On Tue, 17 Mar 2009 14:20:56 -0700, Sarah Sharp <sarah.a.sharp@xxxxxxxxxxxxxxx> wrote: > I recently got a question about companion controllers, and their pros > and cons. I've heard from various people that companion controllers are > a PITA, but I don't know the details. Why are companion controllers > difficult to deal with from a software perspective? Big question, I > know. :0) They are ostensibly independent, yet not really. In Linux in particular things work much better if EHCI is loaded first. Naturally it's a pain to ensure it actually happens. Fortunately I don't deal with the suspend, Alan may fill you in on that. When I tried to implement a root-only suspend for RHEL-5, it was a hell on a stick, which never worked right. No matter what you do there are tradeoffs. Companions were in theory intended so an old OS could still drive the bus without having drivers for EHCI. The question is, is it worth the trouble? Intel had a reasonable success with an alternative approach in AHCI and combined mode. Actually you can say, AHCI has a "companion" taskfile (there's a bit that enables it). The difference is, we don't expect a newer OS to keep talking to the IDE taksfile, whereas in case of USB we do. That's where the problem is, I think. -- Pete -- 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
