The driver only has runtime but no build time dependency with SH_MIGOR, so
it can be built for testing purposes if the COMPILE_TEST option is enabled.

This is useful to have more build coverage and make sure that drivers are
not affected by changes that could cause build regressions.
Signed-off-by: NJavier Martinez Canillas <javier@osg.samsung.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

There is a typo in the driver for the I2C device table that is passed to
the MODULE_DEVICE_TABLE() macro, this leads to the following build error:

  CC [M]  drivers/input/touchscreen//migor_ts.o
drivers/input/touchscreen//migor_ts.c:233:21: error: 'migor_ts' undeclared here (not in a function)
 MODULE_DEVICE_TABLE(i2c, migor_ts);
                     ^
drivers/input/touchscreen//migor_ts.c:233:31: error: '__mod_i2c__migor_ts_device_table' aliased to undefined symbol 'migor_ts'
 MODULE_DEVICE_TABLE(i2c, migor_ts);
                               ^
scripts/Makefile.build:295: recipe for target 'drivers/input/touchscreen//migor_ts.o' failed
make[1]: *** [drivers/input/touchscreen//migor_ts.o] Error 1
Makefile:1446: recipe for target '_module_drivers/input/touchscreen/' failed
make: *** [_module_drivers/input/touchscreen/] Error 2
Signed-off-by: NJavier Martinez Canillas <javier@osg.samsung.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

fix checkpatch.pl warning about 'Prefer 'unsigned int' to bare use of
'unsigned''
Signed-off-by: NClifton Barnes <clifton.a.barnes@gmail.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

Complete() will be run with interrupt enabled, so change to
spin_lock_irqsave().
Signed-off-by: NMing Lei <ming.lei@canonical.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

The CM109 driver reported key press events of volume up / down and
record / playback mute buttons, but no release events. Report those events
properly by handling volume and mute keys seperately. For the record and
playback mute buttons, only presses are registered by the CM109, therefore
simulate press-n-release. This fixes the volume control buttons of various
USB headsets.
Signed-off-by: NFlorian Euchner <florian.euchner@gmail.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

Since memory for a private data is allocated by kzalloc() there is no
need to fill it with zeroes immediately after the allocation.
Signed-off-by: NVladimir Zapolskiy <vz@mleia.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

The bitwise shift operator has lower priority than plus operator. So the
values on macros should be enclosed in parentheses.

For example, "(1 << 4 + 1)" means "(1 << (4 + 1))", but it is not expected
by the macros.

And also fix other two coding style problems reported by
scripts/checkpatch.pl.
Signed-off-by: NRui Teng <rui.teng@linux.vnet.ibm.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

The mutex does not seem to be needed. twl4030-vibra doesn't
use one either.
Signed-off-by: NH. Nikolaus Schaller <hns@goldelico.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

In Cygnus SOC touch screen controller registers are shared with ADC and
flex timer. Using readl/writel could lead to race condition. So touch
screen driver is enhanced to support register access using syscon framework
API's to take care of mutually exclusive access.
Signed-off-by: NRaveendra Padasalagi <raveendra.padasalagi@broadcom.com>
Reviewed-by: NRay Jui <ray.jui@broadcom.com>
Reviewed-by: NScott Branden <scott.branden@broadcom.com>
Acked-by: NRob Herring <robh@kernel.org>
Acked-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

Instead of doing the dance with macro that either resolves to a pointer or
NULL, let's switch to using SIMPLE_DEV_PM_OPS().

Also let's mark suspend and resume methods as __maybe_unused instead of
guarding them with an #ifdef and rely on linker to drop unused code. Doing
so should allow better compile coverage.
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

Commit f799a3d8 ("Input: omap-keypad:
Remove dependencies to mach includes") removed users of the above
functions, but left them in the code.
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

There is no reason to have empty suspend and resume stubs.
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

As of 35f8679f ("Input: omap-keypad -
remove dead check") we no longer declare keypresses as spurious, therefore
we can use constant delay between scans.
Suggested-by: NJanusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

Drop #address-cells and #size-cells, which are not required by the
gpio-keys binding documentation, as button sub-nodes are not devices.

Rename sub-nodes to avoid new dtc unit address warnings when copied.

While at it, adopt the dashes convention for the node name.
Reported-by: NJulien Chauveau <chauveau.julien@gmail.com>
Signed-off-by: NAndreas Färber <afaerber@suse.de>
Reviewed-by: NJavier Martinez Canillas <javier@osg.samsung.com>
Reviewed-by: NJulien Chauveau <chauveau.julien@gmail.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

The device can now easily be derived from the interface.
Stop leaving a private copy.
Signed-off-by: NOliver Neukum <ONeukum@suse.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

The device can now easily be derived from the interface.
Stop leaving a private copy.
Signed-off-by: NOliver Neukum <oneukum@suse.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

The device can now easily be derived from the interface.
Stop leaving a private copy.
Signed-off-by: NOliver Neukum <oneukum@suse.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

The device can now easily be derived from the interface.
Stop leaving a private copy.
Signed-off-by: NOliver Neukum <ONeukum@suse.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

During the initialisation the driver uses a buffer on the stack for DMA.
That violates the cache coherency rules. The fix is to allocate the buffer
with kmalloc().
Signed-off-by: NOliver Neukum <ONeukum@suse.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

The ati_remote2 driver expects at least two interfaces with one
endpoint each. If given malicious descriptor that specify one
interface or no endpoints, it will crash in the probe function.
Ensure there is at least two interfaces and one endpoint for each
interface before using it.

The full disclosure: http://seclists.org/bugtraq/2016/Mar/90Reported-by: NRalf Spenneberg <ralf@spenneberg.net>
Signed-off-by: NVladis Dronov <vdronov@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

Looks like the fimware 8.2 still has the extra buttons spurious release
bug.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=114321
Cc: stable@vger.kernel.org
Signed-off-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

Originally the irq_mask member of rmi_function was a pointer. Then it was
switched to being a zero length array. However, the checks for a NULL
pointer where not removed.
Signed-off-by: NAndrew Duggan <aduggan@synaptics.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

The Windows driver's settings dialog contains a visualization of the
regions for the hardware edge scrolling capability, which uses a
temporarily-enabled limited-resolution absolute mode.

This patch enables this during normal operation, and combines the
absolute packets with the existing relative packets to provide
accurate absolute position and touch reporting.

It also adds documentation for all known gesture packets and
initialization commands.
Reviewed-by: NChris Diamand <chris@diamand.org>
Signed-off-by: NRichard Pospesel <pospeselr@gmail.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

A malicious device missing interface can make the driver oops.
Add sanity checking.
Signed-off-by: NOliver Neukum <ONeukum@suse.com>
CC: stable@vger.kernel.org
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

Add hw_version sysfs attribute for retrieving hardware information
from firmware.
Signed-off-by: NSangwon Jee <jeesw@melfas.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

gcc-6.0 notices that the use of the property_entry in this file that
was recently introduced cannot work right, as we initialize the wrong
field:

raumfeld.c:387:3: error: the address of 'raumfeld_rotary_encoder_steps' will always evaluate as 'true' [-Werror=address]
   DEV_PROP_U32, 1, &raumfeld_rotary_encoder_steps, },
   ^~~~~~~~~~~~
raumfeld.c:389:3: error: the address of 'raumfeld_rotary_encoder_axis' will always evaluate as 'true' [-Werror=address]
   DEV_PROP_U32, 1, &raumfeld_rotary_encoder_axis, },
   ^~~~~~~~~~~~
raumfeld.c:391:3: error: the address of 'raumfeld_rotary_encoder_relative_axis' will always evaluate as 'true' [-Werror=address]
   DEV_PROP_U32, 1, &raumfeld_rotary_encoder_relative_axis, },
   ^~~~~~~~~~~~

The problem appears to stem from relying on an old definition of
'struct property', but it has changed several times since the code
could have last been correct.

This changes the code to use the PROPERTY_ENTRY_INTEGER() macro instead,
which works fine for the current definition and is a safer way of doing
the initialization.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Fixes: a9e340dc ("Input: rotary_encoder - move away from platform data structure")
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

Merge with Linux 4.5 to get PROPERTY_ENTRY_INTEGER() that is needed to
fix pxa/raumfeld rotary encoder properties.

There is a typo so we have && instead of &.

Fixes: ff8f8370 ('Input: synaptics-rmi4 - add support for 2D sensors and F11')
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Acked-by: NAndrew Duggan <aduggan@synaptics.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

The powermate driver expects at least one valid USB endpoint in its
probe function.  If given malicious descriptors that specify 0 for
the number of endpoints, it will crash.  Validate the number of
endpoints on the interface before using them.

The full report for this issue can be found here:
http://seclists.org/bugtraq/2016/Mar/85Reported-by: NRalf Spenneberg <ralf@spenneberg.net>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: NJosh Boyer <jwboyer@fedoraproject.org>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>

Pull MIPS fixes from Ralf Baechle:
 "Another round of MIPS fixes for 4.5:

   - Fix JZ4780 build with DEBUG_ZBOOT and MACH_JZ4780
   - Fix build with DEBUG_ZBOOT and MACH_JZ4780
   - Fix issue with uninitialised temp_foreign_map
   - Fix awk regex compile failure with certain versions of awk.  At
     this time, the sole user, ld-ifversion, is only used on MIPS"

* 'upstream' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus:
  MIPS: smp.c: Fix uninitialised temp_foreign_map
  MIPS: Fix build error when SMP is used without GIC
  ld-version: Fix awk regex compile failure
  MIPS: Fix build with DEBUG_ZBOOT and MACH_JZ4780

When calculate_cpu_foreign_map() recalculates the cpu_foreign_map
cpumask it uses the local variable temp_foreign_map without initialising
it to zero. Since the calculation only ever sets bits in this cpumask
any existing bits at that memory location will remain set and find their
way into cpu_foreign_map too. This could potentially lead to cache
operations suboptimally doing smp calls to multiple VPEs in the same
core, even though the VPEs share primary caches.

Therefore initialise temp_foreign_map using cpumask_clear() before use.

Fixes: cccf34e9 ("MIPS: c-r4k: Fix cache flushing for MT cores")
Signed-off-by: NJames Hogan <james.hogan@imgtec.com>
Cc: Paul Burton <paul.burton@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/12759/Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

The MIPS_GIC_IPI should only be selected when MIPS_GIC is also
selected, otherwise it results in a compile error. smp-gic.c uses some
functions from include/linux/irqchip/mips-gic.h like
plat_ipi_call_int_xlate() which are only added to the header file when
MIPS_GIC is set. The Lantiq SoC does not use the GIC, but supports SMP.
The calls top the functions from smp-gic.c are already protected by
some #ifdefs

The first part of this was introduced in commit 72e20142 ("MIPS:
Move GIC IPI functions out of smp-cmp.c")
Signed-off-by: NHauke Mehrtens <hauke@hauke-m.de>
Cc: Paul Burton <paul.burton@imgtec.com>
Cc: stable@vger.kernel.org # v3.15+
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/12774/Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

The ld-version.sh script fails on some versions of awk with the
following error, resulting in build failures for MIPS:

awk: scripts/ld-version.sh: line 4: regular expression compile failed (missing '(')

This is due to the regular expression ".*)", meant to strip off the
beginning of the ld version string up to the close bracket, however
brackets have a meaning in regular expressions, so lets escape it so
that awk doesn't expect a corresponding open bracket.

Fixes: ccbef167 ("Kbuild, lto: add ld-version and ld-ifversion ...")
Reported-by: NGeert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: NJames Hogan <james.hogan@imgtec.com>
Tested-by: NMichael S. Tsirkin <mst@redhat.com>
Acked-by: NMichael S. Tsirkin <mst@redhat.com>
Tested-by: NSudip Mukherjee <sudip.mukherjee@codethink.co.uk>
Cc: Michal Marek <mmarek@suse.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: linux-mips@linux-mips.org
Cc: linux-kbuild@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Cc: stable@vger.kernel.org # 4.4.x-
Patchwork: https://patchwork.linux-mips.org/patch/12838/Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Ingenic SoC declares ZBOOT support, but debug definitions are missing
for MACH_JZ4780 resulting in a build failure when DEBUG_ZBOOT is set.
The UART addresses are same as with JZ4740, so fix by covering JZ4780
with those as well.
Signed-off-by: NAaro Koskinen <aaro.koskinen@iki.fi>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/12830/Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Pull block merge fix from Jens Axboe.

This fixes the block segment counting bug and resulting sg overrun
reported by Kent Overstreet, introduced with the last block pull.

* 'for-linus' of git://git.kernel.dk/linux-block:
  block: don't optimize for non-cloned bio in bio_get_last_bvec()

Pull x86 fixes from Ingo Molnar:
 "This fixes 3 FPU handling related bugs, an EFI boot crash and a
  runtime warning.

  The EFI fix arrived late but I didn't want to delay it to after v4.5
  because the effects are pretty bad for the systems that are affected
  by it"

[ Actually, I don't think the EFI fix really matters yet, because we
  haven't switched to the separate EFI page tables in mainline yet ]

* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/efi: Fix boot crash by always mapping boot service regions into new EFI page tables
  x86/fpu: Fix eager-FPU handling on legacy FPU machines
  x86/delay: Avoid preemptible context checks in delay_mwaitx()
  x86/fpu: Revert ("x86/fpu: Disable AVX when eagerfpu is off")
  x86/fpu: Fix 'no387' regression

Pull SCSI target bug fix from Nicholas Bellinger:
 "Here is an outstanding target-core bug-fix for v4.5 code."

  This patch addresses a recent Task Management (TMR) regression related
  to larger set of multi-port LUN_RESET bug-fixes in v4.5-rc5.

  It drops a left-over target_put_sess_cmd() of se_cmd->cmd_kref within
  ABORT_TASK failure path, once a se_cmd descriptor has already
  completed posting response to fabric driver logic, and must be skipped
  during normal ABORT_TASK se_cmd->tag lookup"

* git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending:
  target: Drop incorrect ABORT_TASK put for completed commands

For !BIO_CLONED bio, we can use .bi_vcnt safely, but it
doesn't mean we can just simply return .bi_io_vec[.bi_vcnt - 1]
because the start postion may have been moved in the middle of
the bvec, such as splitting in the middle of bvec.

Fixes: 7bcd79ac(block: bio: introduce helpers to get the 1st and last bvec)
Cc: stable@vger.kernel.org
Reported-by: NKent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: NMing Lei <ming.lei@canonical.com>
Signed-off-by: NJens Axboe <axboe@fb.com>