../devices/pci0000:00/0000:00:1d.7/usb_host/usb_host1/ > > After this change the target of the symlink would be > > ../../devices/pci0000:00/0000:00:1d.7/ > > Does that clarify the situation? Hm, I just deleted the whole usb_host structure, and symlinks, as no one uses them anymore. I'll send the patches next. thanks, greg k-h -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
