idle_bias_on was set because cs42l42 has a "VMID" type pseudo-midrail
supply (named FILT+), and these typically take a long time to charge.
But the driver never enabled pm_runtime so it would never have powered-
down the cs42l42 anyway.
In fact, FILT+ can charge to operating voltage within 12.5 milliseconds
of enabling HP or ADC. This time is already covered by the startup
delay of the HP/ADC.
The datasheet warning about FILT+ taking up to 1 second to charge only
applies in the special cases that either the PLL is started or
DETECT_MODE set to non-zero while both HP and ADC are off. The driver
never does either of these.
Removing idle_bias_on allows the Soundwire host controller to suspend
if there isn't a snd_soc_jack handler registered.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Signed-off-by: Stefan Binding <sbinding@opensource.cirrus.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20230127165111.3010960-8-sbinding@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
This adds support for using CS42L42 as a SoundWire device.
SoundWire-specifics are kept separate from the I2S implementation as
much as possible, aiming to limit the risk of breaking the I2C+I2S
support.
There are some important differences in the silicon behaviour between
I2S and SoundWire mode that are reflected in the implementation:
- ASP (I2S) most not be used in SoundWire mode because the two interfaces
share pins.
- The SoundWire capture (record) port only supports 1 channel. It does
not have left-to-right duplication like the ASP.
- DP2 can only be prepared if the HP has powered-up. DP1 can only be
prepared if the ADC has powered-up. (This ordering restriction does
not exist for ASPs.) The SoundWire core port-prepare step is
triggered by the DAI-link prepare(). This happens before the
codec DAI prepare() or the DAPM sequence so these cannot be used
to enable HP/ADC. Instead the HP/ADC enable/disable are done during
the port_prep callback.
- The SRCs are an integral part of the audio chain but in silicon their
power control is linked to the ASP. There is no equivalent power link
to SoundWire DPs so the driver must take "manual" control of SRC power.
- The SoundWire control registers occupy the lower part of the SoundWire
address space so cs42l42 registers are offset by 0x8000 (non-paged) in
SoundWire mode.
- Register addresses are 8-bit paged in I2C mode but 16-bit unpaged in
SoundWire.
- Special procedures are needed on register read/writes to (a) ensure
that the previous internal bus transaction has completed, and
(b) handle delayed read results, when the read value could not be
returned within the SoundWire read command.
There are also some differences in driver implementation between I2S
and SoundWire operation:
- CS42L42 I2S does not runtime_suspend, but runtime_suspend/resume support
has been added into the driver in SoundWire mode as the most convenient
way to power-up the bus manager and to handle the unattach_request
condition, though the CS42L42 chip does not itself suspend or resume.
- Intel SoundWire host controllers have a low-power clock-stop mode that
requires resetting all peripherals when resuming. This means that the
interrupt registers will be reset in between the interrupt being
generated and the interrupt being handled, and since the interrupt
status is debounced, these values may not be accurate immediately,
and may cause spurious unplug events before settling.
- As in I2S mode, the PLL is only used while audio is active because
of clocking quirks in the silicon. For SoundWire the cs42l42_pll_config()
is deferred until the DAI prepare(), to allow the cs42l42_bus_config()
callback to set the SCLK.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Signed-off-by: Stefan Binding <sbinding@opensource.cirrus.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20230127165111.3010960-7-sbinding@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Export functions that will be needed by a SoundWire module.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Signed-off-by: Stefan Binding <sbinding@opensource.cirrus.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20230127165111.3010960-6-sbinding@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Setup of the ASP (audio serial port) was being done as a side-effect of
cs42l42_pll_config() and forces a restriction on the ratio of sample_rate
to bit_clock that is invalid for Soundwire.
Move the ASP setup into a dedicated function.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Signed-off-by: Stefan Binding <sbinding@opensource.cirrus.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20230127165111.3010960-5-sbinding@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The chosen clocking configuration must give an internal MCLK (MCLKint)
that is an integer multiple of the sample rate.
On I2S each of the supported bit clock frequencies can only be generated
from one sample rate group (either the 44100 or the 48000) so the code
could use only the bitclock to look up a PLL config.
The relationship between sample rate and bitclock frequency is more
complex on Soundwire and so it is possible to set a frame shape to
generate a bitclock from the "wrong" group. For example 2*147 with a
48000 sample rate would give a bitclock of 14112000 which on I2S
could only be derived from a 44100 sample rate.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Signed-off-by: Stefan Binding <sbinding@opensource.cirrus.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20230127165111.3010960-4-sbinding@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The SOFT_RESET_REBOOT register is needed to recover CS42L42 state after
a Soundwire bus reset.
This is required to be set whenever there is severe/hard bus reset.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Signed-off-by: Stefan Binding <sbinding@opensource.cirrus.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20230127165111.3010960-3-sbinding@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
After tip sense detects a jack insertion, if automatic
type detection, and manual type detection fails, then
fall back to assume the jack connected belongs to
headphones.
Signed-off-by: Stefan Binding <sbinding@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20220927121440.2506632-1-sbinding@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The driver wants to know the bit rate on the serial bus and takes that
to be the value set by 'set_sysclk'. The 'set_bclk_ratio' op is a better
fit for figuring out the clocking parameters of the serial bus, so
implement that and give it precedence over the prior methods.
Signed-off-by: Martin Povišer <povik+lin@cutebit.org>
Link: https://lore.kernel.org/r/20220915094444.11434-11-povik+lin@cutebit.org
Signed-off-by: Mark Brown <broonie@kernel.org>
The CS42L83 part is a headphone jack codec found in recent Apple
machines. It is a publicly undocumented part but as far as can be told
it is identical to CS42L42 except for two points:
* The chip ID is different.
* Of those registers for which we have a default value in the existing
CS42L42 kernel driver, one register (MCLK_CTL) differs in its reset
value on CS42L83.
To address those two points (and only those), add to the CS42L42 driver
a separate CS42L83 front.
Signed-off-by: Martin Povišer <povik+lin@cutebit.org>
Link: https://lore.kernel.org/r/20220915094444.11434-10-povik+lin@cutebit.org
Signed-off-by: Mark Brown <broonie@kernel.org>
Export the regmap callbacks for indicating readable/volatile registers,
also the range structure, to the CS42L42 core namespace. This is in
advance of reusing these bits in a CS42L83 driver frontend.
Signed-off-by: Martin Povišer <povik+lin@cutebit.org>
Link: https://lore.kernel.org/r/20220915094444.11434-9-povik+lin@cutebit.org
Signed-off-by: Mark Brown <broonie@kernel.org>
Split the I2C bus driver definition and probe()/remove() into a
separate module so that a SoundWire build of CS42L42 support does
not have a spurious dependency on I2C.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Signed-off-by: Martin Povišer <povik+lin@cutebit.org>
Link: https://lore.kernel.org/r/20220915094444.11434-8-povik+lin@cutebit.org
Signed-off-by: Mark Brown <broonie@kernel.org>
Pass pointers to snd_soc_component_driver and snd_soc_dai_driver
objects into cs42l42_common_probe().
This is in preparation for adding SoundWire support.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Signed-off-by: Martin Povišer <povik+lin@cutebit.org>
Link: https://lore.kernel.org/r/20220915094444.11434-7-povik+lin@cutebit.org
Signed-off-by: Mark Brown <broonie@kernel.org>
On SoundWire the system resume cannot restore registers until the
host controller has re-enumerated the peripheral.
This patch splits cs42l42_resume() into two functions, one to
power up and the other to restore registers, ready for adding
SoundWire support.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Signed-off-by: Martin Povišer <povik+lin@cutebit.org>
Link: https://lore.kernel.org/r/20220915094444.11434-6-povik+lin@cutebit.org
Signed-off-by: Mark Brown <broonie@kernel.org>
To prepare for adding SoundWire the probe must be split into three
parts:
1) The bus-specific probe
2) Common bus-agnostic probe steps
3) Initialization of the peripheral registers
Step (3) must be separate because on SoundWire devices the probe must
enable power supplies and release reset so that the peripheral can be
enumerated by the bus, but it isn't possible to access registers until
enumeration has completed.
The call to devm_snd_soc_register_component() must be done at stage (2)
so that it can EPROBE_DEFER if necessary. In SoundWire systems stage (3)
is not a probe event so a deferral at this stage would not result in
re-probing dependencies.
A new init_done flag indicates that the chip has been identified and
initialized. This is used to prevent cs42l42_remove(), cs42l42_suspend(),
cs42l42_restore() and cs42l42_irq_thread() from attempting register
accesses if the chip was not successfully initialized. Although this
cannot happen on I2C, because the entire probe would fail, it is
possible on SoundWire if probe succeeds but the cs42l42 is never
enumerated.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Signed-off-by: Martin Povišer <povik+lin@cutebit.org>
Link: https://lore.kernel.org/r/20220915094444.11434-5-povik+lin@cutebit.org
Signed-off-by: Mark Brown <broonie@kernel.org>
In preparation for splitting cs42l42_i2c_probe() into multiple functions
replace use of &i2c_client->dev with cs42l42->dev. This reduces diff
clutter in the patch that splits the function.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Signed-off-by: Martin Povišer <povik+lin@cutebit.org>
Link: https://lore.kernel.org/r/20220915094444.11434-4-povik+lin@cutebit.org
Signed-off-by: Mark Brown <broonie@kernel.org>
Clean up the handling of bitclock frequency by keeping all the logic
in cs42l42_pcm_hw_params(), which then simply passes the frequency as
an argument to cs42l42_pll_config().
The previous code had become clunky as a legacy of earlier versions of
the clock handling. The logic was split across cs42l42_pcm_hw_params()
and cs42l42_pll_config(), with the params-derived bclk stashed in
struct cs42l42_private only to pass it to cs42l42_pll_config().
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Signed-off-by: Martin Povišer <povik+lin@cutebit.org>
Link: https://lore.kernel.org/r/20220915094444.11434-3-povik+lin@cutebit.org
Signed-off-by: Mark Brown <broonie@kernel.org>
Use the new snd_soc_tdm_params_to_bclk() helper function to
calculate the bclk. This function handles most of the previous
manipulation and makes the code tidier.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20220817122347.1356773-6-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Fix "much change together" to "must change together".
It's probably obvious what was meant but it's nice to fix it.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20220817122347.1356773-5-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Only report a button state change if the interrupt status shows that
there was a button event.
Previously the code would always drop into the button reporting at the
end of interrupt handling if the jack was present. If neither of the
button report interrupts were pending it would report all buttons
released. This could then lead to a button being reported as released
while it is still pressed.
Fixes: c5b8ee0879 ("ASoC: cs42l42: Report jack and button detection")
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20220815123138.3810249-1-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The ASoC core has now been changed to default to the non-legacy DAI
naming, as such drivers using the new scheme no longer need to specify
the non_legacy_dai_naming flag.
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20220623125250.2355471-50-ckeepax@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The i2c probe functions here don't use the id information provided in
their second argument, so the single-parameter i2c probe function
("probe_new") can be used instead.
This avoids scanning the identifier tables during probes.
Signed-off-by: Stephen Kitt <steve@sk2.org>
Reviewed-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Reviewed-by: Lucas Tanure <tanureal@opensource.cirrus.com>
Acked-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20220325170734.1216018-1-steve@sk2.org
Signed-off-by: Mark Brown <broonie@kernel.org>
DETECT_MODE and PLL_START must be zero while HP_PDN and ADC_PDN are
both 1. If this condition is broken it can discharge FILT+ and it
can then take up to 1 second for FILT+ to recharge.
There is no workaround required for this, simply avoiding settings
and sequences that would break the requirement. The driver already
meets the requirement.
But it is not obvious from reading the code that this requirement
exists, or what is ensuring it is met. So it would not currently be
obvious to someone changing the code that there is certain special
behaviour that must be maintained.
To avoid accidental breakage in the future:
- Add comments into the register definitions to warn about this so
that anyone changing the code around DETECT_MODE and PLL_START is
aware of this requirement.
- Add a comment where PLL_START is written to 1 to highlight the
requirement and why it is satisfied.
- Add a comment in cs42l42_setup_hs_type_detect() when DETECT_MODE is
initialized.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20220304144015.398656-1-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Add system suspend functions to handle clean power-down on suspend and
restoring registers on resume.
The jack state could change during suspend. Plug->unplug and unplug->plug
are straightforward because this looks no different from any other plug
state change - there will be a plugged or unplugged interrupt pending.
The jack could be unplugged and a different type of jack plugged, and on
resume the plug state would not have changed. Setting plug_state back to
TS_TRANS (transitioning) will make the next plug interrupt after resume
run a type detection.
During system suspend any jack plug/unplug and button events will not be
reported or generate a system wakeup. If the plug state or headset type
has changed it will be reported after resume.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20220121120412.672284-4-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Rename jack_detect_mutex to irq_lock and make it lock the entire IRQ
handling.
The jack_detect_mutex was introduced to synchronize registering an
ALSA jack handler, via cs42l42_set_jack(), with the jack state
processing in the IRQ handler, and was taken only around the
relevant part of the IRQ handling code.
System suspend will need to synchronize with the IRQ handler thread
so will need a similar mutex that surrounds all of the IRQ handling.
Repurposing the existing jack_detect_mutex is the simplest option.
It does no harm for a call to cs42l42_set_jack() to additionally
block the first few lines of IRQ handling, and the only interrupts
used by the driver are all for jack handling.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20220121120412.672284-3-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
When a plug event is detect report the full state of all status
bits, don't assume that there will have been a previous unplug
event to clear all the bits. Report the state of both HEADPHONE
and MICROPHONE bits according to detected type, and clear all the
button status bits. The current button status is already checked
and reported at the end of the function.
During a system suspend the jack could be unplugged and plugged,
possibly changing the jack type. On resume the interrupt status will
indicate a plug event - there will not be an unplug event to clear
the bits.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20220121120412.672284-2-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
When a jack handler is registered in cs42l42_set_jack() the
initial state should be reported if an attached headphone/headset
has already been detected.
The jack detect sequence takes around 1 second: typically long
enough for the machine driver to probe and register the jack handler
in time to receive the first report from the interrupt handler. So
it is possible on some systems that the correct initial state was seen
simply because of lucky timing. Modular builds were more likely to
miss the reporting of the initial state.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Fixes: 4ca239f337 ("ASoC: cs42l42: Always enable TS_PLUG and TS_UNPLUG interrupts")
Link: https://lore.kernel.org/r/20211119124854.58939-1-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Now that struct cs42l42_private has pll_config, the current PLL
configuration can be looked up directly in pll_ratio_table. This
makes the pll_divout member of cs42l42_private redundant since it
was only a copy of the value from pll_ratio_table.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211116163901.45390-5-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
When reporting a jack unplug there's no need to make the reported
flags conditional on which flags were reported during the plug
event. It's perfectly safe to report all flags and buttons as
not-present and let the higher code filter for changes.
There's also no need to make two separate snd_soc_jack_report()
calls for presence flags and button flags. It can all be done in
one report.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211116163901.45390-4-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The RS_PLUG and RS_UNPLUG interrupt masks are always written as 1 so
those writes are redundant and can be deleted.
This makes it completely clear in the code that only the TS_PLUG and
TS_UNPLUG masks are being changed.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211116163901.45390-3-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
There are multiple places where DETECT_MODE is included in a register
write, but in every case it is written as 0. Removing these redundant
writes makes the code less cluttered and also makes it obvious that
DETECT_MODE is never changed.
A single initialization to 0 is added to cs42l42_setup_hs_type_detect().
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211116163901.45390-2-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
This adds an ALSA control so that the slow-start audio ramp feature
can be disabled. This is useful for high-definition audio applications.
The register field is unusual in that it is a 3-bit field with only
two valid values, 000=off and 111=on.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211101101006.13092-1-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The setting from the cirrus,ts-inv property should be applied to the
TIP_SENSE_INV bit, as this is the one that actually affects the jack
detect block. The TS_INV bit only swaps the meaning of the PLUG and
UNPLUG interrupts and should always be 1 for the interrupts to have
the normal meaning.
Due to some misunderstanding the driver had been implemented to
configure the TS_INV bit based on the jack switch polarity. This made
the interrupts behave the correct way around, but left the jack detect
block, button detect and analogue circuits always interpreting an open
switch as unplugged.
The signal chain inside the codec is:
SENSE pin -> TIP_SENSE_INV -> TS_INV -> (invert) -> interrupts
|
v
Jack detect,
button detect and
analog control
As the TIP_SENSE_INV already performs the necessary inversion the
TS_INV bit never needs to change. It must always be 1 to yield the
expected interrupt behaviour.
Some extra confusion has arisen because of the additional invert in the
interrupt path, meaning that a value applied to the TS_INV bit produces
the opposite effect of applying it to the TIP_SENSE_INV bit. The ts-inv
property has therefore always had the opposite effect to what might be
expected (0 = inverted, 1 = not inverted). To maintain the meaning of
the ts-inv property it must be inverted when applied to TIP_SENSE_INV.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Fixes: 2c394ca796 ("ASoC: Add support for CS42L42 codec")
Link: https://lore.kernel.org/r/20211028140902.11786-3-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Relying on devm to free the irq handler on probe failure leaves a
small window of opportunity for an interrupt to become pending and
then the handler to run after the chip has been reset and powered
off.
For safety cs42l42_probe() should free the irq in the error path.
As the irq is now disabled by the driver in probe() and remove()
there is no point allocating it as a devres-managed item, so
convert to plain non-devres.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211026125722.10220-3-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Driver remove() should assert RESET and disable the supplies.
probe() fail was disabling supplies but it didn't assert reset or
put the codec into a power-down state.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211026125722.10220-2-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The interrupt handling code was getting the struct device* from a
struct snd_soc_component* stored in struct cs42l42_private. If the
interrupt was asserted before ASoC calls component_probe() the
snd_soc_component* will be NULL.
The stored snd_soc_component* is not actually used for anything other
than indirectly getting the struct device*. Remove it, and store the
struct device* in struct cs42l42_private.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211025112258.9282-1-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The driver has runtime_suspend and runtime_resume callbacks, but
pm_runtime is never enabled so these functions won't be called. They
could not be used anyway because the runtime_suspend would cause jack
detect to stop working.
These functions are unused - delete them.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211018164431.5871-1-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The headset type detection must run to set the analogue switches
correctly for the attached headset type. Without this only headsets
with wiring matching the chip default will have a functioning mic.
commit c26a5289e8 ("ASoC: cs42l42: Add support for set_jack calls")
moved the interrupt unmasking to the component set_jack() callback.
But it's not mandatory for a machine driver to register a struct
snd_soc_jack handler. Without a registered handler the type detection
would not have run and so the mic would not work on some types of
headset.
This patch restores the unmasking of TS_PLUG and TS_UNPLUG interrupts
during probe.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211015133619.4698-17-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The driver must free the IRQ in remove() to prevent the potential race
where an IRQ starts to be handled while the driver is being removed but
devres has not yet called free_irq(). However, the driver can run without
an interrupt but devm_free_irq() will hit a WARN() if no devres-managed
interrupt was ever created.
Fix this by only attempting to create the interrupt handler if the hardware
config specified an interrupt, and failing probe() if the interrupt could
not be created. This means that in cs42l42_remove() an interrupt must have
been registered if the irq number is valid and therefore it is safe to call
devm_free_irq().
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211015133619.4698-16-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
According to the datasheet the SRC MCLK must be as near as possible to
(125 * sample rate). This means it should be ~6MHz for rates up to 48k
and ~12MHz for rates above that. As per datasheet table 4-21.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211015133619.4698-14-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
After enabling the HP or ADC by writing the corresponding PDN=0,
it takes around 20 milliseconds for it to power up and the midrail
supply to be stable. Add this wait into a DAPM widget callback.
If HP and ADC are both powering up in a DAPM sequence, there's no
need to do the wait twice. The widget will perform one wait in the
POST_PMU if there was a PRE_PMU for one or both.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211015133619.4698-13-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
It isn't possible to switch MCLK between 12MHz and 24MHz rate groups
on-the-fly - this can only be done when cs42l42 is powered-down.
All "normal" SCLK rates use an MCLK in the 12MHz group, so change the
configs for SCLK > 12.288 MHz to use the PLL to generate an MCLK in
the 12MHz group.
As this means MCLK_DIV is always 0 it can be removed from the pll
configuration setup.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211015133619.4698-12-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The driver currently only supports configuring for sample rates <= 96k
and it isn't possible to setup a configuration that will support all
sample rates up to 192k.
For sample rates up to 96k MCLK is in the 12MHz group.
However, although 192k only requires an I2S clock in the 12MHz group,
the cs42l42 audio path is not natively 192k so the audio must be
resampled. But for 192k the SRC requires a 24MHz MCLK.
It is not possible to switch MCLK between 12MHz and 24MHz groups
on-the-fly. The 12MHz group supports all sample rates up to 96k.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211015133619.4698-11-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The driver can run without an interrupt so if devm_request_threaded_irq()
failed, the probe() just carried on. But if this was EPROBE_DEFER the
driver would continue without an interrupt instead of deferring to wait
for the interrupt to become available.
Fixes: 2c394ca796 ("ASoC: Add support for CS42L42 codec")
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211015133619.4698-6-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
