This is done as simple code transformation, the semantics of the
IRQ API provided by the core are are still very different to those
of genirq (mainly with regard to masking).
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@linux.intel.com>

Rather than open coding individual IRQs in each function which
manipulates them store data for IRQs in a table which is then
referenced in the users.

This is a substantial code shrink and should be a performance win in
cases where only a single IRQ goes off at once since instead of
reading four of the second level IRQ registers for each interrupt
we read only the sub-registers which have had an interrupt flagged.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@linux.intel.com>

In preparation for refactoring - it's over 700 lines of well-isolated
code and having it in a file by itself makes things more managable.

While we're at it make sure that we clean up the IRQ if we fail after
acquiring it on init.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@linux.intel.com>

Instead of hand rolling our own variant.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@linux.intel.com>

This driver provides reporting of the status supply voltage rails
of the WM835x series of PMICs via the hwmon API.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NJean Delvare <khali@linux-fr.org>
Signed-off-by: NSamuel Ortiz <sameo@linux.intel.com>

Due to the way that the WM8350 audio driver handles CODEC_ENA many of
the WM8350 audio registers are marked as volatile when they aren't
actually so. Allow the audio driver to see a cache of these values for
inspection during interrupt context.

To do this we need to stop satisfying any bits from volatile registers
from cache - there's no real benefit from doing so anyway, we did the
read already.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@linux.intel.com>

Reverse the order of the tests for loop exit so we use a valid value
before we time out.  Vanishingly unlikely to happen since we retry for
several times the expected conversion time.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@openedhand.com>

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

Some I2C controllers have high overheads for setting up I2C operations
which makes the register cache setup on startup excessively slow since
it does a lot of small transactions. Reduce this overhead by doing a
bulk read of the entire register bank and filtering out what we don't
need later.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@openedhand.com>

No software visible difference from revision A.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Cc: Samuel Ortiz <sameo@openedhand.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

The IRQs might have been left enabled in hardware, generating spurious
IRQs before the drivers have registered.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@openedhand.com>

With a postfix decrement tries will reach -1 rather than 0,
so the warning will not be issued even upon timeout.
Signed-off-by: NRoel Kluin <roel.kluin@gmail.com>
Acked-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@openedhand.com>

Check the return value of the device I/O functions when reading the
ID registers so we can provide a more useful diagnostic when we're
having trouble talking to the device.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@openedhand.com>

Ensure that the interrupt handling is configured before we do platform
specific init. This allows the platform specific initialisation to
configure things which use interrupts safely.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@openedhand.com>

The voltage and current regulators on the WM8350 AudioPlus PMIC can be
used in concert to provide a power efficient LED driver.  This driver
implements support for this within the standard LED class.

Platform initialisation code should configure the LED hardware in the
init callback provided by the WM8350 core driver.  The callback should
use wm8350_isink_set_flash(), wm8350_dcdc25_set_mode() and
wm8350_dcdc_set_slot() to configure the operating parameters of the
regulators for their hardware and then then use wm8350_register_led() to
instantiate the LED driver.

This driver was originally written by Liam Girdwood, though it has been
extensively modified since then.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NRichard Purdie <rpurdie@linux.intel.com>

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

The WM8351 is a WM8350 variant. As well as register default changes the
WM8351 has fewer voltage and current regulators than the WM8350.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@openedhand.com>

Some WM8350 variants have fewer DCDCs and ISINKs. Identify these at
probe and refuse to use the absent DCDCs when running on these chips.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@openedhand.com>

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

The WM8352 is a variant of the WM8350. Aside from the register defaults
there are no software visible differences to the WM8350.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@openedhand.com>

Since the WM8350 driver was originally written the semantics for the
identification registers of the chip have been clarified, allowing
us to do an exact match on all the fields. This avoids mistakenly
running on unsupported hardware.

Also change to using the datasheet names more consistently for
legibility and fix a printk() that should be dev_err().
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@openedhand.com>

Rather than check for chip revisions in the WM8350 drivers have the core
code set flags for relevant differences.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@openedhand.com>

The auxiliary ADC in the WM8350 is shared between several subdevices
so access to it needs to be arbitrated by the core driver.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@openedhand.com>

No other software changes are required.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@openedhand.com>

Hopefully this will make the purpose of these functions a bit clearer,
it's not immediately obvious that the lock is a hardware feature.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@openedhand.com>

This makes the contents of the cache clearer and fixes incorrect
initialisation of the cache for partially volatile registers.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@openedhand.com>

October 10th linux-next build (powerpc allyesconfig) failed like this:

drivers/mfd/wm8350-core.c:1131: error: __ksymtab_wm8350_create_cache causes a section type conflict

Caused by commit 89b4012b ("mfd: Core
support for the WM8350 AudioPlus PMIC"). wm8350_create_cache is not used
elsewhere, so remove the EXPORT_SYMBOL.
Signed-off-by: NStephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@openedhand.com>

NO_IRQ is only defined on some architectures - the general way to test
for an invalid IRQ in the modern kernel is by comparing with zero.
Reported-by: NStephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@openedhand.com>

Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@openedhand.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

In order to avoid merge problems further down the line add placeholders
for several of the WM8350 client devices and register them, otherwise
the patches adding the client devices will all try to update the same
code.

Also remove redundant checks for null regulator platform devices while
we're at it.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@openedhand.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

Most of the subdevices for the WM8350 code are registered in the same
fashion so factor out the code to do the initial registration.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@openedhand.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

The WM8350 features six DCDC convertors (four buck and two boost), four
LDO voltage regulators and two constant current sinks. This driver adds
support for these through the regulator API.

This driver was written by Liam Girdwood with updates for submission
from Mark Brown.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

The WM8350 has an interrupt line to the CPU which is shared by the
devices on the CPU. This patch adds support for the interrupt
controller within the WM8350 which identifies which identifies the
interrupt cause. In common with other similar chips this is done
outside the standard interrupt framework due to the need to access
the interrupt controller over an interrupt-driven bus.

This code was all originally written by Liam Girdwood with updates for
submission by me.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@openedhand.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

Some functions of the WM8350 require board-specific initialisation on
startup. Provide a callback to the WM8350 driver in platform data
for platforms to use to configure the chip.  Use of a callback allows
platforms to control the ordering of initialisation which can be
important.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@openedhand.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>

The WM8350 is an integrated audio and power management subsystem
intended for use as the primary PMIC in mobile multimedia applications.
The WM8350 can be controlled via either I2C or SPI - the control
interface is provided by a separate module in order to allow greatest
flexibility in configuring the kernel.

This driver was originally written by Liam Girdwood and has since been
updated to current kernel APIs and split up for submission by me.  All
the heavy lifting here was done by Liam.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: NSamuel Ortiz <sameo@openedhand.com>
Signed-off-by: NLiam Girdwood <lrg@slimlogic.co.uk>