ARM: OMAP2+: McBSP: Correct CLKR/FSR clock source mux configuration

On OMAP2/3 McBSP1 port has 6 pin setup, while on OMAP4 the port is McBSP4.
Implement the CLKR/FSR clock mux selection for OMAP4, and make sure that
we add the correct callback for the correct port across supported OMAP
versions.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Jarkko Nikula <jarkko.nikula@bitmer.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Liam Girdwood <lrg@ti.com>

ASoC: omap McBSP: Clear rx_irq at probe time for OMAP4

On OMAP4 we have one interrupt line per McBSP port.
At probe time tx, and rx irq value will be -ENXIO,
and only the tx irq will get corrected.
In omap_mcbsp_request if the rx_irq is not 0 we proceed,
and try to request the interrupt, which will fail on
OMAP4 (rx_irq == -6).
To avoid this error, clear the rx_irq at probe time
on OMAP4.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Jarkko Nikula <jarkko.nikula@bitmer.com>
Signed-off-by: Liam Girdwood <lrg@ti.com>

ASoC: omap-mcbsp: When closing the port select PRCM source for CLKS signal

If external source for the CLKS signal selection kept after the port is no
longer in use the system might refuse to go suspend.
There is also a chance that the external clock is not running when next
time the McBSP port is started which can result errors when we try to
access McBSP registers.
Reset the CLKS source back to PRCM source unconditionally.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Tested-by: Grazvydas Ignotas <notasas@gmail.com>
Signed-off-by: Liam Girdwood <lrg@ti.com>

ASoC: omap-mcbsp: Configure wakeup in later phase

Configure the WAKEUPEN register at the same time we configure the rest of
the McBSP registers.
In case of OMAP3+, if the sysclock has been reconfigured we are going to
disable McBSP for the duration of the clock change, which will reset the
McBSP registers. The WAKEUPEN register need to be configured later, so
the changes will be effective during runtime.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicrocom>
Signed-off-by: Liam Girdwood <lrg@ti.com>

ASoC: omap: mcbsp: Remove redundant checks for the st_data pointer

The parent functions of omap_st_start/stop also checks the validity of the
st_data pointer so we do not need to do it again inside of omap_st_start/stop

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Jarkko Nikula <jarkko.nikula@bitmer.com>
Signed-off-by: Liam Girdwood <lrg@ti.com>

ASoC: omap: mcbsp: Use uniform st_data pointer initialization

In this way we can save few lines, and have uniform way of initializing the
st_data in all functions.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Jarkko Nikula <jarkko.nikula@bitmer.com>
Signed-off-by: Liam Girdwood <lrg@ti.com>

ASoC: omap-mcbsp: Simplify DMA configuration

Configure the DMA request line, port address, and stream name at
probe time instead of every time we start a stream.
These settings are static in the system.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Tested-by: Grazvydas Ignotas <notasas@gmail.com>
Tested-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Jarkko Nikula <jarkko.nikula@bitmer.com>
Signed-off-by: Liam Girdwood <lrg@ti.com>

ASoC: omap-mcbsp: Merge the omap_mcbsp_data into omap_mcbsp structure

Since the drivers has been merged, merge the two structures together.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Tested-by: Grazvydas Ignotas <notasas@gmail.com>
Tested-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Jarkko Nikula <jarkko.nikula@bitmer.com>
Signed-off-by: Liam Girdwood <lrg@ti.com>

ASoC: omap-mcbsp: Cleanup of module probe/remove code

Use devm_* where it is possible to save on cleanup path.
Start merging the two mcbsp file content.
Move pm_runtime_enable/disable calls to ASoC probe, remove from module
probe/remove time.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Tested-by: Grazvydas Ignotas <notasas@gmail.com>
Tested-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Jarkko Nikula <jarkko.nikula@bitmer.com>
Signed-off-by: Liam Girdwood <lrg@ti.com>

ASoC: OMAP McBSP: Remove redundant accessors

We no longer need accessor functions for max_tx/rx_threshold, dma_op_mode
or for the FIFO size.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Tested-by: Grazvydas Ignotas <notasas@gmail.com>
Tested-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Jarkko Nikula <jarkko.nikula@bitmer.com>
Signed-off-by: Liam Girdwood <lrg@ti.com>

ASoC: OMAP: mcbsp.h: Clean up bit definitions

Use BIT() for bit position, correct field definition by adding mask to
them, and also adding the missing spaces around '<<'

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Tested-by: Grazvydas Ignotas <notasas@gmail.com>
Tested-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Jarkko Nikula <jarkko.nikula@bitmer.com>
Signed-off-by: Liam Girdwood <lrg@ti.com>

ARM: OMAP: Do not register omap-mcbsp-dai device

The driver for omap-mcbsp-dai no longer exist since it has been merged with
the omap-mcbsp driver.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Tested-by: Grazvydas Ignotas <notasas@gmail.com>
Tested-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Acked-by: Jarkko Nikula <jarkko.nikula@bitmer.com>
Signed-off-by: Liam Girdwood <lrg@ti.com>

ASoC: omap-mcbsp: Create a single driver for McBSP

The OMAP McBSP driver stack used to contain two different
drivers. One of them was used as kind low-level access to
the IP, while the other driver was the ASoC DAI driver.
There were global, shared structures, in different places,
the McBSP instances are reffered with id numbers (sometimes
0 based, in other cases 1 based id numbers).

Create one single driver for OMAP McBSP with name: omap-mcbsp.
Convert the old omap-mcbsp driver initially to be a library
for the omap-mcbsp DAI driver. With this change we can get rid
of all global variables, structures.

Further cleanup is coming...

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Tested-by: Grazvydas Ignotas <notasas@gmail.com>
Tested-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Jarkko Nikula <jarkko.nikula@bitmer.com>
Signed-off-by: Liam Girdwood <lrg@ti.com>

ASoC: OMAP: McBSP: Consolidate plat/mcbsp.h content

Move most of the content of the plat/mcbsp.h header file under
sound/soc/omap/ to help further cleanups.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Tested-by: Grazvydas Ignotas <notasas@gmail.com>
Tested-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Jarkko Nikula <jarkko.nikula@bitmer.com>
Signed-off-by: Liam Girdwood <lrg@ti.com>

OMAP: mcbsp: Move core driver under sound/soc/omap

In order to consolidate the McBSP driver move it out from
arch/arm/plat-omap directory under sound/soc/omap/

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Tested-by: Grazvydas Ignotas <notasas@gmail.com>
Tested-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Acked-by: Jarkko Nikula <jarkko.nikula@bitmer.com>
Signed-off-by: Liam Girdwood <lrg@ti.com>

ARM: OMAP: mcbsp: Convert core driver to proper platform driver

Convert the plat-omap/mcbsp.c driver to be proper platform driver.
Remove the omap_mcbsp_init function call which was called from
mach-omap1/2/mcbsp.c to register the platform driver for the just
created platform device in the same function.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Tested-by: Grazvydas Ignotas <notasas@gmail.com>
Tested-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Acked-by: Jarkko Nikula <jarkko.nikula@bitmer.com>
Signed-off-by: Liam Girdwood <lrg@ti.com>

OMAP4: panda: Correct cpu version check for 4430

The cpu_is_omap4430() macro always return with 0. Use the correct
cpu_is_omap443x() to check for Panda revision.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Liam Girdwood <lrg@ti.com>

Merge tag 'asoc-3.4' of git://git./linux/kernel/git/broonie/sound into topic/asoc

A few more ASoC updates, the main one is the move of the audmux driver
from arch/arm into sound/soc.  There's also some general driver specific
tweaks and fixes.

ASoC: fsl: add dt support for imx-audmux

It adds device tree probe support for imx-audmux driver.

Signed-off-by: Richard Zhao <richard.zhao@linaro.org>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: Add missing regmap_init_i2c in wm8804_i2c_probe

commit 891271c "ASoC: Convert wm8804 to direct regmap API usage"
only converts wm8804_spi_probe to use regmap_init_spi.

This patch adds missing regmap_init_i2c in wm8804_i2c_probe.

Signed-off-by: Axel Lin <axel.lin@gmail.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8962: Remove defaults for volatile registers

Save a little RAM.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8962: Remove unneeded pm_runtime_set_active()

The default pm_runtime status is enabled which is what we want.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8994: Use audio mode for jack detection when system is active

When we are out of system sleep always use audio mode for jack detection
in order to avoid potential performance issues handing off between modes.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: imx: add an explicit Kconfig option for imx-ssi driver

Currently ASoC:imx uses menuconfig option SND_IMX_SOC selects imx-ssi
driver, and it works because all the machine driver covered by the
menuconfig need to build imx-ssi driver in.  However, it will not work
any more if we have a imx based machine driver going into the menuconfig
while working with fsl_ssi driver (sound/soc/fsl/fsl_ssi.c) rather than
imx-ssi one.

The patch adds an explicit Kconfig option SND_SOC_IMX_SSI for imx-ssi
driver, so that it can be selected independently from the menuconfig
option SND_IMX_SOC.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Acked-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: imx: separate imx-pcm bits from imx-ssi driver

Currently the imx-ssi.c[h] accommodates the imx-pcm common bits which
are shared between imx-pcm-dma-mx2 and imx-pcm-fiq drivers.  It assumes
that imx-pcm-dma-mx2 and imx-pcm-fiq will always be used together with
imx-ssi driver.  However this becomes untrue when we see that driver
sound/soc/fsl/fsl_ssi could possibly work with imx-pcm-dma-mx2 too.

The patch moves the imx-pcm common bits from imx-ssi.c[h] into new
files imx-pcm.c[h], and let imx-pcm-dma-mx2 and imx-pcm-fiq drivers
build it in, so that imx-pcm-dma-mx2 can work with no dependency on
imx-ssi driver.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Acked-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: imx: rename audmux prefix mxc to imx

It renames the legacy name mxc used in audmux function and macro
to imx.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Acked-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: imx: move audmux driver into sound/soc/imx

As audmux becomes a platform driver and its callers are all ASoC
machine drivers, there is no reason to keep it in arch folder, so
move it to sound/soc/imx.

One bonus point would be those ASoC machine drivers stop including
mach/audmux.h, since it's been moved to sound/soc/imx/imx-audmux.h.
This should be a move to the right direction in terms of single kernel
image goal.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Acked-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ARM: imx: convert audmux to a platform driver

It coverts audmux to a platform driver, so that it can be moved into
sound/soc/imx and adopt device tree support later.

Signed-off-by: Richard Zhao <richard.zhao@linaro.org>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Acked-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ARM: imx: merge audmux-v1 and audmux-v2

It merges audmux-v1 and audmux-v2 under arch/arm/plat-mxc into one.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Acked-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: imx: move phycore audmux call into ASoC machine driver

It moves phycore audmux configuration call from board file into ASoC
machine driver phycore-ac97 to ease converting audmux into a platform
driver later.
It moves phycore audmux configuration call from board file into ASoC
machine driver phycore-ac97, so that it gets aligned with wm1133-ev1
and mx27vis-aic32x4, and more importantly it will ease the moving of
audmux into sound/soc/imx as a platform driver later.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Acked-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: imx: move eukrea audmux call into ASoC machine driver

It moves eukrea audmux configuration call from board file into ASoC
machine driver eukrea-tlv320, so that it gets aligned wm1133-ev1 and
mx27vis-aic32x4, and more importantly it will ease the moving of audmux
into sound/soc/imx as a platform driver later.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Acked-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: dapm: Show if widgets are forced in debugfs

The information was not otherwise visible.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Liam Girdwood <lrg@ti.com>

Merge tag 'asoc-3.4' of git://git./linux/kernel/git/broonie/sound into topic/asoc

This has been a very active release for ASoC, as well as the usual raft
of bugfixes and driver updates there's quite a few framework enhancements.
Most are either small or are laying the groundwork for user visible
features (especially dynamic PCM), the most directly visible change is
the dmaengine library.  There's also a bunch of regmap API enhancements
pulled into the tree so that either the framework or drivers can take
advantage of the new features.

Changes include:

- Support for widgets not associated with a CODEC, an important part of
  the dynamic PCM framework.

- A library factoring out the common code shared by dmaengine based DMA
  drivers contributed by Lars-Peter Clausen.  This will save a lot of
  code and make it much easier to deploy enhancements to dmaengine.

- Support for binary controls, used for providing runtime configuration
  of algorithm coefficients.

- A new DAPM widget type for regulator supplies allowing drivers for
  devices that can power down unused supplies while active to do without
  any per-driver code.

- DAPM widgets for DAIs, initially giving a speed boost for playback
  startup and shutdown and also the basis for CODEC<->CODEC DAI link
  support.

- Support for specifying the number of significant bits on audio
  interfaces, useful for allowing applications to know how much effort to
  put into generating data for a larger sample format.

- Conversion of the FSI driver used on some SH processors to DMAEngine.

- New CODEC drivers for Maxim MAX9768 and Wolfson Microelectronics WM2200.

ASoC: wm8753: fix initialization

Without that fix the wm8753 SPI initialization fails, and then produces
  a kernel panic during boot with the following call trace:
    Unable to handle kernel paging request at virtual address 37386d9b
    [<c01ccafc>] (regmap_get_val_bytes+0x0/0x14) from [<c0243dfc>] (snd_soc_codec_set_cache_io+0x9c/0xcc)
    [<c0243dfc>] (snd_soc_codec_set_cache_io+0x9c/0xcc) from [<c0244a4c>] (wm8753_probe+0x5c/0x1c4)
    [<c0244a4c>] (wm8753_probe+0x5c/0x1c4) from [<c023bb24>] (soc_probe_codec+0x174/0x284)
    [<c023bb24>] (soc_probe_codec+0x174/0x284) from [<c023c2c0>] (snd_soc_instantiate_cards+0x68c/0xe28)
    [<c023c2c0>] (snd_soc_instantiate_cards+0x68c/0xe28) from [<c023d278>] (snd_soc_register_card+0x240/0x2d4)
    [<c023d278>] (snd_soc_register_card+0x240/0x2d4) from [<c023d330>] (soc_probe+0x24/0x40)
    [<c023d330>] (soc_probe+0x24/0x40) from [<c01c3900>] (platform_drv_probe+0x14/0x18)
    [...]

The commit d3398ff05907167f463e119421b053ce043741d1
 ( ASoC: Convert WM8753 to direct regmap API usage ) introduced
 the problem.

Thanks to Lars-Peter Clausen for helping me a bit during the debugging.

Signed-off-by: Denis 'GNUtoo' Carikli <GNUtoo@no-log.org>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: core: Split the union for CODEC/platform in the DAI

There's now core code which falls back to global CODEC operations for
DAI calls that needs to be able to tell if it's dealing with a CPU or
CODEC DAI and given the small number of DAIs in a typical system and
overall memory usage pattern saving a pointer per DAI is really not
worth the effort.

Reported-by: Ian Lartey <ian@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Liam Girdwood <lrg@ti.com>

ASoC: sgtl5000: rename device tree binding document

It moves and renames sgtl5000 device tree binding document to make
it aligned with other codecs.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: imx: move SND_SOC_AC97_BUS selection down to machine driver

SND_SOC_AC97_BUS is selected to enable the AC97 support in soc-core.
Rather than selecting the option under SND_IMX_SOC, it's better to
leave the selection to individual machine driver which knows if AC97
support is needed or not.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: imx: initialize dma_params burstsize just in imx-ssi

It's not necessary for imx-pcm-dma-mx2 to access imx_ssi.dma_params
for burstsize initialization.  Instead, it can just be done in imx-ssi
probe function once.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8994: Factor out WM1811A detection mode setting

Push everything through one function for active use cases, should be
no practical effect.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8994: Don't bother updating the jackdet mode needlessly

If we're not doing jackdet it's not needed.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8994: Disable JACKDET when disabling detecton

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8994: Make sure we sync DAPM on WM8958 detection mode changes

Normally this will have no effect as we set detection up at system startup
before DAPM syncs take effect, this will only be useful if the system
enables and disables detection at runtime.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm_hubs: Disable cache of the DC servo calibration for WM1811

The WM1811 DC servo is able to run much faster than previous devices so
the benefit of skipping calibration is not useful.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8994: Suppress noop updates of FLL K

Using snd_soc_write() means we always write to the register even if it
already contains the newly calculated value.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8994: Push wm8994 private data allocation out into device probe

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: core: Log a warning when machines use soc-audio

snd_soc_register_card() has been available and strongly preferred since
2.6.38 but we're still seeing new drivers using it and the conversion rate
for older machines has been low. Help address both issues by logging a
warning when the soc-audio device probes.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Liam Girdwood <lrg@ti.com>

ASoC: core: cleanup platform debugfs on probe failure.

Make sure we cleanup the platform debugfs when probe fails.

Signed-off-by: Liam Girdwood <lrg@ti.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: mxs-pcm: Use dmaengine PCM helper functions

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Tested-by: Shawn Guo <shawn.guo@linaro.org>
Acked-by: Vinod Koul <vinod.koul@linux.intel.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: imx-pcm-dma: Use dmaengine PCM helper functions

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Tested-by: Shawn Guo <shawn.guo@linaro.org>
Acked-by: Vinod Koul <vinod.koul@linux.intel.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: Add dmaengine PCM helper functions

This patch adds a set of functions which are intended to be used when
implementing a dmaengine based sound PCM driver.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Tested-by: Shawn Guo <shawn.guo@linaro.org>
Acked-by: Vinod Koul <vinod.koul@linux.intel.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8962: Run the headphone in class G mode when sidetone is enabled

Class W mode with sidetone is not fully supported.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8994: Disable debounce of jack detection on inserted jack

Don't debounce jack detection for inserted jacks, giving improved
responsiveness.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8962: Remove register default for PLL2

The initial value can be changed depending on system configuration.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: Samsung: Update email id of the author

I moved on from a great employer and the email-id no longer exists.
Update email-id to a personal one, assuming I don't move on from
myself anytime soon. And when I do, people don't get the eulogies
bounced.

Signed-off-by: Jaswinder Singh <jassisinghbrar@gmail.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: Samsung: Merge two identical if-else clauses

Saves two lines and a hell of a lot of embarrassment looking at the code.

Signed-off-by: Jassi Brar <jaswinder.singh@linaro.org>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8994: Remove stub of register access code

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm_hubs: Bomb out if we can't read back the DC servo result

Should have no practical impact but it's safer than trying to soldier on.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8996: Remove stub register cache

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8996: Fix /RESET bounce ordering

We want to leave the device out of rather than in reset.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8994: Make sure we don't have MICBIAS on during jackdet mode

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: DT: Update digital microphone binding documentation to PAZ00 board.

This patch updates device tree binding documentation to add digital
microphone to PAZ00 board.

Signed-off-by: Leon Romanovsky <leon@leon.nu>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: Remove unnecessary -codec from cs4270 driver name

Similar to what commit 1e3ad57 (ASoC: Remove redundant -codec from
WM8776 driver name) does for wm8776 driver, this patch does the same
thing for cs4270 driver.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Acked-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: fsl: align mpc8610_hpcd with p1022_ds on getting codec node

Align mpc8610_hpcd with p1022_ds on getting codec node by just calling
of_parse_phandle.  The bonus point of doing that is we can save
exporting get_node_by_phandle_name() when we consolidate the common
bits between mpc8610_hpcd and p1022_ds into a module, which can be
shared by more machine drivers added later.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Acked-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: fsl: correct get_dma_channel parameter name

The second parameter of function get_dma_channel is actually a property
name rather than a compatible string, so rename it for less confusing.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Acked-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8753: Convert to devm_kzalloc()

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: fix trivial build error in mpc5200_dma.c

Add the obvious header to fix this:

sound/soc/fsl/mpc5200_dma.c:301: error: implicit declaration of function 'DMA_BIT_MASK'
sound/soc/fsl/mpc5200_dma.c:301: error: initializer element is not constant

Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: imx: let SND_MXC_SOC_FIQ select FIQ

CONFIG_FIQ is only needed when CONFIG_SND_MXC_SOC_FIQ is selected to
build imx-pcm-fiq.c, so let SND_MXC_SOC_FIQ select FIQ.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: core: Don't overwrite .poweroff in snd_soc_pm_ops

SET_SYSTEM_SLEEP_PM_OPS writes .poweroff = *_resume once. Then we overwrite it
again explicitly as .poweroff = snd_soc_poweroff. Even though it works, as the
second one overwrites the first one, this is not the correct way. Fix this by
expanding SET_SYSTEM_SLEEP_PM_OPS in our structure.

Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8962: Convert interrupt handler to direct regmap usage

Avoids potential locking issues with anything that needs the CODEC lock.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8962: Remove mistakenly committed debug logging

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: ak4104: Convert to direct regmap API usage

Since the cache is currently open coded this is more of a win than for
most devices.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Daniel Mack <zonque@gmail.com>

ASoC: ak4104: Use snd_soc_write() rather than internal write function

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Daniel Mack <zonque@gmail.com>

ASoC: ak4104: Use snd_soc_update_bits() for read/modify/write

Don't use the internal I/O functions directly.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Daniel Mack <zonque@gmail.com>

ASoC: mxs-pcm: Request DMA channel early

Request the DMA channel in the PCM open callback instead of the hwparams
callback, this allows us to let open fail if no dma channel is available. This
also fixes a bug where the channel will be requested multiple times if hwparams
is called multiple times.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Tested-by: Shawn Guo <shawn.guo@linaro.org>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: imx-pcm: Request DMA channel early

Request the DMA channel in the pcm open callback. This allows us to let open
fail if there is no dma channel available.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Tested-by: Shawn Guo <shawn.guo@linaro.org>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: imx-ssi: Set dma data early

Move the call to snd_soc_dai_set_dma_data from the hw_params callback to the
startup callback. This allows us to use the dma data in the pcm driver's open
callback.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Tested-by: Shawn Guo <shawn.guo@linaro.org>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8994: Move wm_hubs callback before we start ramping VMID

Allows the generic code to set up for that.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8996: Convert to use DAPM routes for stream connections

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm5100: Implement DRC coefficient configuration

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8996: Implement DRC coefficient configuration

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: core: Add support for masking out parts of coefficient blocks

Chip designers frequently include things like the enable and disable
controls for algorithms in the register blocks which also hold the
coefficients. Since it's desirable to split out the enable/disable
control from userspace the plain SND_SOC_BYTES() isn't optimal for
these devices.

Add a SND_SOC_BYTES_MASK() which allows a bitmask from the first word
of the block to be excluded from the control. This supports the needs
of devices I've looked at and lets us have a reasonably simple API.
Further controls can be added in future if that's needed.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Liam Girdwood <lrg@ti.com>

ASoC: core: Add SND_SOC_BYTES control for coefficient blocks

Allow devices to export blocks of registers to the application layer,
intended for use for reading and writing coefficient data which can't
usefully be worked with by the kernel at runtime (for example, due to
requiring complex and expensive calculations or being the results of
callibration procedures). Currently drivers are using platform data to
provide configurations for coefficient blocks which isn't at all
convenient for runtime management or configuration development.

Currently only devices using regmap are supported, an error will be
generated for any attempt to work with a byte control on a non-regmap
device. There's no fundamental block to other devices so support could
be added if required.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Liam Girdwood <lrg@ti.com>

ASoC: io: Retrieve val_bytes from the regmap API

Allow us to build infrastructure which needs to know the size of a value
without requiring regmap based drivers to supply this information to both
ASoC and regmap by asking regmap for the value.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Liam Girdwood <lrg@ti.com>

Merge tag 'topic/introspection' of git://git./linux/kernel/git/broonie/regmap into HEAD

New interfaces to allow other subsystems to gather information about the
regmap, allowing them to build further subsystem specific generic
features on top of the regmap.

Merged into ASoC in order to allow us to implement SND_SOC_BYTES_MASK()
controls which need to know the word size of the underlying registers.

ASoC: mxs-pcm: Remove unused fields from struct mxs_pcm_runtime_data

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: imx-pcm: Remove unused fields from imx_pcm_runtime_data struct

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: imx-pcm: Remove empty prepare callback

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8994: Support external capacitors on MICBIAS2 with jack detection

When an external capacitor is connected to MICBIAS2 on devices with
jack detection (which is not required but may be done in some systems)
then the loading may mean that better performance is obtained when
the microphone bias is enabled normally rather than using the low power
mode. Provide platform data allowing systems to indicate if they require
this.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8994: Actively discharge idle MICBIAS with jack detect

This minimises the chance of any external capacitors that are fitted
being discharged into headphones as they insert.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8994: Enable headphone startup mode 1 for WM1811 and WM8958

The latest recommendation for optimal performance.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: twl4030: Debug code cleanup

Replace the printk(KERN_ERR* instances with dev_err in the driver.
While we are here clean up some of the debug messages as well.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm_hubs: Convert headphone driver to output driver widget

Mostly for neatness, though it may help with sequencing in some
situations.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: wm8776: Add WM8775 device ID to the WM8776

The WM8775 is register compatible with the WM8776 so can be supported with
the same driver though it is an ADC only part. Add the device ID to the
WM8776 driver, further updates will be added in the future.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: ak4104: Convert to module_spi_driver()

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: ak4104: Remove uninformative print on probe()

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

regmap: Allow users to query the size of register values

Generic infrastructure based on top of regmap may want to operate on
blocks of data and therefore find it useful to find the size of the
register values. Provide an accessor operation for this.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: imx/mxs: remove redundant SND_PCM selection

SND_PCM is already selected by SND_SOC, there is no need for
SND_IMX_SOC and SND_MXS_SOC to select it again.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

ASoC: fsi: Add DMAEngine support

This patch supports DMAEngine to FSI driver.
It supports only Tx case at this point.
If platform/cpu doesn't support DMAEngine, FSI driver will
use PIO transfer.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

Merge tag 'v3.3-rc4' into for-3.4 in order to resolve the conflict
resolved below within the FSI driver and allow the application of the
dmaeengine conversion that depends on this resolution.

Linux 3.3-rc4

Conflicts:
sound/soc/sh/fsi.c

ASoC: Add __devinit annotation for pxa2xx_ac97_probe

This fixes below build warning:
WARNING: vmlinux.o(.text+0x1e632c): Section mismatch in reference from the function pxa2xx_ac97_probe() to the function .devinit.text:pxa2xx_ac97_hw_probe()
The function pxa2xx_ac97_probe() references
the function __devinit pxa2xx_ac97_hw_probe().
This is often because pxa2xx_ac97_probe lacks a __devinit
annotation or the annotation of pxa2xx_ac97_hw_probe is wrong.

Also rename pxa_ac97_dai to pxa_ac97_dai_driver to fix below build warning:

  LD      sound/soc/pxa/built-in.o
WARNING: sound/soc/pxa/built-in.o(.data+0x18c): Section mismatch in reference from the variable pxa_ac97_dai to the function .devinit.text:pxa2xx_ac97_probe()
The variable pxa_ac97_dai references
the function __devinit pxa2xx_ac97_probe()
If the reference is valid then annotate the
variable with __init* or __refdata (see linux/init.h) or name the variable:
*driver, *_template, *_timer, *_sht, *_ops, *_probe, *_probe_one, *_console

Signed-off-by: Axel Lin <axel.lin@gmail.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>