In the case a device can support retaining the firmware memory across
low power states it is useful for the preloader widget to only power up
whilst actually loading/unloading the core, as opposed to the normal
operation where the widget is powered for the entire time a firmware is
preloaded onto the core. Add support for this mode and a flag to enable
it.
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20220105113026.18955-7-ckeepax@opensource.cirrus.comSigned-off-by: Mark Brown <broonie@kernel.org>

If the Halo registers are kept in the register cache the
HALO_CORE_RESET bit will be retained as 1 after reset is triggered in
cs_dsp_halo_start_core. This will cause subsequent writes to reset
the core which is not desired. Apart from this bit the rest of the
register bits are cacheable, so for safety sake clear the bit to
ensure the cache is consistent.
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20220105113026.18955-6-ckeepax@opensource.cirrus.comSigned-off-by: Mark Brown <broonie@kernel.org>

It makes no sense to cache the test/user key registers, since they
require values written at specific times, mark them volatile. It is
probably best if they can't be accessed from user-space either, so
mark them precious as well.

The interrupt force, edge, polarity and debounce are all settings
applied to the IRQ rather than status bits and as such should not be
volatile.

The OTP trim values will require re-application in the event of a
cache sync and as such should not be volatile. The OTPID however
should be volatile.

The DSP scratch registers are used to read back an error/debug code
from the DSP on shutdown, as such these should be marked volatile.

Finally, add some missing defaults, add TST_FS_MON0, and allow the
DSP core control register to be cached.
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20220105113026.18955-5-ckeepax@opensource.cirrus.comSigned-off-by: Mark Brown <broonie@kernel.org>

The wm_adsp_event should be called before the early_event on power
down, event stops the core running and early_event then powers down
the core. Additionally, the core should only be stopped if it was
actually running in the first place.
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20220105113026.18955-4-ckeepax@opensource.cirrus.comSigned-off-by: Mark Brown <broonie@kernel.org>

The IRQ is not used for the PDN_DONE bit, this is polled during the DAPM
sequence, remove the misleading comment.
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20220105113026.18955-3-ckeepax@opensource.cirrus.comSigned-off-by: Mark Brown <broonie@kernel.org>

Add IDs for the CS35L51/53 variants, the functionality is shared with
CS35L41.
Signed-off-by: David Rhodes <david.rhodes@cirrus.com>
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20220105113026.18955-2-ckeepax@opensource.cirrus.comSigned-off-by: Mark Brown <broonie@kernel.org>

modprobe can't handle spaces in aliases.

Fixes: 9e28f653 ("ASoC: fsl_mqs: Add MQS component driver")
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Link: https://lore.kernel.org/r/20220104132218.1690103-1-hi@alyssa.isSigned-off-by: Mark Brown <broonie@kernel.org>

Merge series from Trevor Wu <trevor.wu@mediatek.com>:

This series of patches repairs some problems for pcmif BE dai.
The unexpected control flow is corrected, and the missing playback
support of DPCM is added.

Merge series from Lucas Tanure <tanureal@opensource.cirrus.com>:

Initial refactoring to support use of cs35l41 from the HDA driver stack.

So we can send to Linus.

ASoC and HDA will use the same registers to configure
internal boost for the device
Signed-off-by: Lucas Tanure <tanureal@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211217115708.882525-7-tanureal@opensource.cirrus.comSigned-off-by: Mark Brown <broonie@kernel.org>

ASoC and HDA will use the same register to set channels
for the device
Signed-off-by: Lucas Tanure <tanureal@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211217115708.882525-6-tanureal@opensource.cirrus.comSigned-off-by: Mark Brown <broonie@kernel.org>

ASoC and HDA systems require the same errata patches, so
move it to the shared code using a function the correctly
applies the patches by revision

Also, move CS35L41_DSP1_CCM_CORE_CTRL write to errata
patch function as is required to be written at boot,
but not in regmap_register_patch sequence as will affect
waking up from hibernation
Signed-off-by: Lucas Tanure <tanureal@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211217115708.882525-5-tanureal@opensource.cirrus.comSigned-off-by: Mark Brown <broonie@kernel.org>

ASoC and HDA systems for all revisions of CS35L41 will benefit
from having this initialization, so add it to reg_sequence of
each revision

By moving to reg_sequence all gains are set to zero. And boost,
monitoring parts, and class D amplifier are disabled.
Signed-off-by: Lucas Tanure <tanureal@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211217115708.882525-4-tanureal@opensource.cirrus.comSigned-off-by: Mark Brown <broonie@kernel.org>

ASoC and HDA will do the same cs35l41_otp_unpack, so move it
to shared code
Signed-off-by: Lucas Tanure <tanureal@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211217115708.882525-3-tanureal@opensource.cirrus.comSigned-off-by: Mark Brown <broonie@kernel.org>

To support CS35L41 in HDA systems the HDA driver
for CS35L41 would have to duplicate some functions
that already exist on ASoC driver
So instead of duplicate the code, use the new lib
source as a shared resource for both ASoC and HDA

Also, change the way CONFIG_SND_SOC_CS35L41 is
selected, as reported by Intel Kernel test robot,
it is possible to build SND_SOC_CS35L41_SPI/I2C
without the main driver, which would lead to build
failures.
Signed-off-by: Lucas Tanure <tanureal@opensource.cirrus.com>
Reported-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/r/20211217115708.882525-2-tanureal@opensource.cirrus.comSigned-off-by: Mark Brown <broonie@kernel.org>

PCM1_BE should be a dai_link for both playback and capture.
In the patch, the missing DPCM playback support is added.
Signed-off-by: Trevor Wu <trevor.wu@mediatek.com>
Link: https://lore.kernel.org/r/20211230084731.31372-3-trevor.wu@mediatek.comSigned-off-by: Mark Brown <broonie@kernel.org>

Originally, the conditions for preventing reentry are not correct.
dai->component->active is not the state specifically for pcmif dai, so it
is not a correct condition to indicate the status of pcmif dai.
On the other hand, snd_soc_dai_stream_actvie() in prepare ops for both
playback and capture possibly return true at the first entry when these
two streams are opened at the same time.

In the patch, I refer to the implementation in mt8192-dai-pcm.c.
Clock and enabling bit for PCMIF are managed by DAPM, and the condition
for prepare ops is replaced by the status of dai widget.

Fixes: 1f95c019 ("ASoC: mediatek: mt8195: support pcm in platform driver")
Signed-off-by: Trevor Wu <trevor.wu@mediatek.com>
Link: https://lore.kernel.org/r/20211230084731.31372-2-trevor.wu@mediatek.comSigned-off-by: Mark Brown <broonie@kernel.org>

The mclk might not be registered as a fixed clk name "mclk" on some
platforms.
In those platforms, if the mclk needed to be controlled by codec driver
and acquired by a fixed name, it would be a problem.

This patch to fix the issue that wclk becomes an orphan due to the fixed
mclk's name.
Signed-off-by: Derek Fang <derek.fang@realtek.com>
Link: https://lore.kernel.org/r/20211227055446.27563-1-derek.fang@realtek.comSigned-off-by: Mark Brown <broonie@kernel.org>

Playback pop is observed and the root cause is the reference clock
provided by MT8195 is diabled before RT5682 finishes the control flow.

To ensure the reference clock supplied to RT5682 is disabled after RT5682
finishes all register controls. We replace BCLK with MCLK for RT5682
reference clock, and makes use of set_bias_level_post to handle MCLK
which guarantees MCLK is off after all RT5682 register access.
Signed-off-by: Trevor Wu <trevor.wu@mediatek.com>
Reviewed-by: Tzung-Bi Shih <tzungbi@google.com>
Link: https://lore.kernel.org/r/20211228064821.27865-1-trevor.wu@mediatek.comSigned-off-by: Mark Brown <broonie@kernel.org>

Because of the potential failure of the ioremap(), the buf->area could
be NULL.
Therefore, we need to check it and return -ENOMEM in order to transfer
the error.

Fixes: f09aecd5 ("ASoC: SAMSUNG: Add I2S0 internal dma driver")
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Link: https://lore.kernel.org/r/20211228034026.1659385-1-jiasheng@iscas.ac.cnSigned-off-by: Mark Brown <broonie@kernel.org>

Uses of_device_get_match_data() helper to clean some boilerplate code.
Signed-off-by: Tzung-Bi Shih <tzungbi@google.com>
Link: https://lore.kernel.org/r/20211227062153.3887447-1-tzungbi@google.comSigned-off-by: Mark Brown <broonie@kernel.org>

The Chip ID - Register 01h contains the following description
as per the CS4265 datasheet:

"Bits 7 through 4 are the part number ID, which is 1101b (0Dh)"

The current error message is incorrect as it prints CS4265_CHIP_ID,
which is the register number, instead of printing the expected
part number ID value.

To make it clearer, also do a shift by 4, so that the error message
would become:

[    4.218083] cs4265 1-004f: CS4265 Part Number ID: 0x0 Expected: 0xd
Signed-off-by: Fabio Estevam <festevam@denx.de>
Acked-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20211222141920.1482451-1-festevam@gmail.comSigned-off-by: Mark Brown <broonie@kernel.org>

Merge series from Bard Liao <yung-chuan.liao@linux.intel.com>:

This series contains three topics.
1. SoundWire: Intel: remove pdm support
2. ASoC/SoundWire: dai: expand 'stream' concept beyond SoundWire
3. ASoC/SOF/SoundWire: fix suspend-resume on pause with dynamic pipelines

The topics are independent but the changes are dependent. So please
allow me to send them in one series.

Configure the speaker gpio pin based on power sequence of the DAPM
speaker events.
Enable speaker after widget power up and Disable before widget  powerdown.
Signed-off-by: V sujith kumar Reddy <vsujithkumar.reddy@amd.com>
Link: https://lore.kernel.org/r/20211224150058.2444776-1-vsujithkumar.reddy@amd.comSigned-off-by: Mark Brown <broonie@kernel.org>

return value form directly instead of
taking this in another redundant variable.
Reported-by: Zeal Robot <zealci@zte.com.cm>
Signed-off-by: chiminghao <chi.minghao@zte.com.cn>
Link: https://lore.kernel.org/r/20211209015707.409870-1-chi.minghao@zte.com.cnSigned-off-by: Mark Brown <broonie@kernel.org>

Yes, you are right and now the return code depending on the
init_clks().

Fixes: 6078c651 ("soc: mediatek: Refine scpsys to support multiple platform")
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Link: https://lore.kernel.org/r/20211222015157.1025853-1-jiasheng@iscas.ac.cnSigned-off-by: Mark Brown <broonie@kernel.org>

Device nodes can be released after components have bound.

Shortens the lifecycle of the device nodes.  Releases the reference
counts after snd_soc_register_card.
Signed-off-by: Tzung-Bi Shih <tzungbi@google.com>
Link: https://lore.kernel.org/r/20211224064719.2031210-5-tzungbi@google.comSigned-off-by: Mark Brown <broonie@kernel.org>

Eliminates error messages if snd_soc_register_card() failed.  Kernel
emits messages if device probe error anyway.
Signed-off-by: Tzung-Bi Shih <tzungbi@google.com>
Link: https://lore.kernel.org/r/20211224064719.2031210-4-tzungbi@google.comSigned-off-by: Mark Brown <broonie@kernel.org>

Fixes the device_node leak.
Signed-off-by: Tzung-Bi Shih <tzungbi@google.com>
Link: https://lore.kernel.org/r/20211224064719.2031210-3-tzungbi@google.comSigned-off-by: Mark Brown <broonie@kernel.org>

Fixes the device_node leak.
Signed-off-by: Tzung-Bi Shih <tzungbi@google.com>
Link: https://lore.kernel.org/r/20211224064719.2031210-2-tzungbi@google.comSigned-off-by: Mark Brown <broonie@kernel.org>

While the hardware supports PDM streams, this capability has never
been tested or enabled on any product, so this is dead-code. Let's
remove all this.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Acked-By: Vinod Koul <vkoul@kernel.org>
Link: https://lore.kernel.org/r/20211224021034.26635-8-yung-chuan.liao@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

cppcheck warning:

drivers/soundwire/intel.c:1487:10: style: Variable 'ret' is assigned a
value that is never used. [unreadVariable]
 int ret = 0;
         ^
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Acked-By: Vinod Koul <vkoul@kernel.org>
Link: https://lore.kernel.org/r/20211224021034.26635-7-yung-chuan.liao@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

Overloading the tx_mask with a linear value is asking for trouble and
only works because the codec_dai hw_params() is called before the
cpu_dai hw_params().

Move to the more generic set_stream() API to pass the hdac_stream
information.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@intel.com>
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Link: https://lore.kernel.org/r/20211224021034.26635-6-yung-chuan.liao@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

The HDAudio ASoC support relies on the set_tdm_slots() helper to store
the HDaudio stream tag in the tx_mask. This only works because of the
pre-existing order in soc-pcm.c, where the hw_params() is handled for
codec_dais *before* cpu_dais. When the order is reversed, the
stream_tag is used as a mask in the codec fixup functions:

	/* fixup params based on TDM slot masks */
	if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK &&
	    codec_dai->tx_mask)
		soc_pcm_codec_params_fixup(&codec_params,
					   codec_dai->tx_mask);

As a result of this confusion, the codec_params_fixup() ends-up
generating bad channel masks, depending on what stream_tag was
allocated.

We could add a flag to state that the tx_mask is really not a mask,
but it would be quite ugly to persist in overloading concepts.

Instead, this patch suggests a more generic get/set 'stream' API based
on the existing model for SoundWire. We can expand the concept to
store 'stream' opaque information that is specific to different DAI
types. In the case of HDAudio DAIs, we only need to store a stream tag
as an unsigned char pointer. The TDM rx_ and tx_masks should really
only be used to store masks.

Rename get_sdw_stream/set_sdw_stream callbacks and helpers as
get_stream/set_stream. No functionality change beyond the rename.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@intel.com>
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Acked-By: Vinod Koul <vkoul@kernel.org>
Link: https://lore.kernel.org/r/20211224021034.26635-5-yung-chuan.liao@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

This patch provides both a simplification of the suspend flows and a
better balanced operation during suspend/resume transition, as part of
the transition of Sound Open Firmware (SOF) to dynamic pipelines: the
DSP resources are only enabled when required instead of enabled on
startup.

The exiting code relies on a convoluted way of dealing with suspend
signals. Since there is no .suspend DAI callback, we used the
component .suspend and marked all the component DAI dmas as
'suspended'. The information was used in the .prepare stage to
differentiate resume operations from xrun handling, and only
reinitialize SHIM registers and DMA in the former case.

While this solution has been working reliably for about 2 years, there
is a much better solution consisting in trapping the TRIGGER_SUSPEND
in the .trigger DAI ops. The DMA is still marked in the same way for
the .prepare op to run, but in addition the callbacks sent to DSP
firmware are now balanced.

Normal operation:
hw_params -> intel_params_stream
hw_free   -> intel_free_stream

suspend    -> intel_free_stream
prepare    -> intel_params_stream

This balanced operation was not required with existing SOF firmware
relying on static pipelines instantiated at every boot. With the
on-going transition to dynamic pipelines, it's however a requirement
to keep the use count for the DAI widget balanced across all
transitions.

The component suspend is not removed but instead modified to deal with
a corner case: when a substream is PAUSED, the ALSA core does not
throw the TRIGGER_SUSPEND. This is problematic since the refcount for
all pipelines and widgets is not balanced, leading to issues on
resume. The trigger callback keeps track of the 'paused' state with a
new flag, which is tested during the component suspend called later to
release the remaining DSP resources. These resources will be
re-enabled in the .prepare step.

The IPC used in the TRIGGER_SUSPEND to release DSP resources is not a
problem since the BE dailink is already marked as non-atomic.
Co-developed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Acked-By: Vinod Koul <vkoul@kernel.org>
Link: https://lore.kernel.org/r/20211224021034.26635-4-yung-chuan.liao@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

We don't really need to pass a substream to the callback, we only need
the direction. No functionality change, only simplification to enable
improve suspend with paused streams.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@intel.com>
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Acked-By: Vinod Koul <vkoul@kernel.org>
Link: https://lore.kernel.org/r/20211224021034.26635-3-yung-chuan.liao@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

We have a helper, use it to simplify widget lookup
Suggested-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@intel.com>
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Link: https://lore.kernel.org/r/20211224021034.26635-2-yung-chuan.liao@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

Merge series from Stephan Gerhold <stephan@gerhold.net>:

Some sound card setups might require extra pin switches to allow
turning off certain audio components. simple-card supports this
already using the "pin-switches" and "widgets" device tree property.
This series makes it possible to use the same properties for the Qcom
sound cards.

To implement that, the function that parses the "pin-switches" property
in simple-card-utils.c is first moved into the ASoC core. Then two
simple function calls are added to the common Qcom sound card DT parser.
Finally there is a small patch for the msm8916-wcd-analog codec to make
it possible to model sound card setups used in some MSM8916 smartphones.
(See PATCH 2/4 for an explanation of some real example use cases.)

Using pin switches rather than patching codec drivers with switches was
originally suggested by Mark Brown on a patch for the tfa989x codec:
https://lore.kernel.org/alsa-devel/YXaMVHo9drCIuD3u@sirena.org.uk/

The analog codec has separate output paths for the left headphone channel
(HPH_L) and the right headphone channel (HPH_R). While they are usually
used together for actual headphones output, some devices also have an
analog speaker amplifier connected to one of the headphone channels.

To allow modelling that properly (and to avoid powering on the unneeded
output path), HPH_L and HPH_R should be represented by separate outputs
rather than a shared HEADPHONE output that always activates both paths.

Cc: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Signed-off-by: Stephan Gerhold <stephan@gerhold.net>
Link: https://lore.kernel.org/r/20211214142049.20422-5-stephan@gerhold.netSigned-off-by: Mark Brown <broonie@kernel.org>