Hi,

The following two patches fix race conditions related to ACPI GPEs and PM.

The first one if for a race between wholesale disabling of GPEs (during
system suspend and shutdown) and the handling of level-triggered GPEs.  The
race is that the latter may re-enable a GPE disabled by the former under
specific conditions, which is not good.

The second one is to prevent suspend-to-idle from leaking GPE handling
into the late code (that has been sent once already).

Thanks!

-- 
I speak only for myself.
Rafael J. Wysocki, Intel Open Source Technology Center.