Otherwise there's a race where the DAI might get used without everything
having been set up.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSangbeom Kim <sbkim73@samsung.com>

Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSangbeom Kim <sbkim73@samsung.com>

Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Signed-off-by: NLeon Romanovsky <leon@leon.nu>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Merge branch 'topic/cache' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regmap into for-3.3

Move the initialization regcache related fields of the regmap struct to
regcache_init. This allows us to keep regmap and regcache code better
separated.
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Signed-off-by: NAxel Lin <axel.lin@gmail.com>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Signed-off-by: NAxel Lin <axel.lin@gmail.com>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Signed-off-by: NAxel Lin <axel.lin@gmail.com>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

All users now use regmap directly so delete the ASoC version of the code.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Now that the mfd is using the register map cache there's no need for the
CODEC driver to do any register cache management or any funny dances to
interact with the other drivers using the device so just remove the cache
initialisation and volatility information.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Just go directly to the regmap API, saving code and making integration
that bit more direct.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Merge branch 'mfd/wm8994' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/misc into for-3.3

Factor out the irq_chip implementation, substantially reducing the code
size for the driver.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@linux.intel.com>

Merge branch 'topic/irq' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regmap into wm8994-mfd

For later chip revisions the WM1811 GPIO6 register is always volatile so
store the device revision when initialising the driver and then check at
runtime if we're running on a newer device.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@linux.intel.com>

struct wm8994 includes a mutex so we need to include mutex.h before we
declare it. All current users rely on this being done implicitly.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@linux.intel.com>

The WM1811A is a variant of the WM1811 with pin configuration changes.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@linux.intel.com>

As WM1811 revision C was transparent to software the revision IDs for
subsequent revisions are one less than they would normally be. Correct
for this in log messages.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@linux.intel.com>

The driver has no need to modify the regulator_init_data so declare it
const to allow machine code to do so.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@linux.intel.com>

The different devices handled by the WM8994 can be distinguished using
their ID registers so we don't need to rely on the user having registered
the device correctly. Instead do the initial regmap setup with a minimal
configuration only supporting physical I/O and then configure the cache
once we have identified the device.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@linux.intel.com>

As part of this we provide information about the registers that exist in
the device to the regmap core, drop the small amount of cache that the
core had been using and let regmap do the sync.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@linux.intel.com>

Merge branch 'topic/cache' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regmap into HEAD

Describe the register map to the regmap core so that we can use its
diagnostic features and cache support.  This is split out from the patch
using it due to the size so that the actual code change is a bit clearer.

As the various devices are supersets of each other the access maps are
built up by layering the functions on top of each other, though the
interface for specifying the register defaults isn't currently amenable
to this.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@linux.intel.com>

Add a bunch of definitions for wm8994 registers that are not currently
used by software.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@linux.intel.com>

Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@linux.intel.com>

Disable more pulls by default on WM8994 for a small current saving. Since
some designs do leave SPKMODE floating provide platform data to allow that
to be left enabled.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@linux.intel.com>

Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@linux.intel.com>

Add a placeholder device tree binding for the wm8994 driver. At present
the binding is essentially null as none of the platform data is supported,
and at least some of that will depend on the pending regulator bindings.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@linux.intel.com>

Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Oritz <sameo@linux.intel.com>

This driver can be built as module and the file header indicates that
the driver is published under the GPL.
Thus add MODULE_LICENSE("GPL") for it.
Signed-off-by: NAxel Lin <axel.lin@gmail.com>
Acked-by: NSascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

To support advanced system functionality for additional components; the
actively used clocks will remain the same for current components. Also
factor the rate out to a single #define while we're at it.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Signed-off-by: NAxel Lin <axel.lin@gmail.com>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Signed-off-by: NJonathan Neuschäfer <j.neuschaefer@gmx.net>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Conflicts:
	sound/soc/mxs/mxs-pcm.c

The message was obviously copied from soc_init_codec_debugfs()
Signed-off-by: NLothar Waßmann <LW@KARO-electronics.de>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

Signed-off-by: NLothar Waßmann <LW@KARO-electronics.de>
Acked-by: NWolfram Sang <w.sang@pengutronix.de>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>

The sound driver refuses to load as module, because of the missing
MODULE_LICENSE("GPL").
The file header indicates that the driver is indeed published under
the GPL.
Signed-off-by: NLothar Waßmann <LW@KARO-electronics.de>
Acked-by: NWolfram Sang <w.sang@pengutronix.de>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>