The attached patch simplifies 29df8d8f.  As
the "pnp_xxx" structs are not designed to cope with IORESOURCE_DISABLED, and
hence no code can test for this value, setting this value is actually a "no op"
and can be skipped altogether.  It is sufficient to remove the checks for
"empty" resources and continue processing.

The patch is applied against 3.1.
Signed-off-by: NWitold Szczeponik <Witold.Szczeponik@gmx.net>
Acked-by: NBjorn Helgaas <bhelgaas@google.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

These macros are no longer in module.h and module.h is no longer
present everywhere.  Call out export.h for the users who are
making use of these macros.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

When parsing PnP ACPI resource structures, it may happen that some of
the resources are disabled (in which case "the size" of the resource
equals zero).

The current solution is to skip these resources completely - with the
unfortunate side effect that they are not registered despite the fact
that they exist, after all.  (The downside of this approach is that
these resources cannot be used as templates for setting the actual
device's resources because they are missing from the template.) The
kernel's APM implementation does not suffer from this problem and
registers all resources regardless of "their size".

This patch fixes a problem with (at least) the vintage IBM ThinkPad 600E
(and most likely also with the 600, 600X, and 770X which have a very
similar layout) where some of its PnP devices support options where
either an IRQ, a DMA, or an IO port is disabled.  Without this patch,
the devices can not be configured using the
"/sys/bus/pnp/devices/*/resources" interface.

The manipulation of these resources is important because the 600E has
very demanding requirements.  For instance, the number of IRQs is not
sufficient to support all devices of the 600E.  Fortunately, some of the
devices, like the sound card's MPU-401 UART, can be configured to not
use any IRQ, hence freeing an IRQ for a device that requires one.
(Still, the device's "ResourceTemplate" requires an IRQ resource
descriptor which cannot be created if the resource has not been
registered in the first place.)

As an example, the dependent sets of the 600E's CSC0103 device (the
MPU-401 UART) are listed, with the patch applied, as:

  Dependent: 00 - Priority preferred
    port 0x300-0x330, align 0xf, size 0x4, 16-bit address decoding
    irq <none> High-Edge
  Dependent: 01 - Priority acceptable
    port 0x300-0x330, align 0xf, size 0x4, 16-bit address decoding
    irq 5,7,2/9,10,11,15 High-Edge

(The same result is obtained when PNPBIOS is used instead of PnP ACPI.)
Without the patch, the IRQ resource in the preferred option is not
listed at all:

  Dependent: 00 - Priority preferred
    port 0x300-0x330, align 0xf, size 0x4, 16-bit address decoding
  Dependent: 01 - Priority acceptable
    port 0x300-0x330, align 0xf, size 0x4, 16-bit address decoding
    irq 5,7,2/9,10,11,15 High-Edge

And in fact, the 600E's DSDT lists the disabled IRQ as an option, as can
be seen from the following excerpt from the DSDT:

	Name (_PRS, ResourceTemplate ()
	{
        StartDependentFn (0x00, 0x00)
        {
            IO (Decode16, 0x0300, 0x0330, 0x10, 0x04)
            IRQNoFlags () {}
        }
        StartDependentFn (0x01, 0x00)
        {
            IO (Decode16, 0x0300, 0x0330, 0x10, 0x04)
            IRQNoFlags () {5,7,9,10,11,15}
        }
        EndDependentFn ()
	})

With this patch applied, a user space program - or maybe even the kernel
- can allocate all devices' resources optimally.  For the 600E, this
means to find optimal resources for (at least) the serial port, the
parallel port, the infrared port, the MWAVE modem, the sound card, and
the MPU-401 UART.

The patch applies the idea to register disabled resources to all types
of resources, not just to IRQs, DMAs, and IO ports.  At the same time,
it mimics the behavior of the "pnp_assign_xxx" functions from
"drivers/pnp/manager.c" where resources with "no size" are considered
disabled.

No regressions were observed on hardware that does not require this
patch.

The patch is applied against 2.6.39.

NB: The kernel's current PnP interface does not allow for disabling individual
resources using the "/sys/bus/pnp/devices/$device/resources" file.  Assuming
this could be done, a device could be configured to use a disabled resource
using a simple series of calls:

  echo disable > /sys/bus/pnp/devices/$device/resources
  echo clear > /sys/bus/pnp/devices/$device/resources
  echo set irq disabled > /sys/bus/pnp/devices/$device/resources
  echo fill > /sys/bus/pnp/devices/$device/resources
  echo activate > /sys/bus/pnp/devices/$device/resources

This patch addresses only the parsing of PnP ACPI devices.

ChangeLog (v1 -> v2):
 - extend patch description
 - fix typo in patch itself
Signed-off-by: NWitold Szczeponik <Witold.Szczeponik@gmx.net>
Cc: Len Brown <lenb@kernel.org>
Cc: Adam Belay <abelay@mit.edu>
Cc: Bjorn Helgaas <bjorn.helgaas@hp.com>
Cc: Bjorn Helgaas <bhelgaas@google.com>
Cc: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Cc: Matthew Garrett <mjg@redhat.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Several fixes as well where the +1 was missing.

Done via coccinelle scripts like:

@@
struct resource *ptr;
@@

- ptr->end - ptr->start + 1
+ resource_size(ptr)

and some grep and typing.

Mostly uncompiled, no cross-compilers.
Signed-off-by: NJoe Perches <joe@perches.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

The PNP ACPI driver squirrels the ACPI handles of PNP devices' ACPI
companions, but this isn't correct, because those handles should be
accessed using the DEVICE_ACPI_HANDLE() macro operating on struct
device objects.

Using DEVICE_ACPI_HANDLE() in the PNP ACPI driver instead of the
driver's own copies of the ACPI handles allows us to avoid a problem
with docking stations where a machine docked before suspend to RAM
and undocked while suspended crashes during the subsequent resume (in
that case the ACPI companion of the PNP device in question doesn't
exist any more while the device is being resumed).  It also allows us
to avoid the problem where suspend to RAM fails when the machine was
undocked while suspended before (again, the ACPI companion of the PNP
device is not present any more while it is being suspended).

This change doesn't fix all of the the PNP ACPI driver's problems
with PNP devices in docking stations (generally speaking, the driver
has no idea that devices can come and go and doesn't even attempt to
handle such events), but at least it makes suspend work for the
users of docking stations who don't use the PNP devices located in
there.

References: https://bugzilla.kernel.org/show_bug.cgi?id=15100Reported-and-tested-by: NToralf Förster <toralf.foerster@gmx.de>
Signed-off-by: NRafael J. Wysocki <rjw@sisk.pl>
Acked-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

Signed-off-by: NJan Beulich <jbeulich@novell.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

This is cleanup mostly, nothing urgent.
I came up with it when looking at dynamic debug which can
enable pr_debug messages at runtime or boot param
for a specific module.

Advantages:
  - Any pnp code can make use of the moduleparam.h interface, the modules
    will show up as pnp.param.
  - Passing pnp.ddebug as kernel boot param will enable all pnp debug messages
    with my previous patch and CONFIG_DYNAMIC_DEBUG enabled.
Signed-off-by: NThomas Renninger <trenn@suse.de>
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

Signed-off-by: NThomas Renninger <trenn@suse.de>
Signed-off-by: NLen Brown <len.brown@intel.com>

If primary ID (HID) is invalid try locating first valid ID on compatible
ID list before giving up.

This helps, for example, to recognize i8042 AUX port on Sony Vaio VPCZ1
which uses SNYSYN0003 as HID. Without the patch users are forced to
boot with i8042.nopnp to make use of their touchpads.
Tested-by: NJan-Hendrik Zab <jan@jhz.name>
Signed-off-by: NDmitry Torokhov <dtor@mail.ru>
Signed-off-by: NLen Brown <len.brown@intel.com>

This patch (as1354) adds remote-wakeup support to the pnpacpi driver.
The new can_wakeup method also allows other PNP protocol drivers
(pnpbios or iaspnp) to add wakeup support, but I don't know enough
about how they work to actually do it.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
Reviewed-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NRafael J. Wysocki <rjw@sisk.pl>

ACPI _CRS Address Space Descriptors have _MIN, _MAX, and _LEN.  Linux has
been computing Address Spaces as [_MIN to _MIN + _LEN - 1].  Based on the
tests in the bug reports below, Windows apparently uses [_MIN to _MAX].

Per spec (ACPI 4.0, Table 6-40), for _CRS fixed-size, fixed location
descriptors, "_LEN must be (_MAX - _MIN + 1)", and when that's true, it
doesn't matter which way we compute the end.  But of course, there are
BIOSes that don't follow this rule, and we're better off if Linux handles
those exceptions the same way as Windows.

This patch makes Linux use [_MIN to _MAX], as Windows seems to do.  This
effectively reverts 3162b6f0 and replaces it with simpler code.

    https://bugzilla.kernel.org/show_bug.cgi?id=14337 (round)
    https://bugzilla.kernel.org/show_bug.cgi?id=15480 (truncate)
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

The ACPI spec (sec 6.4.3.5 in v4.0) requires that for Address Space Resource
Descriptors, _LEN <= _MAX - _MIN + 1 in all cases, but there are BIOSes that
violate this.  We experimentally determined that Windows truncates the
resource so it doesn't extend past _MAX, so let's do the same thing in
Linux.

http://bugzilla.kernel.org/show_bug.cgi?id=15480Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Acked-by: NRafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: NLen Brown <len.brown@intel.com>

include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h

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

Add support for bus number resources.  This is for bridges with a range of
bus numbers behind them.  Previously, PNP ignored bus number resources.
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

Add support for resource windows.  This is for bridge resources, i.e.,
regions where a bridge forwards transactions from the primary to the
secondary side.  This does not add support for *setting* windows via
the /proc interface.
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

Add pnp_acpi_device(pnp_dev), which takes a PNP device and returns the
associated ACPI device (or NULL, if the device is not a PNPACPI device).

This allows us to write a PNP driver that can manage both traditional
PNPBIOS and ACPI devices, treating ACPI-only functionality as an optional
extension.
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

Some drivers need to look at things in the acpi_device structure
besides the handle.
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

Every acpi_device has at least one ID (if there's no _HID or _CID, we
give it a synthetic or default ID).  So there's no longer a need to
check whether an ID exists; we can just use it.
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

There's no need to treat _HID and _CID differently.  Keeping them in
a single list makes code that uses the IDs a little simpler because it
can just traverse the list rather than checking "do we have a HID?",
"do we have any CIDs?"
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Reviewed-by: NAlex Chiang <achiang@hp.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

Use acpi_device_hid() rather than accessing acpi_device.pnp.hardware_id
directly.
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

Completed a major update for the acpi_get_object_info external interface.
Changes include:
 - Support for variable, unlimited length HID, UID, and CID strings
 - Support Processor objects the same as Devices (HID,UID,CID,ADR,STA, etc.)
 - Call the _SxW power methods on behalf of a device object
 - Determine if a device is a PCI root bridge
 - Change the ACPI_BUFFER parameter to ACPI_DEVICE_INFO.
These changes will require an update to all callers of this interface.
See the ACPICA Programmer Reference for details.

Also, update all invocations of acpi_get_object_info interface
Signed-off-by: NBob Moore <robert.moore@intel.com>
Signed-off-by: NLin Ming <ming.m.lin@intel.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

Extended Address Space Descriptors are new in ACPI 3.0 and allow the
BIOS to communicate device resource cacheability attributes (write-back,
write-through, uncacheable, etc) to the OS.

Previously, PNPACPI ignored these descriptors, so if a BIOS used them,
a device could be responding at addresses the OS doesn't know about.
This patch adds support for these descriptors in _CRS and _PRS.  We
don't attempt to encode them for _SRS (just like we don't attempt to
encode the existing 16-, 32-, and 64-bit Address Space Descriptors).

Unfortunately, I don't have a way to test this.
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

6328a574
"Enable PNPACPI _PSx Support, v3"

added a call to acpi_bus_set_power(handle, ACPI_STATE_D3)
to pnpacpi_disable_resource() before the existing call
to evaluate _DIS on the device.

This caused suspend to fail on the system in
http://bugzilla.kernel.org/show_bug.cgi?id=13243
because the sanity check to verify we entered _PS3
failed on the serial port.

As a work-around, that sanity check can be disabled
system-wide with "acpi.power_nocheck=1"

Or perhaps we should just shrug off the _PS3 failure
and carry on with _DIS like we used to -- which is
what this patch does.
Signed-off-by: NLen Brown <len.brown@intel.com>

We want to use dev_to_node() later on, to be aware of the 'home node'
of the GSI in question.

[ Impact: cleanup, prepare the IRQ code to be more NUMA aware ]
Signed-off-by: NYinghai Lu <yinghai@kernel.org>
Acked-by: NLen Brown <lenb@kernel.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Len Brown <lenb@kernel.org>
Cc: Bjorn Helgaas <bjorn.helgaas@hp.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: linux-acpi@vger.kernel.org
Cc: linux-ia64@vger.kernel.org
LKML-Reference: <49F65560.20904@kernel.org>
Signed-off-by: NIngo Molnar <mingo@elte.hu>

(This is an update to the patch presented earlier in
http://lkml.org/lkml/2008/12/8/284, with new error handling.)

This patch sets the power of PnP ACPI devices to D0 when they
are activated and to D3 when they are disabled.  The latter is
in correspondence with the ACPI 3.0 specification, whereas the
former is added in order to be able to power up a device after
it has been previously disabled (or when booting up a system).
(As a consequence, the patch makes the PnP ACPI code more ACPI
compliant.)

Section 6.2.2 of the ACPI Specification (at least versions 1.0b
and 3.0a) states: "Prior to running this control method [_DIS],
the OS[PM] will have already put the device in the D3 state."
Unfortunately, there is no clear statement as to when to put
a device in the D0 state. :-( Therefore, the patch executes the
method calls as _PS3/_DIS and _SRS/_PS0. What is clear: "If the
device is disabled, _SRS enables the device at the specified
resources." (From the ACPI 3.0a Specification.)

The patch fixes a problem with some IBM ThinkPads (at least the
600E and the 600X) where the serial ports have a dedicated
power source that needs to be brought up before the serial port
can be used.  Without this patch, the serial port is enabled
but has no power. (In the past, the tpctl utility had to be
utilized to turn on the power, but support for this feature
stopped with version 5.9 as it did not support the more recent
kernel versions.)

The error handlers that handle any errors that can occur during
the power up/power down phases return the error codes to the
caller directly.  Comments welcome! :-)

No regressions were observed on hardware that does not require
this patch.

The patch is applied against 2.6.27.x.
Signed-off-by: NWitold Szczeponik <Witold.Szczeponik@gmx.net>
Acked-by: NZhao Yakui <yakui.zhao@intel.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

External driver files should not include any private acpica headers.
Signed-off-by: NLin Ming <ming.m.lin@intel.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

According to ACPI spec when the status of some device is not present
but functional, the device is valid and the children of this device
should be enumerated. It means that the device should be added to
linux acpi device tree. But the device driver for this device should not
be loaded.
    The detailed info can be found in the section 6.3.7 of ACPI 3.0b spec.
    _STA may return bit 0 clear (not present) with bit 3 set (device is
functional). This case is used to indicate a valid device for which no
device driver should be loaded (for example, a bridge device.).
Children of this device may be present and valid. OS should continue
enumeration below a device whose _STA returns this bit combination

http://bugzilla.kernel.org/show_bug.cgi?id=3358Signed-off-by: NZhao Yakui <yakui.zhao@intel.com>
Signed-off-by: NLi Shaohua <shaohua.li@intel.com>
Signed-off-by: NZhang Rui <rui.zhang@intel.com>
Signed-off-by: NAndi Kleen <ak@linux.intel.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

CONFIG_PNP_DEBUG is no longer used to turn on dev_dbg() in PNP,
since we have pnp_dbg() which can be enabled at boot-time, so
this patch removes the config option.

Note that pnp_dock_event() checks "#ifdef DEBUG".  But there's
never been a clear path for enabling that via configgery.  It
happened that CONFIG_PNP_DEBUG enabled it after 1bd17e63,
but that was accidental and only in 2.6.26.
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NAndi Kleen <ak@linux.intel.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

pnp_dbg() is equivalent to dev_dbg() except that we can turn it
on at boot-time with the "pnp.debug" kernel parameter, so we don't
have to build a new kernel image.
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NAndi Kleen <ak@linux.intel.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

There are only a few remaining uses of pnp_info(), so I just
converted them to printk and removed the pnp_err(), pnp_info(),
pnp_warn(), and pnp_dbg() wrappers.

I also removed a couple debug messages that don't seem useful any
more ("driver registered", "driver unregistered", "driver attached").
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NAndi Kleen <ak@linux.intel.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

Use dev_printk() when possible for more informative error messages.
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NAndi Kleen <ak@linux.intel.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

We already did that a long time ago for pnp_system_init, but
pnpacpi_init and pnpbios_init remained as subsys_initcalls, and get
linked into the kernel before the arch-specific routines that finalize
the PCI resources (pci_subsys_init).

This means that the PnP routines would either register their resources
before the PCI layer could, or would be unable to check whether a PCI
resource had already been registered.  Both are problematic.

I wanted to do this before 2.6.27, but every time we change something
like this, something breaks.  That said, _every_ single time we trust
some firmware (like PnP tables) more than we trust the hardware itself
(like PCI probing), the problems have been worse.
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

The Extended Interrupt descriptor has a producer/consumer bit, but
it's not clear what that would mean, and existing BIOSes use the bit
inconsistently.  This patch makes Linux PNPACPI ignore the bit.

The ACPI spec contains examples of PCI Interrupt Link devices marked
as ResourceProducers, but many BIOSes mark them as ResourceConsumers.

I also checked with a Windows contact, who said:

    Windows uses only "resource consumer" when dealing with
    interrupts.  There's no useful way of looking at a resource
    producer of interrupts.

    ... NT-based Windows largely infers the producer/consumer stuff
    from the device type and ignores the bits in the namespace.  This
    was necessary because Windows 98 ignored them and early namespaces
    contained random junk.

The reason I want to change this is because if PNPACPI devices exclude
ResourceProducer IRQ resources, we can't write PNP drivers for those
devices.

For example, on machines such as the the HP rx7620, rx7640, rx8620,
rx8640, and Superdome, HPET interrupts are ResourceProducers.  The
HPET driver currently has to use acpi_bus_register_driver() and do its
own _CRS parsing, even though it requires absolutely no ACPI-specific
functionality.

It would be better if the HPET driver were a PNP driver and took
advantage of the _CRS parsing built into PNPACPI.

This producer/consumer check was originally added here:
    http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=2b8de5f50e4a302b83ebcd5b0120621336d50bd6

to fix this bug:
    http://bugzilla.kernel.org/show_bug.cgi?id=6292

However, the bug was related only to memory and I/O port resources,
where the distinction is sensible and important to Linux.  Given that
the distinction is muddled for IRQ resources, I think it was a mistake
to add the check there.
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NAndi Kleen <ak@linux.intel.com>

The HP CCSR descriptor describes MMIO address space that should appear
as a MEM resource.  This patch adds support for parsing these descriptors
in the _CRS data.

The visible effect of this is that these MEM resources will appear
in /sys/devices/pnp0/.../resources, which means that "lspnp -v" will
report it, user applications can use this to locate device CSR space,
and kernel drivers can use the normal PNP resource accessors to
locate them.
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NLen Brown <len.brown@intel.com>
Signed-off-by: NAndi Kleen <ak@linux.intel.com>

ISAPNP, PNPBIOS, and ACPI describe the "possible resource settings" of
a device, i.e., the possibilities an OS bus driver has when it assigns
I/O port, MMIO, and other resources to the device.

PNP used to maintain this "possible resource setting" information in
one independent option structure and a list of dependent option
structures for each device.  Each of these option structures had lists
of I/O, memory, IRQ, and DMA resources, for example:

  dev
    independent options
      ind-io0  -> ind-io1  ...
      ind-mem0 -> ind-mem1 ...
      ...
    dependent option set 0
      dep0-io0  -> dep0-io1  ...
      dep0-mem0 -> dep0-mem1 ...
      ...
    dependent option set 1
      dep1-io0  -> dep1-io1  ...
      dep1-mem0 -> dep1-mem1 ...
      ...
    ...

This data structure was designed for ISAPNP, where the OS configures
device resource settings by writing directly to configuration
registers.  The OS can write the registers in arbitrary order much
like it writes PCI BARs.

However, for PNPBIOS and ACPI devices, the OS uses firmware interfaces
that perform device configuration, and it is important to pass the
desired settings to those interfaces in the correct order.  The OS
learns the correct order by using firmware interfaces that return the
"current resource settings" and "possible resource settings," but the
option structures above doesn't store the ordering information.

This patch replaces the independent and dependent lists with a single
list of options.  For example, a device might have possible resource
settings like this:

  dev
    options
      ind-io0 -> dep0-io0 -> dep1->io0 -> ind-io1 ...

All the possible settings are in the same list, in the order they
come from the firmware "possible resource settings" list.  Each entry
is tagged with an independent/dependent flag.  Dependent entries also
have a "set number" and an optional priority value.  All dependent
entries must be assigned from the same set.  For example, the OS can
use all the entries from dependent set 0, or all the entries from
dependent set 1, but it cannot mix entries from set 0 with entries
from set 1.

Prior to this patch PNP didn't keep track of the order of this list,
and it assigned all independent options first, then all dependent
ones.  Using the example above, that resulted in a "desired
configuration" list like this:

  ind->io0 -> ind->io1 -> depN-io0 ...

instead of the list the firmware expects, which looks like this:

  ind->io0 -> depN-io0 -> ind-io1 ...
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NAndi Kleen <ak@linux.intel.com>
Acked-by: NRene Herman <rene.herman@gmail.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

When building resource options, ISAPNP and PNPBIOS set the priority
to something like "0x100 | PNP_RES_PRIORITY_ACCEPTABLE", but we
immediately mask off the 0x100 again in pnp_build_option(), so that
bit looks superfluous.

Thanks to Rene Herman <rene.herman@gmail.com> for pointing this out.
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NAndi Kleen <ak@linux.intel.com>
Acked-by: NRene Herman <rene.herman@gmail.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

ACPI Extended Interrupt Descriptors can encode 32-bit interrupt
numbers, so an interrupt number may exceed the size of the bitmap
we use to track possible IRQ settings.

To avoid corrupting memory, complain and ignore too-large interrupt
numbers.

There's similar code in pnpacpi_parse_irq_option(), but I didn't
change that because the small IRQ descriptor can only encode
IRQs 0-15, which do not exceed bitmap size.

In the future, we could handle IRQ numbers greater than PNP_IRQ_NR
by replacing the bitmap with a table or list.
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NAndi Kleen <ak@linux.intel.com>
Acked-by: NRene Herman <rene.herman@gmail.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

This patch moves all the option allocations (pnp_mem, pnp_port, etc)
into the pnp_register_{mem,port,irq,dma}_resource() functions.  This
will make it easier to rework the option data structures.

The non-trivial part of this patch is the IRQ handling.  The backends
have to allocate a local pnp_irq_mask_t bitmap, populate it, and pass
a pointer to pnp_register_irq_resource().
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NAndi Kleen <ak@linux.intel.com>
Acked-by: NRene Herman <rene.herman@gmail.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

This adds a typedef for the IRQ bitmap, which should cause
no functional change, but will make it easier to pass a
pointer to a bitmap to pnp_register_irq_resource().
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NAndi Kleen <ak@linux.intel.com>
Acked-by: NRene Herman <rene.herman@gmail.com>
Signed-off-by: NLen Brown <len.brown@intel.com>

PNP previously defined PNP_PORT_FLAG_16BITADDR and PNP_PORT_FLAG_FIXED
in a private header file, but put those flags in struct resource.flags
fields.  Better to make them IORESOURCE_IO_* flags like the existing
IRQ, DMA, and MEM flags.
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NAndi Kleen <ak@linux.intel.com>
Acked-by: NRene Herman <rene.herman@gmail.com>
Signed-off-by: NLen Brown <len.brown@intel.com>