10 Jun, 2011
1 commit
-
Currently the regulator supply implementation is somewhat complex and
fragile as it doesn't look like standard consumers but is instead a
parallel implementation. This causes issues with locking and reference
counting.Move the implementation over to using standard consumers to address this.
Rather than only notifying the supply on the first enable/disable we do so
every time the regulator is enabled or disabled, simplifying locking as we
don't need to hold a lock on the consumer we are about to enable.Signed-off-by: Mark Brown
Signed-off-by: Liam Girdwood
30 May, 2011
1 commit
-
In order to reduce the impact of ramp times rather than enabling the
regulators for a device in series use async tasks to run the actual
enables. This means that the delays which the enables implement can all
run in parallel, though it does mean that the order in which the
supplies come on may be unstable.For super bonus fun points if any of the regulators are shared between
multiple supplies on the same device (as is rather likely) then this
will test our locking. Note that in this case we only delay once for
each physical regulator so the threads shouldn't block each other while
delaying.It'd be even nicer if we could coalesce writes to a shared enable registers
in PMICs but that's definitely future work, and it may also be useful
and is certainly more achievable to optimise out the parallelism if none
of the regulators implement ramp delays.Signed-off-by: Mark Brown
Signed-off-by: Liam Girdwood
27 May, 2011
2 commits
-
Some systems, particularly physically large systems used for early
prototyping, may experience substantial voltage drops between the regulator
and the consumers as a result of long traces in the system. With these
systems voltages may need to be set higher than requested in order to
ensure reliable system operation.Allow systems to work around such hardware issues by allowing constraints
to supply an offset to be applied to any requested and reported voltages.
This is not ideal, especially since the voltage drop may be load dependant,
but is sufficient for most affected systems, it is not expected to be used
in production hardware. The offset is applied after all constraint
processing so constraints should be specified in terms of consumer values
not physically configured values.Signed-off-by: Mark Brown
Signed-off-by: Liam Girdwood -
supply_regulator_dev (using a struct pointer) has been deprecated in favour
of supply_regulator (using a regulator name) for quite a few releases
now with a warning generated if it is used and there are no current in tree
users so just remove the code.Signed-off-by: Mark Brown
Signed-off-by: Liam Girdwood
25 May, 2011
1 commit
-
The DB8500 has ePOD:s (electronic power domains) which are possible
to switch on/off to deactivate silicon blocks on the DB8500 SoC
by cutting their power without retention. We model these as simple
regulators with one bit on/off settings.Acked-by: Liam Girdwood
Acked-by: Mark Brown
Signed-off-by: Bengt Jonsson
Signed-off-by: Sundar Iyer
Signed-off-by: Jonas Aberg
Signed-off-by: Virupax Sadashivpetimath
Signed-off-by: Martin Persson
Signed-off-by: Linus Walleij
26 Mar, 2011
5 commits
-
This exposes the functionality for rise/fall fime when setting
voltage to the consumers.Cc: Bengt Jonsson
Signed-off-by: Linus Walleij
Acked-by: Mark Brown
Signed-off-by: Liam Girdwood -
This makes it possible to set the stabilization time for voltage
regulators in the same manner as enable_time(). The interface
only supports regulators that implements fixed selectors.Cc: Bengt Jonsson
Signed-off-by: Linus Walleij
Acked-by: Mark Brown
Signed-off-by: Liam Girdwood -
The regulators on the AB8500 have a lot of custom
hardware control settings pertaining to 8 external
signals, settings which are board-specific and need
be provided from the platform at startup.Initialization added for regulators Vana, VextSupply1,
VextSupply2, VextSupply3, Vaux1, Vaux2, Vaux3, VTVout,
Vintcore12, Vaudio, Vdmic, Vamic1, Vamic2, VrefDDR.Signed-off-by: Bengt Jonsson
Reviewed-by: Rickard Andersson
Reviewed-by: Jonas Aberg
Signed-off-by: Linus Walleij
Acked-by: Mark Brown
Signed-off-by: Liam Girdwood -
Signed-off-by: Bengt Jonsson
Signed-off-by: Linus Walleij
Acked-by: Mark Brown
Signed-off-by: Liam Girdwood -
The regulator core had suspend-prepare that turns off the regulators
when entering a system-wide suspend. However, it did not have
suspend-finish that pairs with suspend-prepare and the regulator core
has assumed that the regulator devices and their drivers support
autonomous recover at resume.This patch adds regulator_suspend_finish that pairs with the
previously-existed regulator_suspend_prepare. The function
regulator_suspend_finish turns on the regulators that have always_on set
or positive use_count so that we can reset the regulator states
appropriately at resume.In regulator_suspend_finish, if has_full_constraints, it disables
unnecessary regulators.Signed-off-by: MyungJoo Ham
Signed-off-by: Kyungmin Park
Acked-by: Mark Brown
--
Updates
v3
comments corrected (Thanks to Igor)
v2
disable unnecessary regulators (Thanks to Mark)
Signed-off-by: Liam Girdwood
12 Jan, 2011
7 commits
-
We only expose the use and open counts to userspace, providing a tiny
bit of insight into what the API is up to.Signed-off-by: Mark Brown
Signed-off-by: Liam Girdwood -
When cooperating with an external control source the regulator setup
may be changed underneath the API. Currently consumers can just redo
the regulator_set_voltage() to restore a previously set configuration
but provide an explicit API for doing this as optimsations in the
regulator_set_voltage() implementation will shortly prevent that.Signed-off-by: Mark Brown
Signed-off-by: Liam Girdwood -
Many regulator drivers implement voltage setting by looping through a
table of possible values, normally because the set of available voltages
can't be mapped onto selectors with simple calcuation. Factor out these
loops by providing a variant of set_voltage() which takes a selector rather
than a voltage range as an argument and implementing a loop through the
available selectors in the core.This is not going to be suitable for use with all devices as when the
regulator voltage can be mapped onto selector values with a simple
calculation the linear scan through the available values will be more
expensive than just doing the calculation, especially for regulators
that provide fine grained voltage control.Signed-off-by: Mark Brown
Signed-off-by: Liam Girdwood -
The define for number of regulators is moved from ab8500-core to
ab8500-regulator so that the regulator driver can be updated
independently of ab8500-core. This also changes the platform
configuration structure of ab8500-core so that it contains a
pointer to the regulator_init_data array plus number of
regulators instead of an fixed size array of pointers to
regulator_init_data.Signed-off-by: Bengt Jonsson
Acked-by: Linus Walleij
Acked-by: Mark Brown
Signed-off-by: Liam Girdwood -
Since drivers already have to provide an API for translating selectors
into voltages they may as well just report the selector values directly
to the core API rather than implement the lookup themselves. The old
interface is left in place for now, but may be removed in future.Signed-off-by: Mark Brown
Signed-off-by: Liam Girdwood -
Currently the regulator API uses the constraints structure passed in to
the core throughout the lifetime of the object. This means that it is not
possible to mark the constraints as __initdata so if the kernel supports
many boards the constraints for all of them are kept around throughout the
lifetime of the system, consuming memory needlessly. By copying constraints
that are actually used we allow the use of __initdata, saving memory when
multiple boards are supported.This also means the constraints can be const.
Signed-off-by: Mark Brown
Signed-off-by: Liam Girdwood -
Change the interface used by set_voltage() to report the selected value
to the regulator core in terms of a selector used by list_voltage().
This allows the regulator core to know the voltage that was chosen
without having to do an explict get_voltage(), which would be much more
expensive as it will generally access hardware.Signed-off-by: Mark Brown
Signed-off-by: Liam Girdwood
29 Oct, 2010
3 commits
-
Allow machine drivers to explicitly enable the use of the dummy regulator,
enabling simpler support for systems with only a few specific supplies
visible to software.It is strongly recommended that this is not used on systems with
substantial software control over their PMICs, for maximum functionality
constrints should be as fully specified as possible.Signed-off-by: Mark Brown
Signed-off-by: Liam Girdwood -
This patch adds regulator drivers for National Semiconductors LP3972 PMIC.
This LP3972 PMIC controller has 3 DC/DC voltage converters and 5 low drop-out
(LDO) regulators. LP3972 PMIC controller uses I2C interface.Signed-off-by: Axel Lin
Acked-by: Mark Brown
Signed-off-by: Liam Girdwood -
MAX8952 PMIC is used to provide voltage output between 770mV - 1400mV
with DVS support. In this initial release, users can set voltages for
four DVS modes, RAMP delay values, and SYNC frequency.
Controlling FPWM/SYNC_MODE/Pull-Down/Ramp Modes and reading CHIP_ID
is not supported in this release.If GPIO of EN is not valid in platform data, the driver assumes that it
is always-on. If GPIO of VID0 or VID1 is invalid, the driver pulls down
VID0 and VID1 to fix DVS mode as 0 and disables DVS support.We assume that V_OUT is capable to provide every voltage from 770mV to
1.40V in 10mV steps although the data sheet has some ambiguity on it.Signed-off-by: MyungJoo Ham
Signed-off-by: Kyungmin Park
Acked-by: Mark Brown
--
v2:
- Style correction
- Can accept platform_data with invalid GPIOs
- Removed unnecessary features
- Improved error handling
Signed-off-by: Liam Girdwood
11 Aug, 2010
1 commit
-
Acked-by: Linus Walleij
Acked-By: Mattias Wallin
Acked-By: Bengt JONSSON
Signed-off-by: Sundar R Iyer
Acked-by: Mark Brown
Acked-by: Samuel Ortiz
Signed-off-by: Liam Girdwood
28 Jul, 2010
1 commit
-
Acked-by: Mark Brown
In TPS6507x, depending on the status of DEFDCDC{2,3} pin either
DEFDCDC{2,3}_LOW or DEFDCDC{2,3}_HIGH register needs to be read or
programmed to change the output voltage.The current driver assumes DEFDCDC{2,3} pins are always tied low
and thus operates only on DEFDCDC{2,3}_LOW register. This need
not always be the case (as is found on OMAP-L138 EVM).Unfortunately, software cannot read the status of DEFDCDC{2,3} pins.
So, this information is passed through platform data depending on
how the board is wired.Signed-off-by: Anuj Aggarwal
Signed-off-by: Sekhar Nori
Signed-off-by: Liam Girdwood
25 May, 2010
1 commit
-
When one regulator supplies another allow the relationship to be specified
using names rather than struct regulators, in a similar manner to that
allowed for consumer supplies. This allows static configuration at compile
time, reducing the need for dynamic init code.Also change the references to LINE supply to be system supply since line
is sometimes used for actual supplies and therefore potentially confusing.Signed-off-by: Mark Brown
Signed-off-by: Liam Girdwood
19 Apr, 2010
1 commit
-
Have the stub variant of regulator_get() return NULL, so that drivers
can (but still don't have to) handle this case specifically.Signed-off-by: Jean Delvare
Cc: Mark Brown
Cc: Jerome Oufella
Acked-by: Mark Brown
Signed-off-by: Liam Girdwood
03 Mar, 2010
4 commits
-
Signed-off-by: Haojian Zhuang
Acked-by: Mark Brown
Signed-off-by: Liam Girdwood -
Add a field to specify a delay for the start-up time of
a fixed voltage regulator.Signed-off-by: Adrian Hunter
Acked-by: Mark Brown
Signed-off-by: Liam Girdwood -
Regulators may sometimes take longer to enable than the control operation
used to do so, either because the regulator has ramp rate control used to
limit inrush current or because the control operation is very fast (GPIO
being the most common example of this). In order to ensure that consumers
do not rely on the regulator before it is enabled provide an enable_time()
operation and have the core delay for that time before returning to the
caller.This is implemented as a function since the ramp rate may be specified in
voltage per unit time and therefore the time depend on the configuration.
In future it would be desirable to allow the bulk operations to run the
delays for multiple enables in parallel but this is not currently supported.Signed-off-by: Mark Brown
-
The intended use case is for drivers which disable regulators to save
power but need to do some work to restore the hardware state when
restarting. If the supplies are not actually disabled due to board
limits or sharing with other active devices this notifier allows the
driver to avoid unneeded reinitialisation, particularly when used with
runtime PM.Signed-off-by: Mark Brown
17 Dec, 2009
3 commits
-
consumer.h requires device.h for stand alone build.
Signed-off-by: Liam Girdwood
-
Since some regulators in the system may not support suspend mode
configuration we need to allow some regulators to have a missing
suspend mode configuration. Do this by requiring that disabled
regulators are explicitly flagged and then skip over regulators
that have no state specified.Try to avoid surprises by warning the if we could set the state
but no configuration is provided. This also ensures that an all
zeros configuration generates a warning rather than silently
disabling the regulator.Reported-by: Joonyoung Shim
Signed-off-by: Mark Brown
Signed-off-by: Liam Girdwood -
Tested with a MX25-based custom board.
Signed-off-by: Wolfram Sang
Acked-by: Mark Brown
Signed-off-by: Liam Girdwood
22 Sep, 2009
8 commits
-
Document the possibility that is_enabled may also return with negative
errorcodes.Signed-off-by: Wolfram Sang
Acked-by: Mark Brown
Signed-off-by: Liam Girdwood -
Fix a couple of typos I found while working with this subsystem.
Signed-off-by: Wolfram Sang
Acked-by: Mark Brown
Signed-off-by: Liam Girdwood -
Now fixed regulators that have their enable pin connected to a GPIO line
can use the fixed regulator driver for regulator enable/disable control.
The GPIO number and polarity information is passed through platform data.
GPIO enable control is achieved using gpiolib.Signed-off-by: Roger Quadros
Reviewed-by: Philipp Zabel
Reviewed-by: Felipe Balbi
Acked-by: Mark Brown
Signed-off-by: Liam Girdwood -
Simplify checking of support for voltage ranges by providing an API which
wraps the existing count and list operations.Signed-off-by: Mark Brown
Signed-off-by: Liam Girdwood -
Some consumers require complete control of the regulator and can't
tolerate sharing it with other consumers, most commonly because they need
to have the regulator actually disabled so can't have other consumers
forcing it on. This new regulator_get_exclusive() API call allows these
consumers to explicitly request this, documenting the assumptions that
they are making.In order to simplify coding of such consumers the use count for regulators
they request is forced to match the enabled state of the regulator when
it is requested. This is not possible for consumers which can share
regulators due to the need to keep track of the ownership of use counts.A new API call is used rather than an additional argument to the existing
regulator_get() in order to avoid merge headaches with driver code in
other trees.Signed-off-by: Mark Brown
Signed-off-by: Liam Girdwood -
Signed-off-by: Haojian Zhuang
Acked-by: Mark Brown
Signed-off-by: Liam Girdwood -
This allows machine drivers to build without ifdefs if they have
full constraints. Suggested by machine drivers contributed by
Haojian Zhuang .Signed-off-by: Mark Brown
Signed-off-by: Liam Girdwood -
Follow the approach suggested by Russell King and implemented by him in
the clkdev API and allow consumer device supply mapings to be set up
using the dev_name() for the consumer instead of the struct device.
In order to avoid making existing machines instabuggy and creating merge
issues the use of struct device is still supported for the time being.This resolves problems working with buses such as I2C which make the
struct device available late providing that the final device name is
known, which is the case for most embedded systems with fixed setups.Consumers must still use the struct device when calling regulator_get().
Signed-off-by: Mark Brown
Signed-off-by: Liam Girdwood
17 Sep, 2009
1 commit
-
This is useful for implementing get_status() in terms of get_mode().
Signed-off-by: Mark Brown
Signed-off-by: Samuel Ortiz