A previous patch added the ac_draw_failed goto and label to
pda_power_probe(). The goto would be invoked after a failed call to
regulator_get().

However the way ac_draw is used - always after a check for NULL - suggests
that a failed call to regulator_get() was not fatal.

This patch removes the ac_draw_failed goto and label, partly reverting the
previous patch.

This patch also removes the assignment of an error code to ret after a
failed call to regulator_get(), since the error code is now never used.
Signed-off-by: NPaul Parsons <lost.distance@yahoo.com>
Cc: Philipp Zabel <philipp.zabel@gmail.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

This patch add support sysfs entry for each charger(regulator).
Charger-manager use one or more chargers for charging battery but some
charger isn't necessary on specific scenario. So, if some charger isn't
needed, can disable specific charger through 'externally_control' entry
while system is on state and confirm the information(name, state) of
charger.

The list of added sysfs entry
- /sys/class/power_supply/battery/chargers/charger.[index]/name
  show name of charger(regulator)
- /sys/class/power_supply/battery/chargers/charger.[index]/state
  show either enabled or disabled state of charger
- /sys/class/power_supply/battery/chargers/charger.[index]/externally_control

If 'externally_control' of specific charger is 1, Charger-manager cannot
enable regulator for charging when charger cable is attached and charger
must be maintained with disabled state. If 'externally_control' is zero,
Charger-manager usually can control to enable/disable regulator.
Signed-off-by: NChanwoo Choi <cw00.choi@samsung.com>
Signed-off-by: NMyungjoo Ham <myungjoo.ham@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

This patch check maximum possible duration of charging/discharging.

If whole charging duration exceed 'desc->charging_max_duration_ms', cm
stop charging to prevent overcharge/overheat. And if discharging duration
exceed, charger cable is attached, after full-batt, cm start charging to
maintain fully charged state for battery.
Signed-off-by: NChanwoo Choi <cw00.choi@samsung.com>
Signed-off-by: NMyungjoo Ham <myungjoo.ham@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

This patch check periodically fully charged state of battery to protect
overcharge and overheat. If battery is fully charged, stop charging and
check droped voltage with 'fullbatt_vchkdrop_ms' period. When voltage of
battery is more droped than 'fullbatt_vchkdrop_uV' voltage,
charger-manager will restart charging for battery.
Signed-off-by: NChanwoo Choi <cw00.choi@samsung.com>
Signed-off-by: NMyungjoo Ham <myungjoo.ham@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

This is really minor, but it improves the readability.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

Fix the warning on MODULE_AUTHOR.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

Add description and relocate data.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

Declare a variable at one line and align lines.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

Declare a variable at one line. Just return when no charger exists to make
code simpler.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

If the charger is pulled out, just return as DISCHARGING. Then no need for
additional 'else' statement.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

Charger has only one valid property - ONLINE. If the property is not
ONLINE, then just return quickly.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

Just return with lp8727_write_byte(), no need to check its value.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

Change return type to boolean. Remove unnecessary check routine for NULL
string. (Power supply name is always valid when the function is executed.)
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

Add new LP8727_ICHG_SHIFT definition and replace a magic number. Reuse
definition for the size of interrupt register buffer.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

All definitions should be unique, since they're in the gloabl namespace.
So the prefix LP8727_ are added. Additionally, use BIT() macro for bit
masks. Remove unnecessary definitions such as SW_DM1_U1 and SW_DP2_U2.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

Enum lp8727_chg_state can be removed because only one charger status is
used - EOC(End Of Charge).

To check whether the EOC is reached or not, use simple comparison rather
than shift-operation.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

For better understanding, use specific function and definitions rather
than magic numbers. New enum type for die temperature is matched with hex
codes. Specific temperature names are better than 0x1, 0x2 and 0x3.

And lp8727_is_high_temperature() function has better readability than
comparing raw register values.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

To initialize the device, previous interrupts need to be cleared while
loading the driver.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

For better understanding, function name is changed. (lp8727_intr_config()
is replaced with lp8727_setup_irq().)

The private IRQ number is set when the IRQ is allocated successfully. This
data is used for releasing the IRQ on unloading the driver. Even the IRQ
number is not defined, the driver should be operated. In this case, just
return as 0.

In additional function lp8727_release_irq(), the workqueue is canceled and
the allocated IRQ is released.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

LP8727 has two IRQ threads. One is the I2C HW IRQ pin, the other is for
delayed interrupt processing. But this delayed processing can be handled
without additional single thread by using schedule_delayed_work() with
jiffies time value.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

Debounce time is configurable in the platform side. If it is not defined,
the default value is 270ms.

Platform data is msec unit, and this time is converted to jiffies
internally. The workqueue uses this jiffies time in the interrupt
handling. So debounce_jiffies is added in the private data.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

LP8727 platform data is optional, so the driver should work even the
platform data is NULL.

To check the platform data, charging parameter data should be changed to
the pointer type.

Fix NULL point access problem when getting the battery properties. When
the data is NULL, just return as invalid value.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

If the lp8727_register_psy() gets failed, registered interrupt handler
should be freed, but this is not complete solution. It has still problem.
Assume that the IRQ occurs while unregistering power supply devices. Then
the ISR will access to freed IRQ.

From Anton's opinion, it can be resolved if re-ordering the call sequence.

Register power supplies first, then create interrupt handler. Then no need
to free the IRQ in _probe(). Additionally goto statements can be removed
because those can be replaced with return statements.

The _remove() should be changed the sequence - in reverse order of
_probe()
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

Use devm_kzalloc() rather than kzalloc()/kfree() to make
allocating/freeing the private data simpler.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

Generated by: scripts/coccinelle/misc/irqf_oneshot.cocci

Make sure threaded IRQs without a primary handler are always request with IRQF_ONESHOT
Signed-off-by: NFengguang Wu <fengguang.wu@intel.com>
Acked-by: NFelipe Balbi <balbi@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

Generated by: scripts/coccinelle/misc/irqf_oneshot.cocci

ERROR: Threaded IRQ with no primary handler requested without IRQF_ONESHOT

Make sure threaded IRQs without a primary handler are always request
with IRQF_ONESHOT
Signed-off-by: NFengguang Wu <fengguang.wu@intel.com>
Acked-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

Sparse complains:

  CHECK   drivers/power/88pm860x_battery.c
drivers/power/88pm860x_battery.c:128:5: warning: symbol 'array_soc' was not declared. Should it be static?
  CHECK   drivers/power/88pm860x_charger.c
drivers/power/88pm860x_charger.c:640:3: warning: symbol 'pm860x_irq_descs' was not declared. Should it be static?
  CHECK   drivers/mfd/88pm860x-core.c
drivers/mfd/88pm860x-core.c:803:53: warning: incorrect type in assignment (different base types)
drivers/mfd/88pm860x-core.c:803:53:    expected struct charger_regulator *charger_regulators
drivers/mfd/88pm860x-core.c:803:53:    got struct regulator_bulk_data static [toplevel] *

The issues are minor, except for the last one. We seemed to use
'regulator_bulk_data' struct (just as charger manager documentation
wrongly tells us), but in real it should have been
'struct charger_regulator'. The only reason that it worked is
because both 'supply' and 'regulator_name' struct members are the
first in these structs. :-)
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

There are charger and battery measurement feature for 88pm860x PMIC.

For charger, it can support pre-charge with small current when battery is
nearly exausted and then changed into fast-charge with CC&CV mode.

For battery monitor, it can support battery measurement such as
vbat,vsys,vchg and ibat etc,it can aslo accumulating the Coulomb value
charged or discharged from battery based on Conlomb Counter, we use it
to estimate battery capacity.
Signed-off-by: NJett.Zhou <jtzhou@marvell.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

TI LP8788 PMU supports regulators, battery charger, RTC, ADC, backlight
driver and current sinks.

This patch supports the charger operations including the charger
interrupt handling.

The LP8788 charger driver provides configurable platform data.  The
charger platform data includes ADC input ID, maximum battery voltage,
charging parameters and etc.

The ADC input is used for getting the battery voltage and temperature.
Charging parameters are used for updating the charger operations such
like setting charging current and setting End-of-charge conditions.

When the charging interrupt occurs, power supply uevents are generated
in order to update the user-space information.  For platform specific
charger action, the charger_event() function can be used in the
platform side.
Signed-off-by: NMilo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

We should decrement "i" before doing the free_irq().  If we call this
because request_threaded_irq() failed then we don't want to free the
thing which failed.  Or in the case where we get here because
power_supply_register() failed then the original codes does a read past
the end of the array.
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

Turns out this driver doesn't actually try talking to the device at
probe time, so if it's incorrectly configured in the device tree or
platform data (or if the battery has been removed from the system),
then probe will succeed and every access will sit there and time out.
The end result is a possibly laggy system that thinks it has a battery
but can never read status, which isn't very useful.

Instead, just read any register (I chose status) at probe, and if that
fails, don't register the device.
Signed-off-by: NOlof Johansson <olof@lixom.net>
Acked-by: NRhyland Klein <rklein@nvidia.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

This patch checks for charger status register for determining the
battery charging status and reports Discharing/Charging/Not Charging/Full
accordingly.

This patch also adds the interrupt support for Safety Timer Expiration.
This interrupt is helpful in debugging the cause for charger fault.
Signed-off-by: NRamakrishna Pallala <ramakrishna.pallala@intel.com>
Acked-by: NMika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

When I reboot my iPAQ hx4700 in its cradle, the battery will not begin
to charge even though the AC supply is connected. Charging will start
only after the PDA power driver is tickled by some other power event,
such as reseating the iPAQ in its cradle or connecting the USB cable.

The problem lies in pda_power_probe(), where ac_draw is used by the call
to update_charger() before being set by the call to regulator_get().

Moving the regulator_get() call to before the update_charger() call
fixes the problem.
Signed-off-by: NPaul Parsons <lost.distance@yahoo.com>
Cc: Philipp Zabel <philipp.zabel@gmail.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

Currently the MFD core supports remapping MFD cell interrupts using an
irqdomain but only if the MFD is being instantiated using device tree
and only if the device tree bindings use the pattern of registering IPs
in the device tree with compatible properties.  This will be actively
harmful for drivers which support non-DT platforms and use this pattern
for their DT bindings as it will mean that the core will silently change
remapping behaviour and it is also limiting for drivers which don't do
DT with this particular pattern.  There is also a potential fragility if
there are interrupts not associated with MFD cells and all the cells are
omitted from the device tree for some reason.

Instead change the code to take an IRQ domain as an optional argument,
allowing drivers to take the decision about the parent domain for their
interrupts.  The one current user of this feature is ab8500-core, it has
the domain lookup pushed out into the driver.
Signed-off-by: NMark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: NSamuel Ortiz <sameo@linux.intel.com>

There are many reports (including 2 of my machines) that iTCO_wdt watchdog
driver fails to be initialized in 3.5 kernel with error message like:

[    5.265175] ACPI Warning: 0x00001060-0x0000107f SystemIO conflicts with Region \_SB_.PCI0.LPCB.TCOI 1 (20120320/utaddress-251)
[    5.265192] ACPI: If an ACPI driver is available for this device, you should use it instead of the native driver
[    5.265206] lpc_ich: Resource conflict(s) found affecting iTCO_wdt

The root cause the iTCO_wdt driver in 3.4 probes the HW IO resource from
LPC's PCI config space, while in 3.5 kernel it relies on lpc_ich driver
for the probe, which adds a new acpi_check_resource_conflict() check, and
give up the probe if there is any conflict with ACPI.

Fix it by removing all the checks for iTCO_wdt to keep the same behavior as
3.4 kernel.
https://bugzilla.kernel.org/show_bug.cgi?id=44991

Actually the same check could be removed for the gpio-ich in lpc_ich.c,
but I'm not sure if it will cause problems.
Signed-off-by: NFeng Tang <feng.tang@intel.com>
Cc: Aaron Sierra <asierra@xes-inc.com>
Cc: Wim Van Sebroeck <wim@iguana.be>
Cc: Len Brown <len.brown@intel.com>
Cc: Bob Moore <robert.moore@intel.com>
Signed-off-by: NSamuel Ortiz <sameo@linux.intel.com>

There are different types of chargers avalibale like AC, Solar, USB,
etc..  Even in USB we have different types SDP/DCP/CDP/ACA and all these
chargers have different o/p ratings. For example SDP supports only 500mA
of charge current whereas AC charger can support upto 8A or more.

Similarly batteries also come with charge current and voltage ratings
and these ratings vary depending on its capacity and the technology
used.

This patch adds two new power supply properties
CONSTANT_CHARGE_CURRENT_MAX and CONSTANT_CHARGE_CURRENT_MAX.
Signed-off-by: NRamakrishna Pallala <ramakrishna.pallala@intel.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

This patch remove unnecessary variable(cm->fullbatt_vchk_uV) by using
'desc->fullbatt_uV' field directly in fullbatt_handler() function
to check the state of battery.
Signed-off-by: NChanwoo Choi <cw00.choi@samsung.com>
Signed-off-by: NMyungjoo Ham <myungjoo.ham@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>

This patch disable regulator for charging when charger cable is detached
before stopping charging forcibly on abnormal battery state and check
return value of regulator_enable/disable() function to confirm correct
operation of enabling or disabling regulator for charging.
Signed-off-by: NChanwoo Choi <cw00.choi@samsung.com>
Signed-off-by: NMyungjoo Ham <myungjoo.ham@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: NAnton Vorontsov <anton.vorontsov@linaro.org>