Lochnagar is an evaluation and development board for Cirrus
Logic Smart CODEC and Amp devices. It allows the connection of
most Cirrus Logic devices on mini-cards, as well as allowing
connection of various application processor systems to provide a
full evaluation platform. This driver supports the board
controller chip on the Lochnagar board.
The Lochnagar board provides power supplies for the attached
CODEC/Amp device. Currently this driver supports the microphone
supplies and the digital core voltage for the attached
device. There are some additional supplies that will be
added in time but these supplies are sufficient for most
systems/use-cases.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Link to the generic GPIO specifier bindings now that the second cell of
the binding has some support in the driver.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Since a copy of the pdata was added into the core struct in
commit f6dd8449cd ("mfd: wm831x: Add basic device tree binding")
the pdata pointer in probe can no longer be NULL. As such remove
the redundant checks for this case.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Currently, when looking up a regulator supply, the regulator name
takes priority over the consumer mappings. As there are a lot of
regulator names that are in fairly common use (VDD, MICVDD, etc.) this
can easily lead to obtaining the wrong supply, when a system contains
two regulators that share a name.
The explicit consumer mappings contain much less ambiguity as they
specify both a name and a consumer device. As such prioritise those if
one exists and only fall back to the regulator name if there are no
matching explicit mappings.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Very fast systems may violate the minimum constraints for time the reset
line needs to remain low, or communicate with the device too soon after
releasing the reset. Fix this by adding some delays in to allow the chip
to properly reset, also factor out the reset into a function as it is
likely it will be re-used in later additions to the driver.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Acked-by: Paul Handrigan <Paul.Handrigan@cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
A hardcoded register is accidentally used instead of the register
address passed into the function. Correct this and use the appropriate
variable. This would cause minor issues on wm5102, but all other
devices using this driver would have been unaffected.
Fixes: commit ef84f885e0 ("mfd: arizona: Refactor arizona_poll_reg")
Reported-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
The Arizona devices only maintain the state of output GPIOs whilst the
CODEC is active, this can cause issues if the CODEC suspends whilst
something is relying on the state of one of its GPIOs. However, in
many systems the CODEC GPIOs are used for audio related features
and thus the state of the GPIOs is unimportant whilst the CODEC is
suspended. Often keeping the CODEC resumed in such a system would
incur a power impact that is unacceptable.
Allow the user to select whether a GPIO output should keep the
CODEC resumed, by adding a flag through the second cell of the GPIO
specifier in device tree.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Commit 4c0facddb7 ("gpio: core: Decouple open drain/source flag with
active low/high") decoupled the open collector outputs from active
low/high but did not update the documentation.
Update the device tree documentation to correctly reflect this new
separation between the two concepts.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Acked-by: Laxman Dewangan <ldewangan@nvidia.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
The function already defines a ret variable at the top and makes
no particular use of the shadowed definition, as such remove the
redundant definition.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Acked-by: Paul Handrigan <Paul.Handrigan@cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
A somewhat overdue update of the address for sending patches on Wolfson
parts to since our acquision a couple of years ago by Cirrus Logic.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Zero is a totally valid value to specify for the bst_ipk, as such we
should append CS35L35_VALID_PDATA to ensure that it actually makes it
into the register value.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Currently, we specify the timeout in terms of the number of polls but it
is more clear from a user of the functions perspective to specify the
timeout directly in milliseconds, as such update the function to these new
semantics.
Additionally, arizona_poll_reg essentially hard-codes
regmap_read_poll_timeout, update the implementation to use
regmap_read_poll_timeout. We still keep arizona_poll_reg around as
regmap_read_poll_timeout is a macro so rather than expand this for each
caller keep it wrapped in arizona_poll_reg.
Whilst we are doing this make the timeouts a little more generous as
the previous system had a bit more slack as it was done as a delay per
iteration of the loop whereas regmap_read_poll_timeout compares ktime's.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Register addresses are normally displayed in hex throughout the Arizona
driver. Update the arizona_poll_reg function to follow this convention.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
arizona_poll_reg already returns ETIMEDOUT if we don't see the expected
register changes before the time out, so remove pointless local setting of
ETIMEDOUT.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
The logic to check if the pin is an input or output whilst testing if we
need to read the register value from the hardware or not is currently
inverted. Remove the erroneous not from the if statement.
Fixes: 11598d1740 ("gpio: arizona: Correct handling for reading input GPIOs")
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
On the chip the IMON signal is a full 24-bits however normally only
some of the bits will be sent over the bus. The chip provides a field
to select which bits of the IMON will be sent back, this is the only
feedback signal that has this feature.
Add an additional entry to the cirrus,imon device tree property to
allow the IMON scale parameter to be passed.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Acked-by: Brian Austin <brian.austin@cirrus.com>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
The GPIO register is cached since all the configuration resides within
it, however, this means for input GPIOs the driver will not return the
actual state but the last value written to the register cache.
To correct this in the case of reading an input GPIO resume the CODEC
and drop the cache for the input register to ensure an actual hardware
read takes place.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Add a helper function regulator_set_pull_down_regmap to allow regmap
based regulators to easily enable pull down.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Add a helper function regulator_set_soft_start_regmap to allow regmap
based regulators to easily enable soft start.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Shorten the time it takes to power down the amp by disabling the volume
ramp whilst doing the final shutdown. The driver has already muted the
amplifier at this stage so doing the volume ramp serves no purpose.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Acked-by: Brian Austin <brian.austin@cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
As the HPDET can't be aborted mid way through we should not allow any new
insertion to be processed until the previous HPDET has finished. It is very
unlikely but with low enough debounce settings you could start a new HPDET
before the old one has completed, which results in an erroneous reading.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Chanwoo Choi <cw00.choi@samsung.com>
The maintainers entry for the Wolfson parts seems to be missing
an entry that covers the Arizona regulator drivers, correct this by
adding one.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The device argument passed to of_get_regulator_init_data is used to
do some devres memory allocation. Currently the driver passes the MFD
device pointer to this function, this could result in the init_data
allocation being leaked if the regulator is unbound but the MFD isn't.
Correct this issue by correctly passing the local platform device.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The device argument passed to of_get_regulator_init_data is used to
do some devres memory allocation. Currently the driver passes the MFD
device pointer to this function, this could result in the init_data
allocation being leaked if the regulator is unbound but the MFD isn't.
Correct this issue by correctly passing the local platform device.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The error path in probe attempts to put the device back into reset.
Should we fail to get the reset_gpio (such as a probe defer) we will
leave the error value in there, which the gpiod_set_value_cansleep on
the error path will attempt to deference.
Fix this issue by clearing reset_gpio before we head into the error
path.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Acked-by: Brian Austin <brian.austin@cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Use devm_gpiochip_add_data to simplify the error path in
samsung_gpiolib_register.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
If we request a GPIO hog, then gpiochip_add_data will attempt
to request some of its own GPIOs. The driver also uses
gpiochip_generic_request which means that for any GPIO request to
succeed the pinctrl needs to be registered. Currently however the
driver registers the GPIO and then the pinctrl meaning all GPIO hog
requests will fail, which then in turn causes the whole driver to fail
probe.
Fix this up by ensuring we register the pinctrl first. This
does require us to manually set the GPIO base for the
pinctrl. Fortunately the driver already assigns a fixed GPIO base, in
samsung_gpiolib_register, and uses the same calculation it does for
the pin_base. Meaning the two will always be the same and allowing us
to reuse the pinbase and avoid the issue.
Although currently there are no users of GPIO hogs in mainline
there are plenty of Samsung based boards that are widely used for
development purposes of other hardware. Indeed we hit this issue
whilst attaching some additional hardware to an Arndale system.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Add a device tree binding document for the wm831x series of PMICs.
Currently only support for the registering the device and the GPIOs are
actually implemented in the driver.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Add the basic ability to register the device through device tree, more
work is needed to get each individual sub-driver functioning correctly
but this is enough to get the device to probe from device tree.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
The driver stashes a CODEC pointer in the cs35l35_private structure,
which is used to obtain a struct device pointer for error messages in the
interrupt handler.
However, doing so is not very safe as the interrupt is registered, as it
should be in bus probe, but the CODEC pointer can't be safely stored until
the ASoC level probe. This leaves a window between the two probes where if
any interrupts are received a NULL pointer will be deferenced in the IRQ
handler.
Fix this issue by saving a pointer to the device directly and passing that
to the error messages in the interrupt handler rather than using the CODEC
pointer to access the device pointer.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
As it is quite common to use a stereo pair of amps but share the IRQ
line between them both add the IRQF_SHARED flag whilst requesting
cs35l35's IRQ.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Acked-by: Brian Austin <brian.austin@cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Add support for setting how the I2S pins are driven in unused slots,
currently the chip will just use the default of drive 0, however this
causes issues when multiple devices are attached to the same bus.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Acked-by: Brian Austin <brian.austin@cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Add a binding for setting how the I2S pins are driven in unused slots,
currently the chip will just use the default of drive 0, however
this causes issues when multiple devices are attached to the same bus.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Acked-by: Brian Austin <brian.austin@cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
A return statement is missing just before the error paths at the end of
probe. This causes us to fall straight into the error path and disable
the supplies and re-enable reset, as these are only controlled during
probe this causes the part to no longer function.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Acked-by: Brian Austin <brian.austin@cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
On a specific audio system an interrupt input of an audio CODEC is used as a
shared interrupt. That interrupt input is handled by a CODEC specific irq
chip driver and triggers a CPU interrupt via the CODEC irq output line.
The CODEC interrupt handler demultiplexes the CODEC interrupt inputs and
the interrupt handlers for these demultiplexed inputs run nested in the
context of the CODEC interrupt handler.
The demultiplexed interrupts use handle_nested_irq() as their interrupt
handler, which unfortunately has no support for shared interrupts. So the
above hardware cannot be supported.
Add shared interrupt support to handle_nested_irq() by iterating over the
interrupt action chain.
[ tglx: Massaged changelog ]
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Cc: patches@opensource.wolfsonmicro.com
Link: http://lkml.kernel.org/r/1488904098-5350-1-git-send-email-ckeepax@opensource.wolfsonmicro.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
We should not be writing acknowledge controls until the firmware is
running, as in the case of preloaded firmwares the DSP memory may be
unaccessible to whilst in the preloaded state. This means a write to the
control during this time could be lost.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Volatile controls should only be accessed when the firmware is active,
currently however writes to these controls will succeed, but the data
will be lost, if the firmware is powered down. Update this behaviour such
that an error is returned the same as it is for reads.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
This declaration has never been used and is likely some left over from
early prototypes of the code, just remove it.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Currently we leak a lot of things when tearing down the IRQs this patch
fixes this cleaning up both the IRQ mappings and the IRQ domain itself.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Currently we are not disabling MEM_ENA on the error path, we should
really do this to unwind the state back to how it was. This patch adds a
clear of MEM_ENA on the error path, again there is no major issues
caused by this minor fix.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The booted and running flags should really only be set once all the
steps at that power level have been complete. Currently operations can
fail after the flags have been set, which would leave us in an
inconsistent state where the flags are set but the things expected to
reach that level have not happened. Whilst there isn't really any major
impact from this it is best to clean it up.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The recent refactoring overlooked some places which should be covered by
the pwr_lock, all code that affects or depends on the power status of
the DSP should be covered, this patch adds the missing coverage.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
arizona_spk_init uses snd_soc_dapm_new_control which since
commit 37e1df8c95 ("ASoC: dapm: handle probe deferrals") will
occasionally request a probe deferral. Which means we should propagate the
error out of our driver from it.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Most of the kernel-doc comments in regmap don't actually generate
correctly. This patch fixes up a few common issues, corrects some typos
and adds some missing argument descriptions.
The most common issues being using a : after the function name which
causes the short description to not render correctly and not separating
the long and short descriptions of the function. There are quite a few
instances of arguments not being described or given the wrong name as
well.
This patch doesn't fixup functions/structures that are currently missing
descriptions.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Currently we handle both old style GPIO and new style GPIOD differently
simplify this slightly by converting the old style GPIO to a GPIOD and
just using that from then on.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Chanwoo Choi <cw00.choi@samsung.com>
As requirements to bring up audio paths are continuous getting tighter
and the DSP download to most ADSP devices happens over an external bus
it can become an important factor in the path bring up time. As such
sometimes it is a reasonable trade off to download the firmware ahead of
when it will be required and take a small hit on power consumption for
keeping the core powered up.
This "preloading" adds an additional control for each DSP core "DSPx
Preload Switch" that when set to true will power up the DSP core and
download the firmware currently selected in the "DSPx Firmware" control.
Whilst the core is preloaded the current firmware can not be changed and
the CODEC will be kept powered up and SYSCLK held on. Although future
improvements may allow the SYSCLK to be powered down as well because
the hardware only requires SYSCLK whilst the download is actually taking
place, but this is not covered in this series.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
There is some unnecessary complexity in the error path which now things
are converted to devm is actually very simple. This patch simplifies
things.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Every function handling the mode within the regulator core uses an unsigned
int for mode, except for regulator_mode_constrain. This patch changes the
type of mode within regulator_mode_constrain which fixes several instances
where we are passing pointers to unsigned ints then treating them as an int
within this function.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Functions are given external linkage by default making the extern's
unnecessary, as such remove them.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Functions are given external linkage by default making the extern's
unnecessary, as such remove them.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
As DCVDD will often be supplied by a child node of the MFD, we
can't call mfd_remove_devices as the first step in arizona_dev_exit
as might be expected (tidy up the children before we tidy up the
MFD). We need to disable and put the DCVDD regulator before we call
mfd_remove_devices, to prevent PM runtime from turning this back on we
also need to disable the PM runtime before we do this. Finally we can
not clean up the IRQs until all the MFD children have been removed, as
they may have registered IRQs themselves.
This creates a window of time where the interrupts are enabled but
the PM runtime, on which the IRQ handler depends, is not available,
any interrupts in this window will go unhandled and fill the log with
failed to resume device messages. To avoid this we simply disable the
main IRQ at the start of arizona_dev_exit, we don't need to actually
handle any IRQs in this window as we are removing the driver.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
No need to use irq_create_mapping (although there is no issue with
doing so) when we are only looking up an existing mapping. Just to
streamline things a little and make the code a little more clear
change some calls from irq_create_mapping to irq_find_mapping.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
We are not checking the return value from a call to wm_adsp_buffer_init
it looks like this used to be returned at the bottom of the function but
probably got missed when more error paths were added. This patch adds
the appropriate error check.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The control list may contain shutdown controls for firmwares that are
not currently active, attempting to write this will at best fail. To
avoid this issue we skip any control that is not active.
Fixes: commit f4f0c4c60c ("ASoC: wm_adsp: Signal firmware shutdown
through event control")
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The handling of bypass_val_on that was added in
regulator_get_bypass_regmap is done unconditionally however
several drivers don't define a value for bypass_val_on. This
results in those drivers reporting bypass being enabled when
it is not. In regulator_set_bypass_regmap we use bypass_mask
if bypass_val_on is zero. This patch adds similar handling in
regulator_get_bypass_regmap.
Fixes: commit dd1a571dae ("regulator: helpers: Ensure bypass register field matches ON value")
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The call to arizona_init_notifiers was only added for CODECs that are
generating voice trigger events, however, this is somewhat annoying
for machine drivers that might be used with multiple CODECs as they
need to conditionally register for the notifier, depending on the
CODEC being attached.
As the cost of initialising the notifier is so minimal, and we may
well add other events in the future that apply to more CODECs, simply
do this for all Arizona CODECs.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
These functions are very thin wrappers around core functions, so they
make sense as inline functions. Also making them inline avoids build
issues in the case where the machine driver is built in but the CODEC
is built as a module.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The core expects "const char * const" and "unsigned int" for enum
controls, the mixer control definitions in wm2200 use "const char *"
and "int". This patch corrects the type of these arrays.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Support was added to allow location of both CPU and CODEC components
of a DAI link from their parent's of_node if they did not have an
of_node themselves in this commit:
commit 3e0aa8d83b ("ASoC: core: If component doesn't have of_node
use parent's node instead")
However this leaves platforms as something of a special case as the
major DAI component that doesn't do this. Since this is useful for MFD
devices which often utilise a single device tree entry for the whole
device, add support for looking up platforms from the parent's of_node
as well.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
As almost all of the callers of the regmap_read_poll_timeout macro
will include a local ret variable we will always get a Sparse warning
about the duplication of the ret variable:
warning: symbol 'ret' shadows an earlier one
Simply rename the ret variable in the marco to pollret to make this
significantly less likely to happen.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
As the compressed ops run on the platform side of things we should
really access the driver data through the platform pointer rather than
the CODEC pointer. As the compressed DAIs in our systems always connect
our CODEC to our platform this has never been an issue, but should still
be corrected. Additionally it clears the way for future core refactoring
work.
Reported-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
It is more idiomatic to request all resources in the bus level probe,
this patch moves the request of the DSP compressed data IRQ from the
ASoC level probe into the bus level probe.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
It is more idiomatic to request all resources in the bus level probe,
this patch moves the request of the speaker thermal event IRQs from the
ASoC level probe into the bus level probe.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Don't manually enable cache_bypass for reading the ID registers they
don't have a default anyway so the first read will always hit the
hardware. The old code worked this is simply the more standard way
to implement this.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Acked-by: Brian Austin <brian.austin@cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Rather than manually enabling cache bypass when reading the ID registers
simply remove the default which will cause the first read to go to the
hardware. The old code worked this is simply the more standard way to
implement this.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Acked-by: Brian Austin <brian.austin@cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Rather than manually enabling cache bypass when reading the ID registers
simply remove the default which will cause the first read to go to the
hardware. The old code worked this is simply the more standard way to
implement this. There is a comment included in the code that claims the
chip ID register also contains the right input volume, however this is
clearly not the case from the rest of the driver. Further investigation
reveals exactly the same comment in the wm8962 driver, where this is the
case, so this is almost certainly a copy and paste error from when the
driver was created.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Acked-by: Brian Austin <brian.austin@cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Unfortunately, I seem to have missed a case where an IRQ safe spinlock was
required, in samsung_i2s_dai_remove, when I fixed up the other calls in
this patch:
316fa9e09a ("ASoC: samsung: Use IRQ safe spin lock calls")
This causes a lockdep warning when unbinding and rebinding the audio card:
[ 104.357664] CPU0 CPU1
[ 104.362174] ---- ----
[ 104.366692] lock(&(&pri_dai->spinlock)->rlock);
[ 104.371372] local_irq_disable();
[ 104.377283] lock(&(&substream->self_group.lock)->rlock);
[ 104.385259] lock(&(&pri_dai->spinlock)->rlock);
[ 104.392469] <Interrupt>
[ 104.395072] lock(&(&substream->self_group.lock)->rlock);
[ 104.400710]
[ 104.400710] *** DEADLOCK ***
Fixes: ce8bcdbb61 ("ASoC: samsung: i2s: Protect more registers with a spinlock")
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Reviewed-by: Krzysztof Kozlowski <krzk@kernel.org>
Reviewed-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Whilst ultimately we would like to move all the clocking over to the
clock framework, as an intermediate step to get people going for now
enable the source clocks for FLLs as they are powered up.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Tested-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Whilst ultimately we would like to move all the clocking over to the
clock framework, as an intermediate step to get people going for now
gating the source clocks for SYSCLK/ASYNCCLK when they are configured
to come directly from an MCLK pin.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Tested-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The Arizona CODECs will generally function correctly without a reset line
although it is strongly advised to have one, as such we do allow the system
to boot if the reset gpio is missing or incorrectly specified. However
we should fail probe if we get a probe deferral request, this patch adds
handling for this case.
Reported-by: Richard Fitzgerald <rf@opensource.wolfsonmicro.com>
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
This function is only used in a single place and no new users will be
added as all the devices other required GPIOs are already handled. As
such just merge the code back into the calling function.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
As runtime PM doesn't function whilst processing system suspend/resume
operations and the Arizona IRQ handlers need runtime PM to function
we must disable IRQs during these operations. Whilst this is
already done in the driver we are using suspend/suspend_late and
resume/resume_noirq to do so which has two problems. Firstly, as
suspend_late is before suspend_noirq that means we still have a
small window where an IRQ can cause issues. Secondly, if another
suspend_late handler fails after ours has run then (as resume_noirq
will not run) we will make unbalanced calls to enable_irq.
This is all simply fixed by using the suspend_noirq callback rather
than suspend_late. Whilst we are doing this tidy the code up a little,
and use the appropriate helper macros.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
There are some cases in arizona_dev_init, such as where we don't
recognise the chip ID, in which we head to the error path without
setting a sensible error code in ret. This would lead to the chip
silently failing probe, as it would still return 0. Fix this up by
adding appropriate sets of the return value.
Whilst adding these update the existing paths that do return an error
when the chip is not recognised to use ENODEV, which seems like a better
fit.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
The later Arizona parts do run write sequences to power up and down the
speaker path as such a delay needs to be inserted into the DAPM sequence
to allow this to run. This patch adds appropriate delays into the
existing coalesced delay scheme.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
When debugging it is useful to check the total power up/down delay that
is executed as part of the coalesced output delay. This patch adds some
debug prints for this.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The booting process for the DSP is clearly separated into two parts, the
preloader brings up the core and downloads code, then the main widget
starts the code actually executing. However the shutdown sequence is all
handled with the main widget.
To allow the preloading to be run independently of the main audio bring
up it makes sense, and is generally just cleaner, for the preloader
widget to shutdown those things it initialised. This patch moves the
appropriate parts of the shutdown process into the preloader widget.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Between when we load the DSP and when it actually starts running put the
core into a lower power state where the memory is retained but nothing
is clocked.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Replace the 1ms msleep in wm_adsp2_ena with a usleep_range, as per
normal guidance on delay functions. Also tighten up the delay a little
as 1ms was quite generous.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Currently the DSP loading is split into two widgets, the preloader that
is a snd_soc_dapm_dai_link widget which starts a thread to download
the firmware, and the DSP itself which is a snd_soc_dapm_out_drv and
synchronises the thread back in to the DAPM sequence. This allows the
firmware download to be overlapped with the rest of the path bring up.
The use of a snd_soc_dapm_dai_link widget requires the preloader to be part
of the audio path in DAPM, really a supply widget is a better fit for the
preloader. The preloader is something that needs to be done for the DSP to
function, not a part of the audio path itself.
This change makes the DSP preloader widget a supply widget, which as well
as probably being a better fit will also make it much simpler to power up
the preloader widget to trigger firmware download to the core independently
of the audio path coming up.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Currently SYSCLK is attached to every compressed DAI as this follows the
pattern of attaching clocks to the chips inputs and outputs, however, it is
really the DSP that requires the clock here. As firmware download can be a
significant part of the path startup time for these devices occasionally it
would be desirable to download the firmware in advance of the path being
brought up.
To help facilitate this early firmware loading this patch attaches the
SYSCLK to the DSP preloader widget. This also saves us adding a new route
to SYSCLK every time a new compressed DAI is created.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
As part of the work to download firmware before the audio path is brought
up the DSP will be put into a low power state between downloading firmware
to the core and starting it running. This will mean that the firmware ALSA
controls are not accessible in the hardware during this period of time.
To prepare for this change we gate access to the hardware in the ALSA
control handlers on the DSP being running rather than simply booted and
move the synchronisation of the control caches out of the preloader delayed
work and into the main DAPM thread after the DSP will have been brought out
of its low power state.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Currently the wm_adsp driver has a flag that indicates the DSP is
"running", this flag is used to gate access to the hardware. However this
flag is actually set in the firmware download thread after the firmware has
been downloaded, but this is before the core is actually started running,
so really it currently indicates that the core has been booted and is
perhaps running.
This patch clearly separates out the concepts of booted (firmware is
downloaded) and running (code is executing on the DSP) within the wm_adsp
driver.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
For best performance changing the synchroniser state whilst the FLL is
running should be avoided. As this has been done fairly regularly in
practice rather than hard preventing this, simply improve the FLL enable
sequence and give a warning if the user changes the synchroniser state.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
In preparation for future improvements allow a base to be passed to
arizona_is_enabled_fll, this will allow it to be used to check the state
of the synchroniser path as well.
Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Charles Keepax