Correct issue where transmission complete interrupt was not disabled when
using interrupt driven serial API with power management enabled,
resulting in continuous spurious interrupts that effectively locked
the system.
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
If the transaction of write or read is divided into two transfers
(not two messages), the command queue mode does not support.
Signed-off-by: Tim Lin <tim2.lin@ite.corp-partner.google.com>
Rework the <BUS>_emul_register calls to not pass the name param. The
name param is only used for logging and we can get it from the
struct <BUS>_emul instead.
Signed-off-by: Kumar Gala <galak@kernel.org>
Having a per-instance init function makes code cluttered and hard to
read. Just create a per-instance IRQ connect function (required to
resolve IRQ_CONNECT parameters at compile time).
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Stop using NRFX_SPI(S|M)_INSTANCE helper so that peripheral address from
Devicetree is used. We should not rely on HAL for hardware description
but Devicetree.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Move the RX (MISO) delaying capability information to Devicetree. It is
done using 2 properties:
- rx-delay-supported: enabled on SPI nodes that support delaying RX.
This property can be used by the driver to determine if this
capability is supported or not on a given instance.
- rx-delay: the actual RX delay value
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The maximum transfer buffer length is SoC specific, not instance
specific. This patch defines MAX_BUF_LEN at driver level in a SoC
specific manner instead of using HAL values that are instance specific
but that always take the same value depending on the SoC.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The maximum transfer chunk length is SoC specific, not instance
specific. This patch defines MAX_CHUNK_LEN at driver level in a SoC
specific manner instead of using HAL values that are instance specific
but that always take the same value depending on the SoC.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Specify the overrun character in Devicetree. Since 0xFF is the most
common value, DT property contains such default.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Devicetree is the natural place to describe hardware, so move the
maximum frequency the SPI can work with to Devicetree instead of relying
on values from HAL.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
There are a few registers that map nicely onto structs. Switch from
using index constants and byte arrays in those cases.
Signed-off-by: Andreas Sandberg <andreas@sandberg.uk>
Some panels using this driver don't provide tcon/cdi/pwr/softstart
values in their reference code. It is normally expected that the right
values will be loaded from OTP in such cases. Make these values
optional to support such panels.
Signed-off-by: Andreas Sandberg <andreas@sandberg.uk>
The GD7965 driver uses malloc to allocate a single line buffer when
clearing the display. This buffer is used to clear the display line by
line. This as two problems. First, it means that the driver introduces
an unnecessary requirement to support heap allocations. Second, it
causes a lot of weird and unnecessary SPI transactions that look like
partial updates without the actual refresh. This is very inefficient
since the same action can be performed in a single SPI transaction.
Add a gd7965_write_cmd_pattern() helper function that writes a pattern
of a specified length to a register in the device.
Signed-off-by: Andreas Sandberg <andreas@sandberg.uk>
The protocol used by the GD7965 driver requires that SPI transactions
are split into two phases, a command phase and a data phase, to change
the state of a GPIO signaling commands/data.
We currently don't lock the bus which means that other drivers could
potentially take over the bus and introduce unpredictable delays
between the command and data phase.
Add the necessary locking.
Signed-off-by: Andreas Sandberg <andreas@sandberg.uk>
DT nodes aren't guaranteed to define a label property. But emulated bus
controllers currently make use of this property to dispatch to the
associated emulator.
Have emulated bus controllers use DEVICE_DT_GET(node_id) to dispatch to
right target peripheral emulator. This also change makes emul_get_binding
and device_get_binding synonymous in behavior with respect to their
parameters.
This also strictly enforces a 1:1 correspondence between invocations of
DEVICE_DT_DEFINE and EMUL_DEFINE.
Signed-off-by: Aaron Massey <aaronmassey@google.com>
In several locations of the emulator code there are unused function
arguments that were never caught.
Declare these as unused or remove the unused function parameters entirely.
Signed-off-by: Aaron Massey <aaronmassey@google.com>
Allow emulator creators to write an init function that can be used
across multiple busses so as to reduce the boilerplate and cognitive
load in creating an emulator.
Part of this change includes allowing access to the emul struct from a
field in a {bus}_struct api (e.g. i2c_struct), which removes the need for
sporadic usages of CONTAINER_OF to access the emul struct.
Overall, this change simplifies and reduces the amount of boilerplate
code to get a device emulator up and running, thus reducing excise work
to writing tests.
TEST=twister on accel,espi, and eeprom drivers tests
Signed-off-by: Aaron Massey <aaronmassey@google.com>
Run clang-format on all files touched by improved emulator API pull-request
that allowed access to the target device emulator from its bus api without
CONTAINER_OF usage.
drivers/i2c/i2c_emul.c
drivers/spi/spi_emul.c
include/zephyr/drivers/emul.h
include/zephyr/drivers/espi_emul.h
include/zephyr/drivers/i2c_emul.h
include/zephyr/drivers/spi_emul.h
subsys/emul/emul.c
subsys/emul/emul_bmi160.c
subsys/emul/espi/emul_espi_host.c
subsys/emul/i2c/emul_atmel_at24.c
TEST=twister on accel,espi, and eeprom drivers tests
Signed-off-by: Aaron Massey <aaronmassey@google.com>
Update video drivers to use DT_HAS_<compat>_ENABLED Kconfig symbol
to expose the driver and enable it by default based on devicetree.
Signed-off-by: Kumar Gala <galak@kernel.org>
Fix building this sample on bl5340_dvk_cpuapp_ns and
pinnacle_100_dvk. On these boards the NORDIC_QSPI_NOR
driver needs to be enabled for the sample to build.
Signed-off-by: Kumar Gala <galak@kernel.org>
The twim_config structure is no longer modified at runtime, so it can be
placed in driver's config (const).
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
TWIM peripheral needs to be disabled for bus recovery, otherwise SCL/SDA
lines will not be released. This patch makes sure to disable peripheral
if active, and, restore its state afterwards (including pin
configuration).
It is worth to note that a better solution would be to:
1. Define scl/sda pins as `-gpios` in DT
2. Use GPIO API in the driver to perform recovery (as some other drivers
do)
3. Potentially use a "gpio" pinctrl state for this case
Unfortunately HAL is doing everything under the hood, so we have little
options to improve this unless we don't use it for such case. GPIO based
recovery should likely be generalized as many drivers seem to replicate
such _algorithm_.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Enable device runtime PM on the I2C driver. Note that this mechanism
replaces the old nrfx_twim_enable/disable calls with
pm_device_runtime_get/put, which means that driver will not save power
unless CONFIG_PM_DEVICE_RUNTIME=y.
Some quick measurements on thingy_52 running the samples/sensor/hts221
show a decrease of ~2-3uA in average when enabling device runtime PM.
Note that the driver already had implicit PM before, so the change for
users will be ~none, except that they now must enable the PM subsystem
features. While this case is not the best example, the PM subsystem will
turn to be beneficial as a whole when all devices in the board implement
it.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The device definition macro defined per-device init function, resulting
in a cluttered and hard to extend init function. Create a per-instance
IRQ connect function instead, which is called at init time by a common
function.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The driver already implemented some sort of runtime PM by
unconditionally calling nrfx_twim_disable/enable. Perform the same
operations in the PM callback instead of doing a full init/deinit every
time.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The driver re-initialized itself on the twim_transfer call if it wasn't.
This condition was likely added because of some manual/custom PM
schemes. Drop it in preparation for runtime PM.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
nrfx HAL does not support updating frequency at runtime, which means we
need to do a full init/deinit to update it when it is a matter of a
simple register write. Fix this by using the lower level HAL.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Move driver to use {NET_}DEVICE_DT_INST_DEFINE. This lets us
remove the IEEE802154_CC2520_DRV_NAME Kconfig symobl.
We also update the ieee802154 build_all test to actually enable
the CC2520 driver.
Signed-off-by: Kumar Gala <galak@kernel.org>
Move driver to use {NET_}DEVICE_DT_INST_DEFINE. This lets us
remove the IEEE802154_B91_DRV_NAME Kconfig symobl.
Signed-off-by: Kumar Gala <galak@kernel.org>
Any project with Kconfig option CONFIG_LEGACY_INCLUDE_PATH set to n
couldn't be built because some files were missing zephyr/ prefix in
includes
Re-run the migrate_includes.py script to fix all legacy include paths
Signed-off-by: Tomislav Milkovic <milkovic@byte-lab.com>
