Mode Control - Register 02h
Digital Interface Format (DIF[2:0]) Bits 6-4
DIF2 DIF1 DIF0 Description
0 0 0 Left-Justified, up to 24-bit data
0 0 1 I²S, up to 24-bit data
0 1 0 Right-Justified, 16-bit data
0 1 1 Right-Justified, 24-bit data
1 0 0 TDM slot 0
1 0 1 TDM slot 1
1 1 0 TDM slot 2
1 1 1 TDM slot 3
The DIF_MASK is 0x70, so current code does not correctly set the DIFx
setting. Fix it.
Signed-off-by: Axel Lin <axel.lin@ingics.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
rtd->codec does not necessarily point to the CODEC instance for which the
callback was called (e.g. for CODEC<->CODEC or multi-CODEC links). Use
dai->codec instead.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
dev_get_drvdata() will not return the snd_soc_runtime to which this CODEC
is attached, so the current code will result in undefined behavior.
To fix this just use regmap_update_bits(cs4349->regmap, ...) directly
instead of snd_soc_update_bits(rtd->codec, ...).
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
"ASoC: rt5645: Check if codec is initialized in workqueue handler"
adds a check if codec is NULL in rt5645_irq_detection, which
returns an int. However, "ASoC: rt5645: Remove irq_jack_detection
function" removes that function, and moves the code in
jack_detect_work, which returns void.
Remove the return value to fix compilation warning.
Signed-off-by: Nicolas Boichat <drinkcat@chromium.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
The max_register should be the maximum valid register index rather than
number of registers. Also remove unused defines.
Signed-off-by: Axel Lin <axel.lin@ingics.com>
Acked-by: Tim Howe <tim.howe@cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
regmap: Create a new struct reg_sequence for register sequences
In order to allow us to start adding extra annotations for sequences
without bloating register default tables duplicate the structure under
the new name reg_sequence and update the APIs to use that instead of
reg_default.
Conflicts:
sound/soc/codecs/da7210.c
sound/soc/codecs/rt5651.c
sound/soc/codecs/wm8993.c
Introduced by:
commit 8019ff6cfc
("regmap: Use reg_sequence for multi_reg_write / register_patch")
Interacting with:
commit d1acd31883
("ASoC: wm5110: Add special DRE on/off handling for the headphone path")
Signed-off-by: Nariman Poushin <nariman@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
regmap: Create a new struct reg_sequence for register sequences
In order to allow us to start adding extra annotations for sequences
without bloating register default tables duplicate the structure under
the new name reg_sequence and update the APIs to use that instead of
reg_default.
The driver does not free irq when snd_soc_register_codec returns error.
It does not return error when request irq failed, either.
Add return when request irq failed, and free_irq if
snd_soc_register_codec failed.
Signed-off-by: Koro Chen <koro.chen@mediatek.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
"ASoC: rt5645: Simplify rt5645_enable_push_button_irq" removes
the test that accessed rt5645->pdata.jd_mode (that test is now
done in rt5645_jack_detect only), so we do not need that
variable anymore.
Signed-off-by: Nicolas Boichat <drinkcat@chromium.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
The dev_pm_ops variables are not modified after initialization in these
drivers, so make them const.
Signed-off-by: Axel Lin <axel.lin@ingics.com>
Signed-off-by: Mark 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: Nicolas Boichat <drinkcat@chromium.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
Separate the functionality using sequences of register writes from the
functions that take register defaults. This change renames the arguments
in order to support the extension of reg_sequence to take an optional
delay to be applied after any given register in a sequence is written.
This avoids adding an int to all register defaults, which could
substantially increase memory usage for regmaps with large default tables.
This also updates all the clients of multi_reg_write/register_patch.
Signed-off-by: Nariman Poushin <nariman@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The driver does not use any CODEC specific constructs anymore. Convert it
to snd_soc_component.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
The driver does not use any CODEC specific constructs anymore. Convert it
to snd_soc_component.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
Use a stack local variable to handle function local state rather than a
global static variable. The later has a potential for race conditions if
the probe function runs for two devices concurrently.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
LDO2/Mic Det Power pins are already enabled/disabled in rt5645_jack_detect
(the jack out code path previously did not disable those if button function
is enabled: modify it to make it so).
Signed-off-by: Nicolas Boichat <drinkcat@chromium.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
This makes sure the dapm state is consistent when the card is
instantiated.
However, if the card is not instantiated yet, we still update the
registers manually in the "jack in" case, so that we can immediately
report if a mic is present or not.
Disabling "Mic Det Power" after detection, and on jack out, can
wait until the card gets instantiated.
Signed-off-by: Nicolas Boichat <drinkcat@chromium.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
This line doesn't include any information. So, remove it.
Signed-off-by: Bard Liao <bardliao@realtek.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Fixes for compilation warnings and errors reported by Kbuild test robot
for x86_64 and i386.
Signed-off-by: Arnaud Pouliquen <arnaud.pouliquen@st.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
i2c_driver does not need to set an owner because i2c_register_driver()
will set it.
Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
da732x_dai1_ops is identical to da732x_dai2_ops, so merge them to
da732x_dai_ops. The da732x_dai_ops is not modified after initialization,
so make it const.
Signed-off-by: Axel Lin <axel.lin@ingics.com>
Acked-by: Adam Thomson <Adam.Thomson.Opensource@diasemi.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The snd_soc_dai_ops variables are not modified after initialization in
these drivers, so make them const.
Signed-off-by: Axel Lin <axel.lin@ingics.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
irq_jack_detection is only called from rt5645_jack_detect_work:
remove the function to simplify the code.
Signed-off-by: Nicolas Boichat <drinkcat@chromium.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
The control_data field of the snd_soc_codec struct is no longer used by
ASoC drivers using standard regmap IO.
There is no need to initialize the field to point to the drivers regmap
struct, so drop that.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
gpiolib/asoc system already prints enough info if there are any problems.
Signed-off-by: Anatol Pomozov <anatol.pomozov@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Spec does not say anything about DAC called SDMode.
It makes more sense to use DAI input that created automatically by DAPM
and route it to Speaker output.
Signed-off-by: Anatol Pomozov <anatol.pomozov@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The option is not needed if chip is always on or managed by some other part
of system like platform card driver.
Signed-off-by: Anatol Pomozov <anatol.pomozov@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
sound/soc/codecs/sti-sas.c:616:3-8: No need to set .owner here. The core will do it.
Remove .owner field if calls are used which set it automatically
Generated by: scripts/coccinelle/api/platform_no_drv_owner.cocci
Signed-off-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Both devm_regmap_init and syscon_regmap_lookup_by_phandle return ERR_PTR
on failure.
Signed-off-by: Axel Lin <axel.lin@ingics.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Instead of doing the read-modify-update cycle by hand when updating a
register use regmap_update_bits().
This also means we can now remove uda134x_read_reg_cache() and
uda134x_write() since they are unused.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
Use regmap rather then the legacy ASoC IO for the uda134x driver.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
The UDA134X family extends the rather limited L3 register set by using part
of the register value as additional address bits.
These extra address bits are currently stored in the default register cache
and rely on them remaining constant. While this works it is rather
idiomatic and slightly confusing. Change it so that the additional address
bits are explicitly added when writing a register.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
Resource allocations should be done in the bus probe rather than the CODEC
probe. Move the allocation of the drivers state struct there.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
Codec part of the sti platform that supports codec IPs.
This first version does not support HDMI, but only DAC and SPDIF out.
Signed-off-by: Arnaud Pouliquen <arnaud.pouliquen@st.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
If we fail to add some DSPs or fail to add the controls we should call
wm_adsp2_codec_remove for all the cores we have already added. This
patch fixes this up on the wm5102 and wm5110.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
For the best performance the headphone path enable/disable must be
handled specially for the situations of DRE on and DRE off.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
This reverts ed043aebe6 ("ASoC: wm8996: Pass the IRQF_ONESHOT flag").
The coccinelle warnings is false positive because the original code does
set IRQF_ONESHOT by "trigger |= IRQF_ONESHOT;".
Signed-off-by: Axel Lin <axel.lin@ingics.com>
Acked-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
