Greg: The following series addresses a minor problem caused by a change introduced in 3.8-rc1. Because the autosuspend delay for hubs is now set to 0, the PM core will try to runtime-suspend root hubs during an ongoing port resume (the root hub appears to be idle at such times). Although the attempted runtime suspend will fail -- the HCD rejects it because of the resuming port -- the PM core will simply try to do another suspend, over and over again, until the port is finished resuming and the root hub appears to be in use. We don't want all this futile, unnecessary activity to take place. The first patch in the series adds a mechanism whereby HCDs can inform usbcore about port resumes, allowing usbcore to prevent runtime suspends of the root hub. The second and third patches add the appropriate calls to ehci-hcd and uhci-hcd. No analogous change is needed in ohci-hcd because it generates interrupts when port resumes end rather than when they begin. However it's quite possible that xhci-hcd will need a similar change. I haven't looked at it yet. All these patches should be merged by 3.8-final, if timing permits. 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
