ASoC: dt-bindings: ti,pcm1681: Convert to dtschema

Convert the Texas Instruments PCM1681 bindings to DT schema.
Make bindings complete by adding #sound-dai-cells.

Signed-off-by: Animesh Agarwal <animeshagarwal28@gmail.com>
Reviewed-by: Krzysztof Kozlowski <krzk@kernel.org>
Link: https://msgid.link/r/20240406141454.45529-1-animeshagarwal28@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: improve and extend HDaudio-based

Merge series from Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>:

For LunarLake, the SoundWire in-band wake detection is reported with
the HDAudio WAKE_EN/WAKE_STS registers. In the existing code, these
registers are only handled for HDaudio codecs. Now the same registers
have to be handled with care as shared resources.

The in-band wake detection mainly used for jack detection. Without
this patchset, the SoundWire headset codecs signal an event that would
be ignored and not reported.

ASoC: Intel: avs: Fixes and cleanups for 6.10

Merge series from Cezary Rojewski <cezary.rojewski@intel.com>:

Set of changes targeting the avs-driver only. No new features, patchset
either fixes or fortifies existing code.

Patchset starts off with a fix for debugbility on ICL+ platforms which I
have forgotten to fixup when providing support for these initially.
The next two address copier module initialization, most importantly,
silence the gcc 'field-spanning write' false-positive.

The following four:
6/13 ASoC: Intel: avs: Replace risky functions with safer variants
7/13 ASoC: Intel: avs: Fix potential integer overflow
8/13 ASoC: Intel: avs: Test result of avs_get_module_entry()
9/13 ASoC: Intel: avs: Remove dead code

address problems found out by Coverity static analysis tool.

The last two worth mentioning are: recommendation from the firmware team
to wake subsystem from D0ix when starting any pipeline -and- shielding
against invalid period/buffer sizes. Audio format shall be taken into
consideration when calculating either of these.

Amadeusz Sławiński (2):
  ASoC: Intel: avs: Restore stream decoupling on prepare
  ASoC: Intel: avs: Add assert_static to guarantee ABI sizes

Cezary Rojewski (11):
  ASoC: Intel: avs: Fix debug-slot offset calculation
  ASoC: Intel: avs: Silence false-positive memcpy() warnings
  ASoC: Intel: avs: Fix config_length for config-less copiers
  ASoC: Intel: avs: Fix ASRC module initialization
  ASoC: Intel: avs: Replace risky functions with safer variants
  ASoC: Intel: avs: Fix potential integer overflow
  ASoC: Intel: avs: Test result of avs_get_module_entry()
  ASoC: Intel: avs: Remove dead code
  ASoC: Intel: avs: Wake from D0ix when starting streaming
  ASoC: Intel: avs: Init debugfs before booting firmware
  ASoC: Intel: avs: Rule invalid buffer and period sizes out

 sound/soc/intel/avs/avs.h      |  1 +
 sound/soc/intel/avs/cldma.c    |  2 +-
 sound/soc/intel/avs/core.c     |  4 +--
 sound/soc/intel/avs/icl.c      | 12 ++++++---
 sound/soc/intel/avs/loader.c   |  6 +++--
 sound/soc/intel/avs/messages.h | 47 ++++++++++++++++++++++++++++++++--
 sound/soc/intel/avs/path.c     | 13 ++++------
 sound/soc/intel/avs/pcm.c      | 34 +++++++++++++++++++++++-
 sound/soc/intel/avs/probes.c   | 14 ++++++----
 9 files changed, 109 insertions(+), 24 deletions(-)

--
2.25.1

ASoC: Intel: avs: Rule invalid buffer and period sizes out

While HDAudio controller supports buffer packets up to 128 bytes low,
audio format shall be taken into consideration when calculating buffer
and period sizes to avoid undesired xruns.

As *_size in ALSA terms means frames (channels times bit-depth-bytes),
hw_rules can calculate minimal buffer and period sizes solely from
sample rate and the number of milliseconds commonly used on the
AudioDSP firmware side.

Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://msgid.link/r/20240405090929.1184068-14-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: Add assert_static to guarantee ABI sizes

In order to make sure that IPC interface is stable use assert_static to
check union and struct sizes that describe communication interface.

Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski@linux.intel.com>
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://msgid.link/r/20240405090929.1184068-13-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: Init debugfs before booting firmware

When bringing up setups it's vital to have access to debug functionality
even if firmware boot fails. As order of probe()ing operations is
changed, update remove() procedure accordingly.

Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://msgid.link/r/20240405090929.1184068-12-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: Wake from D0ix when starting streaming

It is recommended to keep the DSP domain in full-power when starting DMA
engines.

Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://msgid.link/r/20240405090929.1184068-11-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: Remove dead code

The result of list_next_entry()/list_last_entry() is never null.

Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://msgid.link/r/20240405090929.1184068-10-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: Test result of avs_get_module_entry()

While PROBE_MOD_UUID is always part of the base AudioDSP firmware
manifest, from maintenance point of view it is better to check the
result.

Fixes: dab8d000e25c ("ASoC: Intel: avs: Add data probing requests")
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://msgid.link/r/20240405090929.1184068-9-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: Fix potential integer overflow

While stream_tag for CLDMA on SKL-based platforms is always 1, function
hda_cldma_setup() uses AZX_SD_CTL_STRM() macro which does:
stream_tag << 20

what combined with stream_tag type of 'unsigned int' generates a
potential overflow issue. Update the field type to fix that.

Fixes: 45864e49a05a ("ASoC: Intel: avs: Implement CLDMA transfer")
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://msgid.link/r/20240405090929.1184068-8-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: Replace risky functions with safer variants

strscpy() and snprintf() are the recommended equivalents of their
riskier friends.

Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://msgid.link/r/20240405090929.1184068-7-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: Fix ASRC module initialization

The ASRC module configuration consists of several reserved fields. Zero
them out when initializing the module to avoid sending invalid data.

Fixes: 274d79e51875 ("ASoC: Intel: avs: Configure modules according to their type")
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://msgid.link/r/20240405090929.1184068-6-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: Fix config_length for config-less copiers

Copier's config_length shall always be at least one even if there is no
configuration payload to carry. While the firmware treats
config_length=0 or 1 in the same manner, the driver shall initialize the
module properly.

Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://msgid.link/r/20240405090929.1184068-5-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: Silence false-positive memcpy() warnings

Commit df8fc4e934c1 ("kbuild: Enable -fstrict-flex-arrays=3") enforced
strict flex array declarations. This generates false-positive in form of:
"memcpy: detected field-spanning write". Avoid it by utilizing the
DECLARE_FLEX_ARRAY() macro.

Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://msgid.link/r/20240405090929.1184068-4-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: Fix debug-slot offset calculation

For resources with ID other than 0 the current calculus is incorrect.

Fixes: 275b583d047a ("ASoC: Intel: avs: ICL-based platforms support")
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://msgid.link/r/20240405090929.1184068-3-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: Restore stream decoupling on prepare

Revert changes from commit b87b8f43afd5 ("ASoC: Intel: avs: Drop
superfluous stream decoupling") to restore working streaming during S3.

Fixes: b87b8f43afd5 ("ASoC: Intel: avs: Drop superfluous stream decoupling")
Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski@linux.intel.com>
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://msgid.link/r/20240405090929.1184068-2-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda-ctrl: only clear WAKESTS for HDaudio codecs

When a PME wake happens due to a SoundWire wake, we currently clear
all WAKESTS bits during the resume operation initiated by the PCI
subsystem. As a result, we are unable to identify which SoundWire
links need to be resumed and don't properly handle jack detection.

This patch only clears the WAKESTS bits for the HDaudio codecs
detected earlier.

Note that we still clear all WAKESTS bits unconditionally in
hda_dsp_ctrl_stop_chip(). The existing behavior is potentially racy if
e.g. a jack event happens during a suspend routine, but there's a risk
of breaking shutdown or reboot sequences so the code is left as is for
now.

Closes: https://github.com/thesofproject/linux/issues/4687
Co-developed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Keqiao Zhang <keqiao.zhang@intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Link: https://msgid.link/r/20240404190357.138073-5-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda-codec: preserve WAKEEN values

Since LunarLake, we use the HDadio WAKEEN/WAKESTS to detect wakes for
SoundWire codecs. Unfortunately, the existing code in
hda_codec_jack_wake_enable() unconditionally resets the WAKEEN bits.

This patch changes the initialization to preserve SoundWire WAKEEN
bits. For HDAudio codecs the same strategy is used, WAKEEN is only set
when the jacktbl.used property is set.

Closes: https://github.com/thesofproject/linux/issues/4687
Co-developed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Keqiao Zhang <keqiao.zhang@intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Link: https://msgid.link/r/20240404190357.138073-4-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: lnl: add helper to detect SoundWire wakes

The global STATESTS register will provide information on all
links. Rather than iterate on all possible links, the helpers only
filters the range of possible bits for a quick lookup. The
process_wakeen() helper will walk through all the links and deal with
wakes.

Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Link: https://msgid.link/r/20240404190357.138073-3-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda-ctrl: add missing WAKE_STS clear

For some reason, the programming sequences in the SOF driver do not
include a clear of the WAKE_STS bits before resetting the controller.

This clear is not formally required by the HDaudio specification, but
was added to harden the snd-hda-reset back in 2007. Adding this
sequence back avoids an issue reported by the Intel CI.

Closes: https://github.com/thesofproject/linux/issues/4889
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240404190357.138073-2-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda-stream: clarify comment

The Yoda grammar and multiple negatives are unclear.

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>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://msgid.link/r/20240404185448.136157-5-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: don't ignore IOC interrupts for non-audio transfers

The HDaudio stream interrupts are ignored unless the stream is PCM or
compressed audio. For alternate non-audio usages, such as code loader
or SoundWire BPT case, the IOC interrupt on the last buffer
transferred is silently ignored.

This patch adds a 'struct completion' for each HDaudio stream. This
capability helps detect if the non-audio data transfers
completed. There is no performance impact for audio streams.

In the code loader case, the code currently starts the DMA and
directly checks if the firmware status changes, without checking if
the DMA succeeded. With a first pass waiting for the DMA to complete,
system validation engineers can gather more precise timing information
on firmware boot time or root-cause boot failures more accurately.

A timeout of 500ms was selected for the code loader DMA. This is an
experimental value which should be more than enough - higher values
would certainly be problematic from a usage/latency perspective.

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>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://msgid.link/r/20240404185448.136157-4-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda-loader: change signature of code loader helpers

We need to reuse cl_prepare, cl_trigger and cl_cleanup helpers from a
SoundWire context where only the device information is
available. Rather than pass the 'sdev' argument, use get_drvdata() to
retrieve the required structure.

For consistency, rename hda_cl_stream_prepare() as hda_cl_prepare().

These three helpers are also exported so that they can be referenced
from another module.

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>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://msgid.link/r/20240404185448.136157-3-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda-loader: add flag to tag ICCMAX stream

The existing code conflates capture stream and ICCMAX stream. This
isn't going to be true any longer when we add support for SDW BPT RX
streams. Add a boolean tag to flag ICCMAX streams.

No functionality change, only future-proofing change.

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>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://msgid.link/r/20240404185448.136157-2-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Drop soc-topology ABI v4 support

Merge series from Cezary Rojewski <cezary.rojewski@intel.com>:

This patchset impacts UAPI.

The only known users of the soc-topology ABI v4 are Chromebook
configurations. Starting from kernel v5.4, all of them are making use of
soc-topology ABI v5. The patchset first removes obsolete code from the
Intel's skylake-driver - the driver of choice for the mentioned
Chromebooks - and then proceeds with removal of relevant soc-topology.c
and uapi bits.

Cezary Rojewski (4):
  ASoC: Intel: Skylake: Remove soc-topology ABI v4 support
  ASoC: topology: Remove ABI v4 support
  ASoC: topology: Cleanup after ABI v4 support removal
  ASoC: topology: Remove obsolete ABI v4 structs

 include/uapi/sound/asoc.h               |  56 ------
 include/uapi/sound/skl-tplg-interface.h |  74 --------
 sound/soc/intel/skylake/skl-topology.c  | 169 -----------------
 sound/soc/soc-topology.c                | 241 ++----------------------
 4 files changed, 18 insertions(+), 522 deletions(-)

--
2.25.1

ASoC: SOF: Clean up sof_ipc_flood_dfs_write()

This function doesn't support partial writes so using
simple_write_to_buffer() doesn't really make sense.  It's better to
just use copy_from_user().

Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Acked-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Link: https://msgid.link/r/a35dded2-392b-4ccb-9dbb-d782ac9b6547@moroto.mountain
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Disable pointless writes to debugfs file

The permissions on this debugfs file are 0444 so it can't be written to.
And writing to the file hasn't done anything since commit 6e9548cdb30e
("ASoC: SOF: Convert the generic IPC flood test into SOF client").
Delete the write function.

Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Acked-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Link: https://msgid.link/r/9d2477cf-25aa-4d12-818f-fdafc9aaa28a@moroto.mountain
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: topology: Remove obsolete ABI v4 structs

There are no users of soc-topology ABI v4 since kernel v5.4 and no
kernel code makes use of them.

Cc: Curtis Malainey <cujomalainey@chromium.org>
Cc: Łukasz Majczak <lmajczak@google.com>
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://msgid.link/r/20240403091629.647267-5-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: topology: Cleanup after ABI v4 support removal

Flag 'abi_match' along with several other local variables has no purpose
with v4 related members removed.

Cc: Curtis Malainey <cujomalainey@chromium.org>
Cc: Łukasz Majczak <lmajczak@google.com>
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://msgid.link/r/20240403091629.647267-4-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: topology: Remove ABI v4 support

There are no users of soc-topology ABI v4 since kernel v5.4 so remove
all v4 -> v5 converters.

Cc: Curtis Malainey <cujomalainey@chromium.org>
Cc: Łukasz Majczak <lmajczak@google.com>
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://msgid.link/r/20240403091629.647267-3-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: Skylake: Remove soc-topology ABI v4 support

The only known users are Chromebook configurations. Starting from
kernel v5.4, all of them are making use of soc-topology ABI v5.

Cc: Curtis Malainey <cujomalainey@chromium.org>
Cc: Łukasz Majczak <lmajczak@google.com>
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://msgid.link/r/20240403091629.647267-2-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: imx: stop using the imx8_*_clocks API

Merge series from Laurentiu Mihalcea <laurentiumihalcea111@gmail.com>:

The imx8_*_clocks API requires keeping track of all of the clocks used
by the IMX SOF driver via an array. This is unnecessary and doesn't
scale well. As such, remove it altogether and replace it with
devm_clk_bulk_get_all() and friends.

ASoC: nau8822: Remove unused member from struct

Remove unused mclk_idx struct member.

Signed-off-by: Francesco Dolcini <francesco.dolcini@toradex.com>
Link: https://msgid.link/r/20240404111953.99287-1-francesco@dolcini.it
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: codecs: rk3308: fix building without CONFIG_OF

The newly added codec has an extraneous of_match_ptr() annotation for
the ID table:

sound/soc/codecs/rk3308_codec.c:956:34: error: 'rk3308_codec_of_match' defined but not used [-Werror=unused-const-variable=]
  956 | static const struct of_device_id rk3308_codec_of_match[] = {

Remove it to avoid introducing a warning when -Wunused-const-variable
gets enabled by default:

Fixes: 4ed0915f5bc4 ("ASoC: codecs: Add RK3308 internal audio codec driver")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Link: https://msgid.link/r/20240404095755.650364-1-arnd@kernel.org
Signed-off-by: Mark Brown <broonie@kernel.org>

ALSA: cirrus: Tidy up of firmware control read/write

Merge series from Richard Fitzgerald <rf@opensource.cirrus.com>:

This set of patches factors out some repeated code to clean up
firmware control read/write functions, and removes some redundant
control notification code.

base-commit: f193957b0fbbba397c8bddedf158b3bf7e4850fc

ASoC: codecs: rk3308: fix build warning without OF

Merge series from Luca Ceresoli <luca.ceresoli@bootlin.com>:

This small series fixes a build warning reported by kernel test robot
<lkp@intel.com> and improves the Kconfig entry.

Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
---
Changes in v2:
- Replaced v1 implementation with __maybe_unused
- Added patch to depend on || COMPILE_TEST
- Link to v1: https://lore.kernel.org/r/20240329-rk3308-audio-codec-fix-warning-v1-1-b9d177fcd6c9@bootlin.com

---
Luca Ceresoli (2):
      ASoC: codecs: rk3308: fix "defined but not used" warning on !OF
      ASoC: codecs: rk3308: depend on ARM64 || COMPILE_TEST

 sound/soc/codecs/Kconfig        | 1 +
 sound/soc/codecs/rk3308_codec.c | 2 +-
 2 files changed, 2 insertions(+), 1 deletion(-)
---
base-commit: 4ed0915f5bc4bcc81bca783a5b984f3d81e9764e
change-id: 20240329-rk3308-audio-codec-fix-warning-51bb572ebd96

Best regards,
--
Luca Ceresoli <luca.ceresoli@bootlin.com>

ASoC: SOF: Intel: mtl/lnl: Improve firmware boot state

Merge series from Peter Ujfalusi <peter.ujfalusi@linux.intel.com>:

This series improves the firmware/boot state handling which will allow
failed IMR boot recovery and human readable boot failure decoding.

Additionally a new debugfs file is added to force a purge/clean boot
of the DSP for developers.

ALSA: hda: hda_cs_dsp_ctl: Remove notification of driver write

Any control that the driver is updating should be marked as SYSTEM and
therefore will not have an ALSA control to notify.

Signed-off-by: Simon Trimmer <simont@opensource.cirrus.com>
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Link: https://msgid.link/r/20240325113127.112783-6-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: wm_adsp: Remove notification of driver write

Any control that the driver is updating should be marked as SYSTEM and
therefore will not have an ALSA control to notify.

Signed-off-by: Simon Trimmer <simont@opensource.cirrus.com>
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Link: https://msgid.link/r/20240325113127.112783-5-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ALSA: hda: hda_cs_dsp_ctl: Use cs_dsp_coeff_lock_and_[read|write]_ctrl()

Using the cs_dsp_coeff_lock_and_[read|write]_ctrl() wrappers tidies
the calling functions as it does not need to manage the DSP pwr_lock.

Signed-off-by: Simon Trimmer <simont@opensource.cirrus.com>
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Link: https://msgid.link/r/20240325113127.112783-4-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: wm_adsp: Use cs_dsp_coeff_lock_and_[read|write]_ctrl()

Using the cs_dsp_coeff_lock_and_[read|write]_ctrl() wrappers tidies
the calling functions as it does not need to manage the DSP pwr_lock.

Signed-off-by: Simon Trimmer <simont@opensource.cirrus.com>
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Link: https://msgid.link/r/20240325113127.112783-3-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>

firmware: cs_dsp: Add locked wrappers for coeff read and write

It is a common pattern for functions to take and release the DSP
pwr_lock over the cs_dsp calls to read and write firmware controls.
Add wrapper functions to do this sequence so that the calling code can
be simplified to a single function call..

Signed-off-by: Simon Trimmer <simont@opensource.cirrus.com>
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Link: https://msgid.link/r/20240325113127.112783-2-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Merge up fixes

Some of these, particularly the wm_adsp one in the immediate case, are
needed as a basis for new work.

ASoC: Merge up left over v6.8 fix

This v6.8 change didn't make it into the release, send it as a fix for
v6.9.

ASoC: SOF: imx: remove imx8_*_clocks API

Since the i.MX drivers no longer use the imx8_*_clocks API
this can be removed.

Signed-off-by: Laurentiu Mihalcea <laurentiu.mihalcea@nxp.com>
Reviewed-by: Iuliana Prodan <iuliana.prodan@nxp.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Daniel Baluta <daniel.baluta@nxp.com>
Link: https://msgid.link/r/20240328221201.24722-3-laurentiumihalcea111@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: imx: drop usage of the imx8_*_clocks API

Currently, the driver has to keep track of all the clocks
it uses via an array of "struct clk_bulk_data", which doesn't
scale well and is unnecessary. As such, replace the usage of
the imx8_*_clocks with "devm_clk_bulk_get_all()" and friends.

Signed-off-by: Laurentiu Mihalcea <laurentiu.mihalcea@nxp.com>
Reviewed-by: Iuliana Prodan <iuliana.prodan@nxp.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Daniel Baluta <daniel.baluta@nxp.com>
Link: https://msgid.link/r/20240328221201.24722-2-laurentiumihalcea111@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: codecs: rk3308: depend on ARM64 || COMPILE_TEST

This codec is only known to exist in the RK3308 ARM64 SoC, so depend on it
except for compile test cases. Note that the driver won't probe without
CONFIG_OF, but ARM64 selects OF already so it is not needed.

Suggested-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Link: https://msgid.link/r/20240403-rk3308-audio-codec-fix-warning-v2-2-816bae4c1dc5@bootlin.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: codecs: rk3308: fix "defined but not used" warning on !OF

Building with CONFIG_OF=n triggers:

  warning: 'rk3308_codec_of_match' defined but not used [-Wunused-const-variable=]
  warning: unused variable 'rk3308_codec_of_match' [-Wunused-const-variable]

Even though OF is needed for probing, fix by declaring as __maybe_unused to
still allow building on non-OF configurations for build testing.

Fixes: 9fdd7b45da18 ("arm64: defconfig: enable Rockchip RK3308 internal audio codec driver")
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202403271905.BYbGJiPi-lkp@intel.com/
Closes: https://lore.kernel.org/oe-kbuild-all/202403271907.0z0uuG5I-lkp@intel.com/
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Link: https://msgid.link/r/20240403-rk3308-audio-codec-fix-warning-v2-1-816bae4c1dc5@bootlin.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: mtl: Correct the mtl_dsp_dump output

The ROM/firmware state handling has changed between CAVS and ACE
architecture:
CAVS: ROM and firmware uses the SRAM window for the state and status/error
code reporting

ACE: ROM code is using two registers to report the state and error while
the firmware is using the SRAM window to report states and status/error
codes.

Use the generic hda_dsp_get_state() to decode ROM state and error codes and
print out the firmware state and status/error code only if the SRAM
window is accessible - the firmware is booted and the Status readout is
not 0xffffffff.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Liam Girdwood <liam.r.girdwood@intel.com>
Link: https://msgid.link/r/20240403105210.17949-8-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda-dsp/mtl: Add support for ACE ROM state codes

The ROM state codes differ between CAVS and ACE architecture, there is a
slight overlap.
Add the ACE related state defines to mtl.h, introduce new table and
use it on case the function is called when running on ACE architecture.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Liam Girdwood <liam.r.girdwood@intel.com>
Link: https://msgid.link/r/20240403105210.17949-7-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: mtl: Implement firmware boot state check

With the corrected rom_status_reg values we can now add a check for target
boot status for firmware booting.
With the check now we can identify failed firmware boots (IMR boots) and
we can use the fallback to purge boot the DSP.

Fixes: 064520e8aeaa ("ASoC: SOF: Intel: Add support for MeteorLake (MTL)")
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Liam Girdwood <liam.r.girdwood@intel.com>
Link: https://msgid.link/r/20240403105210.17949-6-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: mtl: Disable interrupts when firmware boot failed

In case of error during the firmware boot we need to disable the interrupts
which were enabled as part of the boot sequence.

Fixes: 064520e8aeaa ("ASoC: SOF: Intel: Add support for MeteorLake (MTL)")
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Liam Girdwood <liam.r.girdwood@intel.com>
Link: https://msgid.link/r/20240403105210.17949-5-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: lnl: Correct rom_status_reg

ACE2 architecture changed the place where the ROM updates the status code
from the shared SRAM window (and HFFLGP1QW0 in ACE1) to HFDSC register for
the status and HFDEC (HFDSC + 4) for the error code.

The rom_status_reg is not used on LNL because it was wrongly assigned based
on older platform convention (SRAM window) and it was giving inconsistent
readings.

Add new header file for lnl specific register definitions.

Fixes: 64a63d9914a5 ("ASoC: SOF: Intel: LNL: Add support for Lunarlake platform")
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Liam Girdwood <liam.r.girdwood@intel.com>
Link: https://msgid.link/r/20240403105210.17949-4-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: mtl: Correct rom_status_reg

ACE1 architecture changed the place where the ROM updates the status code
from the shared SRAM window to HFFLGP1QW0 register for the status and
HFFLGP1QW0 + 4 for the error code.

The rom_status_reg is not used on MTL because it was wrongly assigned based
on older platform convention (SRAM window) and it was giving inconsistent
readings.

Fixes: 064520e8aeaa ("ASoC: SOF: Intel: Add support for MeteorLake (MTL)")
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Liam Girdwood <liam.r.girdwood@intel.com>
Link: https://msgid.link/r/20240403105210.17949-3-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda: Create debugfs file to force a clean DSP boot

When IMR boot is supported on a platform it is always going to be used to
boot the DSP unless some catastrophic event happens.
There is no way for a developer to force a clean DSP boot without removing
and re-inserting the modules.

Create a 'skip_imr_boot' debugfs file which can be used to force the
next DSP boot as clean (prune) boot.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240403105210.17949-2-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: rt5682: Constify card_headset_pins

card_headset_pins is never modified, mark it const.

Reviewed-by: Cezary Rojewski <cezary.rojewski@intel.com>
Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski@linux.intel.com>
Link: https://msgid.link/r/20240403093145.3375857-11-amadeuszx.slawinski@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: rt5663: Constify card_headset_pins

card_headset_pins is never modified, mark it const.

Reviewed-by: Cezary Rojewski <cezary.rojewski@intel.com>
Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski@linux.intel.com>
Link: https://msgid.link/r/20240403093145.3375857-10-amadeuszx.slawinski@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: rt298: Constify card_headset_pins

card_headset_pins is never modified, mark it const.

Reviewed-by: Cezary Rojewski <cezary.rojewski@intel.com>
Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski@linux.intel.com>
Link: https://msgid.link/r/20240403093145.3375857-9-amadeuszx.slawinski@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: rt286: Constify card_headset_pins

card_headset_pins is never modified, mark it const.

Reviewed-by: Cezary Rojewski <cezary.rojewski@intel.com>
Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski@linux.intel.com>
Link: https://msgid.link/r/20240403093145.3375857-8-amadeuszx.slawinski@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: rt274: Constify card_headset_pins

card_headset_pins is never modified, mark it const.

Reviewed-by: Cezary Rojewski <cezary.rojewski@intel.com>
Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski@linux.intel.com>
Link: https://msgid.link/r/20240403093145.3375857-7-amadeuszx.slawinski@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: nau8825: Constify card_headset_pins

card_headset_pins is never modified, mark it const.

Reviewed-by: Cezary Rojewski <cezary.rojewski@intel.com>
Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski@linux.intel.com>
Link: https://msgid.link/r/20240403093145.3375857-6-amadeuszx.slawinski@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: es8336: Constify card_headset_pins

card_headset_pins is never modified, mark it const.

Reviewed-by: Cezary Rojewski <cezary.rojewski@intel.com>
Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski@linux.intel.com>
Link: https://msgid.link/r/20240403093145.3375857-5-amadeuszx.slawinski@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: hdaudio: Constify probing_link

probing_link is passed to devm_kmemdup, and is never modified, may as
well mark it const.

Reviewed-by: Cezary Rojewski <cezary.rojewski@intel.com>
Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski@linux.intel.com>
Link: https://msgid.link/r/20240403093145.3375857-4-amadeuszx.slawinski@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: Use devm_kstrdup_const

FW name is constant and we just duplicate it, use const variant of
devm_kstrdup to possibly save a bit of memory.

Reviewed-by: Cezary Rojewski <cezary.rojewski@intel.com>
Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski@linux.intel.com>
Link: https://msgid.link/r/20240403093145.3375857-3-amadeuszx.slawinski@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: avs: hdaudio: Use devm_kstrdup_const

HDA codec name is constant and we just duplicate it, use const variant
of devm_kstrdup to possibly save a bit of memory.

Reviewed-by: Cezary Rojewski <cezary.rojewski@intel.com>
Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski@linux.intel.com>
Link: https://msgid.link/r/20240403093145.3375857-2-amadeuszx.slawinski@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rsnd: reg cleanup

Merge series from Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>:

These are Renesas Sound driver cleanups for Gen3/Gen4.

ASoC: SOF: cppcheck fixes and debugfs addition

Merge series from Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>:

Small changes with 3 cppcheck fixes and the firmware version now
visible with debugfs instead of only via dmesg logs.

ASoC: SOF: Intel: improve SoundWire support for

Merge series from Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>:

This patchset contains important updates for SoundWire support.

We initially implemented support for multiple amplifiers on different
links using a single HDaudio DMA transfer. To align with the other OS,
the 'aggregation' is now supported by the firmware. This change in
directions has kernel impacts, since we now have multiple HDaudio DMAs
to program and start, but since there are no platforms released so far
there's no end-user impact.

In addition, the behavior in case of xruns is improved by clearing the
PCM states and better handling of the hw_free case.

Note that the hw_free support will compile but will only be functional
with the companion patch "soundwire: intel: add intel_free_stream() back"
already applied in the SoundWire tree.

ASoC: Use snd_ctl_find_id_mixer() instead of

Merge series from Richard Fitzgerald <rf@opensource.cirrus.com>:

The first two patches change snd_soc_card_get_kcontrol() to use the
core snd_ctl_find_id_mixer() functionality instead of open-coding its
own list walk.

The last patch adds a KUnit test for this, which was tested on the
original and modified code.

ASoC: codecs: ES8326: solve some hp issues and

Merge series from Zhang Yi <zhangyi@everest-semi.com>:

We solved some issues related to headphone detection.And for using
the same configuration in different power conditions,we modified the
clock table

ASoC: dt-bindings: fsl-asoc-card: convert to YAML

Convert the fsl-asoc-card binding to YAML.

When testing dtbs_check, found below compatible strings
are not listed in document:

fsl,imx-sgtl5000
fsl,imx53-cpuvo-sgtl5000
fsl,imx51-babbage-sgtl5000
fsl,imx53-m53evk-sgtl5000
fsl,imx53-qsb-sgtl5000
fsl,imx53-voipac-sgtl5000
fsl,imx6-armadeus-sgtl5000
fsl,imx6-rex-sgtl5000
fsl,imx6-sabreauto-cs42888
fsl,imx6-wandboard-sgtl5000
fsl,imx6dl-nit6xlite-sgtl5000
fsl,imx6q-ba16-sgtl5000
fsl,imx6q-nitrogen6_max-sgtl5000
fsl,imx6q-nitrogen6_som2-sgtl5000
fsl,imx6q-nitrogen6x-sgtl5000
fsl,imx6q-sabrelite-sgtl5000
fsl,imx6q-sabresd-wm8962
fsl,imx6q-udoo-ac97
fsl,imx6q-ventana-sgtl5000
fsl,imx6sl-evk-wm8962
fsl,imx6sx-sdb-mqs
fsl,imx6sx-sdb-wm8962
fsl,imx7d-evk-wm8960
karo,tx53-audio-sgtl5000
tq,imx53-mba53-sgtl5000

So add them in yaml file to pass the test.

Also correct the 'dai-format' to 'format' in document.

For 'audio-routing', the items are not listed. Because
this fsl-asoc-card is generic driver, which supports several
codecs, if list all the items, there will be a long list.

Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Link: https://msgid.link/r/1711976056-19884-3-git-send-email-shengjiu.wang@nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: ipc4-topology: Save the ALH DAI index during hw_params

The node_id for multi-gateway ALH DAI's get overwritten with the
group_id during the DAI copier's ipc_prepare op. So, save the ALH
dai_index during the BE DAI hw_params in the dai_index field of struct
ipc4_copier and use that to set the device ID in the configuration blob.
This will avoid errors during copier init after an xrun.

Note that the dai_index is typically set in topology for DMIC/SSP, but
it's not used for ALH. Reclaiming this dai_index field to store the
node_id does not generate a conflict with topology-defined values.

Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240402151828.175002-18-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: pcm: reset all PCM sources in case of xruns

With the delayed stops, the xrun handling is problematic: the
applications expects everything to be reset but the firmware and DMA
are still in a PAUSED state.

This patch makes sure the prepare while pending_stop is set is
special-cased.

Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Co-developed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240402151828.175002-17-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: pcm: add pending_stop state variable

Add a state variable to keep track of delayed stops, in case
pcm_ops->platform_stop_during_hw_free is set.

This patch should be iso-functionality, possibly removing no-op
cases. The main purpose of this new state variable is to prepare a
follow-up patch to reset all PCM and DMAs in case of stop/prepare xrun
sequences.

Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240402151828.175002-16-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: pcm: simplify sof_pcm_hw_free() with helper

The same sequence is used twice, use common helper.

Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240402151828.175002-15-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: sof-audio: revisit sof_pcm_stream_free() error handling and logs

For some reason the existing code stops on the first error, which
potentially leaves the DMA and widgets in a weird state.

Change to free-up all resources even in case of errors.

Also add a more consistent error handling and logs, with the first
error code returned to the caller.

Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240402151828.175002-14-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda: Clear Soundwire node ID during BE DAI hw_free

When an xrun happens, the BE DAI hw_params doesn't get invoked before
the stream restarts with a prepare. In this case, clearing the node ID
when the DAI widget is freed and unprepared will result in an error when
it is re-initialized. In order to avoid this, move the code to clear the
node ID to the BE DAI hw_free op to keep it balanced with the BE DAI
hw_params.

Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240402151828.175002-13-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda: move helper to static inline

To allow using widget_to_sdev() in other files, move it as static inline
in shared header file.

Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240402151828.175002-12-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: IPC4: extend dai_data with node_id

The node_id value needs to be handled specifically for ALH.

Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240402151828.175002-11-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda: extend signature of sdw_hda_dai_hw_params()

Add intel_alh_id to set the expected gateway node_id in a follow-up
patch.

Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240402151828.175002-10-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: set the DMA TLV device as dai_index

We've already defined the value for dai_index, let's use it instead of
open-coding the same thing. No functionality change.

Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240402151828.175002-9-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda-dai: add helpers to set dai config

We need to be able to set the dai config differently for SoundWire.

Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240402151828.175002-8-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda-dai: set dma_stream_channel_map device

sof_ipc4_dma_config_tlv{} is required for ACE2.x. The patch follow the
convention to set the dma_stream_channel_map.mapping device as
"link_id << 8 | pdi_id".
And the mapping in sof_ipc4_alh_configuration_blob{} should be the same
as dma_stream_channel_map.mapping in sof_ipc4_dma_config{}.
The purposes of device id is to map DMA tlv.

Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240402151828.175002-7-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda-dai: set lowest N bits in ch_mask

We always use the lowest N channels of stream. So, set ch_mask to
GENMASK(params_channels(params) - 1, 0).

Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240402151828.175002-6-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: make dma_config_tlv be an array

Each stream needs a dma_config_tlv. We will handle multi dma_config_tlv
in the follow up commits.

Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240402151828.175002-5-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

Revert "ASoC: SOF: Intel: hda-dai-ops: only allocate/release streams for first CPU DAI"

This reverts commit f8ba62ac863c33fc0d8ac3f1270985c2b77f4377.

The SoundWire aggregated solution was to use one DMA on multiple links.
But, the solution changed to use one DMA for each link. It means that
we should assign HDaudio stream_tag for each cpu_dai.

Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240402151828.175002-4-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

Revert "ASoC: SOF: Intel: hda-dai-ops: reset device count for SoundWire DAIs"

This reverts commit 699e146d9ebf42ee2a5d4e4e28f7a49c4aef0105.
Don't reset device_count as we will use the multi-gateway firmware
configuration.

Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240402151828.175002-3-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda-dai: fix channel map configuration for aggregated dailink

The existing code derives the channel map used to program the HDaudio
link DMA from the hw_params, but that is not quite right in the case
of aggregation. The code in soc-pcm.c splits the hw_params depending
on the codec_ch_map, and we need to reconstruct the channel-map to
insert the data in the right places.

This issue is seen only on amplifier feedback capture where the data
from the second amplifier was replaced by that of the first amplifier.

Note that the loop iterator of the macro for_each_rtd_cpu_dais() is
reused in a following loop. This is different to all existing usages
of that macro, hence the use of a boolean flag to avoid an access to
an uninitialized variable.

Fixes: 2960ee5c4814 ("ASoC: SOF: Intel: hda-dai: add helpers for SoundWire callbacks")
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240402151828.175002-2-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: ipc4-topology: remove shadowed variable

Cppcheck reports this:

sound/soc/sof/ipc4-topology.c:569:23: style: Local variable 'sdev'
shadows outer variable [shadowVariable]
  struct snd_sof_dev *sdev = snd_soc_component_get_drvdata(scomp);
                      ^
sound/soc/sof/ipc4-topology.c:512:22: note: Shadowed declaration
 struct snd_sof_dev *sdev = snd_soc_component_get_drvdata(scomp);
                     ^
sound/soc/sof/ipc4-topology.c:569:23: note: Shadow variable
  struct snd_sof_dev *sdev = snd_soc_component_get_drvdata(scomp);
                      ^

Remove shadowed variable.

Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Link: https://msgid.link/r/20240402145959.172619-5-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: ipc4-priv: align prototype and function declaration

Somehow instance_id and id were mixed. Align on instance_id for
consistency.

Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Link: https://msgid.link/r/20240402145959.172619-4-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: amd: acp-loader: abort firmware download on write error

cppcheck reports this issue:

sound/soc/sof/amd/acp-loader.c:299:6: style: Variable 'ret' is
reassigned a value before the old one has been
used. [redundantAssignment]
 ret = request_firmware(&adata->fw_dbin, fw_filename, sdev->dev);
     ^
sound/soc/sof/amd/acp-loader.c:289:6: note: ret is assigned
 ret = snd_sof_dsp_block_write(sdev, SOF_FW_BLK_TYPE_IRAM, 0,
     ^
sound/soc/sof/amd/acp-loader.c:299:6: note: ret is overwritten
 ret = request_firmware(&adata->fw_dbin, fw_filename, sdev->dev);
     ^

This behavior is probably unintentional, there's no reason to return
an error for the DRAM but not the IRAM.

Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Link: https://msgid.link/r/20240402145959.172619-3-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: ipc4-loader: save FW version info to debugfs

To check loaded FW version:

$ hexdump -C /sys/kernel/debug/sof/fw_version
00000000  02 00 07 00 63 00 01 00  00 00 00 00 00 00 00 00  |....c...........|

Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: Yong Zhi <yong.zhi@intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://msgid.link/r/20240402145959.172619-2-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: nau8325: Revise soc_enum structure definition.

This patch is modified nau8325_dac_oversampl_enum enumerated.
Use SOC_VALUE_ENUM instead to hide empty (presumably invalid) options.

Signed-off-by: Seven Lee <wtli@nuvoton.com>
Link: https://msgid.link/r/20240401104753.3432507-1-wtli@nuvoton.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rsnd: no exception for SCU

R-Car Gen1 and Gen4 doesn't have SCU, and current driver checks Gen1
(only) on each devices (A). But these drivers are assuming it might be
not enabled after that (B).

/* This driver doesn't support Gen1 at this point */
(A) if (rsnd_is_gen1(priv))
return 0;

(B) node = rsnd_src_of_node(priv);
if (!node)
return 0; /* not used is not error */

Gen4 DT doesn't have these device settings. (A) check doesn't work for
it, but (B) check handling it.
Gen1 DT doesn't have these device settings either, and (A) is handling
it, but (B) only is very enough. (A) is no longer needed.

This patch removes (A)

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://msgid.link/r/87v850a4dl.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rsnd: R-Car Gen1/Gen2 exception

Renesas Sound driver is using rsnd_is_genX() macro to handling
difference. We can use "grep rsnd_is_gen" to find-out difference for
each SoC today.

In general, driver uses flag to handling SoC difference. But in Renesas
Sound case, the differences are very specific, if you try to make it
general, the code will become be very complicated.

Some of them can use flag to handling, but we can find-out all diffence
by using "grep rsnd_is_gen" today. It is not so good, but not so bad so
far. So we will keep current style for a while.

This patch cares Gen1/Gen2 exception by using negative handling to makes no
more update for this part when we support Gen5 or later.

- if (rsnd_is_gen3() || rsnd_is_gen4() || rsnd_is_gen5() || ...)
+ if (!rsnd_is_gen1() && !rsnd_is_gen2())

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://msgid.link/r/87wmpga4ds.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rsnd: rename rsnd_is_e3() to rsnd_is_gen3_e3()

Renesas Sound driver is using rsnd_is_genX() macro to handling
difference. We can use "grep rsnd_is_gen" to find-out difference for
each SoC except rsnd_is_e3(). Let's put same rule for E3 as well.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://msgid.link/r/87y19wa4dy.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rsnd: don't get resource from ID

All SoC has reg-name, no need to support getting resource from ID
any more. Remove it. To get physical address for DMA settings,
it still need to know the index, but it is no longer fixed.
This patch tidyup it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://msgid.link/r/87zfuca4e6.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rsnd: cleanup regmap table

R-Car Sound is using almost same register mapping on each generation,
therefore we can share it. Here indicates some memo/comments.

SSI/SSIU/SCU have multi channels (SSI0 - SSI9 etc), but ADG is single,
so, we don't need to have multi channel settings for it

- rsnd_gen_regmap_init(..., 10, ..ADG, ...);
+ rsnd_gen_regmap_init(...,  1, ..ADG, ...);

Gen4 added new SSI_BUSIF, but it is not used

Gen4 calls rsnd_gen_regmap_init() for SDMC, but the address is calculated
by rsnd_gen4_dma_addr(). It needs physical address, but other info are
not needed. use NULL conf for it.

- rsnd_gen_regmap_init(..., 10, ..SDMC, conf_sdmc);
+ rsnd_gen_regmap_init(...,  1, ..SDMC, conf_null);

Gen4 have only SSI0/SSIU0, we don't need to have multi channel settings.

- rsnd_gen_regmap_init(..., 10, ..SSI, ...);
+ rsnd_gen_regmap_init(...,  1, ..SSI, ...);

SSI_MODE2 address is different between Gen2/3 and Gen4, but it is not
used on Gen4. Thus we can share same regmap table.

Gen2/3 : 0x808
Gen4   : 0xa0c

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Tested-by: Khanh Le <khanh.le.xr@renesas.com>
Link: https://msgid.link/r/871q7obiyt.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: soc-card: Add KUnit test case for snd_soc_card_get_kcontrol

Add a new snd-soc-card KUnit test with a simple test case for
snd_soc_card_get_kcontrol() and snd_soc_card_get_kcontrol_locked().

Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Link: https://msgid.link/r/20240401100210.61277-4-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: soc-card: Use snd_ctl_find_id_mixer() instead of open-coding

Use the snd_ctl_find_id_mixer[_locked]() wrapper in
snd_soc_card_get_kcontrol[_locked]() instead of open-coding a custom
list walk of the card controls list.

Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Link: https://msgid.link/r/20240401100210.61277-3-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>