On Wed, 21 Oct 2009, Oliver Neukum wrote: > Am Mittwoch, 21. Oktober 2009 19:16:29 schrieb Alan Stern: > > > Just like that case, here the only modules which have all the knowledge > > > to make an informed decision are the upper layer modules. > > > > This is, of course, correct. But the infrastructure for the higher > > layers to signal us isn't in place yet. I was wondering if in the > > meantime, we could implement a stopgap approach that wouldn't be 100% > > right but would be better than nothing. > > It sounds seductive, but Matthew is right, it's not worth doing it twice. All right. I accept your and Matt's advice. > > Or should I just forget about it and work on implementing support for > > the new runtime PM framework in both USB and SCSI? > > Do you need to do it for USB? It seems to me you could implement > the hooks only for storage and use the new framework only for SCSI. > This is not as nice, but faster to do and very little work to reverse > and leads to a nicer migration path. Doing it for USB is likely to be a lot easier than changing SCSI. Much of the support already exists; most of the changes will be confined to usbcore and it's pretty clear what needs to be done there. With SCSI, adding runtime PM support will be a fundamental change and it's likely to provoke objections from some of the SCSI developers. 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
