PCA9420 PMIC offers of multiple operation states, or DVS (Dynamic
Voltage Scaling). Such states may be automatically changed by hardware
using MODESEL0/1 pins. Certain MCUs allow to automatically configure
certain output pins when entering low power modes so that PMIC state is
changed without software intervention. This means that application just
needs to configure the voltages for each state using
`nxp,modeN-microvolt`, set `nxp,enable-modesel-pins` in devicetree and
forget about configuring regulators.
This patch introduces a new _parent_ API to expose such functionality in
a vendor agnostic way. Consider this API as experimental for now, until
we have other usecases.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Most of devicetree properties for regulator, such as:
- regulator-min/max-microvolt
- regulator-min/max-microamp
- regulator-allowed-modes
- etc.
Are meant to specify limits on what consumers may set. They are **NOT**
meant to describe the hardware capabilities. For example, I could have a
BUCK converter that supports 0-5V output voltage, but my circuit may
only allow working on the 2.7-3.3V range.
This patch reworks the API so that the API class layer manages this
information. This is done by drivers collecting all such fields in a
common configuration structure that is later accessed by the class
layer. This simplifies drivers implementation. For example, if A
consumer calls regulator_set_voltage() with a voltage that is supported
but not allowed, driver code won't be called. Similarly, if a regulator
is configured to be `always-on`, enable/disable driver code will never
be called.
Drivers have been adjusted. PCA9420 mode settings have been removed from
devicetree in this commit as they are not actual modes but PMIC states.
This will be refactored in a follow-up commit.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
- Remove redundant comments
- Remove internal *_mode functions, as they were only used once.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Define regulator mode as an opaque type, same as we do in other
subsystems like GPIO (e.g. gpio_flags_t).
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Clarify the API for regulators that have the option to set mode
externally, such as PCA9420. Adjust the PCA9420 driver to comply with
the interface.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
- Function now returns error, value is obtained by reference. This
allows to propagate potential bus errors.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
- Mode specific APIs repeat the same functionality offered by non-mode
specific APIs
- The same functionality can be achieved by the non-mode APIs, since
they apply to the active mode which can be set using
regulator_set_mode() first.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The function can be implemented by using regulator_count_voltages() +
regulator_list_voltage(), so there's no need to defer the job to each
driver.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
- Rename to regulator_list_voltage (it is listing a single voltage)
- Function returns the value via a parameters, so that we can indicate
wether the given index is valid or not. If a driver doesn't implement
this call, function returns -EINVAL (as it should be).
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The regulator_count_modes was an API that was not useful. Unlike with
voltages where the interface guarantees they are zero-indexed, modes can
take an arbitrary identifier. So counting supported modes doesn't
provide any useful information such as if a mode is allowed.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
- Interface guarantees that voltage indices start from zero to
regulator_count_voltages()
- Function always return an unsigned number. If the device doesn't
implement the API returned value is zero, as expected.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
In the case that IN endpoint is not busy there will be
no notification from the HAL driver.
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
Adds an attribute that describes the maximum number of scatter gather
blocks that the DMA will accept. Useful in cases where DMA is used as
the abstraction for moving data around (e.g. SoF audio streams) and
variances of the hardware need to inform decisions about how to use
the DMA API.
Co-authored-by: Adrian Bonislawski <adrian.bonislawski@intel.com>
Co-authored-by: Jaska Uimonen <jaska.uimonen@intel.com>
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
The SBS fuel gauge driver did not return a count of the number of
properties that failed.
Fix this and add a test that verifies the sbs_gauge returns the number of
failing properties as well as a test that verifies a negative return code
if all properties failed.
Signed-off-by: Aaron Massey <aaronmassey@google.com>
zephyr kernel header file should be included otherwise gcc will report
the warning: implicit declaration of function 'k_aligned_alloc' and
the return value of 'k_aligned_alloc' will be treated as an int type,
which will cause an error on the 64 bits platform.
Signed-off-by: Huifeng Zhang <Huifeng.Zhang@arm.com>
According to section 5.1.9.3.5 and 5.1.9.3.6 of the I3C Specification
v1.1.1. This CCC is may be optionally supported if the target device
has no settable limit.
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
Rename is_primary to is_secondary. The justification for this is
because it is less likely to have something configured to be
secondary, and the 0 value would be if it is primary.
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
Previously, a call to set the net_context state was added here to
accommodate a new KConfig scenario that permitted disabling offload
sockets. This code path is used by both local and offloaded sockets.
While it is necessary to update the state of the net_context for a
locally managed socket, setting the net_context for an offloaded socket
is an error, as the net_context is invalid (in fact, it points to a
hard-coded dummy socket), as observed in #52346 and #38544.
A prior commit (a9ac0a88) attempting to remedy the offload socket
scenario removed this line, which resulted in problems for the local
socket scenario. The socket never got updated to the connected state,
and thus was unusable.
Adding this guard allows the eswifi socket to set the socket net_context
only if the socket is NOT offloaded.
Signed-off-by: Brian Dunlay <brian@nubix.io>
In Infineon XMC4XXX SoCs, gpio interrupts are triggered via an
Event Request Unit (ERU) module. A subset of the gpios are
connected to the ERU. The ERU monitors edge triggers and creates
a SR.
This driver configures the ERU for a target port/pin combination
for rising/falling edge events. Note that the ERU module does
not generate SR based on the gpio level. Internally the ERU
tracks the *status* of an event. The status is set on a positive
edge and unset on a negative edge (or vice-versa depending on
the configuration). The value of the status is used to implement
a level triggered interrupt; The ISR checks the status flag and
calls the callback function if the status is set.
The ERU configurations for supported port/pin combinations are
stored in a devicetree file dts/arm/infineon/xmc4xxx_x_x-intc.dtsi.
The configurations are stored in the opaque array
uint16 port_line_mapping[].
Signed-off-by: Andriy Gelman <andriy.gelman@gmail.com>
Config NOCACHE_MEMORY depend on ARCH_HAS_NOCACHE_MEMORY_SUPPORT. Enable
ARCH_HAS_NOCACHE_MEMORY_SUPPORT for Cortex-R52 to run NXP S32Z/E with
nocache attibute.
Enable nocache in each driver use it.
Signed-off-by: Duong Vu Nam <duong.vunam@nxp.com>
Adds support for GD32 timer
Note: Currently, it is not supporting RISC-V(GD32V) devices.
It needs some work on the interrupt controller first.
Signed-off-by: TOKITA Hiroshi <tokita.hiroshi@gmail.com>
The timer_ids contain timers that belong to any bus.
So, It should recognize with entire id, not only the CLOCK_ID_BIT part.
Signed-off-by: TOKITA Hiroshi <tokita.hiroshi@gmail.com>
DT_COMPAT_GET_ANY_STATUS_OKAY is not suited for the node's existing check.
(This macro returns the stem of the DTS macro name,
the stem part is not a defined symbol.)
Instead, it should use the DT_HAS_COMPAT_STATUS_OKAY macro.
Signed-off-by: TOKITA Hiroshi <tokita.hiroshi@gmail.com>
Change Summary:
Update Base addr to set direction to use Aligned IP base address instead
of Port Base Address. Port Base Address + offset to direction register
will set incorrect value for Port B,C & D. For ex: In cases when more than
1 port is configured on the same IP, the DTS node for port B will start
at offset 0xC. Calculating the port using Port Base will yield offset of
PORTB DIR register which is at offset 0x10 from Aligned Base, and as a
result will result in setting DIR register of PORTC instead.
Signed-off-by: Ravik Hasija <ravikh@fb.com>
Change Summary:
Moving the setting of Pin direction before setting/clearing the pin
configured as output for the change to correctly take place.
Signed-off-by: Ravik Hasija <ravikh@fb.com>
For architectures that rely on a PCIe controller (for example, ARM64),
scanning the PCI space will only succeed after the controller has
initialized. Therefore, in the presence of PCIe controller, the PCIe
initialization is bumped to the next system init level.
In the past, drivers like ivshmem would do a late scan of the PCI space
in case the early scan failed; however, the cited commit removed this
feature and ivshmem fails for ARM64. This commit fix this by making the
early scan succeed.
Fixes: a96016d747 ("drivers: ivshmem: Remove unnecessary BDF lookup ...")
Signed-off-by: Rodrigo Cataldo <rodrigo.cataldo@huawei.com>
Co-authored-by: Henri Xavier <datacomos@huawei.com>
Only include the MSI header when ivshmem-doorbell is employed, as the
ivshmem data structures already use the same protection for msi-related
structures.
Signed-off-by: Rodrigo Cataldo <rodrigo.cataldo@huawei.com>
Co-authored-by: Henri Xavier <datacomos@huawei.com>
The ivshmem code does not use any platform-specific code; therefore,
remove the dependency to the soc interface.
No functional change intended.
Signed-off-by: Rodrigo Cataldo <rodrigo.cataldo@huawei.com>
Co-authored-by: Henri Xavier <datacomos@huawei.com>
Add a Kconfig option allowing users to configure the transfer timeout
value, as the default 500 ms may not be sufficient in specific cases.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
In multi-image environment, after jump to the image we can have UART in
unexpected state. Reset UART to default state to make sure that UART is
initialized properly and won't cause system to crash or hang.
Signed-off-by: Patryk Duda <pdk@semihalf.com>
In multi-image environment, after jump to the image we can have running
timer with interrupts enabled. If interrupt is triggered, the asserts
in the driver can cause a crash.
This patch also adds 'resets' property for all timer nodes.
Signed-off-by: Patryk Duda <pdk@semihalf.com>
To deassert reset in STM32MP1 RCC the driver needs to set the bit in
reset clear register.
This patch extends existing implementation to support this type of
register.
Signed-off-by: Patryk Duda <pdk@semihalf.com>
This driver exposes STM32 RCC reset functionality through reset API.
Information about RCC register offset and bit is encoded just like GD32.
The first 5 least significant bits contains register bit number.
Next 12 bits are used to keep RCC register offset. Remaining bits are
unused.
Signed-off-by: Patryk Duda <pdk@semihalf.com>
This function is responsible for checking if combination of msize,
mburst and FIFO level is allowed. Possible combinations can be found in
ST documentation, eg. Table 36. FIFO threshold configurations, RM0402
9.3.13 FIFO chapter.
Previously there was no 'break' or '__fallthrough' in msize switch which
caused compilation errors. Since we are confirming that combination is
correct, 'break' statements should be used.
Besides of introducing missing 'break' statements, this patch moves
'return false' from switch to the end of the function. This makes code
shorter and easier to understand, because we have only correct
combinations.
Signed-off-by: Patryk Duda <pdk@semihalf.com>
The poll_in function of the NuMicro UART driver was using the UART_Read
function from the Nuvoton HAL, which is blocking. Replace it with a
non-blocking implementation.
Signed-off-by: Filip Brozovic <fbrozovic@gmail.com>
Add the resume API function for the dma driver of the stm32U5 serie.
That completes the suspend API function.
Controlling the SUSPF bit of the GPDMA CR register is enough
to suspend/resume the channel.
Signed-off-by: Francois Ramu <francois.ramu@st.com>
There is no need to reset the channel else DMA config is lost and
channel should be enabled again in case of resume.
Signed-off-by: Francois Ramu <francois.ramu@st.com>
