Supply platform data for speaker driver with info of additional
speaker inputs supported. The pmic pm8038 supports one additional
input path to Speaker Power amps. Add this to pdata as part
of PMIC sub devices platform data.

Change-Id: I8045cc517a29853590b436f86cac5504557de049
Signed-off-by: Asish Bhattacharya <asishb@codeaurora.org>

* changes:
  msm: gemini: Pass device pointers to regulator_get()
  msm: rotator: Pass device pointers to regulator_get()
  msm: vcap: Pass device pointers to regulator_get()

* changes:
  defconfig: 9615: Rename CONFIG_ATH6K_LEGACY to CONFIG_ATH6K_LEGACY_EXT
  net: wireless: Rename CONFIG_ATH6K_LEGACY to CONFIG_ATH6K_LEGACY_EXT

* changes:
  msm: pm-8x60: Use Qtimer for copper
  msm: pm-boot: Add devicetree support for pm-boot

* changes:
  msm: copper: Add Low Power Management drivers for CPUIdle
  msm: board-copper: Initialize SPM device driver ahead of Device Tree

The graceful argument allows clients to specify whether they want
to flush now, or wait until the end of the current scatter gather or
box mode descriptor. This also helps us unify msm_dmov_flush and
msm_dmov_stop_cmd.

Change-Id: Ifdfe5da804e79343782cc2999ad3624d2afbec4d
Signed-off-by: Jeff Ohlstein <johlstei@codeaurora.org>

Currently all HDMI/HDCP timers are initialized in the driver's
probe function with a maximum possible expiration value. However,
the 32-bit jiffies value is initialized in the kernel so as to
wrap 5 minutes after boot. This results in the HDCP timer being
fired 5 minutes after boot even though reauthentication is not
necessary. This change fixes this issue by not starting the HDCP
reauth timer during probe. The timer is only started when an
authentication failure is detected from an existing valid HDCP
session.

Change-Id: I0a28437daa2e95d24a02131322305f3b21a6445c
CRs-Fixed: 352080
Signed-off-by: Aravind Venkateswaran <aravindh@codeaurora.org>

CONFIG_ATH6K_LEGACY already exists in kernel tree so renaming to
CONFIG_ATH6K_LEGACY_EXT to support QCA AR6003 (Non cfg80211) driver
which will be compiled outside of kernel tree.

Change-Id: I00406009808896c8bd7d44fcc42db6c031dbf988
Signed-off-by: Sridhar Parasuram <sridhar@codeaurora.org>

CONFIG_ATH6K_LEGACY already exists in kernel tree so renaming to
CONFIG_ATH6K_LEGACY_EXT to support QCA AR6003 (Non cfg80211) driver
which will be compiled outside of kernel tree. AR6003 driver needs
WIRELESS_EXT support. WIRELESS_EXT is going to be deprecated so
adding it to defconfig will not work.

Change-Id: I171a3e4dbdf9bd2cf7e7eaa3741ec2fa9730a114
Signed-off-by: Sridhar Parasuram <sridhar@codeaurora.org>

Passing a device pointer instead of NULL as the first argument
to regulator_get() allows the device to be taken into account
when finding a matching regulator.

Change-Id: Ib5cd5858cd0f61ea1b699385763adec9ca9ae72e
Signed-off-by: Matt Wagantall <mattw@codeaurora.org>

Passing a device pointer instead of NULL as the first argument
to regulator_get() allows the device to be taken into account
when finding a matching regulator.

Change-Id: Idc9990eff2c6c78fc9fc9a36d638eef4fd1b632a
Signed-off-by: Matt Wagantall <mattw@codeaurora.org>

Passing a device pointer instead of NULL as the first argument
to regulator_get() allows the device to be taken into
account when finding a matching regulator. This removes the need
to specify a unique regulator_name in platform data.

Change-Id: Id16f85866c94b3f3b57aa720cdfbdcf3e2fd241d
Signed-off-by: Matt Wagantall <mattw@codeaurora.org>

Passing a device pointer instead of NULL as the first argument
to regulator_get() allows the device to be taken into account
when finding a matching regulator.

Change-Id: I8cb02e5dc3294090f2f81371fca64b82139756d7
Signed-off-by: Matt Wagantall <mattw@codeaurora.org>

Passing a device pointer instead of NULL as the first argument
to regulator_get() allows the device to be taken into account
when finding a matching regulator.

Change-Id: I34917d18e1216be01c0cf8ccfd0b6038e5474ffd
Signed-off-by: Matt Wagantall <mattw@codeaurora.org>

Passing a device pointer instead of NULL as the first argument
to regulator_get() allows the device to be taken into account
when finding a matching regulator. This removes the need to
specify a unique regulator_name in struct kgsl_pwrctrl.

Change-Id: I6f6c299d5ad39b8b5b24a465ecabab484ea6d18e
Signed-off-by: Matt Wagantall <mattw@codeaurora.org>

regulator_get() expects two arguments: a consumer device pointer and
a regulator name.  For footswitch regulators, users of this API have
usually passed NULL in place of the device pointer, which was treated
like a wildcard for the purposes of matching. Those that passed
pointers received no advantage from it anyway, since it was not
possible to specify consumer device names in the devices-* files using
the macros in footswitch.h.

Improve on this by expanding the footswitch macros to take a consumer
name strings. This allows regulator_get() to find matches using a
stricter criteria that considers the consumer device name and ID,
as well as the regulator name.

Change-Id: I77618007b674480aff031329757b41cc00014478
Signed-off-by: Matt Wagantall <mattw@codeaurora.org>

Fixed the mask to enabled power counters during device start.
These counters are used for performance measurements and DCVS.
DCVS later in the code reinitializes the counters correctly.
This fix is required for performance measurements.

Change-Id: I08d3a184b5120c5251060bb9c5f080af0167732d
Signed-off-by: Tarun Karra <tkarra@codeaurora.org>

In order to support synchronization in a process with a single
gralloc handle we require the ability to write lock a buffer
while it is already read locked by the same handle. This change
extends the concept of an exclusive write lock or recursive read
locks to a genlock handle (rather than the genlock lock).
Genlock cannot provide deadlock protection because the same
handle can be used simultaneously by a producer and consumer.
In practice an error will still be generated when the timeout
expires.

CRs-fixed: 356263
Change-Id: I322e7fadc8b43287f53b211242b176d3de731db2
Signed-off-by: Jeff Boody <jboody@codeaurora.org>

Remove holding the AMP manger list lock when sending the
A2MP Change Notify message, as this uses L2CAP and will
lock the socket and possibly sleep.

This removes a potential deadlock and a "BUG: sleeping
function called from invalid context" message from the
kernel log.

Change-Id: Icf8c5afc1a5cde36517bb979820c83a20ee207fd
CRs-fixed: 358736
Signed-off-by: Peter Krystad <pkrystad@codeaurora.org>

iommu map functions requires clients to specify whether
the mappings should be read only or read-write (write only
is not supported). Ensure clients specify the appropriate
flags when calling map functions.

Change-Id: If42b4b0367b4311aa36d3063979ef04af3a43fc6
Signed-off-by: Olav Haugan <ohaugan@codeaurora.org>

Add board file configuration, clock configuration and
clock setting to enable MT9E013 sensor on 8x60 v4l2
architecture

Change-Id: I90eb9c7055ec5e3223f5c608769af7c363a49f54
Signed-off-by: Sreesudhan Ramakrish Ramkumar <srramku@codeaurora.org>

msm_timer is no longer available for copper. Use alternate Qtimer API
for copper.

Change-Id: I2df4f6d90090d4c751331eee9bb768629b745f09
Signed-off-by: Praveen Chidambaram <pchidamb@codeaurora.org>

pm-boot driver abstracts warm boot configurations for the power
management drivers. Add device tree support to read the boot
configuration information from DTS file.

Change-Id: I1135f796ddb054bb1ea4b64b9ff323736157bdc8
Signed-off-by: Praveen Chidambaram <pchidamb@codeaurora.org>

When the core is in idle, CPUIdle framework calls into the architecture
specific idle driver. The idle driver uses RPM resources (L2, VDD, XO)
and the time available to determine the lowest power mode possible and
uses the SPM driver to put the core in low power mode.

Add LPM levels that specify the latency and the energy required to enter
different low power modes. These values are used to determine how deep a
sleep can be achieved before the next scheduled wakeup.

Change-Id: Ia2ba25e618aa72cb02f49fbd0028a0615e465dc8
Signed-off-by: Praveen Chidambaram <pchidamb@codeaurora.org>