On Wednesday, September 28, 2011, Alan Stern wrote: > On Wed, 28 Sep 2011, Rafael J. Wysocki wrote: > > > > > That's possible. Moreover, it is possible that the BIOS removes power > > > > from the bus segment the device is on entirely, in which case it would > > > > need the auxiliary power source to generate wakeup events. > > > > > > Is there any way to find out whether this is happening? > > > > Frankly, I don't know. > > > > We may be entering an area that's never been tested by any vendors > > in this case. > > Probably not. Still, I thought perhaps a little special-purpose > debugging code at the right spot in the appropriate PCI early-resume > routine might be able to read the PME# status before the device gets > reset or its config space gets overwritten. (Although if the BIOS > messes around with the controller, it might destroy what I'm looking > for. Maybe disabling USB support in the BIOS will help...) > > What's the earliest spot in the resume path where I can read the > PME# status? In pci_restore_standard_config(), before pci_restore_state() is called. Thanks, Rafael -- To unsubscribe from this list: send the line "unsubscribe linux-pci" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
