ASoC: test-component: add Test Component YAML bindings

This patch adds test-component sound device YAML bindings.
It can be used for Sound Test/Debug.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/878ryyn8s4.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rt5682: fix a little pop while playback

A little pop can be heard obviously from HP while playing a silent.
This patch fixes it by using two functions:
1. Enable HP 1bit output mode.
2. Change the charge pump switch size during playback on and off.

Signed-off-by: Derek Fang <derek.fang@realtek.com>
Link: https://lore.kernel.org/r/20211014094054.811-1-derek.fang@realtek.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rt1011: Fix 'I2S Reference' enum control

There are several things the patch adding the support for 'I2S Reference'
got wrong:
- "None" selection is in fact equals to last selected reference
- The custom put overrides RX/TX len, TDM slot sizes, etc
- the enum is useless in most part for the reference tracking
- there is no need for EXT control as there is a single bit in
  RT1011_TDM1_SET_1 register (bit 7) which selects the reference
- it was using ucontrol->value.integer.value[0] in the put/get callbacks
  which causesed access to 'I2S Reference' enum with alsamixer to fail

Complements: c3de683c4d1d ("ASoC: rt1011: Fix 'I2S Reference' enum control caused error")
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Link: https://lore.kernel.org/r/20211013123300.11095-1-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: dt-bindings: rockchip: i2s-tdm: Fix rockchip,i2s-[rt]x-route

make dt_binding_check:

    Documentation/devicetree/bindings/sound/rockchip,i2s-tdm.yaml: properties:rockchip,i2s-rx-route: {'required': ['maxItems']} is not allowed for {'$ref': '/schemas/types.yaml#/definitions/uint32-array', 'description': 'Defines the mapping of I2S RX sdis to I2S data bus lines. By default, they are mapped one-to-one. rockchip,i2s-rx-route = <3> would mean sdi3 is receiving from data0.', 'maxItems': 4, 'items': [{'enum': [0, 1, 2, 3]}]}
    hint: "maxItems" is not needed with an "items" list
    from schema $id: http://devicetree.org/meta-schemas/items.yaml#
    Documentation/devicetree/bindings/sound/rockchip,i2s-tdm.yaml: properties:rockchip,i2s-tx-route: {'required': ['maxItems']} is not allowed for {'$ref': '/schemas/types.yaml#/definitions/uint32-array', 'description': 'Defines the mapping of I2S TX sdos to I2S data bus lines. By default, they are mapped one-to-one. rockchip,i2s-tx-route = <3> would mean sdo3 is sending to data0.', 'maxItems': 4, 'items': [{'enum': [0, 1, 2, 3]}]}
    hint: "maxItems" is not needed with an "items" list
    from schema $id: http://devicetree.org/meta-schemas/items.yaml#

Drop the "-" from the enums to fix this.

Fixes: 510f1c133aedcf69 ("ASoC: dt-bindings: rockchip: add i2s-tdm bindings")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/606809e10db02f92b1e7f90c491cc72dd8e16f79.1634132907.git.geert+renesas@glider.be
Signed-off-by: Mark Brown <broonie@kernel.org>

Merge series "ASoC: rt9120: Add Richtek RT9120 supprot" from cy_huang <u0084500@gmail.com>
ChiYuan Huang <cy_huang@richtek.com>:

From: ChiYuan Huang <cy_huang@richtek.com>

This patch series Add the Richtek RT9120 support.

In v4:
- Add 'classd_tlv' for 'SPK Gain Volume' control item.
- Unify the tlv declaration to the postfix '_tlv'.
- Fix 'digital_tlv' mute as 1 to declare the minimum is muted.

In v3:
- Add dvdd regulator binding to check the dvdd voltage domain.
- Refine sdo_select_text.
- Use switch case in 'internal_power_event' function.
- Remove the volume and mute initially write in component probe.
- Remove the mute API. It's no need by HW design.

In v2:
- Add missing #sound-dai-cells property.

ChiYuan Huang (2):
  ASoC: dt-bindings: rt9120: Add initial bindings
  ASoC: rt9120: Add rt9210 audio amplifier support

 .../devicetree/bindings/sound/richtek,rt9120.yaml  |  59 +++
 sound/soc/codecs/Kconfig                           |  10 +
 sound/soc/codecs/Makefile                          |   2 +
 sound/soc/codecs/rt9120.c                          | 495 +++++++++++++++++++++
 4 files changed, 566 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/sound/richtek,rt9120.yaml
 create mode 100644 sound/soc/codecs/rt9120.c

--
2.7.4

ASoC: rt9120: Add rt9210 audio amplifier support

Add Richtek rt9120 audio amplifier support.

Signed-off-by: ChiYuan Huang <cy_huang@richtek.com>
Link: https://lore.kernel.org/r/1634088519-995-3-git-send-email-u0084500@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: dt-bindings: rt9120: Add initial bindings

Add initial bindings for Richtek rt9120 audio amplifier.

Signed-off-by: ChiYuan Huang <cy_huang@richtek.com>
Link: https://lore.kernel.org/r/1634088519-995-2-git-send-email-u0084500@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: amd: vangogh: constify static struct snd_soc_dai_ops

The only usage of acp5x_i2s_dai_ops is to assign its address to the ops
field in the snd_soc_dai_driver struct, which is a pointer to const.
Make it const to allow the compiler to put it in read-only memory.

Signed-off-by: Rikard Falkeborn <rikard.falkeborn@gmail.com>
Link: https://lore.kernel.org/r/20211012211506.21159-1-rikard.falkeborn@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: mediatek: Constify static snd_soc_ops

These are only assigned to the ops field in the snd_soc_dai_link struct
which is a pointer to const struct snd_soc_ops. Make them const to allow
the compiler to put them in read-only memory.

Signed-off-by: Rikard Falkeborn <rikard.falkeborn@gmail.com>
Link: https://lore.kernel.org/r/20211012205521.14098-1-rikard.falkeborn@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rt9120: Drop rt9210 audio amplifier support

This drops the rt9210 support due to a race with a new version being
sent out for some incremental changes.

Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rt1011: Fix 'I2S Reference' enum control caused error

Access to 'I2S Reference' enum causes alsamixer to fail to load:
$ alsamixer
cannot load mixer controls: Invalid argument

cml_rt1011_rt5682 cml_rt1011_rt5682: control 2:0:0:TL I2S Reference:0: access overflow

The reason is that the original patch adding the code was using
ucontrol->value.integer.value[0]
instead the correct
ucontrol->value.enumerated.item[0]

for an ENUM control.

Fixes: 87f40af26c262 ("ASoC: rt1011: add i2s reference control for rt1011")
Reported-by: Seppo Ingalsuo <seppo.ingalsuo@linux.intel.com>
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Link: https://lore.kernel.org/r/20211011144518.2518-1-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

Merge series "ASoC: Intel: bytcr_rt5651: few cleanups" from Andy Shevchenko <andriy.shevchenko@linux.intel.com>:

The small set of cleanups against bytcr_rt5651 board file.

In v2:
- added commit message to patch 2 (Joe, Pierre)
- added cover letter (Pierre)
- added Hans to Cc list (Hans)

Andy Shevchenko (4):
  ASoC: Intel: bytcr_rt5651: Get platform data via dev_get_platdata()
  ASoC: Intel: bytcr_rt5651: Use temporary variable for struct device
  ASoC: Intel: bytcr_rt5651: use devm_clk_get_optional() for mclk
  ASoC: Intel: bytcr_rt5651: Utilize dev_err_probe() to avoid log
    saturation

 sound/soc/intel/boards/bytcr_rt5651.c | 118 +++++++++++---------------
 1 file changed, 50 insertions(+), 68 deletions(-)

--
2.33.0

ASoC: Intel: bytcr_rt5651: Utilize dev_err_probe() to avoid log saturation

dev_err_probe() avoids printing into log when the deferred probe is invoked.
This is possible when clock provider is pending to appear.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20211007170250.27997-5-andriy.shevchenko@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: bytcr_rt5651: use devm_clk_get_optional() for mclk

The devm_clk_get_optional() helper returns NULL when devm_clk_get()
returns -ENOENT. This makes things slightly cleaner. The added benefit
is mostly cosmetic.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20211007170250.27997-4-andriy.shevchenko@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: bytcr_rt5651: Use temporary variable for struct device

Use temporary variable for struct device to make code neater.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20211007170250.27997-3-andriy.shevchenko@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: bytcr_rt5651: Get platform data via dev_get_platdata()

Access to platform data via dev_get_platdata() getter to make code cleaner.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20211007170250.27997-2-andriy.shevchenko@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: bytcr_rt5640: Utilize dev_err_probe() to avoid log saturation

dev_err_probe() avoids printing into log when the deferred probe is invoked.
This is possible when clock provider is pending to appear.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Tested-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20211007165715.27463-5-andriy.shevchenko@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: bytcr_rt5640: use devm_clk_get_optional() for mclk

The devm_clk_get_optional() helper returns NULL when devm_clk_get()
returns -ENOENT. This makes things slightly cleaner. The added benefit
is mostly cosmetic.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Tested-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20211007165715.27463-4-andriy.shevchenko@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: bytcr_rt5640: Use temporary variable for struct device

Use temporary variable for struct device to make code neater.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Tested-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20211007165715.27463-3-andriy.shevchenko@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: bytcr_rt5640: Get platform data via dev_get_platdata()

Access to platform data via dev_get_platdata() getter to make code cleaner.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Tested-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20211007165715.27463-2-andriy.shevchenko@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

sh: Use modern ASoC DAI format terminology

The SH machine drivers have some ASoC DAI format specifications that use
older defines based on outdated terminology which we're trying to retire,
update to the new bindings.

Signed-off-by: Mark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/20210915172302.36677-1-broonie@kernel.org
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: amd: Kconfig: Select fch clock support with machine driver

We are using fch clock controller as parent mclk source for rt5682
codec. Add config to enable clock framework support for 48MHz fixed
clock when machine driver config is selected.

Signed-off-by: Ajit Kumar Pandey <AjitKumar.Pandey@amd.com>
Link: https://lore.kernel.org/r/20211011055354.67719-1-AjitKumar.Pandey@amd.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rockchip: i2s-tdm: Fix error handling on i2s_tdm_prepare_enable_mclk failure

In the case where the call to i2s_tdm_prepare_enable_mclk fails the
function returns before the error handling goto is executed. Fix this
by removing the return do perform the intended error handling exit.

Fixes: 081068fd6414 ("ASoC: rockchip: add support for i2s-tdm controller")
Addresses-Coverity: ("Structurally dead code")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Message-Id: <20211008095430.62680-2-colin.king@canonical.com>
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rockchip: i2s-tdm: Remove call to rockchip_i2s_ch_to_io

The call to rockchip_i2s_ch_to_io is only useful for its return
value which is not being used. The function call also has no
side effects, the call is effectively useless and can be removed.

Addresses-Coverity: ("Useless call")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Message-Id: <20211008095430.62680-1-colin.king@canonical.com>
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: mediatek: mt8195: update audsys clock parent name

Because clock names are modified in mediatek CCF driver, sync the updated
clock names to audsys driver.

Signed-off-by: Trevor Wu <trevor.wu@mediatek.com>
Message-Id: <20211008070424.14347-1-trevor.wu@mediatek.com>
Signed-off-by: Mark Brown <broonie@kernel.org>

Merge series "ASoC: rt9120: Add Richtek RT9120 supprot" from cy_huang <u0084500@gmail.com>
ChiYuan Huang <cy_huang@richtek.com>:

From: ChiYuan Huang <cy_huang@richtek.com>

This patch series Add the Richtek RT9120 support.

In v3:
- Add dvdd regulator binding to check the dvdd voltage domain.
- Refine sdo_select_text.
- Use switch case in 'internal_power_event' function.
- Remove the volume and mute initially write in component probe.
- Remove the mute API. It's no need by HW design.

In v2:
- Add missing #sound-dai-cells property.

ChiYuan Huang (2):
  ASoC: dt-bindings: rt9120: Add initial bindings
  ASoC: rt9120: Add rt9210 audio amplifier support

 .../devicetree/bindings/sound/richtek,rt9120.yaml  |  59 +++
 sound/soc/codecs/Kconfig                           |  10 +
 sound/soc/codecs/Makefile                          |   2 +
 sound/soc/codecs/rt9120.c                          | 489 +++++++++++++++++++++
 4 files changed, 560 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/sound/richtek,rt9120.yaml
 create mode 100644 sound/soc/codecs/rt9120.c

--
2.7.4

Merge series "ASoC: Intel: bytcht_es8316: few cleanups" from Andy Shevchenko <andriy.shevchenko@linux.intel.com>:

The small set of cleanups against bytcht_es8316 board file.

In v4:
- fixed Pierre's email (Pierre)
- added Hans to the Cc list

In v3:
- actually added a Pierre's tag (Mark)

In v2:
- added tag (Pierre)
- added commit message to the patch 2 (Joe)

Andy Shevchenko (4):
  ASoC: Intel: bytcht_es8316: Get platform data via dev_get_platdata()
  ASoC: Intel: bytcht_es8316: Use temporary variable for struct device
  ASoC: Intel: bytcht_es8316: Switch to use gpiod_get_optional()
  ASoC: Intel: bytcht_es8316: Utilize dev_err_probe() to avoid log
    saturation

 sound/soc/intel/boards/bytcht_es8316.c | 37 +++++++++-----------------
 1 file changed, 12 insertions(+), 25 deletions(-)

--
2.33.0

ASoC: SOF: prepare code to allocate IPC messages in fw_ready

The fixed maximum size of IPC message does not allow for large
transfers, e.g. for filter data. Currently such messages will
be divided into smaller pieces and sent to firmware in multiple
chunks. For future IPC, this strategy is not suitable.

The maximum IPC message size is limited by host box size which
can be known when firmware is ready, so the fw_ready callback
can allocate IPC messages with platform-specific sizes instead
of the current fixed-size.

To be compatible with released firmware, current platforms will
still use SOF_IPC_MSG_MAX_SIZE. For future platforms, there will
be a new fw_ready function and the platform-specific allocation
will take place there.

Signed-off-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Bard Liao <bard.liao@intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Message-Id: <20211008093836.28210-1-peter.ujfalusi@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: dt-bindings: mediatek: mt8192: re-add audio afe document

The document was merged as commit 1afc60e00de3 ("dt-bindings:
mediatek: mt8192: add audio afe document").

However, [1] revealed that the commit 1afc60e00de3 breaks
dt_binding_check due to dt-bindings/clock/mt8192-clk.h doesn't
exist.

As a temporary fix, commit 7d94ca3c8acd ("ASoC: mt8192: revert
add audio afe document") reverted commit 1afc60e00de3.

dt-bindings/clock/mt8192-clk.h is in mainline per commit
f35f1a23e0e1 ("clk: mediatek: Add dt-bindings of MT8192 clocks").
Re-adds the document back.

[1]: https://mailman.alsa-project.org/pipermail/alsa-devel/2020-November/176873.html

Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Tzung-Bi Shih <tzungbi@google.com>
Message-Id: <20211008025523.1852319-1-tzungbi@google.com>
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: bytcht_es8316: Utilize dev_err_probe() to avoid log saturation

dev_err_probe() avoids printing into log when the deferred probe is invoked.
This is possible when clock provider is pending to appear.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Message-Id: <20211007164523.27094-5-andriy.shevchenko@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: bytcht_es8316: Switch to use gpiod_get_optional()

First of all, replace indexed API by plain one since we have index 0.
Second, switch to optional variant and drop duplicated code.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Message-Id: <20211007164523.27094-4-andriy.shevchenko@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: bytcht_es8316: Use temporary variable for struct device

Use temporary variable for struct device to make code neater.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Message-Id: <20211007164523.27094-3-andriy.shevchenko@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: bytcht_es8316: Get platform data via dev_get_platdata()

Access to platform data via dev_get_platdata() getter to make code cleaner.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Message-Id: <20211007164523.27094-2-andriy.shevchenko@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rt9120: Add rt9210 audio amplifier support

Add Richtek rt9120 audio amplifier support.

Signed-off-by: ChiYuan Huang <cy_huang@richtek.com>
[Fix a Gain->Volume -- broonie]
Message-Id: <1633668612-25524-3-git-send-email-u0084500@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: dt-bindings: rt9120: Add initial bindings

Add initial bindings for Richtek rt9120 audio amplifier.

Signed-off-by: ChiYuan Huang <cy_huang@richtek.com>
Message-Id: <1633668612-25524-2-git-send-email-u0084500@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>

Merge series "Add reset-gpios handling for max98927" from Alejandro Tafalla <atafalla@dnyon.com>:

The max98927 codec on some devices (i.e. Xiaomi Mi A2 Lite phone) requires
hardware-resetting the codec by driving a reset-gpio. This series adds
support for it through an optional reset-gpios property.

v4:
* Correctly assert/deassert the GPIO states
* Wait for the i2c port to be ready after reset
* Reset device when removed

v3:
* Fix indentation on the dev_err_probe line

v2:
* Use dev_err_probe instead of dev_err

Alejandro Tafalla (2):
  ASoC: max98927: Handle reset gpio when probing i2c
  dt-bindings: sound: max98927: Add reset-gpios optional property

 .../devicetree/bindings/sound/max9892x.txt    |  3 +++
 sound/soc/codecs/max98927.c                   | 25 +++++++++++++++++++
 sound/soc/codecs/max98927.h                   |  1 +
 3 files changed, 29 insertions(+)

--
2.33.0

Merge series "Introduce new SOF helpers" from Daniel Baluta <daniel.baluta@oss.nxp.com>
Daniel Baluta <daniel.baluta@nxp.com>:

From: Daniel Baluta <daniel.baluta@nxp.com>

This patchseries adds new helpers in order to reduce code duplication
and prepare for compress audio support with SOF.

Bud Liviu-Alexandru (1):
  ASoC: SOF: Make Intel IPC stream ops generic

Daniel Baluta (3):
  ASoC: SOF: Introduce snd_sof_mailbox_read / snd_sof_mailbox_write
    callbacks
  ASoC: SOF: imx: Use newly introduced generic IPC stream ops
  ASoC: SOF: Introduce fragment elapsed notification API

 sound/soc/sof/Kconfig                         |  4 ++
 sound/soc/sof/Makefile                        |  3 +-
 sound/soc/sof/compress.c                      | 51 +++++++++++++++++++
 sound/soc/sof/imx/Kconfig                     |  2 +
 sound/soc/sof/imx/imx8.c                      | 39 +++++++-------
 sound/soc/sof/imx/imx8m.c                     | 26 ++++------
 sound/soc/sof/intel/Makefile                  |  3 --
 sound/soc/sof/intel/apl.c                     |  4 ++
 sound/soc/sof/intel/bdw.c                     | 12 +++--
 sound/soc/sof/intel/byt.c                     | 24 ++++++---
 sound/soc/sof/intel/cnl.c                     |  4 ++
 sound/soc/sof/intel/icl.c                     |  4 ++
 sound/soc/sof/intel/pci-tng.c                 | 12 +++--
 sound/soc/sof/intel/tgl.c                     |  4 ++
 sound/soc/sof/ipc.c                           |  6 ++-
 sound/soc/sof/ops.h                           | 15 ++++++
 sound/soc/sof/pcm.c                           |  7 ++-
 sound/soc/sof/sof-audio.h                     | 11 +++-
 sound/soc/sof/sof-priv.h                      | 28 ++++++----
 .../sof/{intel/intel-ipc.c => stream-ipc.c}   | 50 +++++++++---------
 sound/soc/sof/topology.c                      |  6 ++-
 21 files changed, 218 insertions(+), 97 deletions(-)
 create mode 100644 sound/soc/sof/compress.c
 rename sound/soc/sof/{intel/intel-ipc.c => stream-ipc.c} (58%)

--
2.27.0

Merge series "ASoC: SOF: Improvements for debugging" from Peter Ujfalusi <peter.ujfalusi@linux.intel.com>:

Hi,

The aim of this series is to clean up, make it easier to interpret and less
'chatty' prints aimed for debugging errors.

For example currently the DSP/IPC dump is printed every time we have an IPC
timeout and it is posible to lost the first and more indicative dump to find the
rootcause.

Regards,
Peter
---
Peter Ujfalusi (18):
  ASoC: SOF: debug: Swap the dsp_dump and ipc_dump sequence for
    fw_exception
  ASoC: SOF: ipc and dsp dump: Add markers for better visibility
  ASoC: SOF: Print the dbg_dump and ipc_dump once to reduce kernel log
    noise
  ASoC: SOF: loader: Print the DSP dump if boot fails
  ASoC: SOF: intel: atom: No need to do a DSP dump in atom_run()
  ASoC: SOF: debug/ops: Move the IPC and DSP dump functions out from the
    header
  ASoC: SOF: debug: Add SOF_DBG_DUMP_OPTIONAL flag for DSP dumping
  ASoC: SOF: intel: hda-loader: Use snd_sof_dsp_dbg_dump() for DSP dump
  ASoC: SOF: Drop SOF_DBG_DUMP_FORCE_ERR_LEVEL and sof_dev_dbg_or_err
  ASoC: SOF: debug: Print out the fw_state along with the DSP dump
  ASoC: SOF: ipc: Re-enable dumps after successful IPC tx
  ASoC: SOF: ops: Force DSP panic dumps to be printed
  ASoC: SOF: Introduce macro to set the firmware state
  ASoC: SOF: intel: hda: Drop 'error' prefix from error dump functions
  ASoC: SOF: core: Clean up snd_sof_get_status() prints
  ASoC: SOF: loader: Drop SOF_DBG_DUMP_REGS flag when firmware start
    fails
  ASoC: SOF: Intel: hda-loader: Drop SOF_DBG_DUMP_REGS flag from
    dbg_dump calls
  ASoC: SOF: Intel: hda: Dump registers and stack when SOF_DBG_DUMP_REGS
    is set

Pierre-Louis Bossart (1):
  ASoC: SOF: core: debug: force all processing on primary core

 sound/soc/sof/core.c             | 24 ++++++-------
 sound/soc/sof/debug.c            | 61 ++++++++++++++++++++++++++++++--
 sound/soc/sof/intel/atom.c       |  5 +--
 sound/soc/sof/intel/hda-loader.c | 11 +++---
 sound/soc/sof/intel/hda.c        | 16 +++------
 sound/soc/sof/ipc.c              | 10 ++++--
 sound/soc/sof/loader.c           | 11 ++++--
 sound/soc/sof/ops.c              |  3 ++
 sound/soc/sof/ops.h              | 12 +------
 sound/soc/sof/pm.c               |  6 ++--
 sound/soc/sof/sof-priv.h         | 31 ++++++++++------
 sound/soc/sof/topology.c         |  6 ++++
 12 files changed, 131 insertions(+), 65 deletions(-)

--
2.33.0

Merge series "Rockchip I2S/TDM controller" from Nicolas Frattaroli <frattaroli.nicolas@gmail.com>:

Hello,

this is version 5 of the I2S/TDM driver patchset. A big thanks
to everyone who has provided their valuable feedback so far.

Changes in v5:
 driver:
 - change comment style of the first comment to C++ style
 - make refcount non-atomic, as it's only ever used inside
   a spinlock
 - use newer SND_SOC_DAIFMT_CB* defines
 - change ternary statements to if/else conditions
 - make _clk_compensation_put return 1 if clock changed
 - implement set_bclk_ratio callback
 - always set half frame sync mode in TDM mode
 - automatically enable mclk-calibrate mode when the clocks for
   it are specified in the device tree
 bindings:
 - add Reviewed-by: Rob Herring
 - drop rockchip,frame-width property (done by set_bclk_ratio)
 - drop rockchip,fsync-half-frame property
 - drop rockchip,mclk-calibrate property
 dts:
 - drop empty codec block from Quartz64 device tree

Changes in v4:
 driver:
 - factor TDE/RDE enable/disable into their own inlined functions
 - add an RDE disable in a location where it looks like it was
   forgotten (rxctrl else), judging by corresponding TDE code
 - remove parentheses around CLK_PPM_MIN/MAX values
 - wording + titlecasing in the clock compensation control
 - use if statement in precious_reg instead
 - refactor rockchip_i2s_io_multiplex to have the switch statements
   in a function call to make the function less unwieldy
 - get rid of IS_ERR checks around clk enable/disable calls where
   already checked before by the probe
 - reworded some error message strings
 - fix potential deadlock in txrxctrl found by Sugar Zhang
   using spin_lock_irqsave
 - fix potential deadlock in trcm_mode found by Sugar Zhang
   using spin_lock_irqsave
 - use devm_platform_get_and_ioremap_resource in probe
 - only set DMA things if controller has capture/playback ability.
   Did not move this into init_dai because I'd then need to pass in
   the res and probe it earlier in the function, and it's also used
   elsewhere in the probe function
 - use _get_optional_exclusive for reset controls, as some controllers
   only have capture or playback capability
 bindings:
 - remove status = "okay" since that's the default
 - change the path configs to be an enum
 - rename "foo" to "bus"
 - make resets optional as controller may lack either playback or
   capture capability, and therefore also doesn't have a reset.
   At least one reset is still required, because a controller with
   no playback and no capture is not very useful

Changes in v3:
 driver:
 - alphabetically sort includes
 - check pm_runtime_get_sync return value, act on it
 - remove unnecessary initialisers in set_fmt
 - use udelay(15) in retry code: 10 retries * 15 = 150, so at worst
   we wait the full i2s register access delay
 - fix some weird returns to return directly
 - use __maybe_unused instead of #ifdef CONFIG_PM_SLEEP, also put
   __maybe_unused on the runtime callbacks
 - use (foo) instead of foo in header macros for precedence reasons
 - when using mclk-calibrate, also turn off/on those clocks during
   suspend and resume operations
 - remove mclk_tx and mclk_rx reenablement code in remove
 - move hclk enablement further down the probe, and disable it
   on probe failure
 - make reset controls mandatory, since the bindings state this too
 - use _exclusive for getting the reset controls
 - change reset assert/deassert delays to both be 10 usec
   (thank you Sugar Zhang!)
 - properly prepare and enable all mclks in probe, especially before
   calling clk_get_rate on them
 - if registering PCM fails, also use the cleanup error path instead of
   returning directly
 - bring back playback and capture only but in the way Sugar Zhang
   suggested it: set those modes depending on dma-names
 - rework clock enablement in general. Probe now always enables these,
   instead of relying on the pm resume thing
 - add myself to MAINTAINERS for this driver
 dt bindings:
 - fix a description still mentioning clk-trcm in the schema
 - document rockchip,io-multiplex, a property that describes the
   hardware as having multiplexed I2S GPIOs so direction needs to
   be changed dynamically
 - document rockchip,mclk-calibrate, which allows specifying
   different clocks for the two sample rate bases and switch between
   them as needed
 - dma-names now doesn't have a set order and items can be absent to
   indicate that the controller doesn't support this mode
 - add myself to MAINTAINERS for these bindings

Changes in v2:
 - remove ad-hoc writeq and needless (and broken) optimisation in
   reset assert/deassert. This wouldn't have worked on Big Endian,
   and would've been pointless on any other platform, as the
   overhead for saving one write was comparatively big
 - fix various checkpatch issues
 - get rid of leftover clk-trcm in schema
 - set status = "okay" in example in schema instead of "disabled"
 - change dma-names so rx is first, adjust device trees as necessary
 - properly reference uint32-array for rx-route and tx-route
   instead of uint32
 - replace trcm-sync with two boolean properties, adjust DT changes
   accordingly and also get rid of the header file
 - get rid of rockchip,no-dmaengine. This was only needed for
   some downstream driver and shouldn't be in the DT
 - get rid of rockchip,capture-only/playback-only. Rationale being
   that I have no way to test whether they're needed, and
   unconditionally setting channels_min to 0 breaks everything
 - change hclk description in "clocks"

Nicolas Frattaroli (4):
  ASoC: rockchip: add support for i2s-tdm controller
  ASoC: dt-bindings: rockchip: add i2s-tdm bindings
  arm64: dts: rockchip: add i2s1 on rk356x
  arm64: dts: rockchip: add analog audio on Quartz64

 .../bindings/sound/rockchip,i2s-tdm.yaml      |  198 ++
 MAINTAINERS                                   |    7 +
 .../boot/dts/rockchip/rk3566-quartz64-a.dts   |   31 +-
 arch/arm64/boot/dts/rockchip/rk356x.dtsi      |   26 +
 sound/soc/rockchip/Kconfig                    |   11 +
 sound/soc/rockchip/Makefile                   |    2 +
 sound/soc/rockchip/rockchip_i2s_tdm.c         | 1848 +++++++++++++++++
 sound/soc/rockchip/rockchip_i2s_tdm.h         |  398 ++++
 8 files changed, 2520 insertions(+), 1 deletion(-)
 create mode 100644 Documentation/devicetree/bindings/sound/rockchip,i2s-tdm.yaml
 create mode 100644 sound/soc/rockchip/rockchip_i2s_tdm.c
 create mode 100644 sound/soc/rockchip/rockchip_i2s_tdm.h

--
2.33.0

ASoC: soc-pcm: restore mixer functionality

Recent changes in soc-pcm completely broke basic support for mixers on
Intel systems: the filters on BE states prevent the connection of a
second mixer input while the back-end is already active.

Rather than reverting the changes, which would be problematic for
Tegra systems, this patch suggests an additional filter which will
only apply to Tegra systems. This is a temporary solution which will
have to be revisited - additional issues have been reported with DPCM.

Fixes: 0c25db3f7621 ('ASoC: soc-pcm: Don't reconnect an already active BE')
Suggested-by: Sameer Pujar <spujar@nvidia.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Link: https://lore.kernel.org/r/20211004212141.193136-1-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rt5682s: Fix hp pop produced immediately after resuming

When the system plays a sound immediately after resuming from S3,
it could hear a little pop from headphones.
It is due to the HP was unmuted before the completion of
jack re-detection finished in parallel.

This patch adds a lock to make sure the HP unmute after jack detect handler,
and adds a few depop changes.

Signed-off-by: Derek Fang <derek.fang@realtek.com>
Link: https://lore.kernel.org/r/20211007085519.12543-1-derek.fang@realtek.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: dt-bindings: rockchip: add i2s-tdm bindings

This adds the YAML bindings for the Rockchip I2S/TDM audio driver.

Signed-off-by: Nicolas Frattaroli <frattaroli.nicolas@gmail.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/20211001171531.178775-3-frattaroli.nicolas@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rockchip: add support for i2s-tdm controller

This commit adds support for the rockchip i2s-tdm controller,
which enables audio output on the following rockchip SoCs:
- px30
- rk1808
- rk3308
- rk3566
- rk3568
- rv1126

This is a cleaned up version of the downstream vendor kernel's
driver. It can be enabled through the SND_SOC_ROCKCHIP_I2S_TDM
configuration option.

Signed-off-by: Nicolas Frattaroli <frattaroli.nicolas@gmail.com>
Link: https://lore.kernel.org/r/20211001171531.178775-2-frattaroli.nicolas@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Introduce fragment elapsed notification API

This patch prepares the introduction of the compress API with SOF.

After each fragment is accepted by the DSP we need to inform
the userspace applications that they can send the next fragment.
This is done via snd_compr_fragment_elapsed.

Similar with the PCM case, in order to avoid sending an IPC before
the previous IPC is handled we need to schedule a delayed work to
call snd_compr_fragment_elapsed().

See snd_sof_pcm_period_elapsed.

To sum up this patch offers the following API to SOF code:
* snd_sof_compr_init_elapsed_work
* snd_sof_compr_fragment_elapsed

Note that implementation for compressed function is in a new file
selected via CONFIG_SND_SOC_SOF_COMPRESS invisible config option.
This option is automatically selected for platforms that support
the compress interface. For now only i.MX8 platforms support this.

For symmetry we introduce snd_sof_pcm_init_elapsed_work to setup
the work struct for PCM case.

Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Signed-off-by: Bud Liviu-Alexandru <budliviu@gmail.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Paul Olaru <paul.olaru@oss.nxp.com>
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Link: https://lore.kernel.org/r/20211004152147.1268978-5-daniel.baluta@oss.nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: imx: Use newly introduced generic IPC stream ops

This makes IMX use the newly introduced generic IPC ops
instead of imx specific ones, and removes the old IMX
ipc ops, as they are no longer needed.

Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Signed-off-by: Bud Liviu-Alexandru <budliviu@gmail.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Link: https://lore.kernel.org/r/20211004152147.1268978-4-daniel.baluta@oss.nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Make Intel IPC stream ops generic

This operations should be generic as there is nothing Intel
specific. This works well for NXP i.MX8 stream IPC ops.

We start by moving sof/intel/intel-ipc.c into sof/stream-ipc.c and
rename the functions to be generic.

Notice that we use newly introduced snd_sof_dsp_mailbox_read
instead of sof_mailbox_read, to make sure that we are not
bound to existing MMIO memory access, and we allow platform
to implement their own memory access routines.

Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Signed-off-by: Bud Liviu-Alexandru <budliviu@gmail.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Daniel Baluta <daniel.baluta@gmail.com>
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Link: https://lore.kernel.org/r/20211004152147.1268978-3-daniel.baluta@oss.nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Introduce snd_sof_mailbox_read / snd_sof_mailbox_write callbacks

We need to introduce snd_sof_mailbox_{read/write} in order to provide
a generic way for mailbox access. These routines are optional, each
platform can implement their own specific routines.

So far, all platforms use mmapped I/O thus they can use custom made
routines sof_mailbox_read / sof_mailbox_write that use MMIO.

Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Signed-off-by: Bud Liviu-Alexandru <budliviu@gmail.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Link: https://lore.kernel.org/r/20211004152147.1268978-2-daniel.baluta@oss.nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: pipelines: Harmonize all functions to use struct snd_sof_dev

First thing the pipelines function which have "struct device *dev" as
parameter do is:
struct snd_sof_dev *sdev = dev_get_drvdata(dev);

and in all cases the passed dev is actually coming from sdev->dev.

Skip this steps and pass directly the sdev to all pipelines related
functions as few of them already does this.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20211006111651.10027-1-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: topology: do not power down primary core during topology removal

When removing the topology components, do not power down
the primary core. Doing so will result in an IPC timeout
when the SOF PCI device runtime suspends.

Fixes: 0dcdf84289fb ("ASoC: SOF: add a "core" parameter to widget loading functions")

Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Link: https://lore.kernel.org/r/20211006104041.27183-1-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda: Dump registers and stack when SOF_DBG_DUMP_REGS is set

Instead of checking the fw_state to decide what information should be
printed, use the SOF_DBG_DUMP_REGS bit in the flags to dump registers and
stack.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20211006110645.26679-20-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda-loader: Drop SOF_DBG_DUMP_REGS flag from dbg_dump calls

In cl_dsp_init() we are powering up the DSP, register dump is not valid.
In hda_dsp_cl_boot_firmware() we are downloading the firmware to DSP, again
the register dump is not a valid concept.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20211006110645.26679-19-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: loader: Drop SOF_DBG_DUMP_REGS flag when firmware start fails

snd_sof_dsp_run() failure indicates that the DSP did not even booted up,
thus asking for dumping registers at this point is not valid.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20211006110645.26679-18-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: core: Clean up snd_sof_get_status() prints

Clean up the error prints when decoding the status in snd_sof_get_status():
Drop the "error:" prefixes from the prints,
Use %# to print hexadecimal numbers,
Reword some of the messages to be more precise,
For a known error print out the panic code as well,
For unknown error print only the panic code without the magic

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20211006110645.26679-17-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: intel: hda: Drop 'error' prefix from error dump functions

Drop the 'error' prefix printed in hda_dsp_dump_ext_rom_status(),
hda_ipc_irq_dump() and hda_ipc_dump() as it gives no value to the
information we print.

The DSP and IPC dump is marked now, which makes the 'error' prefix more
redundant.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20211006110645.26679-16-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Introduce macro to set the firmware state

Add sof_set_fw_state() macro to wrap the sdev->fw_state management to allow
actions to be taken when certain state is set or when state is changing.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20211006110645.26679-15-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: ops: Force DSP panic dumps to be printed

If a DSP panic happens we want to see the dumps.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20211006110645.26679-14-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: ipc: Re-enable dumps after successful IPC tx

The dumps are silenced after an IPC tx timeout by default.
The IPC timeout can indicate severe error (firmware crash) or in some cases
it is less devastating and the firmware remains operational, the timeout
was due to a scheduling spike or other anomaly.

In any case consequent IPC timeouts will not print dumps but if any IPC do
succeed than we should re-enable the dumps to print dumps the next time
a timeout might happen.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20211006110645.26679-13-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: debug: Print out the fw_state along with the DSP dump

The fw state can be an important information along with the DSP dump.
Print it out before the dump.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20211006110645.26679-12-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Drop SOF_DBG_DUMP_FORCE_ERR_LEVEL and sof_dev_dbg_or_err

The sof_dev_dbg_or_err() is only used by intel/hda.c when dumping dsp
debug information.
It was used to print the extended rom status in either dev_dbg (during
retries) and finally with dev_err, but other lines were printed with
dev_err regardless.

Since we now only print the dump once, the flag and the macros is no longer
needed.

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://lore.kernel.org/r/20211006110645.26679-11-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: intel: hda-loader: Use snd_sof_dsp_dbg_dump() for DSP dump

Do not call directly the hda_dsp_dump(), use the generic wrapper instead
to provide consistent output.

Mark the DSP dumps as optional to not spam the kernel log with the
exception of the last dump in case the DSP fails to run.

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://lore.kernel.org/r/20211006110645.26679-10-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: debug: Add SOF_DBG_DUMP_OPTIONAL flag for DSP dumping

The new SOF_DBG_DUMP_OPTIONAL flag can be used to mark a DSP dump that
should only be printed when the SOF_DBG_PRINT_ALL_DUMPS sof_core_debug
flag is set, otherwise it should be ignored and not printed.

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://lore.kernel.org/r/20211006110645.26679-9-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: debug/ops: Move the IPC and DSP dump functions out from the header

To be usable in platform code, move the IPC and DSP dump function to
debug.c and export it in a similar way as the snd_sof_handle_fw_exception()

Make the snd_sof_ipc_dump() static as it is only used in debug.c

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://lore.kernel.org/r/20211006110645.26679-8-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: intel: atom: No need to do a DSP dump in atom_run()

The core already prints a dump if the DSP failed to start in
snd_sof_run_firmware(), there is no need to print it locally as well.

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://lore.kernel.org/r/20211006110645.26679-7-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: loader: Print the DSP dump if boot fails

It can be useful to print the DSP dump from the core in case the DSP boot
failed.

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://lore.kernel.org/r/20211006110645.26679-6-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Print the dbg_dump and ipc_dump once to reduce kernel log noise

Do not print the dump more than once to keep the kernel log cleaner in case
of a firmware failure.

When the DSP is rebooted due to suspend or runtime_suspend reset the flags
to re-enable the dump prints.

Add also a debug flag to print all dumps to get more coverage if needed.

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://lore.kernel.org/r/20211006110645.26679-5-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: ipc and dsp dump: Add markers for better visibility

Add markers to identify the start and end of the IPC and DSP dumps in the
kernel log.

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://lore.kernel.org/r/20211006110645.26679-4-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: debug: Swap the dsp_dump and ipc_dump sequence for fw_exception

snd_sof_dsp_panic() only prints dsp_dump followed by flushing the DMA trace
buffer.

To retain similar 'sequence' first do an ipc_dump then the dsp_dump and
finally flush the trace buffer in case of fw_exception.

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://lore.kernel.org/r/20211006110645.26679-3-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: core: debug: force all processing on primary core

The topology file currently provides information on which
pipeline/processing is to be scheduled on which DSP core.

To help diagnose potential issues, this patch provides an override of
the 'core' tokens to use the primary core (typically core0). Of course
this may result in a Core0 activity that exceeds hardware
capabilities, so this should only be used when the total processing
fits on DSP - possibly using firmware mockup processing and stubs.

No new dmesg log was added to avoid adding noise during topology
parsing, but the existing logs will show the primary core being used.

This is strictly for validation/debug, products should NEVER use this
override, the topology is assumed to be the description of the
firmware graph.

Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Bard Liao <bard.liao@intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Link: https://lore.kernel.org/r/20211006110645.26679-2-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: max98927: Add reset-gpios optional property

Signed-off-by: Alejandro Tafalla <atafalla@dnyon.com>
Link: https://lore.kernel.org/r/5004d52da527bf1dd9e5b17e20b2ce50a0b57b5a.1633572679.git.atafalla@dnyon.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: max98927: Handle reset gpio when probing i2c

The max98927 codec on some devices requires pulling a reset gpio before
responding to any i2c command. This commit adds support for it through
an optional reset-gpios property.

Signed-off-by: Alejandro Tafalla <atafalla@dnyon.com>
Link: https://lore.kernel.org/r/d74b12a79ae9ca728d5d9e64c55b3e59e8c0e509.1633572679.git.atafalla@dnyon.com
Signed-off-by: Mark Brown <broonie@kernel.org>

Merge series "ASoC: Intel: machine driver updates for 5.16" from Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>:

One quirk for a Dell TigerLake/SoundWire device, and initial support
for platforms based on the ES8336 codec (aka ES8316). For full
functionality, an update of the codec driver will be needed.

Pierre-Louis Bossart (5):
  ASoC: Intel: soc-acpi: apl/glk/tgl: add entry for devices based on
    ES8336 codec
  ALSA: intel-dsp-config: add quirk for APL/GLK/TGL devices based on
    ES8336 codec
  ASoC: Intel: add machine driver for SOF+ES8336
  ASoC: Intel: soc-acpi: add missing quirk for TGL SDCA single amp
  ASoC: Intel: sof_sdw: add missing quirk for Dell SKU 0A45

 sound/hda/intel-dsp-config.c                  |  22 +-
 sound/soc/intel/boards/Kconfig                |  14 +
 sound/soc/intel/boards/Makefile               |   2 +
 sound/soc/intel/boards/sof_es8336.c           | 569 ++++++++++++++++++
 sound/soc/intel/boards/sof_sdw.c              |  10 +
 .../intel/common/soc-acpi-intel-bxt-match.c   |   6 +
 .../intel/common/soc-acpi-intel-glk-match.c   |   7 +-
 .../intel/common/soc-acpi-intel-tgl-match.c   |  47 ++
 8 files changed, 674 insertions(+), 3 deletions(-)
 create mode 100644 sound/soc/intel/boards/sof_es8336.c

--
2.25.1

Merge series "ASoC: SOF: Intel: add flags to turn on SSP clocks early" from Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>:

With the chip shortage, some GeminiLake Intel-based designs were
respun and now rely on codecs that need the SSP bit clock turned on in
the hw_params stage, not the trigger stage. This patchset mirrors the
flags added in the SOF DAI_CONFIG IPC, and sets the flags when this
capability is indicated as necessary in the topology files where the
SSP configuration is stored.

We initially considered a more generic solution with an on-demand SSP
clock activation using the common clock framework. This would be a
more elegant solution indeed, but it would have required more
intrusive changes that would conflict with the SOF multi-client
support (in-development), and more backport hassles on product
branches. The on-demand activation of clocks is still a desired
feature that will be enabled at a later point.

Bard Liao (1):
  ASoC: SOF: dai-intel: add SOF_DAI_INTEL_SSP_CLKCTRL_MCLK/BCLK_ES bits

Pierre-Louis Bossart (4):
  ASoC: SOF: dai: mirror group_id definition added in firmware
  ASoC: SOF: dai: include new flags for DAI_CONFIG
  ASoC: SOF: Intel: hda: add new flags for DAI_CONFIG
  ASoC: SOF: Intel: hda-dai: improve SSP DAI handling for dynamic
    pipelines

 include/sound/sof/dai-intel.h |  4 ++
 include/sound/sof/dai.h       | 10 ++++-
 sound/soc/sof/intel/hda-dai.c | 82 ++++++++++++++++++++++++++++++++++-
 sound/soc/sof/intel/hda.c     |  6 +++
 sound/soc/sof/sof-audio.c     |  4 ++
 5 files changed, 103 insertions(+), 3 deletions(-)

--
2.25.1

Merge series "ASoC: SOF: topology: minor updates" from Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>:

Two patches to visualize settings and override dynamic pipelines + one
correction for errors on connections.

Pierre-Louis Bossart (2):
  ASoC: SOF: topology: show clks_control value in dynamic debug
  ASoC: SOF: topology: allow for dynamic pipelines override for debug

Ranjani Sridharan (1):
  ASoC: SOF: topology: return error if sof_connect_dai_widget() fails

 sound/soc/sof/sof-priv.h |  6 ++++++
 sound/soc/sof/topology.c | 24 +++++++++++++++---------
 2 files changed, 21 insertions(+), 9 deletions(-)

--
2.25.1

ASoC: SOF: OF: Add fw_path and tplg_path parameters

This allows specifying an alternate path for SOF firmware or
SOF topology.

This is particularly useful for i.MX when running Linux vs Android.

Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/20211005071949.1277613-1-daniel.baluta@oss.nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: amd: acp-rt5645: Constify static snd_soc_ops

The struct cz_aif1_ops is only assigned to the ops field in the
snd_soc_dai_link struct which is a pointer to const struct snd_soc_ops.
Make it const to allow the compiler to put it in read-only memory.

Signed-off-by: Rikard Falkeborn <rikard.falkeborn@gmail.com>
Link: https://lore.kernel.org/r/20211004224514.8783-1-rikard.falkeborn@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: topology: return error if sof_connect_dai_widget() fails

Return the error if sof_connect_dai_widget() fails to
abort topology loading and prevent card registration.

Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@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://lore.kernel.org/r/20211004212729.199550-4-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: topology: allow for dynamic pipelines override for debug

For debug and community support, it's useful to expose a kernel
parameter to prevent the use of dynamic pipelines exposed in a
topology file, or conversely to force an existing topology to use
dynamic pipelines.

Add an override bit and an enable bit which is valid only when the
override is set.

For products, the intent is that the topology file defines the
behavior, these two bits are only intended for diagnosis and
performance checks.

Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Bard Liao <bard.liao@intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/20211004212729.199550-3-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: topology: show clks_control value in dynamic debug

We log most of the SSP configurations except the clks_control. This
will be used to enable bclk/mclk early start so it's useful to show
the information to the user.

Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Bard Liao <bard.liao@intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20211004212729.199550-2-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda-dai: improve SSP DAI handling for dynamic pipelines

In order to keep the widget use_count balanced, make sure the DAI
widgets are allocated once in hw_params and released in hw_free. A
'setup' status flag is used to deal with cases where the .hw_params
callback is invoked multiple times, and likewise with cases where
hw_free is invoked without hw_params being called first (which can
happen if the FE hw_params fails).

In addition, this patch frees the widgets in the suspend transition,
and reallocates them in the .prepare callback. The 'setup' flag helps
in this case differentiate between resume (setup needed) and
xruns (setup not needed).

This balanced operation was not needed previously but will be required
when SOF dynamic pipelines are enabled.

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>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Bard Liao <bard.liao@intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Link: https://lore.kernel.org/r/20211004171430.103674-6-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: dai-intel: add SOF_DAI_INTEL_SSP_CLKCTRL_MCLK/BCLK_ES bits

Add two clks_control bits. MCLK and/or BCLK will start during hw_params
and stop during hw_free if the corresponding bit is set.

While the kernel does not do anything with these bitfields, this is
also tagged as part of the ABI 3.19 changes.

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: Bard Liao <bard.liao@intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Brent Lu <brent.lu@intel.com>
Link: https://lore.kernel.org/r/20211004171430.103674-5-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Intel: hda: add new flags for DAI_CONFIG

The DAI_CONFIG is used for both hw_params and hw_free. Use flags to
specify what stage the configuration applies to.

the DAI_CONFIG IPC may be sent also during the widget setup so each
flag is cleared after the IPC to restore the state.

Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Bard Liao <bard.liao@intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Brent Lu <brent.lu@intel.com>
Link: https://lore.kernel.org/r/20211004171430.103674-4-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: dai: include new flags for DAI_CONFIG

Mirror changes done in SOF tree. The changes do not rely on
BIT/GENMASK on purpose to keep the structure and flags common with the
firmware tree.

The DAI_CONFIG IPC is currently used in multiple ways. It is sent to
the DSP firmware when enabling static or dynamic pipelines, in
hw_params or prepare callbacks for Intel SSP, HDaudio and ALH, on
trigger_stop and hw_free.

This IPC has been abused a bit in the past, i.e. the values used for
some of the DAI-specific fields are used to either allocate or free
resources. Two typical examples are Intel HDaudio and SoundWire/ALH
DAIs, where using a zero DMA channel number or stream tag signals to
the firmware the DMA channels or tags allocated earlier can be freed.

Rather than add a new IPC for 'hw_params' and 'hw_free', this patch
suggests supporting a 2-bit value conveying the 'stage' information in
an existing IPC structure. Only 3 possible values are used.

The mapping between HW_PARAMS and HW_FREE flags and ALSA definitions
is not strictly 1:1, e.g. in some cases the HW_PARAM flag might be set
during the .prepare callback, while the HW_FREE might be sent during the
ALSA .trigger for stop/suspend.

The semantics of the flags is to reserve and start/stop all needed
resources, typically hardware related such as DMAs or clocks, when the
HW_PARAMS is set, while the HW_FREE flag allows the firmware to
release the resources allocated. The data transfers are still
controlled within the firmware through the propagation of the trigger
command.

The driver can then pass information that the DAI_CONFIG was invoked
in e.g. a pipeline/DAI setup, hw_params or hw_free stage without
having to use a special DAI-specific encoding. Unfortunately we can't
remove old encodings due to backwards-compatibility requirements but
for new cases, such as the SSP in follow-up patches, we can make the
IPC less cryptic.

This change is tagged as ABI 3.19 and is completely backwards compatible.

Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Bard Liao <bard.liao@intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Brent Lu <brent.lu@intel.com>
Link: https://lore.kernel.org/r/20211004171430.103674-3-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: dai: mirror group_id definition added in firmware

This was added in ABI 3.17 but never added to the kernel tree. The
group_id is not currently used but this patch is required before
additional changes.

Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Bard Liao <bard.liao@intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Brent Lu <brent.lu@intel.com>
Link: https://lore.kernel.org/r/20211004171430.103674-2-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: sof_sdw: add missing quirk for Dell SKU 0A45

This device is based on SDCA codecs but with a single amplifier
instead of two.

BugLink: https://github.com/thesofproject/linux/issues/3161
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Bard Liao <bard.liao@intel.com>
Link: https://lore.kernel.org/r/20211004213512.220836-6-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: soc-acpi: add missing quirk for TGL SDCA single amp

We don't have a configuration for a single amp on link1.

BugLink: https://github.com/thesofproject/linux/issues/3161
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Reviewed-by: Bard Liao <bard.liao@intel.com>
Link: https://lore.kernel.org/r/20211004213512.220836-5-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: add machine driver for SOF+ES8336

Add machine driver to support APL/GLK/TGL platforms.
The TGL platform supports DMIC, APL and GLK do not.

Co-developed-by: Huajun Li <huajun.li@intel.com>
Signed-off-by: Huajun Li <huajun.li@intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20211004213512.220836-4-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ALSA: intel-dsp-config: add quirk for APL/GLK/TGL devices based on ES8336 codec

These devices are based on an I2C/I2S device, we need to force the use
of the SOF driver otherwise the legacy HDaudio driver will be loaded -
only HDMI will be supported.

Co-developed-by: Huajun Li <huajun.li@intel.com>
Signed-off-by: Huajun Li <huajun.li@intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20211004213512.220836-3-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: soc-acpi: apl/glk/tgl: add entry for devices based on ES8336 codec

A number of devices, such as the "Chuwi HI10x" and "UNIQCELL Q15.6",
are based on APL/GLK with an I2C/I2S ES8336 codec.

Add table to find topology and firmware files.

Co-developed-by: Huajun Li <huajun.li@intel.com>
Signed-off-by: Huajun Li <huajun.li@intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20211004213512.220836-2-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: dt-bindings: uniphier: Add description of each port number

To eliminate FIXME, add descriptions of what each port number is
associated with.

Cc: Rob Herring <robh@kernel.org>
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Link: https://lore.kernel.org/r/1633313546-23258-1-git-send-email-hayashi.kunihiko@socionext.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: Intel: cht_bsw_nau8824: Set card.components string

Set the card.components string using the new nau8824_components() helper
which returns a components string based on the DMI quirks inside the
nau8824 codec driver.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20211002211459.110124-4-hdegoede@redhat.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: nau8824: Add a nau8824_components() helper

Some devices using the NAU8824 have only one speaker. To still have things
working properly this requires the left + right channels to both be mixed
to the left speaker output.

This mixer setup is done by userspace based on UCM profiles. But this
requires userspace to know that there is a mono-speaker. Add a helper
function (for the machine driver) to get a components string providing
this info.

This is done inside the codec driver because the codec driver already
has a DMI quirk table.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20211002211459.110124-3-hdegoede@redhat.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: nau8824: Add DMI quirk mechanism for active-high jack-detect

Add a quirk mechanism to allow specifying that active-high jack-detection
should be used on platforms where this info is not available in devicetree.

And add an entry for the Cyberbook T116 tablet to the DMI table, so that
jack-detection will work properly on this tablet.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20211002211459.110124-2-hdegoede@redhat.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: nau8824: Fix NAU8824_JACK_LOGIC define

The NAU8824_JACK_LOGIC define was wrong, for active high jack-detect
to work bit 1 needs to be set, rather then bit 0.

The correct bit was found in the Android kernel source dump for
a Cyberbook T116 tablet; and this was also tested on that same tablet.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20211002211459.110124-1-hdegoede@redhat.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: rt5651: Use IRQF_NO_AUTOEN when requesting the IRQ

Use the new IRQF_NO_AUTOEN flag when requesting the IRQ, rather then
disabling it immediately after requesting it.

This fixes a possible race where the IRQ might trigger between requesting
and disabling it; and this also leads to a small code cleanup.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20211003132255.31743-2-hdegoede@redhat.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: es8316: Use IRQF_NO_AUTOEN when requesting the IRQ

Use the new IRQF_NO_AUTOEN flag when requesting the IRQ, rather then
disabling it immediately after requesting it.

This fixes a possible race where the IRQ might trigger between requesting
and disabling it; and this also leads to a small code cleanup.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20211003132255.31743-1-hdegoede@redhat.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: nau8821: new driver

The driver is for codec NAU88L21 of Nuvoton Technology Corporation.
The NAU88L21 is an ultra-low power high performance audio codec that
supports both analog and digital audio functions.

Signed-off-by: Seven Lee <wtli@nuvoton.com>
Link: https://lore.kernel.org/r/20211001103108.3297848-1-wtli@nuvoton.com
Signed-off-by: Mark Brown <broonie@kernel.org>

Merge series "Add support for on demand pipeline setup/destroy" from Peter Ujfalusi <peter.ujfalusi@linux.intel.com>:

Hi,

The previous, v2 of this series was sent by Daniel Baluta:
https://lore.kernel.org/alsa-devel/20210917143659.401102-1-daniel.baluta@oss.nxp.com/

We have agreed that it might be better that someone from Intel is going to take it
from here as we already have the infrastructure up to test and verify the
dynamic pipelines support.

Changes since v2 (sent by Daniel Baluta):
- patch 10: Fix NULL point dereference in hda_dai_update_config()
- I have kept Daniel's SoB for the series.

Changes since v1:
- Signed-off-by tag added by Daniel

This series implements initial support for dynamic pipelines to setup/teardown
pipeline as needed when a PCM is open/closed.

Initially dynamic pipelines are only supported with single core setup which will
be expanded with a follow-up series.

Review with SOF community at
https://github.com/thesofproject/linux/pull/2794

The feature has been merged on 1st of April to sof-dev, all issues found since
has been fixed and squashed to this upstream series.

Regards,
Peter
---
Ranjani Sridharan (12):
  ASoC: topology: change the complete op in snd_soc_tplg_ops to return
    int
  ASoC: SOF: control: Add access field in struct snd_sof_control
  ASoC: SOF: topology: Add new token for dynamic pipeline
  ASoC: SOF: sof-audio: add helpers for widgets, kcontrols and dai
    config set up
  AsoC: dapm: export a couple of functions
  ASoC: SOF: Add new fields to snd_sof_route
  ASoC: SOF: restore kcontrols for widget during set up
  ASoC: SOF: Don't set up widgets during topology parsing
  ASoC: SOF: Introduce widget use_count
  ASoC: SOF: Intel: hda: make sure DAI widget is set up before IPC
  ASoC: SOF: Add support for dynamic pipelines
  ASoC: SOF: topology: Add kernel parameter for topology verification

 include/sound/soc-dpcm.h               |   1 +
 include/sound/soc-topology.h           |   2 +-
 include/uapi/sound/sof/tokens.h        |   1 +
 sound/soc/intel/skylake/skl-topology.c |   6 +-
 sound/soc/soc-dapm.c                   |   2 +
 sound/soc/soc-pcm.c                    |   4 +-
 sound/soc/soc-topology.c               |  10 +-
 sound/soc/sof/intel/hda-dai.c          | 174 +++---
 sound/soc/sof/intel/hda.c              | 177 ++++--
 sound/soc/sof/intel/hda.h              |   5 +
 sound/soc/sof/ipc.c                    |  22 +
 sound/soc/sof/pcm.c                    |  58 +-
 sound/soc/sof/pm.c                     |   4 +-
 sound/soc/sof/sof-audio.c              | 709 +++++++++++++++++++------
 sound/soc/sof/sof-audio.h              |  32 +-
 sound/soc/sof/sof-priv.h               |   1 +
 sound/soc/sof/topology.c               | 362 +++++--------
 17 files changed, 1032 insertions(+), 538 deletions(-)

--
2.33.0

ASoC: fsl_spdif: implement bypass mode from in to out

Implement SPDIF bypass mode. It implies internal SoC
routing of SPDIF input signal to SPDIF output signal. The
test bed requires two boards: B1 configured in bypass mode,
and B2 to feed B1 SPDIF RX port and read B1 SPDIF TX port:
   B2 TX -> B1 RX,
   B2 RX <- B1 TX.
The test procedure:
 a) Boot both boards
 b) B2: start "arecord <spdifcard> -r 48kHz | aplay <local DAC>"
 c) B2: start "aplay <spdifcard> -r 48kHz <2ch 48kHz audio file>"
 d) B1: enable bypass mode:
amixer -cimxspdif cset numid=8,iface=PCM,name='Bypass Mode' on
 e) B2: check DAC audio, make sure the same sample rate is used at
steps b) and c), in example above the rate is 48kHz.
 f) B1: try to run "aplay" or "arecord" on imxspdif card while in
bypass mode - both must fail until bypass mode is disabled
 g) B1: disable bypass mode:
amixer -cimxspdif cset numid=8,iface=PCM,name='Bypass Mode' off
 h) B1: check the usual playback and capture on imxspdif card.
During this test try to set bypass mode - must not be allowed
while playback or capture is running.

Signed-off-by: Viorel Suman <viorel.suman@nxp.com>
Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
Link: https://lore.kernel.org/r/1632649760-1651-1-git-send-email-shengjiu.wang@nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: topology: Add kernel parameter for topology verification

Add a kernel debug flag to enable a one-shot topology
verification for all pipelines including the dynamic
ones. If the debug flag is set, all the topology
component loading will be verified during the complete
callback.

Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Link: https://lore.kernel.org/r/20210927120517.20505-13-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>

ASoC: SOF: Add support for dynamic pipelines

Add support for dynamic pipelines by modifying the PCM
hw_params ioctl implementation to determine the widgets
required for a PCM stream by querying the list of
connected DAPM widgets. This list is saved as part of
snd_sof_pcm_stream struct and will be used to setup the widgets.

The sof_widget_list_setup/free routines setup and free connected
DAPM widgets when a PCM is opened/closed. These routines accept
a list of connected DAPM widgets as input and determine the SOF
widgets, their corresponding pipeline widgets and connections
between them that need to be setup before the PCM is triggered.

Please note that the dynamic pipeline feature will only be enabled
for those pipelines whose dynamic_pipeline_widget flag is set in
topologies. Add a new token called SOF_TKN_SCHED_DYNAMIC_PIPELINE
that when set in topology will be applied to the
dynamic_pipeline_widget flag of the pipeline widget.

Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Link: https://lore.kernel.org/r/20210927120517.20505-12-peter.ujfalusi@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>