Certain registers require patching after the SYSCLK has been brought up
add support for this into the CODEC driver.
Signed-off-by: NCharles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: NMark Brown <broonie@linaro.org>
Cc: stable@vger.kernel.org

Add device tree support to tpa6130a2 driver and document the
bindings.
Signed-off-by: NSebastian Reichel <sre@debian.org>
Signed-off-by: NMark Brown <broonie@linaro.org>

The conversion of the si476x to regmap removed locking of the core during
register updates, allowing things like power state changes for the MFD to
happen during a register update. Avoid this by taking the core lock in the
DAI operations (which are the only things that do register updates) as we
used to do in the open coded register I/O functions.
Signed-off-by: NMark Brown <broonie@linaro.org>
Acked-by: NAndrey Smirnov <andrew.smirnov@gmail.com>

This patch adds support for device tree for the CS42L73 CODEC
Signed-off-by: NBrian Austin <brian.austin@cirrus.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

There is no need for the CODEC to go to standby on resume since the core will
power it up as needed and in any case it is an idle_bias_off CODEC so would
normally sit with bias off while idle.
Signed-off-by: NMark Brown <broonie@linaro.org>
Tested-by: NStephen Warren <swarren@nvidia.com>

If we have control over the LDO then disable it during suspend; the device
is already being put into reset so will be non-functional over suspend
anyway and this will save a small amount of power.
Signed-off-by: NMark Brown <broonie@linaro.org>
Tested-by: NStephen Warren <swarren@nvidia.com>

This is more idiomatic and is required for robust operation since we must
ensure that the clocking configuration is valid as rapidly as possible.
Signed-off-by: NMark Brown <broonie@linaro.org>

This is more idiomatic and fixes bugs in the error handling paths.
Signed-off-by: NMark Brown <broonie@linaro.org>

Cleanup to namespace the defines for the cs42l73 driver
Signed-off-by: NBrian Austin <brian.austin@cirrus.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

Add support for RST GPIO and Charge Pump Freq in platform data
Signed-off-by: NBrian Austin <brian.austin@cirrus.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

'of_match_ptr' is defined in linux/of.h. Include it explicitly.
Signed-off-by: NSachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: NMark Brown <broonie@linaro.org>

'of_match_ptr' is defined in linux/of.h. Include it explicitly.
Signed-off-by: NSachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: NMark Brown <broonie@linaro.org>

'of_match_ptr' is defined in linux/of.h. Include it explicitly.
Signed-off-by: NSachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: NMark Brown <broonie@linaro.org>

'of_match_ptr' is defined in linux/of.h. Include it explicitly.
Signed-off-by: NSachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: NMark Brown <broonie@linaro.org>

'of_match_ptr' is defined in linux/of.h. Include it explicitly.
Signed-off-by: NSachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: NMark Brown <broonie@linaro.org>

According to the datasheet, the max_register is register 23.
Signed-off-by: NAxel Lin <axel.lin@ingics.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

According to the datasheet, the max_register is 13h.
ARRAY_SIZE(pcm1681_reg_defaults) + 1 is 18 which is wrong.
Signed-off-by: NAxel Lin <axel.lin@ingics.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

According to the datasheet, the max_register is 13h.
ARRAY_SIZE(pcm1681_reg_defaults) + 1 is 18 which is wrong.
Signed-off-by: NAxel Lin <axel.lin@ingics.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

Add more infrastructure (i.e. routes, muxes, switches) to the mc13783 DAPM.
Signed-off-by: NSteffen Trumtrar <s.trumtrar@pengutronix.de>
Signed-off-by: NMark Brown <broonie@linaro.org>

Add more controls to the alsa mixer infrastructure.
Signed-off-by: NSteffen Trumtrar <s.trumtrar@pengutronix.de>
Signed-off-by: NMark Brown <broonie@linaro.org>

The two paths were not connected in the DAPM route causing the associated
routes to be non working and the following warnings printed in the logs:
tlv320aic3x-codec 0-001b: ASoC: mux Right Line1L Mux has no paths
tlv320aic3x-codec 0-001b: ASoC: mux Left Line1R Mux has no paths
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

By using the new virtual DAPM mixer controls it is possible to remove the
twl4030 specific implementation of virtual controls.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Acked-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

By using the new virtual DAPM mixer controls it is possible to remove the
twl6040 specific implementation of virtual controls.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Acked-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

When going to suspend, shut down all channels and re-do the init
procedure at resume time.
Signed-off-by: NDaniel Mack <zonque@gmail.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

We'll need to call code to initialize and reset the codec again at
resume time, so factor it out first.
Signed-off-by: NDaniel Mack <zonque@gmail.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

We need to access the charge_period and start_mid_z values from other
places later, so move them to the private struct.
Signed-off-by: NDaniel Mack <zonque@gmail.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NMark Brown <broonie@linaro.org>

Reset the device on probe to make sure that the register settings match the
register cache defaults.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NMark Brown <broonie@linaro.org>

There is no such register.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NMark Brown <broonie@linaro.org>

Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NMark Brown <broonie@linaro.org>

The current si476x I/O implementation wraps the regmap for the core with
functions that make the register map cache only when the device is powered
down. This implementation appears to be incomplete since there is no code
to synchronise the cache so writes done while the core is powered down
will be ignored, the device will only be configured if it is powered.

A better and more idiomatic approach would be to have the MFD manage the
cache, making the device cache only when it powers things down. This also
allows ASoC to use the standard regmap helpers for the device which helps
remove the ASoC custom ones so do convert to do that.
Signed-off-by: NMark Brown <broonie@linaro.org>

This moves us towards being able to remove the duplicate register I/O
functionality in ASoC and saves some code.
Signed-off-by: NMark Brown <broonie@linaro.org>

This is more idiomatic and interacts better with deferred probe.
Signed-off-by: NMark Brown <broonie@linaro.org>

This moves us towards being able to remove the duplicated register I/O
functionality in ASoC.
Signed-off-by: NMark Brown <broonie@linaro.org>

As part of a push to remove the register I/O functionality from ASoC (since
it is now duplicated in the regmap API) convert the mc13783 driver to use
regmap directly.
Signed-off-by: NMark Brown <broonie@linaro.org>

Move the workaround for double sending AUDIO_CODEC and AUDIO_DAC writes
into the SPI core, aiding refactoring to eliminate the ASoC custom I/O
functions and avoiding the extra writes for I2C.
Signed-off-by: NMark Brown <broonie@linaro.org>
Signed-off-by: NLee Jones <lee.jones@linaro.org>

This moves us towards being able to remove the duplicated register I/O
code in ASoC.

The datasheet and the driver document the device as having a register map
divided into pages but since the paging is actually done by sending the
page address and the register address with each transaction this is no
different to having a simple register address. The datasheet does also
document the low five bits of the 16 bit "command" as unused which we could
represent as padding but it seems simpler and less confusing to things
that use block transfers or autoincrement to represent these as part of
the register address.
Signed-off-by: NMark Brown <broonie@linaro.org>

Use snd_soc_update_bits() rather than open coding. Since the register cache
is currently only used where update_bits() is used this means the current
register cache can be removed entirely.
Signed-off-by: NMark Brown <broonie@linaro.org>

Signed-off-by: NMark Brown <broonie@linaro.org>

This moves us towards being able to remove the duplicated register I/O
code in ASoC.
Signed-off-by: NMark Brown <broonie@linaro.org>