Re: kern/37279: usb and cdrom don't work after resume on PM branch

["Jared D. McNeill" <jmcneill%invisible.ca@localhost>]

The following reply was made to PR kern/37279; it has been noted by GNATS.

From: "Jared D. McNeill" <jmcneill%invisible.ca@localhost>
To: "Steven M. Bellovin" <smb%cs.columbia.edu@localhost>
Cc: gnats-bugs%NetBSD.org@localhost, kern-bug-people%NetBSD.org@localhost, 
gnats-admin%NetBSD.org@localhost, 
    netbsd-bugs%NetBSD.org@localhost
Subject: Re: kern/37279: usb and cdrom don't work after resume on PM branch
Date: Mon, 28 Jan 2008 11:09:36 -0500 (EST)

 Some more information on the USB issue.
 
 It appears that the Thinkpad's \_WAK method is responsible for powering 
 off the external ports. I haven't had success powering them back on, as 
 the ICH8 spec claims its EHCI doesn't support per-port power controls.
 
 Here are a few theories of what we can try:
 
    1. Before we call AcpiLeaveSleepState on resume (which is responsible
       for invoking \_WAK), look for all EHCI devices and if the legacy
       capability is present, acquire the semaphore (see
       ehci_pci.c:ehci_get_ownership). With any luck, this will prevent
       SMI from turning the ports off on us.
 
    2. Failing #1, we need to decypher what method call in \_WAK is
       triggering the ports to power off and see if there is an AML method
       we can invoke to reverse the effect.
 
 Cheers,
 Jared