After commit 852972acff8f10f3a15679be2059bb94916cba5d (ACPI: Disable
ASPM if the platform won't provide _OSC control for PCIe) control of
the PCIe Capability Structure is unconditionally requested by
acpi_pci_root_add(), which in principle may cause problems to
happen in two ways.  First, the BIOS may refuse to give control of
the PCIe Capability Structure if it is not asked for any of the
_OSC features depending on it at the same time.  Second, the BIOS may
assume that control of the _OSC features depending on the PCIe
Capability Structure will be requested in the future and may behave
incorrectly if that doesn't happen.  For this reason, control of
the PCIe Capability Structure should always be requested along with
control of any other _OSC features that may depend on it (ie. PCIe
native PME, PCIe native hot-plug, PCIe AER).

Rework the PCIe port driver so that (1) it checks which native PCIe
port services can be enabled, according to the BIOS, and (2) it
requests control of all these services simultaneously.  In
particular, this causes pcie_portdrv_probe() to fail if the BIOS
refuses to grant control of the PCIe Capability Structure, which
means that no native PCIe port services can be enabled for the PCIe
Root Complex the given port belongs to.  If that happens, ASPM is
disabled to avoid problems with mishandling it by the part of the
PCIe hierarchy for which control of the PCIe Capability Structure
has not been received.

Make it possible to override this behavior using 'pcie_ports=native'
(use the PCIe native services regardless of the BIOS response to the
control request), or 'pcie_ports=compat' (do not use the PCIe native
services at all).

Accordingly, rework the existing PCIe port service drivers so that
they don't request control of the services directly.

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

It is possible that the BIOS will not grant control of all _OSC
features requested via acpi_pci_osc_control_set(), so it is
recommended to negotiate the final set of _OSC features with the
query flag set before calling _OSC to request control of these
features.

To implement it, rework acpi_pci_osc_control_set() so that the caller
can specify the mask of _OSC control bits to negotiate and the mask
of _OSC control bits that are absolutely necessary to it.  Then,
acpi_pci_osc_control_set() will run _OSC queries in a loop until
the mask of _OSC control bits returned by the BIOS is equal to the
mask passed to it.  Also, before running the _OSC request
acpi_pci_osc_control_set() will check if the caller's required
control bits are present in the final mask.

Using this mechanism we will be able to avoid situations in which the
BIOS doesn't grant control of certain _OSC features, because they
depend on some other _OSC features that have not been requested.

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

There is the assumption in acpi_pci_osc_control_set() that it is
always sufficient to compare the mask of _OSC control bits to be
requested with the result of an _OSC query where all of the known
control bits have been checked.  However, in general, that need not
be the case.  For example, if an _OSC feature A depends on an _OSC
feature B and control of A, B plus another _OSC feature C is
requested simultaneously, the BIOS may return A, B, C, while it would
only return C if A and C were requested without B.

That may result in passing a wrong mask of _OSC control bits to an
_OSC control request, in which case the BIOS may only grant control
of a subset of the requested features.  Moreover, acpi_pci_run_osc()
will return error code if that happens and the caller of
acpi_pci_osc_control_set() will not know that it's been granted
control of some _OSC features.  Consequently, the system will
generally not work as expected.

Apart from this acpi_pci_osc_control_set() always uses the mask
of _OSC control bits returned by the very first invocation of
acpi_pci_query_osc(), but that is done with the second argument
equal to OSC_PCI_SEGMENT_GROUPS_SUPPORT which generally happens
to affect the returned _OSC control bits.

For these reasons, make acpi_pci_osc_control_set() always check if
control of the requested _OSC features will be granted before making
the final control request.  As a result, the osc_control_qry and
osc_queried members of struct acpi_pci_root are not necessary any
more, so drop them and remove the remaining code referring to them.

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

Make acpi_pci_query_osc() use an additional pointer argument to
return the mask of control bits obtained from the BIOS to the
caller.

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

Make acpi_pci_osc_control_set() attempt to find the handle of the
_OSC object under the given PCI root bridge object after verifying
that its second argument is correct and that there is a struct
acpi_pci_root object for the given root bridge handle, which is
more logical than the old code.

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Reviewed-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

The PCI SIG documentation for the _OSC OS/firmware handshaking interface
states:

"If the _OSC control method is absent from the scope of a host bridge
device, then the operating system must not enable or attempt to use any
features defined in this section for the hierarchy originated by the host
bridge."

The obvious interpretation of this is that the OS should not attempt to use
PCIe hotplug, PME or AER - however, the specification also notes that an
_OSC method is *required* for PCIe hierarchies, and experimental validation
with An Alternative OS indicates that it doesn't use any PCIe functionality
if the _OSC method is missing. That arguably means we shouldn't be using
MSI or extended config space, but right now our problems seem to be limited
to vendors being surprised when ASPM gets enabled on machines when other
OSs refuse to do so. So, for now, let's just disable ASPM if the _OSC
method doesn't exist or refuses to hand over PCIe capability control.

Acked-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Matthew Garrett <mjg@redhat.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

The acpi_pci_root structure contains all the individual items (acpi_device,
domain, bus number) we pass to pci_acpi_scan_root(), so just pass the
single acpi_pci_root pointer directly.

This will make it easier to add _CBA support later.  For _CBA, we need the
entire downstream bus range, not just the base bus number.  We have that in
the acpi_pci_root structure, so passing the pointer makes it available to
the arch-specific code.

Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Reviewed-by: Kenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: Len Brown <len.brown@intel.com>

Previously, we only saved the root bus number, i.e., the beginning of the
downstream bus range.  We now support IORESOURCE_BUS resources, so this
patch uses that to keep track of both the beginning and the end of the
downstream bus range.

It's important to know both the beginning and the end for supporting _CBA
(see PCI Firmware spec, rev 3.0, sec 4.1.3) and so we know the limits for
any possible PCI bus renumbering (we can't renumber downstream buses to be
outside the bus number range claimed by the host bridge).

It's clear from the spec that the bus range is supposed to be in _CRS, but
if we don't find it there, we'll assume [_BBN - 0xFF] or [0 - 0xFF].

Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Reviewed-by: Kenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: Len Brown <len.brown@intel.com>

percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files.  percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.

percpu.h -> slab.h dependency is about to be removed.  Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability.  As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.

  http://userweb.kernel.org/~tj/misc/slabh-sweep.py

The script does the followings.

* Scan files for gfp and slab usages and update includes such that
  only the necessary includes are there.  ie. if only gfp is used,
  gfp.h, if slab is used, slab.h.

* When the script inserts a new include, it looks at the include
  blocks and try to put the new include such that its order conforms
  to its surrounding.  It's put in the include block which contains
  core kernel includes, in the same order that the rest are ordered -
  alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
  doesn't seem to be any matching order.

* If the script can't find a place to put a new include (mostly
  because the file doesn't have fitting include block), it prints out
  an error message indicating which .h file needs to be added to the
  file.

The conversion was done in the following steps.

1. The initial automatic conversion of all .c files updated slightly
   over 4000 files, deleting around 700 includes and adding ~480 gfp.h
   and ~3000 slab.h inclusions.  The script emitted errors for ~400
   files.

2. Each error was manually checked.  Some didn't need the inclusion,
   some needed manual addition while adding it to implementation .h or
   embedding .c file was more appropriate for others.  This step added
   inclusions to around 150 files.

3. The script was run again and the output was compared to the edits
   from #2 to make sure no file was left behind.

4. Several build tests were done and a couple of problems were fixed.
   e.g. lib/decompress_*.c used malloc/free() wrappers around slab
   APIs requiring slab.h to be added manually.

5. The script was run on all .h files but without automatically
   editing them as sprinkling gfp.h and slab.h inclusions around .h
   files could easily lead to inclusion dependency hell.  Most gfp.h
   inclusion directives were ignored as stuff from gfp.h was usually
   wildly available and often used in preprocessor macros.  Each
   slab.h inclusion directive was examined and added manually as
   necessary.

6. percpu.h was updated not to include slab.h.

7. Build test were done on the following configurations and failures
   were fixed.  CONFIG_GCOV_KERNEL was turned off for all tests (as my
   distributed build env didn't work with gcov compiles) and a few
   more options had to be turned off depending on archs to make things
   build (like ipr on powerpc/64 which failed due to missing writeq).

   * x86 and x86_64 UP and SMP allmodconfig and a custom test config.
   * powerpc and powerpc64 SMP allmodconfig
   * sparc and sparc64 SMP allmodconfig
   * ia64 SMP allmodconfig
   * s390 SMP allmodconfig
   * alpha SMP allmodconfig
   * um on x86_64 SMP allmodconfig

8. percpu.h modifications were reverted so that it could be applied as
   a separate patch and serve as bisection point.

Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.

Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>