On Wed, 6 May 2009, Oliver Neukum wrote: > Am Mittwoch, 6. Mai 2009 16:45:56 schrieb Alan Stern: > > However we ought to add reset_resume, pre_reset, and post_reset methods > > to usbfs_driver. Even if they don't do anything -- after all, there's > > no way to inform a user program that its device has been reset or > > suspended. Resets shouldn't be a problem, though, since the only > > entity capable of resetting a device that is owned by some program is > > the program itself. > > That I am afraid turns out not to be the case. Khubd resets hubs > as error handling, which propagates to the hub's children. I don't see > how you can virtualise this without kernel support. It doesn't "propagate" to the hub's children. Resetting a hub causes all the children to be disconnected (see hub_quiesce()). This will be handled properly if the owning program monitors kevents and checks for -ENODEV error codes. Alan Stern -- 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
