I2S2_DAC pin can be used for I2S or GPIO. We should set it as GPIO
if we use GPIO5 as DMIC1 data pin.
Signed-off-by: NBard Liao <bardliao@realtek.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

This fixes kernel panic on boot, if rt5645->codec is NULL when
rt5645_jack_detect_work is first called.

rt5645_jack_detect_work needs rt5645->codec to be initialized to setup
dapm pins. Also, reporting jack events is useless, as the jacks cannot
be set before the codec is ready.

Since we manually call the interrupt handler in
rt5645_set_jack_detect, the initial jack state will be reported
correctly, and dapm pins will be setup at that time.
Signed-off-by: NNicolas Boichat <drinkcat@chromium.org>
Signed-off-by: NMark Brown <broonie@kernel.org>

The rate controls are codec-specific, it's not possible to
generically say what the range or the meaning of each control
is (or even if they exist at all) - that depends on the
particular codec.

This is currently being handled for Arizona codecs by putting
an Arizona-specific table of controls inside the wm_adsp driver.
This creates a dependency between wm_adsp and arizona.c, and is an
awkward solution if the ADSP is used in another family of codecs

Fix this by moving the Arizona-specific rate controls into the
Arizona codec drivers.
Signed-off-by: NRichard Fitzgerald <rf@opensource.wolfsonmicro.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Fix trivial typo.
Signed-off-by: NAxel Lin <axel.lin@ingics.com>
Acked-by: NCharles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Developing a driver for an Asus X205TA laptop I get these dmesg
errors:

 rt5645 i2c-10EC5648:00: ASoC: mux RT5650 IF1 ADC1 Swap Mux has no paths
 rt5645 i2c-10EC5648:00: ASoC: mux RT5650 IF1 ADC2 Swap Mux has no paths
 rt5645 i2c-10EC5648:00: ASoC: mux RT5650 IF1 ADC3 Swap Mux has no paths
 rt5645 i2c-10EC5648:00: ASoC: mux RT5650 IF1 ADC Mux has no paths
 rt5645 i2c-10EC5648:00: ASoC: mux RT5650 IF1 DAC1 L Mux has no paths
 rt5645 i2c-10EC5648:00: ASoC: mux RT5650 IF1 DAC1 R Mux has no paths
 rt5645 i2c-10EC5648:00: ASoC: mux RT5650 IF1 DAC2 L Mux has no paths
 rt5645 i2c-10EC5648:00: ASoC: mux RT5650 IF1 DAC2 R Mux has no paths

so, move these muxes to the rt5650_specific_dapm_widgets[] list.
Signed-off-by: NMichele Curti <michele.curti@gmail.com>
Signed-off-by: NOder Chiou <oder_chiou@realtek.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Constify the ACPI device ID array and the register map, no need to have
them writable at runtime. Also drop the unneeded RT5670_INIT_REG_LEN
define.
Signed-off-by: NMathias Krause <minipli@googlemail.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Constify the ACPI device ID array and the register map, no need to have
them writable at runtime.
Signed-off-by: NMathias Krause <minipli@googlemail.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Constify the ACPI device ID array and the register map, no need to have
them writable at runtime. Also drop the unneeded RT5640_INIT_REG_LEN
define.
Signed-off-by: NMathias Krause <minipli@googlemail.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Current code has duplicate code for 16000, 32000 and 48000 sample rates.
get_srate() returns negative error code for unsupported rate, so we can
remove the duplicate code in the swith cases by calling get_srate() first.
Signed-off-by: NAxel Lin <axel.lin@ingics.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

When running in master mode the bclk divider must be configured to generate a
sane bitclock. Pick the smallest fs multiplicator, which can hold all
transmitted bits.
Signed-off-by: NJulian Scheel <julian@jusst.de>
Acked-by: NCharles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

the enum of "DAC Polarity" should be wm8960_enum[1].
Signed-off-by: NZidan Wang <zidan.wang@freescale.com>
Acked-by: NCharles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: NMark Brown <broonie@kernel.org>
Cc: stable@vger.kernel.org

Since hp-detect is optional, use devm_gpiod_get_optional instead.
In additional, it should return error if devm_gpiod_get_optional fails.
Signed-off-by: NAxel Lin <axel.lin@ingics.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Modify the RT5645 driver to parse platform data from device tree. This is
missing from previous patch in sound/soc/codecs/rt5645.c, that was present
in v3.
Signed-off-by: NBard Liao <bardliao@realtek.com>
Signed-off-by: NOder Chiou <oder_chiou@realtek.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

The patch separates the IO function from the rt286. It is prepared to share
for new chips that support the same IO function.
Signed-off-by: NOder Chiou <oder_chiou@realtek.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

commit ea178d14 ("ASoC: tas2552: Make the enable-gpio really optional")
makes enable-gpio optional. devm_gpiod_get_optional() is the better
function for optional gpio, so let's switch to use it.
Signed-off-by: NAxel Lin <axel.lin@ingics.com>
Acked-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

This patch adds some debugfs nodes to get information
about the currently running firmware.
Signed-off-by: NRichard Fitzgerald <rf@opensource.wolfsonmicro.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Now that we have a codec_probe stage initialization in the wm_adsp
driver, we can make the wm_adsp driver create its own ALSA controls
instead of having that responsibility pushed to every codec driver.
Signed-off-by: NRichard Fitzgerald <rf@opensource.wolfsonmicro.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Currently the only init function in wm_adsp is called by the
codec driver early in its probe before the codec has been
registered with SOC.

This patch adds stubs for the codec_probe and codec_remove stages
and calls them from WM5102 and WM5110 codec drivers. This allows us
to hang anything that needs setup during the codec probe stage off
these functions without further modification of the codec drivers.
Signed-off-by: NRichard Fitzgerald <rf@opensource.wolfsonmicro.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Move gpio to gpio_desc and use gpiod APIs in codec driver.
Signed-off-by: NBard Liao <bardliao@realtek.com>
Signed-off-by: NOder Chiou <oder_chiou@realtek.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Register RT5650_TDM_CTRL_4(0x7A) is readable and used for mixer
setting. It should be added in rt5645_readable_register function.
Signed-off-by: NBard Liao <bardliao@realtek.com>
Signed-off-by: NOder Chiou <oder_chiou@realtek.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

rt5645_enable_push_button_irq uses snd_soc_dapm_*_unlocked
functions, so it needs to lock the required dapm mutex.
Signed-off-by: NNicolas Boichat <drinkcat@chromium.org>
Signed-off-by: NMark Brown <broonie@kernel.org>

The set of supported sample rates depends on the master clock supplied
to the codec. Allow the machine driver to set the required master clock
in hw_params().
Signed-off-by: NSergej Sawazki <ce3a@gmx.de>
Acked-by: NCharles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

In tas2552_sw_shutdown() tas_data is used while the rest of the driver uses
tas2552 when dealing with the 'struct tas2552_data'
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Double semicolon was added by the following commit:
ea178d14 ASoC: tas2552: Make the enable-gpio really optional
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

The module can not be loaded again after it has been removed.
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Align the numbers in the header file to the same column.
At the same time change the wrapping of CFG_2 register write in the probe
function to be uniform with the other calls.
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Correct the bit definition so the code will change the bits what it
supposed to change. Also rename the register define to
TAS2552_BOOST_APT_CTRL.
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Do not write to DOUT Tristate register at probe time, specially not write
data which is defined to be used in Output Data Register.
Fix the defines for the Output Data Register and correct the register write
at probe time.
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

'DIN source' enum can be used to select the DIN Source (muted, left, right
or average of left and right channels).
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Do not restrict the sampling rate to 44.1/48KHz. The pll_clk clock should
be (sampling rate * 512) in all cases.
Correct the J.D calculation (the D part was incorrectly calculated).
Restore PLL enable status after we are done with the configuration.
Implement hardware constraint handling towards the pll_clkin:
if D != 0 (in J.D) then 1.1MHz <= pll_clkin <= 9.2MHz needs to be checked.
If the PLL setup does not met with this constraint, fall back to BCLK as
reference clock, if BCLK fails, use the internal 1.8MHz clock.
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Instead of hard wiring the WCLK frequency at probe time do it runtime.
The hard wired 88_96KHz was not even setting the correct bits since it was
defined as (1 << 6) which will  change the I2S_OUT_SEL bit and will leave
the amplifier configured for 8KHz.
At the same time clean up and fix the CFG3 register bits.
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Configure the word length based on the params_width of the stream.
Also configure the clock per frame value which is used when tas2552 is bus
master.
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Certain sequence need to be followed in order to have smooth power up and
power down performance.
Execute this sequence via DAPM_POST widget.
Remove patching the RESERVED_0D register at probe time since it has to be
handled every time when we stop or start the amplifier.
In order to be able to execute the sequence at the correct time, the driver
need to request to ignore the pmdown time.
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

The last parameter for DECLARE_TLV_DB_SCALE() is to tell if the gain will
be muted or not when it is set to raw 0. IN this case it is not muted.
The PGA_GAIN is in 0-4 bits in the register. Fix the offset in the
SOC_SINGLE_TLV() for this.
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

The strings should be: 'static const char * const tas2552_input_texts[]'
SOC_DAPM_ENUM should have "Route" in place of xname and no need to have it
as an array.
Also align the parameters.
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

TDM support is achieved using DSP transfer mode and setting a programmable
offset which specifies where data begins with respect to the frame sync.
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Use names from the datasheet for the definitions.
Correct the data format definitions since they were not correct.
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

DSP_A mode require one bit delay from the FS, DSP_B is without data delay.
When checking the requested format, also match the bit and fs inversion
flag along with the format since it is not possible to change inversion.
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Instead of hard wiring the PLL_CLKIN and PDM_CLK to be sourced from BCLK add
proper clock configuration via the set_dai_sysclk callback.
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Prevents frequent panic on boot, if the irq handler rt5645_irq
gets called before the workqueue rt5645_jack_detect_work is
initialized.
Signed-off-by: NNicolas Boichat <drinkcat@chromium.org>
Signed-off-by: NMark Brown <broonie@kernel.org>
Cc: stable@vger.kernel.org