The changes in "ASoC: pcm: free path list before exiting from error
conditions" actually introduced both double frees (in case where the
path list was allocated but empty) and frees of unallocated memory (in
cases where the error being handled was -ENOMEM.  Drop the commit for
now.

Fixes: e4ad1acc (ASoC: pcm: free path list before exiting from error conditions)
Reported-by: NBen Hutchings <ben@decadent.org.uk>
Signed-off-by: NMark Brown <broonie@linaro.org>
Cc: stable@vger.kernel.org

Fixes: 115f3f8 ("ASoC: mfld_machine: Convert to table based DAPM and control setup")
Reported-by: Nkbuild test robot <fengguang.wu@intel.com>
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NMark Brown <broonie@linaro.org>

Support for loading the Renesas R-Car sound driver via DeviceTree.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

Use table based setup to register the controls and DAPM widgets and routes.
This on one hand makes the code a bit cleaner and on the other hand
the board level DAPM elements get registered in the card's DAPM context rather
than in the CODEC's DAPM context.

The mfld_machine driver is a bit special in that it directly writes to one of
the CODEC registers from one of the control handlers. Previous to this patch it
was able to get a pointer to the CODEC from the control, since the control was
registered with the CODEC. This won't be possible anymore once the control is
registered with the card. Since there are already global variables in the driver
accessed in the same function the patch adds a global variable that holds a
pointer to the CODEC and uses that.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NMark Brown <broonie@linaro.org>

This patch adds support for the Cirrus Logic CS42448/CS42888 Audio CODEC that
has six/four 24-bit AD and eight 24-bit DA converters.

[ CS42448/CS42888 supports both I2C and SPI control ports. As initial patch,
  this patch only adds the support for I2C. ]
Signed-off-by: NNicolin Chen <Guangyu.Chen@freescale.com>
Acked-by: NBrian Austin <brian.austin@cirrus.com>
Acked-by: NPaul Handrigan <Paul.Handrigan@cirrus.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

When using an rbtree cache, there can be allocations the first time a
register is accessed.  This can cause an attempt to schedule while
atomic in the case that the regmap is using a spinlock.  This could be
fixed by either initializing all the registers or using a flat cache.
The register maps for tegra30_ahub and tegra30_i2s are dense and don't
save much from using a tree so convert them to flat.

Tegra30 changes tested on Norrin, Tegra20 changes compile.
Signed-off-by: NDylan Reid <dgreid@chromium.org>
Tested-by: NStephen Warren <swarren@nvidia.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

The shift values for the ADC,PCM, and Analog kcontrols were wrong causing wrong values for the SOC_DOUBLE_R_SX_TLV macros
Fixed the TLV for aout_tlv to show -102dB correctly

Fixes: 1d99f243 (ASoC: core: Rework SOC_DOUBLE_R_SX_TLV add SOC_SINGLE_SX_TLV)
Reported-by: NThomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: NBrian Austin <brian.austin@cirrus.com>
Signed-off-by: NMark Brown <broonie@linaro.org>
Cc: stable@vger.kernel.org

Since all regmap CODECs should be (and are) using the more advance regmap
cache infrastructure remove the code which supports that and just proxy
I/O straight through to regmap.
Signed-off-by: NMark Brown <broonie@linaro.org>

The current simple-card driver separates the daimft for cpu_dai and codec_dai.
So we might get different values for them (0x4003 and 0x1003 for example):

asoc-simple-card sound-cs42888.12: cpu : 2024000.esai / 4003 / 132000000
asoc-simple-card sound-cs42888.12: codec : cs42888 / 1003 / 24576000
asoc-simple-card sound-cs42888.12: cs42888 <-> 2024000.esai mapping ok

This is not allowed at all as we need to keep the DAIFMT settings identical
for both the ends of the link.

Thus this patch fixes it by overwriting the cpu_dai->fmt with codec_dai->fmt
since we defined the DAIFMT_MASTER basing on CODEC at the first place while
the other bits are same.
Signed-off-by: NNicolin Chen <Guangyu.Chen@freescale.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

The mask bits values were wrong for the SOC_VALUE_ENUM_SINGLE for the PCM/ADC Swap controls
Reported-by: NTakashi Iwai <tiwai@suse.de>
Signed-off-by: NBrian Austin <brian.austin@cirrus.com>
Signed-off-by: NMark Brown <broonie@linaro.org>
Cc: stable@vger.kernel.org

The mask bits values were wrong for the SOC_VALUE_ENUM_SINGLE for the mono mix controls.
Reported-by: NTakashi Iwai <tiwai@suse.de>
Signed-off-by: NBrian Austin <brian.austin@cirrus.com>
Signed-off-by: NMark Brown <broonie@linaro.org>
Cc: stable@vger.kernel.org

The Armada 370 DB board not only has analog audio input/output, but
also S/PDIF input/output. This commit adds support for S/PDIF in the
ASoC machine driver of the Armada 370 DB platform, and adjusts the
Device Tree bindings documentation accordingly.
Signed-off-by: NThomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

With CODEC to CODEC links rtd->codec does not necessarily point to the driver's
CODEC. CODEC drivers should always use dai->codec and never even look at the PCM
runtime.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Acked-by: NCharles Keepax <ckeepax@opensource.wolfsonmicro.com>
Acked-by: NJyri Sarha <jsarha@ti.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

The kfree call is not necessary, but we need to call snd_soc_unregister_codec()
in remove().
Signed-off-by: NAxel Lin <axel.lin@ingics.com>
Acked-by: NJyri Sarha <jsarha@ti.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

There was a overlap between the snd_soc_codec_set_cache_io() cleanup and the
addition of the sirf-audio-codec resulting in the sirf-audio-codec driver still
using the old signature of snd_soc_codec_set_cache_io(), which will cause a
compile error. Since the core is able to automatically setup IO for this driver
we can just remove both the snd_soc_set_cache_io() call and the control_data
assignment.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NMark Brown <broonie@linaro.org>

The global DAI format is used only in the function
asoc_simple_card_parse_of(). So, move it from the private data
to the stack.
Signed-off-by: NJean-Francois Moine <moinejf@free.fr>
Reviewed-by: NXiubo Li <Li.Xiubo@freescale.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

When we boot with non-DT mode the damengine will need the channel number and
a filter function in order to get the channel.
The filter_data is filled in the DAI driver while the filter_function will
be provided by the edma-pcm driver.
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

Platform driver glue for SoC using eDMA3 to use dmaengine PCM.
The maximum number of periods need to be limited to 19 since the
edma dmaengine driver limits the paRAM slot use for audio at
in cyclic mode.
Signed-off-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Reviewed-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NMark Brown <broonie@linaro.org>

Regulator code keep count of enables and disables. Double disable
causes an ugly warning.
Signed-off-by: NJyri Sarha <jsarha@ti.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

The reference count of some device nodes is not correctly reset
at end of card probe.
Signed-off-by: NJean-Francois Moine <moinejf@free.fr>
Signed-off-by: NMark Brown <broonie@linaro.org>

SND_SOC_DAIFMT_CBx_CFx means "codec" side master/slave mode.
Then, rcar will be master mode if it was SND_SOC_DAIFMT_CBS_CFS.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

SND_SOC_DAIFMT_CBx_CFx means "codec" side master/slave mode.
Then, FSI will be master mode if it was SND_SOC_DAIFMT_CBS_CFS.

This patch fixup platform settings too.
Then, it tidyups SND_SOC_DAIFMT_INV settings.
Acked-by: NSimon Horman <horms+renesas@verge.net.au>
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

The driver assigns a value to the control_data field of the driver's state
struct, but never reads it again. Which means it is unused and can be removed.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NMark Brown <broonie@linaro.org>

Machine level DAPM widgets and routes should be registered in the card's DAPM
context, rather than in the CODEC's context.

While we are at it also drop the snd_soc_dapm_enable_pin() calls, since pins are
enabled by default and also turn the snd_soc_dapm_disable_pin() calls into
snd_soc_dapm_nc_pin() calls for unconnected pins.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Acked-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

Machine level DAPM widgets and routes should be registered in the card's DAPM
context, rather than in the CODEC's context.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Acked-by: NPeter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

Signed-off-by: NMark Brown <broonie@linaro.org>

Remove snd_soc_codec_set_cache_io(codec, 8, 8, SND_SOC_REGMAP) call
and codec->control_data = aic31xx->regmap assignment since that
already done by core.
Signed-off-by: NJyri Sarha <jsarha@ti.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

Fix "warning: unused variable 'aic31xx'" from function 'aic31xx_clk_off'.
Signed-off-by: NJyri Sarha <jsarha@ti.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

This commit adds a bare bones driver support for TLV320AIC31XX family
audio codecs. The driver adds basic stereo playback trough headphone
and speaker outputs and mono capture trough microphone inputs.

The driver is currently missing support at least for mini DSP features
and jack detection. I have tested the driver only on TLV320AIC3111,
but based on the data sheets TLV320AIC3100, TLV320AIC3110, and
TLV320AIC3120 should work Ok too.

The base for the implementation was taken from:
git@gitorious.org:ti-codecs/ti-codecs.git ajitk/topics/k3.10.1-aic31xx
-branch at commit 77504eba0294764e9e63b4a0c696b44db187cd13.
Signed-off-by: NJyri Sarha <jsarha@ti.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

Use table based setup to register the DAPM widgets and routes. This on one hand
makes the code a bit shorter and cleaner and on the other hand the board level
DAPM elements get registered in the card's DAPM context rather than in the
CODEC's DAPM context.

While we are at it also remove the snd_soc_dapm_enable_pin() in the init
callback, since pins are enabled by default. Also drop the snd_soc_dapm_sync()
calls, since they are ignored by the core anyway until the card has been fully
instantiated.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NMark Brown <broonie@linaro.org>

The rt5640 driver won't compile without I2C enabled. Hence, the Intel
Haswell and Baytrail+RT5640 ASoC drivers must also depend on I2C, since
these select RT5640.

This solves:
sound/soc/codecs/rt5640.c:2220:1: warning: data definition has no type or storage class [enabled by default]
sound/soc/codecs/rt5640.c:2220:1: error: type defaults to ‘int’ in declaration of ‘module_i2c_driver’ [-Werror=implicit-int]
sound/soc/codecs/rt5640.c:2220:1: warning: parameter names (without types) in function declaration [enabled by default]
sound/soc/codecs/rt5640.c:2210:26: warning: ‘rt5640_i2c_driver’ defined but not used [-Wunused-variable]
Reported-by: NJim Davis <jim.epost@gmail.com>
Signed-off-by: NStephen Warren <swarren@nvidia.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

Use table based setup to register the controls and DAPM widgets and routes. This
on one hand makes the code a bit shorter and cleaner and on the other hand the
board level DAPM elements get registered in the card's DAPM context rather than
in the CODEC's DAPM context.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NMark Brown <broonie@linaro.org>

Use table based setup to register the controls and DAPM widgets and routes. This
on one hand makes the code a bit shorter and cleaner and on the other hand the
board level DAPM elements get registered in the card's DAPM context rather than
in the CODEC's DAPM context.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NMark Brown <broonie@linaro.org>

Use table based setup to register the DAPM widgets and routes.  This on one hand
makes the code a bit shorter and cleaner and on the other hand the board level
DAPM elements get registered in the card's DAPM context rather than in the
CODEC's DAPM context.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NMark Brown <broonie@linaro.org>

snd_soc_dapm_mutex_unlock() wants a pointer to the DAPM context, not the CODEC.

Fixes: 03510ca0 ("ASoC: ams-delta: Update locking around use of DAPM pin API")
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NMark Brown <broonie@linaro.org>

regmap should handle any byte ordering issues required, it is looking for
a byte stream from the bus, so don't set 16 bits per word. This is likely
to have tested out OK due to use of an unmerged SPI controller driver.
Signed-off-by: NMark Brown <broonie@linaro.org>
Tested-by: NMax Filippov <jcmvbkbc@gmail.com>

Freeing the current list element while iterating over the list will cause a use
after free since the iterator function will still use the current element to
look up the next. Use list_for_each_safe() and remove the element from the list
before freeing it to avoid this.

Fixes: 1438c2f6 ("ASoC: Add a per component dai list")
Reported-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NMark Brown <broonie@linaro.org>

Since the CODEC driver could specify its own I/O(read and write)
while registering the CODEC for some reason, maybe the MFDs is
used, etc.

So just do check it, if they are not specified by CODEC driver
then try to set up the default regmap I/O if regmap is used.
Signed-off-by: NXiubo Li <Li.Xiubo@freescale.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

Now that all users have been converted to regmap and the config.reg_bits
and config.val_bits can be setted by each user through regmap core API.
So these two params are redundant here.

Since the only control type that left is SND_SOC_REGMAP, so remove it. Drop
the control params and add struct regmap *regmap to simplify the code.
Signed-off-by: NXiubo Li <Li.Xiubo@freescale.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

For some CODEC drivers like who act as the MFDs children are ignored
by this patch.
Signed-off-by: NXiubo Li <Li.Xiubo@freescale.com>
Signed-off-by: NMark Brown <broonie@linaro.org>