Input: iqs7222 - add support for IQS7222D v1.1 and v1.2

The vendor has introduced two new revisions with slightly different
memory maps; update the driver to support them.

Fixes: dd24e202ac72 ("Input: iqs7222 - add support for Azoteq IQS7222D")
Signed-off-by: Jeff LaBundy <jeff@labundy.com>
Link: https://lore.kernel.org/r/ZelTRYX3fenMQuhF@nixie71
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: Explicitly include correct DT includes

The DT of_device.h and of_platform.h date back to the separate
of_platform_bus_type before it as merged into the regular platform bus.
As part of that merge prepping Arm DT support 13 years ago, they
"temporarily" include each other. They also include platform_device.h
and of.h. As a result, there's a pretty much random mix of those include
files used throughout the tree. In order to detangle these headers and
replace the implicit includes with struct declarations, users need to
explicitly include the correct includes.

Signed-off-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/20230714174633.4058096-1-robh@kernel.org
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: iqs7222 - add support for Azoteq IQS7222D

The vendor has introduced a new variant of silicon which is highly
similar to the existing IQS7222A, but with its independent sliders
essentially replaced with a single-contact trackpad.

Update the common driver to support this new device's register map
and report trackpad events. As with the IQS7222A, the new IQS7222D
can report both raw coordinates as well as gestures.

Signed-off-by: Jeff LaBundy <jeff@labundy.com>
Link: https://lore.kernel.org/r/ZKrpRh6RT6+6KrMQ@nixie71
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: iqs7222 - configure power mode before triggering ATI

If the device drops into ultra-low-power mode before being placed
into normal-power mode as part of ATI being triggered, the device
does not assert any interrupts until the ATI routine is restarted
two seconds later.

Solve this problem by adopting the vendor's recommendation, which
calls for the device to be placed into normal-power mode prior to
being configured and ATI being triggered.

The original implementation followed this sequence, but the order
was inadvertently changed as part of the resolution of a separate
erratum.

Fixes: 1e4189d8af27 ("Input: iqs7222 - protect volatile registers")
Signed-off-by: Jeff LaBundy <jeff@labundy.com>
Link: https://lore.kernel.org/r/ZKrpHc2Ji9qR25r2@nixie71
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: Switch i2c drivers back to use .probe()

After commit b8a1a4cd5a98 ("i2c: Provide a temporary .probe_new()
call-back type"), all drivers being converted to .probe_new() and then
03c835f498b5 ("i2c: Switch .probe() to not take an id parameter") convert
back to (the new) .probe() to be able to eventually drop .probe_new() from
struct i2c_driver.

Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Link: https://lore.kernel.org/r/20230517164645.162294-1-u.kleine-koenig@pengutronix.de
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: iqs7222 - add support for IQS7222A v1.13+

IQS7222A revisions 1.13 and later widen the gesture multiplier from
x4 ms to x16 ms. Add a means to scale the gesture timings specified
in the device tree based on the revision of the device.

Fixes: e505edaedcb9 ("Input: add support for Azoteq IQS7222A/B/C")
Signed-off-by: Jeff LaBundy <jeff@labundy.com>
Reviewed-by: Mattijs Korpershoek <mkorpershoek@baylibre.com>
Link: https://lore.kernel.org/r/Y1SRdbK1Dp2q7O8o@nixie71
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: iqs7222 - protect against undefined slider size

Select variants of silicon do not define a default slider size, in
which case the size must be specified in the device tree. If it is
not, the axis's maximum value is reported as 65535 due to unsigned
integer overflow.

To solve this problem, move the existing zero-check outside of the
conditional block that checks whether the property is present.

Fixes: e505edaedcb9 ("Input: add support for Azoteq IQS7222A/B/C")
Signed-off-by: Jeff LaBundy <jeff@labundy.com>
Link: https://lore.kernel.org/r/Y1SRXEi7XMlncDWk@nixie71
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: iqs7222 - report malformed properties

Nonzero return values of several calls to fwnode_property_read_u32()
are silently ignored, leaving no way to know the properties were not
applied in the event of an error.

Solve this problem by evaluating fwnode_property_read_u32()'s return
value, and reporting an error for any nonzero return value not equal
to -EINVAL which indicates the property was absent altogether.

Fixes: e505edaedcb9 ("Input: add support for Azoteq IQS7222A/B/C")
Signed-off-by: Jeff LaBundy <jeff@labundy.com>
Reviewed-by: Mattijs Korpershoek <mkorpershoek@baylibre.com>
Link: https://lore.kernel.org/r/Y1SRRrpQXvkETjfm@nixie71
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: iqs7222 - drop unused device node references

Each call to device/fwnode_get_named_child_node() must be matched
with a call to fwnode_handle_put() once the corresponding node is
no longer in use. This ensures a reference count remains balanced
in the case of dynamic device tree support.

Currently, the driver never calls fwnode_handle_put(). Solve this
problem by moving the node handling from iqs7222_parse_props() to
the new iqs7222_parse_reg_grp(), leaving the former to do nothing
but parse properties. The latter then manages the reference count
in a single location and consistent fashion.

This change drastically simplifies iqs7222_parse_all(), which can
then call iqs7222_parse_reg_grp() on every register group without
having to treat each register group differently.

For nested event nodes, common parsing code has been factored out
to the new iqs7222_parse_event() so as to allow the event node to
be dropped from as few locations as possible.

As part of this refactor, the 'linux,code' property has been made
optional. This enables applications that define an event with the
sole purpose of enabling a GPIO.

Fixes: e505edaedcb9 ("Input: add support for Azoteq IQS7222A/B/C")
Signed-off-by: Jeff LaBundy <jeff@labundy.com>
Link: https://lore.kernel.org/r/Y1SRJIQ3WPwNpC0K@nixie71
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: iqs7222 - set all ULP entry masks by default

Some devices expose an ultra-low-power (ULP) mode entry mask for
each channel. If the mask is set, the device cannot enter ULP so
long as the corresponding channel remains in an active state.

The vendor has advised setting the mask for any disabled channel.
To accommodate this suggestion, initially set all masks and then
clear them only if specified in the device tree.

Fixes: e505edaedcb9 ("Input: add support for Azoteq IQS7222A/B/C")
Signed-off-by: Jeff LaBundy <jeff@labundy.com>
Link: https://lore.kernel.org/r/20220908131548.48120-8-jeff@labundy.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: iqs7222 - avoid sending empty SYN_REPORT events

Add a check to prevent sending undefined events, which ultimately
map to SYN_REPORT.

Fixes: e505edaedcb9 ("Input: add support for Azoteq IQS7222A/B/C")
Signed-off-by: Jeff LaBundy <jeff@labundy.com>
Link: https://lore.kernel.org/r/20220908131548.48120-7-jeff@labundy.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: iqs7222 - trim force communication command

According to the datasheets, writing only 0xFF is sufficient to
elicit a communication window. Remove the superfluous 0x00 from
the force communication command.

Fixes: e505edaedcb9 ("Input: add support for Azoteq IQS7222A/B/C")
Signed-off-by: Jeff LaBundy <jeff@labundy.com>
Link: https://lore.kernel.org/r/20220908131548.48120-6-jeff@labundy.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: iqs7222 - remove support for RF filter

The vendor has marked the RF filter enable control as reserved in
the datasheet; remove it from the driver.

Fixes: e505edaedcb9 ("Input: add support for Azoteq IQS7222A/B/C")
Signed-off-by: Jeff LaBundy <jeff@labundy.com>
Link: https://lore.kernel.org/r/20220626072412.475211-7-jeff@labundy.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: iqs7222 - handle reset during ATI

If the device suffers a spurious reset during ATI, there is no point
in enduring any further retries. Instead, simply return successfully
from the polling loop.

In this case, the interrupt handler will intervene and recognize the
device has been reset. It then proceeds to initialize the device and
trigger ATI once more.

As part of this change, swap the order of status field evaluation to
match that of the interrupt handler, and correct a nearby off-by-one
error that causes an error message to suggest the final attempt will
be retried.

Fixes: e505edaedcb9 ("Input: add support for Azoteq IQS7222A/B/C")
Signed-off-by: Jeff LaBundy <jeff@labundy.com>
Link: https://lore.kernel.org/r/20220626072412.475211-6-jeff@labundy.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: iqs7222 - acknowledge reset before writing registers

If the device suffers a spurious reset while reacting to a previous
spurious reset, the second reset interrupt is preempted because the
ACK_RESET bit is written last.

To solve this problem, write the ACK_RESET bit prior to writing any
other registers. This ensures that any registers written before the
second spurious reset will be rewritten.

Last but not least, the order in which the ACK_RESET bit is written
relative to the second filter beta register is important for select
variants of silicon. Enforce the correct order so as to not clobber
the system status register.

Fixes: e505edaedcb9 ("Input: add support for Azoteq IQS7222A/B/C")
Signed-off-by: Jeff LaBundy <jeff@labundy.com>
Link: https://lore.kernel.org/r/20220626072412.475211-5-jeff@labundy.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: iqs7222 - protect volatile registers

Select variants of silicon silently mirror part of the event mask
register to the system setup register (0xD0), and vice versa. For
the following sequence:

1. Read registers 0xD0 onward and store their contents.
2. Modify the contents, including event mask fields.
3. Write registers 0xD0 onward with the modified contents.
4. Write register 0xD0 on its own again later, using the contents
from step 1 to populate any reserved fields.

...the event mask register (e.g. address 0xDA) has been corrupted
by writing register 0xD0 with contents that were made stale after
step 3.

To solve this problem, read register 0xD0 once more between steps
3 and 4. When register 0xD0 is written during step 4, the portion
which is mirrored to the event mask register already matches what
was written in step 3.

Fixes: e505edaedcb9 ("Input: add support for Azoteq IQS7222A/B/C")
Signed-off-by: Jeff LaBundy <jeff@labundy.com>
Link: https://lore.kernel.org/r/20220626072412.475211-4-jeff@labundy.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: iqs7222 - fortify slider event reporting

The release cycle of any key mapped to a slider gesture relies upon
trailing interrupts generated by other unmasked sources, the timing
and presence of which are inconsistent.

To solve this problem, explicitly report a release cycle to emulate
a full keystroke. Also, unmask touch interrupts if the slider press
event is defined; this ensures the device reports a final interrupt
with coordinate = 0xFFFF once the finger is lifted.

As a result of how the logic has been refactored, the press/release
event can now be mapped to a GPIO. This is more convenient than the
previous solution, which required each channel within the slider to
specify the same GPIO.

As part of this change, use the device's resolution rather than its
number of interrupt status registers to more safely determine if it
is capable of reporting gestures.

Last but not least, make the code a bit simpler by eliminating some
unnecessarily complex conditional statements and a macro that could
be derived using information that is already available.

Fixes: e505edaedcb9 ("Input: add support for Azoteq IQS7222A/B/C")
Signed-off-by: Jeff LaBundy <jeff@labundy.com>
Link: https://lore.kernel.org/r/20220626072412.475211-3-jeff@labundy.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: iqs7222 - correct slider event disable logic

If a positive swipe/flick gesture is defined but the corresponding
negative gesture is not, the former is inadvertently disabled. Fix
this by gently refactoring the logic responsible for disabling all
gestures by default.

As part of this change, make the code a bit simpler by eliminating
a superfluous conditional check. If a slider event does not define
an enable control, the second term of the bitwise AND operation is
simply 0xFFFF.

Fixes: e505edaedcb9 ("Input: add support for Azoteq IQS7222A/B/C")
Signed-off-by: Jeff LaBundy <jeff@labundy.com>
Link: https://lore.kernel.org/r/20220626072412.475211-2-jeff@labundy.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: iqs7222 - avoid dereferencing a NULL pointer

Select callers of iqs7222_parse_props() do not expect a child node
to be derived and returned via pointer. As such, these callers set
**child_node to NULL. However, this pointer is dereferenced in all
cases.

To solve this problem, dereference the pointer only for cases that
expect a child node in the first place. In these cases, the caller
provides a valid pointer.

Fixes: e505edaedcb9 ("Input: add support for Azoteq IQS7222A/B/C")
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Jeff LaBundy <jeff@labundy.com>
Link: https://lore.kernel.org/r/20220417214132.497487-1-jeff@labundy.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: iqs7222 - propagate some error codes correctly

If fwnode_property_count_u32() returns a negative error code then,
because of type promotion, the "count > ARRAY_SIZE(pins)" condition
will be true. The negative "count" is type promoted to a high unsigned
size_t value.

That means the "else if (count < 0)" condition will always be false and
we don't print that error message or propagate the error code from
fwnode_property_count_u32() as intended.

Fix this by re-ordering the checks so that we check for negative first.

Fixes: e505edaedcb9 ("Input: add support for Azoteq IQS7222A/B/C")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Link: https://lore.kernel.org/r/20220412153954.GA15406@kili
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Input: add support for Azoteq IQS7222A/B/C

This patch adds support for the Azoteq IQS7222A/B/C family of
capacitive touch controllers.

Signed-off-by: Jeff LaBundy <jeff@labundy.com>
Link: https://lore.kernel.org/r/20220403221659.865997-3-jeff@labundy.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>