When searching for the trip points that need to be set, the nearest
higher trip point's temperature is used for the high trip, while the
nearest lower trip point's temperature minus the hysteresis is used for
the low trip. The issue with this logic is that when the current
temperature is inside a trip point's hysteresis range, both high and low
trips will come from the same trip point. As a consequence instability
can still occur like this:
* the temperature rises slightly and enters the hysteresis range of a
  trip point
* polling happens and updates the trip points to the hysteresis range
* the temperature falls slightly, exiting the hysteresis range, crossing
  the trip point and triggering an IRQ, the trip points are updated
* repeat

So even though the current hysteresis implementation prevents
instability from happening due to IRQs triggering on the same
temperature value, both ways, it doesn't prevent it from happening due
to an IRQ on one way and polling on the other.

To properly implement a hysteresis behavior, when inside the hysteresis
range, don't update the trip points. This way, the previously set trip
points will stay in effect, which will in a way remember the previous
state (if the temperature signal came from above or below the range) and
therefore have the right trip point already set.

The exception is if there was no previous trip point set, in which case
a previous state doesn't exist, and so it's sensible to allow the
hysteresis range as trip points.

The following logs show the current behavior when running on a real
machine:

[  202.524658] thermal thermal_zone0: new temperature boundaries: -2147483647 < x < 40000
   203.562817: thermal_temperature: thermal_zone=vpu0-thermal id=0 temp_prev=36986 temp=37979
[  203.562845] thermal thermal_zone0: new temperature boundaries: 37000 < x < 40000
   204.176059: thermal_temperature: thermal_zone=vpu0-thermal id=0 temp_prev=37979 temp=40028
[  204.176089] thermal thermal_zone0: new temperature boundaries: 37000 < x < 100000
   205.226813: thermal_temperature: thermal_zone=vpu0-thermal id=0 temp_prev=40028 temp=38652
[  205.226842] thermal thermal_zone0: new temperature boundaries: 37000 < x < 40000

And with this patch applied:

[  184.933415] thermal thermal_zone0: new temperature boundaries: -2147483647 < x < 40000
   185.981182: thermal_temperature: thermal_zone=vpu0-thermal id=0 temp_prev=36986 temp=37872
   186.744685: thermal_temperature: thermal_zone=vpu0-thermal id=0 temp_prev=37872 temp=40058
[  186.744716] thermal thermal_zone0: new temperature boundaries: 37000 < x < 100000
   187.773284: thermal_temperature: thermal_zone=vpu0-thermal id=0 temp_prev=40058 temp=38698

Fixes: 060c034a ("thermal: Add support for hardware-tracked trip points")
Signed-off-by: Nícolas F. R. A. Prado <nfraprado@collabora.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Co-developed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

Modify the governor .throttle() callback definition so that it takes a
trip pointer instead of a trip index as its second argument, adjust the
governors accordingly and update the core code invoking .throttle().

This causes the governors to become independent of the representation
of the list of trips in the thermal zone structure.

This change is not expected to alter the general functionality.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>

Fold update_passive_instance() into thermal_zone_trip_update() that is
its only caller so as to make the code in question easier to follow.

No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>

Modify the power allocator thermal governor to use trip pointers instead
of trip indices everywhere except for the power_allocator_throttle()
second argument that will be changed subsequently along with the
definition of the .throttle() governor callback.

The general functionality is not expected to be changed.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>

Make get_trip_level() use for_each_trip() to iterate over trip points
and make it call thermal_zone_trip_id() to obtain the integer ID of a
given trip point so as to avoid relying on the knowledge of struct
thermal_zone_device internals.

The general functionality is not expected to be changed.

This change causes the governor to use trip pointers instead of trip
indices everywhere except for the fair_share_throttle() second argument
that will be modified subsequently along with the definition of the
governor .throttle() callback.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>

Define a new macro for_each_trip() to be used by the thermal core code
and thermal governors for walking trips in a given thermal zone.

Modify for_each_thermal_trip() to use this macro instead of an open-
coded loop over trips.

No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>

A trip index can be computed right away as a difference between the
value of a trip pointer pointing to the given trip object and the
start of the trips[] table in the given thermal zone, so change
thermal_zone_trip_id() accordingly.

No intentional functional impact (except for some speedup).
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>

The ACPI thermal driver changes include some thermal core modifications
that are depended on by subsequent thermal core changes, so merge them.

* acpi-thermal: (26 commits)
  thermal: trip: Drop lockdep assertion from thermal_zone_trip_id()
  thermal: trip: Remove lockdep assertion from for_each_thermal_trip()
  thermal: core: Drop thermal_zone_device_exec()
  ACPI: thermal: Use thermal_zone_for_each_trip() for updating trips
  ACPI: thermal: Combine passive and active trip update functions
  ACPI: thermal: Move get_active_temp()
  ACPI: thermal: Fix up function header formatting in two places
  ACPI: thermal: Drop list of device ACPI handles from struct acpi_thermal
  ACPI: thermal: Rename structure fields holding temperature in deci-Kelvin
  ACPI: thermal: Drop critical_valid and hot_valid trip flags
  ACPI: thermal: Do not use trip indices for cooling device binding
  ACPI: thermal: Mark uninitialized active trips as invalid
  ACPI: thermal: Merge trip initialization functions
  ACPI: thermal: Collapse trip devices update function wrappers
  ACPI: thermal: Collapse trip devices update functions
  ACPI: thermal: Add device list to struct acpi_thermal_trip
  ACPI: thermal: Fix a small leak in acpi_thermal_add()
  ACPI: thermal: Drop valid flag from struct acpi_thermal_trip
  ACPI: thermal: Drop redundant trip point flags
  ACPI: thermal: Untangle initialization and updates of active trips
  ...

The lockdep assertion in thermal_zone_trip_id() triggers when the
trip point sysfs attribute of a thermal instance is read, because
there is no thermal zone locking in that code path.

This is not verly useful, though, because there is no mechanism by which
the location of the trips[] table in a thermal zone or its size can
change after binding cooling devices to the trips in that thermal
zone and before those cooling devices are unbound from them.  Thus
it is not in fact necessary to hold the thermal zone lock when
thermal_zone_trip_id() is called from trip_point_show() and so the
lockdep asserion in the former is invalid.

Accordingly, drop that lockdep assertion.

Fixes: 2c7b4bfa ("thermal: core: Store trip pointer in struct thermal_instance")
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

The dev->id value comes from ida_alloc() so it's a number between zero
and INT_MAX.  If it's too high then these sprintf()s will overflow.

Fixes: 203d3d4a ("the generic thermal sysfs driver")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

Become designated reviewer and help thermal subsystem in development
process.
Signed-off-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

Amit Kucheria has not been participating in kernel development in any
way or form for quite some time, so it is not useful to list him as a
designated reviewer for the thermal subsystem or as the thermal zone DT
binding maintainer.

Remove him from the THERMAL entry in MAINTAINERS and list Daniel Lezcano
as the new thermal zone DT binding maintainer.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Acked-by: Amit Kucheria <amitk@kernel.org>

The lockdep assertion in for_each_thermal_trip() was added to possibly
catch incorrect usage of that function without the thermal zone lock.
However, it turns out that the ACPI thermal driver has a legitimate
reason to call for_each_thermal_trip() without locking.

Namely, it is called by acpi_thermal_bind_unbind_cdev() in the thermal
zone registration and unregistration paths.  That function cannot acquire
the thermal zone lock by itself, because it calls functions that acquire
it, thermal_bind_cdev_to_trip() or thermal_unbind_cdev_from_trip().
However, it is invoked when the ACPI notify handler for the thermal
zone in question has not been registered yet (in the registration path)
or after that handler has been unregistered (in the unregistration
path).  Therefore, when for_each_thermal_trip() is called by
acpi_thermal_bind_unbind_cdev(), thermal trip changes induced by the
platform firmware cannot take place and so the thermal zone's trips[]
table is effectively immutable.  Hence, it is valid to call
for_each_thermal_trip() from acpi_thermal_bind_unbind_cdev() without
locking and the lockdep assertion in the former is in fact incorrect, so
remove it.

Fixes: d5ea8892 ("ACPI: thermal: Do not use trip indices for cooling device binding")
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

Because thermal_zone_device_exec() has no users any more and there are
no plans to use it anywhere, revert commit 9a99a996 ("thermal: core:
Introduce thermal_zone_device_exec()") that introduced it.

No functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>

Rearrange the code handling notifications from the platform firmware
regarding trip point updates to carry out one loop over trip points
instead of two of them by using thermal_zone_for_each_trip() for that,
which is more straightforward than using a combination of
thermal_zone_device_exec() and for_each_thermal_trip(), each with its
own callback function.

No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>

Combine acpi_thermal_update_passive_trip() and
acpi_thermal_update_active_trip() into one common function called
acpi_thermal_update_trip(), so as to reduce code duplication and
prepare the code in question for subsequent changes.

No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>

Put the get_active_temp() function next to the analogous
get_passive_temp() one to allow subsequent changes to be easier to
follow.

No functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>

Add a wrapper around for_each_thermal_trip(), called
thermal_zone_for_each_trip(), that will invoke the former under the
thermal zone lock and pass its return value to the caller.

Two drivers will be modified subsequently to use this new function.

No functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>

Fix up the following formatting issues:
 * braces following function declarations should be on a new line
 * empty line should be present between function declarations
Signed-off-by: Jonathan Bergh <bergh.jonathan@gmail.com>
[ rjw: Subject and changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

Notice that the list of device ACPI handles in struct acpi_thermal is not
used and drop it.

No functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>

Rename structure fields holding temperature values in deci-Kelvin so as
to avoid temperature units confusion.

No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>

The critical_valid and hot_valid flags in struct acpi_thermal_trips are
only used during initialization and they are only false if the
corresponding trip temperatures are equal to THERMAL_TEMP_INVALID, so
drop them and use THERMAL_TEMP_INVALID checks instead of them where
applicable.

No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>

Rearrange the ACPI thermal driver's callback functions used for cooling
device binding and unbinding, acpi_thermal_bind_cooling_device() and
acpi_thermal_unbind_cooling_device(), respectively, so that they use trip
pointers instead of trip indices which is more straightforward and allows
the driver to become independent of the ordering of trips in the thermal
zone structure.

The general functionality is not expected to be changed.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>

Merge thermal core changes for v6.7 depended on by subsequent ACPI
thermal driver changes:

 - Drop trips_disabled bitmask that is not really useful any more from
   the thermal zone structure.

 - Drop a reduntant NULL pointer check from for_each_thermal_trip().

 - Redefine struct thermal_instance to hold a thernal trip pointer
   instead of a trip index.

 - Add helpers for binding and unbinding cooling devices to and from
   trip points, reslectively, that take trip pointers as arguments.

* thermal-core:
  thermal: core: Allow trip pointers to be used for cooling device binding
  thermal: core: Store trip pointer in struct thermal_instance
  thermal: trip: Drop redundant trips check from for_each_thermal_trip()
  thermal: core: Drop trips_disabled bitmask

After removing the valid flag from struct acpi_thermal_trip, the trip
temperature value is used in validity checks, so it must be
THERMAL_TEMP_INVALID for the active trip entries in struct acpi_thermal_trips
that are not going to be used (because the corresponding objects are not
present in the ACPI tables, for example).

Accordingly, modify acpi_thermal_get_trip_points() to set the temperature
value to THERMAL_TEMP_INVALID for trip point entries skipped by it after
acpi_thermal_init_trip() has returned 'false' for an active trip.

Fixes: 058f5e40 ("ACPI: thermal: Drop valid flag from struct acpi_thermal_trip")
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

Add new helper functions, thermal_bind_cdev_to_trip() and
thermal_unbind_cdev_from_trip(), to allow a trip pointer to be used for
binding a cooling device to a trip point and unbinding it, respectively,
and redefine the existing helpers, thermal_zone_bind_cooling_device()
and thermal_zone_unbind_cooling_device(), as wrappers around the new
ones, respectively.

No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>

Replace the integer trip number stored in struct thermal_instance with
a pointer to the relevant trip and adjust the code using the structure
in question accordingly.

The main reason for making this change is to allow the trip point to
cooling device binding code more straightforward, as illustrated by
subsequent modifications of the ACPI thermal driver, but it also helps
to clarify the overall design and allows the governor code overhead to
be reduced (through subsequent modifications).

The only case in which it adds complexity is trip_point_show() that
needs to walk the trips[] table to find the index of the given trip
point, but this is not a critical path and the interface that
trip_point_show() belongs to is problematic anyway (for instance, it
doesn't cover the case when the same cooling devices is associated
with multiple trip points).

This is a preliminary change and the affected code will be refined by
a series of subsequent modifications of thermal governors, the core and
the ACPI thermal driver.

The general functionality is not expected to be affected by this change.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>

In order to reduce code duplicationeve further, merge
acpi_thermal_init_passive/active_trip() into one function called
acpi_thermal_init_trip() that will be used for initializing both
the passive and active trip points.

No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>

In order to reduce code duplicationeve further, merge
acpi_thermal_update_passive/active_devices() into one function
called acpi_thermal_update_trip_devices() that will be used for
updating both the passive and active trip points.

No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>

In order to reduce code duplication, merge update_passive_devices() and
update_active_devices() into one function called update_trip_devices()
that will be used for updating both the passive and active trip points.

No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>

The device lists present in struct acpi_thermal_passive and struct
acpi_thermal_active can be located in struct acpi_thermal_trip which
then will allow the same code to be used for handling both the passive
and active trip points, so make that change.

No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>

Free "tz" if the "trip" allocation fails.

Fixes: 5fc2189f9335 ("ACPI: thermal: Create and populate trip points table earlier")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

It is invalid to call for_each_thermal_trip() on an unregistered thermal
zone anyway, and as per thermal_zone_device_register_with_trips(), the
trips[] table must be present if num_trips is greater than zero for the
given thermal zone.

Hence, the trips check in for_each_thermal_trip() is redundant and so it
can be dropped.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>

Notice that the valid flag in struct acpi_thermal_trip is in fact
redundant, because the temperature field of invalid trips is always
equal to THERMAL_TEMP_INVALID, so drop it from there and adjust the
code accordingly.

No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>

Trip point flags previously used by the driver need not be used any more
after the preceding changes, so drop them and adjust the code accordingly.

No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>

Separate the code needed to update active trips (in a response to a
notification from the platform firmware) as well as to initialize them
from the code that is only necessary for their initialization and
cleanly divide it into functions that each carry out a specific action.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>

Separate the code needed to update the passive trip (in a response to a
notification from the platform firmware) as well as to initialize it
from the code that is only necessary for its initialization and cleanly
divide it into functions that each carry out a specific action.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>

Notice that the only piece of information regarding the critical and hot
trips that needs to be stored in the driver's local data structures is
whether or not these trips are valid, so drop all of the redundant
information from there and adjust the code accordingly.

Among other things, this requires acpi_thermal_add() to be rearranged
so as to obtain the critical trip temperature before populating the trip
points table and for symmetry, the hot trip temperature is obtained
earlier too.

No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>

Create and populate the driver's trip points table in acpi_thermal_add()
so as to allow the its data structures to be simplified going forward.

No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Daniel Lezcano <daniel.lezcano@linaro.org>