Using GPIOs and gpiolib is optional. If the kernel is compiled without GPIO
support the driver should not fail if it finds the interrupt using normal
methods.

However, commit a485923e ("HID: i2c-hid: Add support for ACPI GPIO
interrupts") did not take into account that acpi_dev_add_driver_gpios()
returns -ENXIO when !CONFIG_GPIOLIB.

Fix this by checking the return value against -ENXIO and 0 and only in that
case fail the probe.
Reported-by: NGabriele Mazzotta <gabriele.mzt@gmail.com>
Signed-off-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

The HID over I2C specification allows to have the interrupt for a HID
device to be GPIO instead of directly connected to the IO-APIC.

Add support for this so that when the driver does not find proper interrupt
number from the I2C client structure we check if it has ACPI GpioInt()
resource listed in _CRS. If it is found we convert it to an interrupt
number and use it instead.
Signed-off-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

d1c7e29e (HID: i2c-hid: prevent buffer overflow in early IRQ)
changed hid_get_input() to read ihid->bufsize bytes, which can be
more than wMaxInputLength. This is the case with the Dell XPS 13
9343, and it is causing events to be missed. In some cases the
missed events are releases, which can cause the cursor to jump or
freeze, among other problems. Limit the number of bytes read to
min(wMaxInputLength, ihid->bufsize) to prevent such problems.

Fixes: d1c7e29e "HID: i2c-hid: prevent buffer overflow in early IRQ"
Signed-off-by: NSeth Forshee <seth.forshee@canonical.com>
Reviewed-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

The Microsoft HID over I2C specification says two things regarding the
interrupt:

 1) The interrupt should be level sensitive
 2) The device keeps the interrupt asserted as long as it has more reports
    available.

We've seen that at least some Atmel and N-Trig panels keep the line low as
long as they have something to send. The current version of the driver only
detects the first edge but then fails to read rest of the reports (as the
line is still asserted).

Make the driver follow the specification and configure the HID interrupt to
be level sensitive.

The Windows HID over I2C driver also seems to do the same.
Signed-off-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Acked-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

When a hid driver that uses i2c-hid as transport is unloaded, the hid core
will call i2c_hid_stop() which releases all the buffers associated with the
device. This includes also the command buffer.

Now, when the i2c-hid driver itself is unloaded it tries to power down the
device by sending it PWR_SLEEP command. Since the command buffer is already
released we get following crash:

 [   79.691459] BUG: unable to handle kernel NULL pointer dereference at           (null)
 [   79.691532] IP: [<ffffffffa05bc049>] __i2c_hid_command+0x49/0x310 [i2c_hid]
 ...
 [   79.693467] Call Trace:
 [   79.693494]  [<ffffffff810424e1>] ? __unmask_ioapic+0x21/0x30
 [   79.693537]  [<ffffffff81042855>] ? unmask_ioapic+0x25/0x40
 [   79.693581]  [<ffffffffa05bc35b>] ? i2c_hid_set_power+0x4b/0xa0 [i2c_hid]
 [   79.693632]  [<ffffffffa05bc3cf>] ? i2c_hid_runtime_resume+0x1f/0x30 [i2c_hid]
 [   79.693689]  [<ffffffff814c08fb>] ? __rpm_callback+0x2b/0x70
 [   79.693733]  [<ffffffff814c0961>] ? rpm_callback+0x21/0x90
 [   79.693776]  [<ffffffff814c0dec>] ? rpm_resume+0x41c/0x600
 [   79.693820]  [<ffffffff814c1e1c>] ? __pm_runtime_resume+0x4c/0x80
 [   79.693868]  [<ffffffff814b8588>] ? __device_release_driver+0x28/0x100
 [   79.693917]  [<ffffffff814b8d90>] ? driver_detach+0xa0/0xb0
 [   79.693959]  [<ffffffff814b82cc>] ? bus_remove_driver+0x4c/0xb0
 [   79.694006]  [<ffffffff810d1cfd>] ? SyS_delete_module+0x11d/0x1d0
 [   79.694054]  [<ffffffff8165f107>] ? int_signal+0x12/0x17
 [   79.694095]  [<ffffffff8165ee69>] ? system_call_fastpath+0x12/0x17

Fix this so that we only free buffers when the i2c-hid driver itself is
removed.

Fixes: 34f439e4 ("HID: i2c-hid: add runtime PM support")
Reported-by: NGabriele Mazzotta <gabriele.mzt@gmail.com>
Signed-off-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

Before ->start() is called, bufsize size is set to HID_MIN_BUFFER_SIZE,
64 bytes. While processing the IRQ, we were asking to receive up to
wMaxInputLength bytes, which can be bigger than 64 bytes.

Later, when ->start is run, a proper bufsize will be calculated.

Given wMaxInputLength is said to be unreliable in other part of the
code, set to receive only what we can even if it results in truncated
reports.
Signed-off-by: NGwendal Grignou <gwendal@chromium.org>
Reviewed-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

After commit b2b49ccb (PM: Kconfig: Set PM_RUNTIME if PM_SLEEP is
selected) PM_RUNTIME is always set if PM is set, so #ifdef blocks
depending on CONFIG_PM_RUNTIME may now be changed to depend on
CONFIG_PM.

Replace CONFIG_PM_RUNTIME with CONFIG_PM in drivers/hid/i2c-hid/i2c-hid.c.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NJiri Kosina <jkosina@suse.cz>
Reviewed-by: NMika Westerberg <mika.westerberg@linux.intel.com>

Current driver uses a common buffer for reading reports either
synchronously in i2c_hid_get_raw_report() and asynchronously in
the interrupt handler.
There is race condition if an interrupt arrives immediately after
the report is received in i2c_hid_get_raw_report(); the common
buffer is modified by the interrupt handler with the new report
and then i2c_hid_get_raw_report() proceed using wrong data.

Fix it by using a separate buffers for synchronous reports.
Signed-off-by: NJean-Baptiste Maneyrol <jmaneyrol@invensense.com>
[Antonio Borneo: cleanup, rebase to v3.17, submit mainline]
Signed-off-by: NAntonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

Report is received in "buffer"; fix the following i2c_hid_dbg()
to dump data from the correct pointer.
Signed-off-by: NJean-Baptiste Maneyrol <jmaneyrol@invensense.com>
[Antonio Borneo: cleanup and rebase to v3.17]
Signed-off-by: NAntonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

Currently, the i2c-hid driver does not call the suspend, resume, and
reset_resume callbacks in the hid_driver struct when those events occur.
This means that HID drivers for i2c-hid devices will not be able to execute
commands which may be needed during suspend or resume. One example is when a
touchpad using the hid-multitouch driver gets reset by i2c-hid coming out of
resume. Since the reset_resume callback never gets called the device is never
put back into the correct input mode. This patch calls the suspend and resume
callbacks and tries to duplicate the functionality of the usb-hid driver.
Signed-off-by: NAndrew Duggan <aduggan@synaptics.com>
Signed-off-by: NVincent Huang <vincent.huang@tw.synaptics.com>
Reviewed-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

Reading the partial HID Descriptor is causing a firmware lockup in some
sensor hubs. Instead of a partial read, this patch implements the
i2c hid fetch using a fixed descriptor size (30 bytes) followed by a
verification of the BCDVersion (V01.00), and value stored in
wHIDDescLength (30 Bytes) for V1.00 descriptors.

As per i2c hid spec, this is the preferred model.

From hid-over-i2c-protocol-spec-v1-0:

  There are a variety of ways a HOST may choose to retrieve
  the HID Descriptor from the DEVICE. The following is a preferred
  implementation but should not be considered the only implementation.
  A HOST may read the entire HID Descriptor in a single read by
  issuing a read for 30 Bytes to get the entire HID Descriptor
  from the DEVICE.However, the HOST is responsible for validating that

  1. The BCDVersion is V01.00 (later revisions may have different
     descriptor lengths), and

  2. The value stored in wHIDDescLength is 30 (Bytes) for V1.00
     descriptors.
Reported-by: NJoe Tijerina <joe.tijerina@st.com>
Signed-off-by: NArchana Patni <archana.patni@intel.com>
Signed-off-by: NSubramony Sesha <subramony.sesha@intel.com>
Reviewed-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

Nobody calls hid_output_raw_report anymore, and nobody should.
We can now remove the various implementation in the different
transport drivers and the declarations.
Reviewed-by: NDavid Herrmann <dh.herrmann@gmail.com>
Signed-off-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

It is better to check them soon enough before triggering any kernel panic.
Signed-off-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Reviewed-by: NDavid Herrmann <dh.herrmann@gmail.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

Having our own .request() implementation does not give anything,
so use the generic binding.
Signed-off-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Reviewed-by: NDavid Herrmann <dh.herrmann@gmail.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

Add output_report and raw_request to i2c-hid.
The current implementation of i2c_hid_output_raw_report decides
by itself if it should use a direct send of the output report
or use the data register (SET_REPORT). Split that by reimplement
the logic in __i2c_hid_output_raw_report() which will be dropped
soon.
Signed-off-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

dev->hid_get_raw_report(X) and hid_hw_raw_request(X, HID_REQ_GET_REPORT)
are strictly equivalent. Switch the hid subsystem to the hid_hw notation
and remove the field .hid_get_raw_report in struct hid_device.
Reviewed-by: NDavid Herrmann <dh.herrmann@gmail.com>
Signed-off-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

When using hid_output_report(), the buffer should be allocated by hid_alloc_report_buf(),
not a custom malloc.
Signed-off-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

This patch adds runtime PM support for the HID over I2C driver. When the
i2c-hid device is first opened we power it on and on the last close we
power it off. This is actually what the driver is already doing but in
addition it allows subsystems, like ACPI power domain to power off the
device during runtime PM suspend, which should save even more power.

The implementation is not the most power efficient because it needs some
interaction from the userspace (e.g close the device node whenever we are
no more interested in getting events), nevertheless it allows us to save
some power and works with devices that are not wake capable.
Signed-off-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

Use helper functions to simplify _DSM related code in i2c-hid driver.
Signed-off-by: NJiang Liu <jiang.liu@linux.intel.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

When an I2C HID device is powered of during system sleep, as a result of
removing its power resources (by the ACPI core) the interrupt line might go
low as well. This results inadvertent interrupt and wakes the system from
sleep immediately.

To prevent this we disable the device interrupt in the drivers suspend
method and enable it on resume. The device can still wake the system up if
it is wake capable (this also means that not all of its power will be
removed to keep the interrupt line high).
Reported-by: NJerome Blin <jerome.blin@intel.com>
Signed-off-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

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

According to specifications, HID over I2C devices are not bound to respond to
query for INPUT REPORTS. Thus dropping the call during init as many devices
does not respond causing error messages during boot.
Signed-off-by: NBibek Basu <bbasu@nvidia.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

acpi_evaluate_integer() is an ACPI API introduced to evaluate an
ACPI control method that is known to have an integer return value.
This API can simplify the code because the calling function does not need to
use the specified acpi_buffer structure required by acpi_evaluate_object();

Convert acpi_evaluate_object() to acpi_evaluate_integer()
in drivers/hid/i2c-hid/i2c-hid.c in this patch.
Signed-off-by: NZhang Rui <rui.zhang@intel.com>
Acked-by: NJiri Kosina <jkosina@suse.cz>
Acked-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

ACPI 5.0 specification requires the fourth parameter to the _DSM (Device
Specific Method) to be of type package instead of integer. Failing to do
that we get following warning on the console:

  ACPI Warning: \_SB_.PCI0.I2C1.TPL0._DSM: Argument #4 type mismatch - Found [Integer],
                ACPI requires [Package] (20130517/nsarguments-95)

Fix this by passing an empty package to the _DSM method. The HID over I2C
specification doesn't require any specific values to be passed with this
parameter.
Signed-off-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

There is no need to pass constants via stack. The width may be explicitly
specified in the format.
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

Add device tree based support for HID over I2C devices.

Tested on an Odroid-X board with a Synaptics touchpad.
Signed-off-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

HID core provides the same functionality, so drop the custom handler.
Besides, the current handler doesn't schedule any outgoing report so it
did not work, anyway.
Signed-off-by: NDavid Herrmann <dh.herrmann@gmail.com>
Reviewed-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

The current i2c hid driver does not support sending HID output reports using
the output register for devices which support receiving reports through this
method. This patch determines which method to use to send output reports based
 on the value of wMaxOutputLength in the device's HID descriptor.
Signed-off-by: NAndrew Duggan <aduggan@synaptics.com>
Reviewed-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

With the current i2c hid driver set/get report does not work
as expected, for e.g sensor hub properties like power state,
frequency etc is not set properly on the device as a result
we do not get events.
The problem is that i2c hid driver in function i2c_hid_request
sets length equal to default buffer size for which the sensor
hub does not respond on get/set commands. Use report length
and calculate it based on report size and id.
Reviewed-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NHuzefa Kankroliwala <huzefa.nomanx.kankroliwala@intel.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

This allows HID drivers to also get/set reports through hid_hw_request().
Tested-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: NBenjamin Tissoires <benjamin.tissoires@gmail.com>
Signed-off-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

i2c_hid_output_raw_report is used by hidraw to forward set_report requests.
The current implementation of i2c_hid_set_report needs to take the
report_id as an argument. The report_id is stored in the first byte
of the buffer in argument of i2c_hid_output_raw_report.

Not removing the report_id from the given buffer adds this byte 2 times
in the command, leading to a non working command.
Reported-by: NAndrew Duggan <aduggan@synaptics.com>
Signed-off-by: NBenjamin Tissoires <benjamin.tissoires@gmail.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

The HID over I2C protocol specification states that when the device is
enumerated from ACPI the HID descriptor address can be obtained by
executing "_DSM" for the device with function 1. Enable this.
Signed-off-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

CONFIG_HOTPLUG is going away as an option.  As a result, the __dev*
markings need to be removed.

This change removes the use of __devinit, __devexit_p, __devinitdata,
__devinitconst, and __devexit from these drivers.

Based on patches originally written by Bill Pemberton, but redone by me
in order to handle some of the coding style issues better, by hand.

Cc: Bill Pemberton <wfp5p@virginia.edu>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

We should not enter close function while someone else is in open.
This mutex prevents this race.

There is also no need to override the ret value with -EIO in case of
a failure of i2c_hid_set_power.
Signed-off-by: NBenjamin Tissoires <benjamin.tissoires@gmail.com>
Reviewed-by: NJean Delvare <khali@linux-fr.org>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

ret_count has to be at least 3, as we have to count the 2 bytes that are used
for the size of the reply. Without this, memcpy() might be called with zero or
negative count.
Reported-by: NJean Delvare <khali@linux-fr.org>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

The previous memcpy implementation relied on the size advertized by the
device. There were no guarantees that buf was big enough.

Some gymnastic is also required with the +2/-2 to take into account
the first 2 bytes of the returned buffer where the total returned
length is supplied by the device.
Signed-off-by: NBenjamin Tissoires <benjamin.tissoires@gmail.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

There is no point in keeping the irq in i2c_hid as it's already
there in client.
Signed-off-by: NBenjamin Tissoires <benjamin.tissoires@gmail.com>
Reviewed-by: NJean Delvare <khali@linux-fr.org>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

Simplifies i2c_hid_alloc_buffers tests, and makes this function
responsible of the assignment of ihid->bufsize.
The condition for the reallocation in i2c_hid_start is then simpler.
Signed-off-by: NBenjamin Tissoires <benjamin.tissoires@gmail.com>
Reviewed-by: NJean Delvare <khali@linux-fr.org>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

HID descriptors contains 4 bytes of reserved field.
The previous implementation was overriding the next fields in struct i2c_hid.
Signed-off-by: NBenjamin Tissoires <benjamin.tissoires@gmail.com>
Reviewed-by: NJean Delvare <khali@linux-fr.org>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

The pointless WARN_ON() has been removed from i2c_hid_remove(),
so we don't need bug.h any more.
Reported-by: NJean Delvare <khali@linux-fr.org>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>