Use SOF defined callbacks (snd_sof_prepare/snd_sof_complete)
in order to update internal SOF system suspend target.
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Reviewed-by: Paul Olaru <paul.olaru@nxp.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Link: https://lore.kernel.org/r/20200924151518.15841-3-daniel.baluta@oss.nxp.comSigned-off-by: Mark Brown <broonie@kernel.org>

SOF boots the DSP at probe and keeps it up all the time.
With this change, after booting if no one is using the DSP
the SOF core will turn off the DSP to save power.
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Reviewed-by: Paul Olaru <paul.olaru@nxp.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Link: https://lore.kernel.org/r/20200924151518.15841-2-daniel.baluta@oss.nxp.comSigned-off-by: Mark Brown <broonie@kernel.org>

Merge series "ASoC: Intel: sdw machine driver updates for 5.10" from Kai Vehmanen <kai.vehmanen@linux.intel.com>:

Series including fixes and improvements for Intel SoundWire
machine drivers.

Bard Liao (1):
  ASoC: Intel: add support for new SoundWire hardware layout on TGL

Pierre-Louis Bossart (4):
  ASoC: Intel: sof_sdw: remove ternary operator
  ASoC: Intel: add codec name prefix to ACPI machine description
  ASoC: Intel: sof_sdw: remove hard-coded codec_conf table
  ASoC: Intel: sof_sdw_rt700: add codec prefix

Rander Wang (1):
  ASOC: Intel: sof_sdw: restore playback functionality with max98373
    amps

 include/sound/soc-acpi.h                      |   2 +
 sound/soc/intel/boards/sof_sdw.c              | 170 +++++++++---------
 sound/soc/intel/boards/sof_sdw_common.h       |   3 +
 sound/soc/intel/boards/sof_sdw_max98373.c     |  36 +++-
 sound/soc/intel/boards/sof_sdw_rt700.c        |   6 +-
 .../intel/common/soc-acpi-intel-cml-match.c   |  10 ++
 .../intel/common/soc-acpi-intel-cnl-match.c   |   1 +
 .../intel/common/soc-acpi-intel-icl-match.c   |   6 +
 .../intel/common/soc-acpi-intel-tgl-match.c   |  67 +++++++
 9 files changed, 216 insertions(+), 85 deletions(-)

--
2.27.0

Remove unused variables in the private struct and the code as these
variables are initially set and then there is no additional code
utilizing these variables.
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200923132600.10652-6-dmurphy@ti.comSigned-off-by: Mark Brown <broonie@kernel.org>

Remove the code to support the asi-format binding property. The code
does nothing except read the property and set a variable. No additional
action is taken except to reset the variable. The property is supposed
to set the rising or falling RX edge detection of the SBCLK but this
edge detection is done by checking the DAI_FMT_INV_MASK.
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200923132600.10652-5-dmurphy@ti.comSigned-off-by: Mark Brown <broonie@kernel.org>

Set the regcache to cache data and mark cache as dirty when the device
is shutdown when suspend is called. When the device is woken up then
sync the cache and set to not caching the data.
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200923132600.10652-3-dmurphy@ti.comSigned-off-by: Mark Brown <broonie@kernel.org>

Add the hardware shutdown mechanism to shutdown and wake up the device
via a GPIO.
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200923132600.10652-2-dmurphy@ti.comSigned-off-by: Mark Brown <broonie@kernel.org>

Add the shutdown-gpios property to the yaml to define the GPIO that can
be used to place the device in shutdown mode or wake the device up.
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200923132600.10652-1-dmurphy@ti.comSigned-off-by: Mark Brown <broonie@kernel.org>

cppcheck reports the following warning:

sound/soc/intel/boards/hda_dsp_common.c:17:0: style: The function
'hda_dsp_hdmi_pcm_handle' is never used. [unusedFunction]

Fix by moving to static inside compilation block.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Jaska Uimonen <jaska.uimonen@intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200923072939.3100468-1-kai.vehmanen@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

The Max98373 amplifier provides I/V feedback information, which keeps
a DAPM path active even when there is no playback happening. This
prevents entry in low-power mode. Rather than adding new controls and
require UCM/user interaction, the method previously applied is to
enable/disable the Speaker pin during the dailink trigger operations.

Recent changes in the SoundWire stream management moved the stream
trigger to the dailink trigger. This change removed the Maxim-specific
pin handling and resulted in a regression. This patch restores
functionality by combining the SoundWire stream trigger with the pin
enable/disable.

Fixes: ae3a3918 ('ASoC: Intel: sof_sdw: add dailink .trigger callback')
Fixes: 06998d49 ('ASoC: Intel: sof_sdw: add dailink .prepare and .hw_free callback')
Signed-off-by: Rander Wang <rander.wang@intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Keyon Jie <yang.jie@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200923080514.3242858-2-kai.vehmanen@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

The creativity of hardware folks is endless, with a complete
permutation of rt711 (was link0 now link1), rt1308 (was link1 now
link2) and rt715 (was link3 now link0).

Someday we will get all this information from platform firmware, for
now let's add the mapping table.
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 <yung-chuan.liao@linux.intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200923080514.3242858-7-kai.vehmanen@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

Somehow for this codec we never used any prefix for the controls,
likely because the test platform has a single SoundWire device.

Follow the convention and use the codec prefix across the board to
avoid possible conflicts.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200923080514.3242858-6-kai.vehmanen@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

Now that the ACPI machine params provide all the information needed,
allocate the card codec_conf dynamically and set .dlc and
.prefix_name.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200923080514.3242858-5-kai.vehmanen@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

The current SOF machine driver adds a name prefix for each codec,
mainly to differentiate ALSA controls for left and right amplifiers.

This is a good idea, but the machine driver duplicates some of the
information that already exists in ACPI descriptors, so add those
prefixes there. Follow-up patches will make use of the information
encoded in these tables and remove duplication.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200923080514.3242858-4-kai.vehmanen@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

cppcheck reports the following warning:

sound/soc/intel/boards/sof_sdw.c:866:46: style: Clarify calculation
precedence for '&' and '?'. [clarifyCalculation]
 hdmi_num = sof_sdw_quirk & SOF_SDW_TGL_HDMI ?
                                             ^

There's no reason to use the ternary operator here, we might as well
use a regular if-else construct.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Jaska Uimonen <jaska.uimonen@intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200923080514.3242858-3-kai.vehmanen@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

When SND_SOC_CROS_EC_CODEC is enabled and CRYPTO is disabled, it results
in the following Kbuild warning:

WARNING: unmet direct dependencies detected for CRYPTO_LIB_SHA256
  Depends on [n]: CRYPTO [=n]
  Selected by [y]:
  - SND_SOC_CROS_EC_CODEC [=y] && SOUND [=y] && !UML && SND [=y] && SND_SOC [=y] && CROS_EC [=y]

The reason is that SND_SOC_CROS_EC_CODEC selects CRYPTO_LIB_SHA256 without
depending on or selecting CRYPTO while CRYPTO_LIB_SHA256 is subordinate to
CRYPTO.

Honor the kconfig menu hierarchy to remove kconfig dependency warnings.

Fixes: 93fa0af4 ("ASoC: cros_ec_codec: switch to library API for SHA-256")
Signed-off-by: Necip Fazil Yildiran <fazilyildiran@gmail.com>
Link: https://lore.kernel.org/r/20200917141803.92889-1-fazilyildiran@gmail.comSigned-off-by: Mark Brown <broonie@kernel.org>

Remove duplicate code for programming the I/V sense the call to update
the register was duplicated in commit 09ed395b ("ASoC: tas2562:
Add voltage sense slot configuration").

Fixes: 09ed395b ("ASoC: tas2562: Add voltage sense slot configuration")
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200922142411.10364-1-dmurphy@ti.comSigned-off-by: Mark Brown <broonie@kernel.org>

Use set_jack ops to set jack so machine drivers do not need to include
hdmi-codec.h explicitly.
Signed-off-by: Cheng-Yi Chiang <cychiang@chromium.org>
Reviewed-by: Tzung-Bi Shih <tzungbi@google.com>
Link: https://lore.kernel.org/r/20200922062316.1172935-1-cychiang@chromium.orgSigned-off-by: Mark Brown <broonie@kernel.org>

Merge series "ASoC: SOF: fix kcontrol size checks" from Kai Vehmanen <kai.vehmanen@linux.intel.com>:

Series that fixes checks for 'size' in kcontrol get/put ext_bytes methods
for SOF. The gaps in these checks were discovered via cppcheck warnings
on unused variable values.

Pierre-Louis Bossart (5):
  ASoC: SOF: control: fix size checks for ext_bytes control .get()
  ASoC: SOF: control: fix size checks for volatile ext_bytes control
    .get()
  ASoC: SOF: control: add size checks for ext_bytes control .put()
  ASoC: SOF: control: remove const in sizeof()
  ASoC: SOF: topology: remove const in sizeof()

 sound/soc/sof/control.c  | 53 +++++++++++++++++++++++++++++++---------
 sound/soc/sof/topology.c |  2 +-
 2 files changed, 43 insertions(+), 12 deletions(-)

--
2.27.0

To provide backward compatibility to older systems, the SOF HDA driver
allows user to specify which HDMI codec driver to use at runtime via
kernel parameter. This mechanism has a subtle flaw in that it assumes
the codec drivers not to be loaded when the SOF PCI driver is loaded.

The problem is rooted in use of the hdev->type field.
snd_hdac_ext_bus_device_init() initializes this field to HDA_DEV_ASOC.
This signals the HDA core that ASoC drivers should be considered in
driver matching (hda_bus_match()). The SOF and SST drivers continue by
overriding this field to HDA_DEV_LEGACY and proceeding to load driver
modules with request_module(). Correct drivers will get loaded and
attached.

If however the codec drivers are already loaded when
snd_hdac_ext_bus_device_init() is called, the matching will not work as
expected as device type is still set to HDA_DEV_ASOC. Specifically if
hdac-hdmi is attached when machine driver is configured to use hdac-hda,
this leads to out-of-bounds memory access in
hda_dsp_hdmi_build_controls().

Fix the issue by adding codec type as a parameter to
snd_hdac_ext_bus_device_init() and ensuring type is set correctly from
the start.

Fixes: 139c7feb ("ASoC: SOF: Intel: add support for snd-hda-codec-hdmi")
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20200921100841.2882662-1-kai.vehmanen@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

Refactor the tas2770_set_samplerate to simplify the code and access the
I2C bus only once per rate request. The ramp rate and sample rate bits
are contained in the same register so a single call to the
snd_soc_update_bits function is all that is needed
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200918190548.12598-9-dmurphy@ti.comSigned-off-by: Mark Brown <broonie@kernel.org>

Fix up the spacing for argument alignment and add new lines to separate
code. Eliminate unneccessary goto statements when the error code could
just be returned.
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200918190548.12598-8-dmurphy@ti.comSigned-off-by: Mark Brown <broonie@kernel.org>

Update the hardcoded masks with the GENMASK macro. Also update some of
the hardcoded bits with the BIT macro
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200918190548.12598-6-dmurphy@ti.comSigned-off-by: Mark Brown <broonie@kernel.org>

Fix the unbalanced call to the pm_runtime_disable when removing the
module.  pm_runtime_enable is not called nor is the pm_runtime setup in
the code.  Remove the i2c_remove function and the pm_runtime_disable.

Fixes: 1a476abc ("tas2770: add tas2770 smart PA kernel driver")
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200918190548.12598-5-dmurphy@ti.comSigned-off-by: Mark Brown <broonie@kernel.org>

The I2C addresses listed in the yaml are not correct. The addresses can
range from 0x41 through 0x48 based on register configurations. Fix the
example and the description.

Fixes: 4b7151da ("dt-bindings: ASoC: Add tas2770 smart PA dt bindings")
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200918190548.12598-3-dmurphy@ti.comSigned-off-by: Mark Brown <broonie@kernel.org>

Add the TAS2110 amplifier to the TAS2562 driver. The TAS2110 is register
and bitmap compatible. The chips differ in that the TAS2110 does not
have the I/V Sense feedback path. Since these features do not exist the
device needs to be registered without these controls.
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200921153820.18357-2-dmurphy@ti.comSigned-off-by: Mark Brown <broonie@kernel.org>

Add the TAS2110 amplifier compatible.
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200921153820.18357-1-dmurphy@ti.comSigned-off-by: Mark Brown <broonie@kernel.org>

Make sure the TLV header and size are consistent before copying from
userspace.

Fixes: c3078f53 ('ASoC: SOF: Add Sound Open Firmware KControl support')
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200921110814.2910477-4-kai.vehmanen@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

Mirror addition of checks for regular ext_bytes controls.

Fixes: 783560d0 ('ASoC: SOF: Implement snd_sof_bytes_ext_volatile_get kcontrol IO')
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200921110814.2910477-3-kai.vehmanen@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

cppcheck complains twice:

sound/soc/sof/control.c:436:2: style: Assignment of function parameter
has no effect outside the function. [uselessAssignmentArg]
 size -= sizeof(const struct snd_ctl_tlv);
 ^

sound/soc/sof/control.c:436:7: style: Variable 'size' is assigned a
value that is never used. [unreadVariable]
 size -= sizeof(const struct snd_ctl_tlv);

Somehow we dropped the checks for the size argument when upstreaming
the code, somewhere between v5 and v6.

Re-add a size check to avoid providing userspace with more data that
it asked for.

Also fix all error codes, we should return -ENOSPC instead of -EINVAL.

Fixes: c3078f53 ('ASoC: SOF: Add Sound Open Firmware KControl support')
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200921110814.2910477-2-kai.vehmanen@linux.intel.comSigned-off-by: Mark Brown <broonie@kernel.org>

At power-up the analog circuits may take up to one full second before
being charged with the default configuration. Using the analog blocks
before they are ready generates a *very* crappy sound.

Enable the fast charge feature, which will require a bit more power
than normal charge but will definitely speed up the starting operation
by shrinking this delay to up to 40 ms.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/r/20200911173140.29984-4-miquel.raynal@bootlin.comSigned-off-by: Mark Brown <broonie@kernel.org>

Current code expects a single channel to be always used. Fix this
situation by forwarding the number of channels used. Then fix the
derivation of the bdiv clock rate.

Fixes: 96c3bb00 ("ASoC: tlv320aic32x4: Dynamically Determine Clocking")
Suggested-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/r/20200911173140.29984-3-miquel.raynal@bootlin.comSigned-off-by: Mark Brown <broonie@kernel.org>

As indicated in the datasheet, a 10ms delay must be observed after
programming the divisors.

The lack of delay prevents the codec to work properly and the playback
appears extremely slow and totally un-audible on a custom sama5 based
board.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/r/20200911173140.29984-2-miquel.raynal@bootlin.comSigned-off-by: Mark Brown <broonie@kernel.org>

To avoid the IO error, we need to cancel the delayed work and wait for it to finish.
Signed-off-by: Shuming Fan <shumingf@realtek.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20200921094308.31921-1-shumingf@realtek.comSigned-off-by: Mark Brown <broonie@kernel.org>

snd_soc_update_bits returns a 1 when the bit was successfully updated,
returns a 0 is no update was needed and a negative if the call failed.
The code is currently failing the case of a successful update by just
checking for a non-zero number. Modify these checks and return the error
code only if there is a negative.

Fixes: 1a476abc ("tas2770: add tas2770 smart PA kernel driver")
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200918190548.12598-7-dmurphy@ti.comSigned-off-by: Mark Brown <broonie@kernel.org>

The devicetree binding indicates that the ti,asi-format, ti,imon-slot-no
and ti,vmon-slot-no are not required but the driver requires them or it
fails to probe. Honor the binding and allow these entries to be optional
and set the corresponding values to the default values for each as defined
in the data sheet.

Fixes: 1a476abc ("tas2770: add tas2770 smart PA kernel driver")
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200918190548.12598-4-dmurphy@ti.comSigned-off-by: Mark Brown <broonie@kernel.org>

Add the BIAS_STANDBY and BIAS_PREPARE to the set_bias_level or else the
driver will return -EINVAL which is not correct as they are valid
states.

Fixes: 1a476abc ("tas2770: add tas2770 smart PA kernel driver")
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200918190548.12598-2-dmurphy@ti.comSigned-off-by: Mark Brown <broonie@kernel.org>

tas2770_reset is called during i2c probe. The reset calls the
snd_soc_component_write which depends on the tas2770->component being
available. The component pointer is not set until codec_probe so move
the reset to the codec_probe after the pointer is set.

Fixes: 1a476abc ("tas2770: add tas2770 smart PA kernel driver")
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200918190548.12598-1-dmurphy@ti.comSigned-off-by: Mark Brown <broonie@kernel.org>

To avoid the IO error, we need to cancel the delayed work and wait for it to finish.
Signed-off-by: Shuming Fan <shumingf@realtek.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20200921094244.31869-1-shumingf@realtek.comSigned-off-by: Mark Brown <broonie@kernel.org>