The current helper functions, snd_soc_of_parse_audio_simple_widgets()
and snd_soc_of_parse_audio_routing(), set dapm_widgets and dapm_routes
without caring if they are already set by using build-in widgets and
routes in the card driver. So there could be one of them, build-in one
or Device Tree one, overrided by the other depending on which one was
assigned later.

This patch adds an extra pair of dapm_widgets and dapm_routes for DT
use only so as to prevent unexpected overriding.
Signed-off-by: NNicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Don't use generic snapshot of trigger_tstamp if low-level driver or
hardware can get a more precise value for better audio/system time
synchronization.
Signed-off-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

For assigning sysfs entries for a card device from the driver,
introduce a new helper function, snd_card_add_dev_attr().  In this
way, we can avoid the possible race between the device registration
and the sysfs addition / removal.

The driver can pass a new attribute group to add freely.  This has to
be called before snd_card_register().

Currently, up to two extra groups can be added.  More than that, it'll
return an error.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Instead of calling device_create_file() manually, assign the static
attribute group entries at the device registration.  This simplifies
the error handling and avoids the possible races.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>
Signed-off-by: NMark Brown <broonie@kernel.org>

Since the device is no longer hidden but embedded into each component,
we no longer need snd_get_device().  Let's drop it and relevant codes.
Reviewed-by: NJaroslav Kysela <perex@perex.cz>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Now that all callers have been replaced with
snd_device_register_for_dev(), let's drop the obsolete device
registration code and concentrate only on the code handling struct
device directly.  That said,

- remove the old snd_device_register(),
- rename snd_device_register_for_dev() with snd_device_register(),
- drop superfluous arguments from snd_device_register(),
- change snd_unregister_device() to pass the device pointer directly
Reviewed-by: NJaroslav Kysela <perex@perex.cz>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Like previous patches, this one embeds the struct device into struct
snd_compr.  As the dev field wasn't used beforehand, it's reused as
the new device struct.
Reviewed-by: NJaroslav Kysela <perex@perex.cz>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Like previous patches, this changes the device management for rawmidi,
embedding the struct device into struct snd_rawmidi.  The required
change is more or less same as hwdep device.

The currently unused dev field is reused as the new embedded struct
field now.
Reviewed-by: NJaroslav Kysela <perex@perex.cz>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Like previous patches, at this time we embed the struct device into
PCM object.  However, this needs a bit more caution: struct snd_pcm
doesn't own one device but two, for both playback and capture!  Thus
not struct snd_pcm but struct snd_pcm_str object contains the device.

Along with this change, pcm->dev field is dropped for avoiding
confusion.  It was meant to point to a non-standard parent.  But,
since now we can touch each struct device directly, we can manipulate
the parent field easily there, too.
Reviewed-by: NJaroslav Kysela <perex@perex.cz>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Like the previous patch, this one embeds the device object into hwdep
object.  For a proper object lifecycle, it's freed in the release
callback.

This also allows us to create sysfs entries via passing to the groups
field of the device without explicit function calls.  Since each
driver can see the device and touch its groups field directly, we
don't need to delegate in hwdep core any longer.  So, remove the
groups field from snd_hwdep, and let the user (in this case only
hda_hwdep.c) modify the device groups.
Reviewed-by: NJaroslav Kysela <perex@perex.cz>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

This patch embeds a struct device for the control device into the card
object and avoid the device creation at registration time.
Reviewed-by: NJaroslav Kysela <perex@perex.cz>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Introduce a new helper function snd_device_initialize() to initialize
the device object for sound devices.
Reviewed-by: NJaroslav Kysela <perex@perex.cz>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Instead of open-coding the search over the control file loop, provide
a helper function for the preferred subdevice assigned to the current
process.
Reviewed-by: NJaroslav Kysela <perex@perex.cz>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

This is a preliminary patch for the further work on embedding struct
device into each sound device instance.  It changes
snd_register_device*() helpers to receive the device object directly
for skipping creating a device there.
Reviewed-by: NJaroslav Kysela <perex@perex.cz>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Protect the call with a mutex, as this may be called in parallel
(either from the PCM rate change and the clock change).
Acked-by: NJaroslav Kysela <perex@perex.cz>
Tested-by: NPavel Hofman <pavel.hofman@ivitera.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Define snd_ak4114_suspend() and snd_ak4114_resume() functions to
handle PM properly, stopping and restarting the work at PM.
Currently only ice1712/juli.c deals with the PM and ak4114, so fix the
calls there appropriately.

The same PM functions are defined in ak4113.c, too, although they
aren't currently called yet (ice1712/quartet.c may be enhanced to
support PM later).
Acked-by: NJaroslav Kysela <perex@perex.cz>
Tested-by: NPavel Hofman <pavel.hofman@ivitera.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

When ak4114 work calls its callback and the callback invokes
ak4114_reinit(), it stalls due to flush_delayed_work().  For avoiding
this, control the reentrance by introducing a refcount.  Also
flush_delayed_work() is replaced with cancel_delayed_work_sync().

The exactly same bug is present in ak4113.c and fixed as well.
Reported-by: NPavel Hofman <pavel.hofman@ivitera.com>
Acked-by: NJaroslav Kysela <perex@perex.cz>
Tested-by: NPavel Hofman <pavel.hofman@ivitera.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Add helper functions to allow drivers to specify several disjoint
ranges for a variable. In particular, there is a codec (PCM512x) that
has a hole in its supported range of rates, due to PLL and divider
restrictions.

This is like snd_pcm_hw_constraint_list(), but for ranges instead of
points.
Signed-off-by: NPeter Rosin <peda@axentia.se>
Reviewed-by: NLars-Peter Clausen <lars@metafoo.de>
Reviewed-by: NTakashi Iwai <tiwai@suse.de>
Signed-off-by: NMark Brown <broonie@kernel.org>

Nowadays it's recommended.  Replace all in a shot.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

make the sta32x driver usable with device tree configs.  Code is heavily based
on the sta350 driver.
Signed-off-by: NThomas Niederprüm <niederp@physik.uni-kl.de>
Signed-off-by: NMark Brown <broonie@kernel.org>

In some cases it is necessary to before additional operations after the
device has been initialized and before the device is registered. This can
for example be resetting the device.

This patch introduces a new function snd_soc_alloc_ac97_codec() which is
similar to snd_soc_new_ac97_codec() except that it does not register the
device. Any users of snd_soc_alloc_ac97_codec() are responsible for calling
device_add() manually.

Fixes: 6794f709 ("ASoC: ac97: Drop delayed device registration")
Reported-by: NManuel Lauss <manuel.lauss@gmail.com>
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Tested-by: NManuel Lauss <manuel.lauss@gmail.com>
Signed-off-by: NMark Brown <broonie@kernel.org>
Cc: stable@vger.kernel.org

Queues are used both for scheduling playback events and for assigning
timestamps to recorded events, so it is easy to need quite a lot of
them, especially on a multi-user system.  Additionally, the actual
queue objects are allocated dynamically, so it does not really make
sense to have a low limit.  Increase it to something still sane.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Signed-off-by: NClemens Ladisch <clemens@ladisch.de>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Signed-off-by: NClemens Ladisch <clemens@ladisch.de>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

There are no more users of this field left so it can finally be removed.
New users should use snd_soc_dapm_to_codec(w->dapm);

The reason why it is removed is because it doesn't fit to well anymore in
the componentized ASoC hierarchy, where DAPM works on the snd_soc_component
level. And the alternative of snd_soc_dapm_to_codec(w->dapm) typically
generates the same amount of code, so there is really no reason to keep it.

For automatic conversion the following coccinelle semantic patch can be used:
// <smpl>
@@
struct snd_soc_dapm_widget *w;
@@
-w->codec
+snd_soc_dapm_to_codec(w->dapm)
// </smpl>
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NMark Brown <broonie@kernel.org>

Call clk_prepare_enable() and clk_disable_unprepare() for cpu dai
clock and codec dai clock in dai statup and shutdown callbacks. This
to make sure the related clock are enabled when the audio device is
used.
Signed-off-by: NJyri Sarha <jsarha@ti.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

Fixes the following compilation error:

   include/sound/compress_driver.h: In function ‘snd_compr_drain_notify’:
   include/sound/compress_driver.h:177:2: error: implicit declaration of function ‘snd_BUG_ON’ [-Werror=implicit-function-declaration]
     if (snd_BUG_ON(!stream))

snd_BUG_ON() is defined in sound/core.h but the file is not included explicitly,
so include it.
Signed-off-by: NQais Yousef <qais.yousef@imgtec.com>
Acked-by: NVinod Koul <vinod.koul@intel.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

L/R channel will be switched if under/over flow error happen on
Renesas R-Car sound device by the HW bugs. Then, HW restart is required
for salvage. This patch add salvage support for SRC.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

The patch adds the MICBIAS VDD setting in the platform data. It can be set to
1V8 or 3V3 in the MICBIAS VDD.
Signed-off-by: NOder Chiou <oder_chiou@realtek.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

For some setups it is necessary to change the DAI link format at runtime.
This patch factors out the code that does the initial static DAI link format
configuration into a separate helper function which can be used board
drivers as well.

This allows board drivers that have to change the DAI link format at runtime
to reuse it instead of having to manually change the format on all DAIs.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NMark Brown <broonie@kernel.org>

Remove function declarations for functions that don't have a matching
implementation.

For snd_pcm_build_linear_format the implementation was removed in
64d27f96 ("[ALSA] Support 3-bytes 24bit format in PCM OSS
emulation"). All the others never had one (as far as git history goes).
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The various PCM and hwdep allocation functions in this driver take a pointer
to a pointer of a PCM/hwdep where if this parameter is provided the newly
allocated object is stored. All callers pass NULL though, so remove the
parameter.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Most callers of snd_wss_pcm(), snd_wss_timer() and snd_cs4236_pcm() pass
NULL as the last parameter, some callers pass a pointer but never use it
after the function has been called and only a few callers pass a pointer and
actually use it. The later is only the case for snd_wss_pcm() for
snd_cs4236_pcm() and it is possible to get the same PCM object by accessing
the pcm field of the snd_wss struct that was passed as the first parameter.

This function removes the last parameters from the functions mentioned above
and updates the callers which used it to use chip->pcm instead. This allows
us to slightly simplify the functions since they don't have to check and set
the last parameter anymore which makes the code slightly shorter and
cleaner.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

All callers of snd_sb16dsp_pcm() always pass the pcm field of the first
parameter as the last parameter. Simplify the function by moving this inside
the function itself. This makes the code a bit shorter and cleaner.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

snd_sb8dsp_pcm() and snd_sb8dsp_midi() take a pointer to a pointer of a
PCM/MIDI where if this parameter is provided the newly allocated object is
stored. All callers pass NULL though, so remove the parameter. This makes
the code a bit cleaner and shorter.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

snd_gf1_pcm_new() and snd_gf1_rawmidi_new() take a pointer to a pointer of a
PCM/MIDI where if this parameter is provided the newly allocated object is
stored. All callers pass NULL though, so remove the parameter. This makes
the code a bit cleaner and shorter.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

snd_es1688_pcm() takes a pointer to a pointer of a PCM where if this
parameter is provided the newly allocated PCM is stored. This PCM is also
available from the pcm field of the snd_es1688 struct that got passed to the
same function. This patch updates all callers which passed a pointer to use
that field instead and then removes the parameter from the function. This
makes the code a bit shorter and cleaner.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

snd_ad1816a_pcm() and snd_ad1816a_timer() take a pointer to a pointer of a
PCM/timer where if this parameter is provided the newly allocated object is
stored. All callers pass NULL though, so remove the parameter. This makes
the code a bit cleaner and shorter.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Add SNDRV_PCM_TRIGGER_DRAIN trigger for pcm drain.

Some audio devices require notification of drain events
in order to properly drain and shutdown an audio stream.
Signed-off-by: NLibin Yang <libin.yang@intel.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

__fls has the same semantics as ld2, so there is no need to re-implement it.
Furthermore a lot of architectures have custom implementations of __fls that
are able to use special hardware instructions to compute the result. This
makes the code slightly shorter and faster.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>