Allows drivers to distinguish which subsequence is being notified when
they get called back.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>

Could just as well live in sysfs but sysfs doesn't have the simple
value export helpers debugfs does.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>

Allow hookup of cards registered directly with the core to the PM
operations by exporting the device power management operations to
modules, also exporting the default PM operations since it is
expected that most cards will end up using exactly the same setup.

Note that the callbacks require that the driver data for the card be
the snd_soc_card.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>

In order to support cards instantiated without using soc-audio remove
the use of the platform device in the card probe() and remove() ops.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>

The platform device for the card is tied closely to the soc-audio
implementation which we're currently trying to remove in favour of
allowing cards to have their own devices. Begin removing it by
replacing it with the card in the suspend and resume callbacks we
give to cards, also taking the opportunity to remove the legacy
suspend types which are currently hard coded anyway.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>

We generally refer to registers as unsigned ints (including in the
underlying CODEC driver operation).
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>

This is primarily needed to avoid writing back to the cache
whenever we are syncing the cache with the hardware.  This gives a
performance benefit especially for large register maps.
Signed-off-by: NDimitris Papastamos <dp@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Many modern devices have features such as DC servos which take time to start.
Currently these are handled by per-widget events but this makes it difficult
to paralleise operations on multiple widgets, meaning delays can end up
being needlessly serialised. By providing a callback to drivers when all
widgets of a given type have been handled during a DAPM sequence the core
allows drivers to start operations separately and wait for them to complete
much more simply.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>

The machine driver can't register the card directly and need to do this thru
soc-audio device creation

This patch allows the register and unregister card to be directly called by
machine drivers
Signed-off-by: NVinod Koul <vinod.koul@intel.com>
Signed-off-by: NHarsha Priya <priya.harsha@intel.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Currently the soc_probe initializes the card hence it does the card list
initialzation. But if machines directly register the card they would need to
do these steps, so putting them as inline would save lot of code

This patch adds an inline to do list initialzation
Signed-off-by: NVinod Koul <vinod.koul@intel.com>
Signed-off-by: NHarsha Priya <harsha.priya@intel.com>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Ensure that all calls to readable_register()/volatile_register() go via
the snd_soc_codec function pointers.

If the default register access table has been given but no functions
for handling readable()/volatile() registers, use the default ones provided
by soc-cache.
Signed-off-by: NDimitris Papastamos <dp@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Signed-off-by: NDimitris Papastamos <dp@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

For common scenarios, device drivers can provide a table of all the
registers that are at least either readable/writable/volatile.  The idea
is that if a register lookup fails, all of its read/write/vol members
will be zero and will be treated as default.  This also reduces the
size of the register access array.
Signed-off-by: NDimitris Papastamos <dp@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Simplify the use of reg_size, by calculating it once and storing it in
the codec structure for later reference.  The value of reg_size is
reg_cache_size * reg_word_size.
Signed-off-by: NDimitris Papastamos <dp@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Everything else is using snd_soc_ so we should use it here too.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>

A couple Tegra ASoC drivers will create debugfs entries. Mark requested
these by under debugfs/asoc/ not just debugfs/. To enable this, export
the dentry representing debugfs/asoc/.

Also, rename debugfs_root -> asoc_debugfs_root now it's exported to
prevent potential symbol name clashes.
Signed-off-by: NStephen Warren <swarren@nvidia.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Power change event like stream start/stop or kcontrol change in a
cross-device path originates from one device but codec bias and widget power
changes must be populated to another devices on that path as well.

This patch modifies the dapm_power_widgets so that all the widgets on a
sound card are checked for a power change, not just those that are specific
to originating device. Also bias management is extended to check all the
devices. Only exception in bias management are widgetless codecs whose bias
state is changed only if power change is originating from that context.

DAPM context test is added to dapm_seq_run to take care of if power sequence
extends to an another device which requires separate register writes.
Signed-off-by: NJarkko Nikula <jhnikula@gmail.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Decoupling widgets from DAPM context is required when extending the ASoC
core to cross-device paths. Even the list of widgets are now kept in
struct snd_soc_card, the widget listing in sysfs and debugs remain sorted
per device.

This patch makes possible to build cross-device paths but does not extend
yet the DAPM to handle codec bias and widget power changes of an another
device.

Cross-device paths are registered by listing the widgets from device A in
a map for device B. In case of conflicting widget names between the devices,
a uniform name prefix is needed to separate them. See commit ead9b919
"ASoC: Add optional name_prefix for kcontrol, widget and route names" for
help.

An example below shows a path that connects MONO out of A into Line In of B:

static const struct snd_soc_dapm_route mapA[] = {
	{"MONO", NULL, "DAC"},
};

static const struct snd_soc_dapm_route mapB[] = {
	{"Line In", NULL, "MONO"},
};
Signed-off-by: NJarkko Nikula <jhnikula@gmail.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Decoupling DAPM paths from DAPM context is a first prerequisite when
extending ASoC core to cross-device paths. This patch is almost a nullop and
does not allow to construct cross-device setup but the path clean-up part in
dapm_free_widgets is prepared to remove cross-device paths between a device
being removed and others.
Signed-off-by: NJarkko Nikula <jhnikula@gmail.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

This patch removes some legacy structure definitions which are not using
in current ASoC drivers.
Signed-off-by: NSeungwhan Youn <sw.youn@samsung.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Added an optional name member to snd_soc_cache_ops to enable more
sensible diagnostic messages during cache init, exit and sync.

Remove redundant newline in source code.
Signed-off-by: NDimitris Papastamos <dp@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Currently the machine driver can only do bias level configuration before
the CODEC bias level is brought up. This means that the machine cannot do
any configuration which depends on the CODEC bias level being maintained.
Provide a post-CODEC callback which allows the machine driver to do things
like enable the FLL on a CODEC which is brought down to BIAS_OFF when idle.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>

Allow the CODEC driver structure to be marked const by making all
the APIs that use it do so.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>

This patch allows machine drivers to override the compression type
provided by the codec driver.
Signed-off-by: NDimitris Papastamos <dp@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Make sure to use codec->reg_def_copy instead of codec_drv->reg_cache_default
wherever necessary.  This change is necessary because in the next patch we
move the cache initialization code outside snd_soc_register_codec() and by that
time any data marked as __devinitconst such as the original reg_cache_default
array might have already been freed by the kernel.
Signed-off-by: NDimitris Papastamos <dp@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

The snd_soc_codec_conf struct now holds codec specific configuration
information.

A new configuration option has been added to allow machine drivers to
override the compression type set by the codec driver.

In the absence of providing an snd_soc_codec_conf struct or when providing
one but not setting the compress_type member to anything, the one supplied
by the codec driver will be used instead.  In all other cases the one
set in the snd_soc_codec_conf struct takes effect.
Signed-off-by: NDimitris Papastamos <dp@opensource.wolfsonmicro.com>
Acked-by: NJarkko Nikula <jhnikula@gmail.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Ensure that the base value of compress_type starts at 1 so that
we know whether the machine driver has provided a compress_type
for overriding the codec supplied one.
Signed-off-by: NDimitris Papastamos <dp@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

We need to keep a copy of the compress_type supplied by the codec driver
so that we can override it if necessary with whatever the machine driver
has provided us with.  The reason for not modifying the codec->driver
struct directly is that ideally we'd like to keep it const.

Adjust the code in soc-cache and soc-core to make use of the compress_type
member in the snd_soc_codec struct.
Signed-off-by: NDimitris Papastamos <dp@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

We shouldn't be assigning to the driver structure (which really ought
to be const, further patch to follow) though there's unlikely to be any
actual problem except in the unlikely case that two devices with the
same driver but different bus types appear in the same system.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>

This makes possible to register auxiliary dailess codecs in a machine
driver. Term dailess is used here for amplifiers and codecs without DAI or
DAI being unused.

Dailess auxiliary codecs are kept in struct snd_soc_aux_dev and those codecs
are probed after initializing the DAI links. There are no major differences
between DAI link codecs and dailess codecs in ASoC core point of view. DAPM
handles them equally and sysfs and debugfs directories for dailess codecs
are similar except the pmdown_time node is not created.

Only suspend and resume functions are modified to traverse all probed codecs
instead of DAI link codecs.

Example below shows a dailess codec registration.

struct snd_soc_aux_dev foo_aux_dev[] = {
	{
		.name = "Amp",
		.codec_name = "codec.2",
		.init = foo_init2,
	},
};

static struct snd_soc_card card = {
	...
	.aux_dev = foo_aux_dev,
	.num_aux_devs = ARRAY_SIZE(foo_aux_dev),
};
Signed-off-by: NJarkko Nikula <jhnikula@gmail.com>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Signed-off-by: NDimitris Papastamos <dp@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

There is a need to prefix codec kcontrol, widget and internal route names in
an ASoC machine that has multiple codecs with conflicting names. The name
collision would occur when codec drivers try to registering kcontrols with
the same name or when building audio paths.

This patch introduces optional prefix_map into struct snd_soc_card. With it
machine drivers can specify a unique name prefix to each codec that have
conflicting names with anothers. Prefix to codec is matched with codec
name.

Following example illustrates a machine that has two same codec instances.
Name collision from kcontrol registration is avoided by specifying a name
prefix "foo" for the second codec. As the codec widget names are prefixed
then second audio map for that codec shows a prefixed widget name.

static const struct snd_soc_dapm_route map0[] = {
	{"Spk", NULL, "MONO"},
};

static const struct snd_soc_dapm_route map1[] = {
	{"Vibra", NULL, "foo MONO"},
};

static struct snd_soc_prefix_map codec_prefix[] = {
	{
		.dev_name = "codec.2",
		.name_prefix = "foo",
	},
};

static struct snd_soc_card card = {
	...
	.prefix_map = codec_prefix,
	.num_prefixes = ARRAY_SIZE(codec_prefix),
};
Signed-off-by: NJarkko Nikula <jhnikula@gmail.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

This patch adds support for rbtree compression when storing the
register cache.  It does this by not adding any uninitialized registers
(those whose value is 0).  If any of those registers is written
with a nonzero value they get added into the rbtree.

Consider a sample device with a large sparse register map.  The
register indices are between [0, 0x31ff].  An array of 12800 registers
is thus created each of which is 2 bytes.  This results in a 25kB
region.  This array normally lives outside soc-core, normally in the
driver itself.  The original soc-core code would kmemdup this region
resulting in 50kB total memory.  When using the rbtree compression
technique and __devinitconst on the original array the figures are
as follows.  For this typical device, you might have 100 initialized
registers, that is registers that are nonzero by default.  We build
an rbtree with 100 nodes, each of which is 24 bytes.  This results
in ~2kB of memory.  Assuming that the target arch can freeup the
memory used by the initial __devinitconst array, we end up using
about ~2kB bytes of actual memory.  The memory footprint will increase
as uninitialized registers get written and thus new nodes created in
the rbtree.  In practice, most of those registers are never changed.
If the target arch can't freeup the __devinitconst array, we end up
using a total of ~27kB.  The difference between the rbtree and the LZO
caching techniques, is that if using the LZO technique the size of
the cache will increase slower as more uninitialized registers get
changed.
Signed-off-by: NDimitris Papastamos <dp@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

This patch adds support for LZO compression when storing the register
cache.  The initial register defaults cache is marked as __devinitconst
and the only change required for a driver to use LZO compression is
to set the compress_type member in codec->driver to SND_SOC_LZO_COMPRESSION.

For a typical device whose register map would normally occupy 25kB or 50kB
by using the LZO compression technique, one can get down to ~5-7kB.  There
might be a performance penalty associated with each individual read/write
due to decompressing/compressing the underlying cache, however that should not
be noticeable.  These memory benefits depend on whether the target architecture
can get rid of the memory occupied by the original register defaults cache
which is marked as __devinitconst.  Nevertheless there will be some memory
gain even if the target architecture can't get rid of the original register
map, this should be around ~30-32kB instead of 50kB.
Signed-off-by: NDimitris Papastamos <dp@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

This patch introduces the new caching API and migrates the
old caching interface into the new one.  The flat register caching
technique does not use compression at all and it is equivalent to
the old caching technique.  One can still access codec->reg_cache
directly but this is not advised as that will not be portable
across different caching strategies.

None of the existing drivers need to be changed to adapt to this
caching technique.  There should be no noticeable overhead associated
with using the new caching API.
Signed-off-by: NDimitris Papastamos <dp@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Based on discussion the dapm_pop_time in debugsfs should be per card rather
than per device. Single pop time value for entire card is cleaner when the
DAPM sequencing is extended to cross-device paths.

debugfs/asoc/{card->name}/{codec dir}/dapm_pop_time
->
debugfs/asoc/{card->name}/dapm_pop_time
Signed-off-by: NJarkko Nikula <jhnikula@gmail.com>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

There will be need to have sound card specific debugfs entries. This patch
introduces a new debugfs/asoc/{card->name}/ directory but does not add yet
any entries there.
Signed-off-by: NJarkko Nikula <jhnikula@gmail.com>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Decoupling Dynamic Audio Power Management (DAPM) from codec devices is
required when developing ASoC further. Such as for other ASoC components to
have DAPM widgets or when extending DAPM to handle cross-device paths.

This patch decouples DAPM related variables from struct snd_soc_codec and
moves them to new struct snd_soc_dapm_context that is used to encapsulate
DAPM context of a device. ASoC core and API of DAPM functions are modified
to use DAPM context instead of codec.

This patch does not change current functionality and a large part of changes
come because of structure and internal API changes.

Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some
minor core changes, codecs and machine driver conversions from
Jarkko Nikula <jhnikula@gmail.com>.
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NJarkko Nikula <jhnikula@gmail.com>
Cc: Nicolas Ferre <nicolas.ferre@atmel.com>
Cc: Manuel Lauss <manuel.lauss@googlemail.com>
Cc: Mike Frysinger <vapier.adi@gmail.com>
Cc: Cliff Cai <cliff.cai@analog.com>
Cc: Kevin Hilman <khilman@deeprootsystems.com>
Cc: Ryan Mallon <ryan@bluewatersys.com>
Cc: Timur Tabi <timur@freescale.com>
Cc: Sascha Hauer <s.hauer@pengutronix.de>
Cc: Lars-Peter Clausen <lars@metafoo.de>
Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org>
Cc: Wan ZongShun <mcuos.com@gmail.com>
Cc: Eric Miao <eric.y.miao@gmail.com>
Cc: Jassi Brar <jassi.brar@samsung.com>
Cc: Daniel Gloeckner <dg@emlix.com>
Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Facilitating adding trace type stuff. For a first pass add some dev_dbg()
statements into them.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>

With generic AC97 ASoC glue driver (codec/ac97.c), we get following warning when
the device is registered (slightly stripped the backtrace):

kobject (c5a863e8): tried to init an initialized object, something is seriously
                    wrong.
[<c00254fc>] (unwind_backtrace+0x0/0xec)
[<c014fad0>] (kobject_init+0x38/0x70)
[<c0171e94>] (device_initialize+0x20/0x70)
[<c017267c>] (device_register+0xc/0x18)
[<bf20db70>] (snd_soc_instantiate_cards+0x924/0xacc [snd_soc_core])
[<bf20e0d0>] (snd_soc_register_platform+0x16c/0x198 [snd_soc_core])
[<c0175304>] (platform_drv_probe+0x18/0x1c)
[<c0174454>] (driver_probe_device+0xb0/0x16c)
[<c017456c>] (__driver_attach+0x5c/0x7c)
[<c0173cec>] (bus_for_each_dev+0x48/0x78)
[<c0173600>] (bus_add_driver+0x98/0x214)
[<c0174834>] (driver_register+0xa4/0x130)
[<c001f410>] (do_one_initcall+0xd0/0x1a4)
[<c0062ddc>] (sys_init_module+0x12b0/0x1454)

This happens because the generic AC97 glue driver creates its codec->ac97 via
calling snd_ac97_mixer(). snd_ac97_mixer() provides own version of
snd_device.register which handles the device registration when
snd_card_register() is called.

To avoid registering the AC97 device twice, we add a new flag to the
snd_soc_codec: ac97_created which tells whether the AC97 device was created by
SoC subsystem.
Signed-off-by: NMika Westerberg <mika.westerberg@iki.fi>
Acked-by: NLiam Girdwood <lrg@slimlogic.co.uk>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>