Multiple race conditions are possible between PCI hotplug and the generic
PCI bus rescan and device removal that can be triggered via sysfs.

To avoid those race conditions make PCI hotplug use global PCI
rescan-remove locking.

Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

Modify struct acpi_dev_node to contain a pointer to struct acpi_device
associated with the given device object (that is, its ACPI companion
device) instead of an ACPI handle corresponding to it.  Introduce two
new macros for manipulating that pointer in a CONFIG_ACPI-safe way,
ACPI_COMPANION() and ACPI_COMPANION_SET(), and rework the
ACPI_HANDLE() macro to take the above changes into account.
Drop the ACPI_HANDLE_SET() macro entirely and rework its users to
use ACPI_COMPANION_SET() instead.  For some of them who used to
pass the result of acpi_get_child() directly to ACPI_HANDLE_SET()
introduce a helper routine acpi_preset_companion() doing an
equivalent thing.

The main motivation for doing this is that there are things
represented by struct acpi_device objects that don't have valid
ACPI handles (so called fixed ACPI hardware features, such as
power and sleep buttons) and we would like to create platform
device objects for them and "glue" them to their ACPI companions
in the usual way (which currently is impossible due to the
lack of valid ACPI handles).  However, there are more reasons
why it may be useful.

First, struct acpi_device pointers allow of much better type checking
than void pointers which are ACPI handles, so it should be more
difficult to write buggy code using modified struct acpi_dev_node
and the new macros.  Second, the change should help to reduce (over
time) the number of places in which the result of ACPI_HANDLE() is
passed to acpi_bus_get_device() in order to obtain a pointer to the
struct acpi_device associated with the given "physical" device,
because now that pointer is returned by ACPI_COMPANION() directly.
Finally, the change should make it easier to write generic code that
will build both for CONFIG_ACPI set and unset without adding explicit
compiler directives to it.

Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Tested-by: Mika Westerberg <mika.westerberg@linux.intel.com> # on Haswell
Reviewed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: Aaron Lu <aaron.lu@intel.com> # for ATA and SDIO part

Pull PCI changes from Bjorn Helgaas:
 "Host bridge hotplug
    - Major overhaul of ACPI host bridge add/start (Rafael Wysocki, Yinghai Lu)
    - Major overhaul of PCI/ACPI binding (Rafael Wysocki, Yinghai Lu)
    - Split out ACPI host bridge and ACPI PCI device hotplug (Yinghai Lu)
    - Stop caching _PRT and make independent of bus numbers (Yinghai Lu)

  PCI device hotplug
    - Clean up cpqphp dead code (Sasha Levin)
    - Disable ARI unless device and upstream bridge support it (Yijing Wang)
    - Initialize all hot-added devices (not functions 0-7) (Yijing Wang)

  Power management
    - Don't touch ASPM if disabled (Joe Lawrence)
    - Fix ASPM link state management (Myron Stowe)

  Miscellaneous
    - Fix PCI_EXP_FLAGS accessor (Alex Williamson)
    - Disable Bus Master in pci_device_shutdown (Konstantin Khlebnikov)
    - Document hotplug resource and MPS parameters (Yijing Wang)
    - Add accessor for PCIe capabilities (Myron Stowe)
    - Drop pciehp suspend/resume messages (Paul Bolle)
    - Make pci_slot built-in only (not a module) (Jiang Liu)
    - Remove unused PCI/ACPI bind ops (Jiang Liu)
    - Removed used pci_root_bus (Bjorn Helgaas)"

* tag 'pci-v3.9-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci: (51 commits)
  PCI/ACPI: Don't cache _PRT, and don't associate them with bus numbers
  PCI: Fix PCI Express Capability accessors for PCI_EXP_FLAGS
  ACPI / PCI: Make pci_slot built-in only, not a module
  PCI/PM: Clear state_saved during suspend
  PCI: Use atomic_inc_return() rather than atomic_add_return()
  PCI: Catch attempts to disable already-disabled devices
  PCI: Disable Bus Master unconditionally in pci_device_shutdown()
  PCI: acpiphp: Remove dead code for PCI host bridge hotplug
  PCI: acpiphp: Create companion ACPI devices before creating PCI devices
  PCI: Remove unused "rc" in virtfn_add_bus()
  PCI: pciehp: Drop suspend/resume ENTRY messages
  PCI/ASPM: Don't touch ASPM if forcibly disabled
  PCI/ASPM: Deallocate upstream link state even if device is not PCIe
  PCI: Document MPS parameters pci=pcie_bus_safe, pci=pcie_bus_perf, etc
  PCI: Document hpiosize= and hpmemsize= resource reservation parameters
  PCI: Use PCI Express Capability accessor
  PCI: Introduce accessor to retrieve PCIe Capabilities Register
  PCI: Put pci_dev in device tree as early as possible
  PCI: Skip attaching driver in device_add()
  PCI: acpiphp: Keep driver loaded even if no slots found
  ...

This changeset is aimed at fixing a few different but related
problems in the ACPI hotplug infrastructure.

First of all, since notify handlers may be run in parallel with
acpi_bus_scan(), acpi_bus_trim() and acpi_bus_hot_remove_device()
and some of them are installed for ACPI handles that have no struct
acpi_device objects attached (i.e. before those objects are created),
those notify handlers have to take acpi_scan_lock to prevent races
from taking place (e.g. a struct acpi_device is found to be present
for the given ACPI handle, but right after that it is removed by
acpi_bus_trim() running in parallel to the given notify handler).
Moreover, since some of them call acpi_bus_scan() and
acpi_bus_trim(), this leads to the conclusion that acpi_scan_lock
should be acquired by the callers of these two funtions rather by
these functions themselves.

For these reasons, make all notify handlers that can handle device
addition and eject events take acpi_scan_lock and remove the
acpi_scan_lock locking from acpi_bus_scan() and acpi_bus_trim().
Accordingly, update all of their users to make sure that they
are always called under acpi_scan_lock.

Furthermore, since eject operations are carried out asynchronously
with respect to the notify events that trigger them, with the help
of acpi_bus_hot_remove_device(), even if notify handlers take the
ACPI scan lock, it still is possible that, for example,
acpi_bus_trim() will run between acpi_bus_hot_remove_device() and
the notify handler that scheduled its execution and that
acpi_bus_trim() will remove the device node passed to
acpi_bus_hot_remove_device() for ejection.  In that case, the struct
acpi_device object obtained by acpi_bus_hot_remove_device() will be
invalid and not-so-funny things will ensue.  To protect agaist that,
make the users of acpi_bus_hot_remove_device() run get_device() on
ACPI device node objects that are about to be passed to it and make
acpi_bus_hot_remove_device() run put_device() on them and check if
their ACPI handles are not NULL (make acpi_device_unregister() clear
the device nodes' ACPI handles for that check to work).

Finally, observe that acpi_os_hotplug_execute() actually can fail,
in which case its caller ought to free memory allocated for the
context object to prevent leaks from happening.  It also needs to
run put_device() on the device node that it ran get_device() on
previously in that case.  Modify the code accordingly.

Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Yinghai Lu <yinghai@kernel.org>

* pci/acpi-scan2:
  ACPI / scan: Drop acpi_bus_add() and use acpi_bus_scan() instead
  ACPI: update ej_event interface to take acpi_device
  ACPI / scan: Add second pass to acpi_bus_trim()
  ACPI / scan: Change the implementation of acpi_bus_trim()
  ACPI / scan: Drop the second argument of acpi_bus_trim()
  ACPI / scan: Drop the second argument of acpi_device_unregister()
  ACPI: Remove the ops field from struct acpi_device
  ACPI: remove unused acpi_op_bind and acpi_op_unbind
  ACPI / scan: Fix check of device_attach() return value.

* pci/yijing-ari:
  PCI: shpchp: Iterate over all devices in slot, not functions 0-7
  PCI: sgihp: Iterate over all devices in slot, not functions 0-7
  PCI: cpcihp: Iterate over all devices in slot, not functions 0-7
  PCI: pciehp: Iterate over all devices in slot, not functions 0-7
  PCI: Consolidate "next-function" functions
  PCI: Rename pci_enable_ari() to pci_configure_ari()
  PCI: Enable ARI if dev and upstream bridge support it; disable otherwise

Iterate through devices in a slot by using the upstream bridge's
"bus->devices" list instead of assuming they are functions 0-7.  It's
possible there are several slots on the same pci_bus, so restrict it to
only devices matching this slot's device number.

ARI (which allows functions 0-255) is a PCIe-only feature, and this is
a PCI hotplug driver, so we shouldn't find anything other than functions
0-7, but it's better to iterate the same way as other hotplug drivers.

[bhelgaas: changelog, check PCI_SLOT, fix disable_slot()]
Signed-off-by: Yijing Wang <wangyijing@huawei.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

The only difference between acpi_bus_scan() and acpi_bus_add() is the
invocation of acpi_update_all_gpes() in the latter which in fact is
unnecessary, because acpi_update_all_gpes() has already been called
by acpi_scan_init() and the way it is implemented guarantees the next
invocations of it to do nothing.

For this reason, drop acpi_bus_add() and make all its callers use
acpi_bus_scan() directly instead of it.  Additionally, rearrange the
code in acpi_scan_init() slightly to improve the visibility of the
acpi_update_all_gpes() call in there.

Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Yinghai Lu <yinghai@kernel.org>

All callers of acpi_bus_trim() pass 1 (true) as the second argument
of it, so remove that argument entirely and change acpi_bus_trim()
to always behave as though it were 1.

Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Toshi Kani <toshi.kani@hp.com>
Acked-by: Yinghai Lu <yinghai@kernel.org>
Acked-by: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>

The callers of acpi_bus_add() usually assume that if it has
succeeded, then a struct acpi_device object has been attached to
the handle passed as the first argument.  Unfortunately, however,
this assumption is wrong, because acpi_bus_scan(), and acpi_bus_add()
too as a result, may return a pointer to a different struct
acpi_device object on success (it may be an object corresponding to
one of the descendant ACPI nodes in the namespace scope below that
handle).

For this reason, the callers of acpi_bus_add() who care about
whether or not a struct acpi_device object has been created for
its first argument need to check that using acpi_bus_get_device()
anyway, so the second argument of acpi_bus_add() is not really
useful for them.  The same observation applies to acpi_bus_scan()
executed directly from acpi_scan_init().

Therefore modify the relevant callers of acpi_bus_add() to check the
existence of the struct acpi_device in question with the help of
acpi_bus_get_device() and drop the no longer necessary second
argument of acpi_bus_add().  Accordingly, modify acpi_scan_init() to
use acpi_bus_get_device() to get acpi_root and drop the no longer
needed second argument of acpi_bus_scan().

Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Yinghai Lu <yinghai@kernel.org>
Acked-by: Toshi Kani <toshi.kani@hp.com>