[PATCH v2 0/3] ACPI: Improve behaviour on Apple hardware

[andreas.noever@gmail.com]

From: Andreas Noever <andreas.noever@gmail.com>

This is a resend of Matthew's patches from https://lkml.org/lkml/2014/6/1/165
which are needed to fully support Thunderbolt on Apple hardware:
> Apple hardware behaves differently depending on whether or not the OS claims
> to be Darwin. Failing to report Darwin results in some hardware being
> disabled.  However, claiming to be Darwin also alters the behaviour of
> battery reporting and PCI handling. These patches add support for reporting
> Darwin support and fixing up the behavioural quirks that are exposed as a
> result.
Without these patches the firmware will cut power to the controller after
suspend (at the latest) and the thunderbolt driver will fail.

I have reordered them such that the two battery fixes/quirks come before the
_OSI change that breaks battery reporting. I have also merged "ACPI: Don't call
PCI OSC on Apple hardware when claiming to be Darwin" into "ACPI: Support
_OSI("Darwin") correctly" to avoid (temporarily) breaking hotplug events and
modified the patch to not touch ACPICA, as requested by Rafael.

Matthew Garrett (3):
  ACPI: Don't assume the existence of an SBS charger
  ACPI: Disable smart battery manager on Apple
  ACPI: Support _OSI("Darwin") correctly

 drivers/acpi/osl.c      | 10 +++++++
 drivers/acpi/pci_root.c | 14 +++++++++
 drivers/acpi/sbs.c      | 80 +++++++++++++++++++++++++++++++++++++++----------
 3 files changed, 89 insertions(+), 15 deletions(-)

-- 
2.1.0