For the stm32 devices that have a HSI48 clock,
the driver enables it, like any other fixed clock,
if needed and supported by the serie.
For stm32L0, SYSCFG VREFINT is also required.
Signed-off-by: Francois Ramu <francois.ramu@st.com>
This patch reworks how fragments are handled in the net_buf
infrastructure.
In particular, it removes the union around the node and frags members in
the main net_buf structure. This is done so that both can be used at the
same time, at a cost of 4 bytes per net_buf instance.
This implies that the layout of net_buf instances changes whenever being
inserted into a queue (fifo or lifo) or a linked list (slist).
Until now, this is what happened when enqueueing a net_buf with frags in
a queue or linked list:
1.1 Before enqueueing:
+--------+ +--------+ +--------+
|#1 node|\ |#2 node|\ |#3 node|\
| | \ | | \ | | \
| frags |------| frags |------| frags |------NULL
+--------+ +--------+ +--------+
net_buf #1 has 2 fragments, net_bufs #2 and #3. Both the node and frags
pointers (they are the same, since they are unioned) point to the next
fragment.
1.2 After enqueueing:
+--------+ +--------+ +--------+ +--------+ +--------+
|q/slist |------|#1 node|------|#2 node|------|#3 node|------|q/slist |
|node | | *flag | / | *flag | / | | / |node |
| | | frags |/ | frags |/ | frags |/ | |
+--------+ +--------+ +--------+ +--------+ +--------+
When enqueing a net_buf (in this case #1) that contains fragments, the
current net_buf implementation actually enqueues all the fragments (in
this case #2 and #3) as actual queue/slist items, since node and frags
are one and the same in memory. This makes the enqueuing operation
expensive and it makes it impossible to atomically dequeue. The `*flag`
notation here means that the `flags` member has been set to
`NET_BUF_FRAGS` in order to be able to reconstruct the frags pointers
when dequeuing.
After this patch, the layout changes considerably:
2.1 Before enqueueing:
+--------+ +--------+ +--------+
|#1 node|--NULL |#2 node|--NULL |#3 node|--NULL
| | | | | |
| frags |-------| frags |-------| frags |------NULL
+--------+ +--------+ +--------+
This is very similar to 1.1, except that now node and frags are
different pointers, so node is just set to NULL.
2.2 After enqueueing:
+--------+ +--------+ +--------+
|q/slist |-------|#1 node|-------|q/slist |
|node | | | |node |
| | | frags | | |
+--------+ +--------+ +--------+
| +--------+ +--------+
| |#2 node|--NULL |#3 node|--NULL
| | | | |
+------------| frags |-------| frags |------NULL
+--------+ +--------+
When enqueuing net_buf #1, now we only enqueue that very item, instead
of enqueing the frags as well, since now node and frags are separate
pointers. This simplifies the operation and makes it atomic.
Resolves#52718.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
- Similar to what was done for other parts of the driver, remove any
register specification from Devicetree (modesel-reg/mask)
- Keep all the information in the driver, and define modes as "numbers",
e.g. PCA9420_MODE0: 0, PCA9420_MODE1: 1, etc.
- Bindings provide IC defaults now (all modes allowed 0/1/2/3 and
initial mode set to 0).
- When mode is controlled via the MODESEL0/1 pins (ie directly by an iMX
MCU using the dedicated PMIC_MODE0/1 pins), the driver will not allow
to select a mode (it is not possible). This mode is now enabled by
setting `nxp,enable-modesel-pins` in Devicetree. When enabled, all the
allowed modes are configured to be selectable via pins. When disabled,
mode can be set via I2C (using TOP_CNTL3 MODE0/1_I2C fields)
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Add support for setting the help description for each entry in a dictionary
command. Currently the syntax string alone may not provide sufficient
description of its entry. This commit also helps keep the help messages
consistent with existing style.
Signed-off-by: Xinyang Tan <xinyang.tan@delve.com>
GCC_VERSION is defined in a few modules, and those headers are often
included first, so replace the one used in zephyr with
TOOLCHAIN_GCC_VERSION. Do the same with CLANG_VERSION, replacing it with
TOOLCHAIN_CLANG_VERSION.
BUILD_ASSERT is also defined in include/toolchain/common.h, which might
get included before gcc.h. We want to use the gcc-specific one instead
of the general one.
Signed-off-by: Keith Packard <keithp@keithp.com>
This header file makes use of INT_MAX, which is defined in limits.h.
Include the header to make sure it can be used.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
Add `i2c_is_ready_dt` function that validates I2C spec bus
is ready and it allows the user to pass `i2c_dt_spec` directly.
Signed-off-by: Bartosz Bilas <b.bilas@grinn-global.com>
Adds fs_mkfs function to fs api. It will allow to perform mkfs operation
in file systems added to Zephyr.
Signed-off-by: Franciszek Zdobylak <fzdobylak@internships.antmicro.com>
Add bindings for Aspeed AST10x0 reset driver. The reset line can be
de-asserted or asserted through the syscon registers.
Signed-off-by: Dylan Hung <dylan_hung@aspeedtech.com>
Add a no-op stub for __ASSERT_POST_ACTION.
This allows code to reference __ASSERT_POST_ACTION even when Zephyr
asserts are off.
Signed-off-by: Rob Barnes <robbarnes@google.com>
Add ARM64 support for `CONFIG_TIMER_READS_ITS_FREQUENCY_AT_RUNTIME`.
Used for the `qemu_kvm_arm64` board. This should have been part
of #51932. Without this, the `qemu_kvm_arm64` board is painfully
slow due to the timer frequency being wrong.
Signed-off-by: Henri Xavier <datacomos@huawei.com>
Declarations for `getopt()` should be in `<unistd.h>`
according to the spec. The extended versions `getopt_long()`
and `getopt_long_only()` are declared in `<getopt.h>`.
Signed-off-by: Chris Friedt <cfriedt@meta.com>
Add `gpio_is_ready_dt` function that validates GPIO DT spec port
is ready and it allows the user to pass `gpio_dt_spec` directly.
Signed-off-by: Bartosz Bilas <b.bilas@grinn-global.com>
Introduce an API mirroring the CMSIS-DSP's basicmath. If CMSIS_DSP is
enabled, then it will by default be used as a backend. Developers may
opt into a custom backend by setting CONFIG_DSP_BACKEND_CMSIS=n. If
done, the application must provide `zdsp_backend/dsp.h` and optionally
implement the functions in its own .c files.
Signed-off-by: Yuval Peress <peress@google.com>
The device supprt brings support for multiple stack instances,
multiple configuration, asynchronous transfer model, ability to
change most of the properties of a device at runtime and
the composition of configuration and classes at runtime.
The stack requires new UDC driver API and is not compatible
with old driver API (usb_dc_). The classes (functions) of old
(current) USB device stack cannot be used with new ones and must
be ported.
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
Expand ifdef by adding new Kconfig option UDC_KINETIS as
preparation for USBFSOTG UDC driver.
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
Add new USB device controller API and nRF USBD controller driver.
The new UDC API brings support for multiple instances and asynchronous
transfer model, transfers use net_buf and store methadata in the
user data area.
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
Add macros to get endpoint address, direction, and attributes.
Remove identical endpoint macros from usb_dc header.
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
When dealing with groups, the specified window in
linear_range_group_get_win_index can span across multiple groups. In the
current implementation, if maximum value was not in the same sub-window
as the minimum value the function would return -EINVAL. This patch fixes
the problem and updates test cases to cover such scenario.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Add clarification about the representation of the data
in the buffer of a multi-channel ADC sequence.
Signed-off-by: Declan Snyder <declan.snyder@nxp.com>
This seems to have caused build failures in spite of CI being
green in PR 52653.
This reverts commit fc92eb1b37.
Signed-off-by: Chris Friedt <cfriedt@meta.com>
Declarations for `getopt()` should be in `<unistd.h>`
according to the spec. The extended versions `getopt_long()`
and `getopt_long_only()` are declared in `<getopt.h>`.
Signed-off-by: Chris Friedt <cfriedt@meta.com>
The cache operations must be quick, optimized and possibly inlined. The
current API is clunky, functions are not inlined and passing parameters
around that are basically always known at compile time.
In this patch we rework the cache functions to allow us to get rid of
useless parameters and make inlining easier.
In particular this changeset is doing three things:
1. `CONFIG_HAS_ARCH_CACHE` is now `CONFIG_ARCH_CACHE` and
`CONFIG_HAS_EXTERNAL_CACHE` is now `CONFIG_EXTERNAL_CACHE`
2. The cache API has been reworked.
3. Comments are added.
Signed-off-by: Carlo Caione <ccaione@baylibre.com>
Drop the async enable function. This feature is rarely/never used,
complicates driver design, and doesn't really follow the sync/async API
design/naming used in other areas. In the future we can introduce
regulator_enable_async if needed, with support from the driver class (no
onoff). Note that drivers like PCA9420 did not implement any
asynchronous behavior. regulator-fixed implemented in the past
asynchronous behavior using work queues, an overkill for most GPIO
driven regulators. Let's keep things simple for now and extend the API
when needed, based on specific usecases.
In the current implementation, reference counting is managed by the
driver class. \isr-ok attribute is dropped, since calls are potentially
blocking. Note that drivers like PCA9420 already violated such rule.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Remove conditionnal declaration of `bt_tbs_dbg_print_calls` in
`subsys/bluetooth/audio/tbs.h` so the function still exist even if the
debug is not enabled.
Signed-off-by: Théo Battrel <theo.battrel@nordicsemi.no>
Microchip XEC GPIO pins support inverting the output of
alternate pin functions. This feature may be useful for
those peripherals that do not implement output inversion
in the peripheral. GPIO control register pad input and
parallel input register values are not affected by the
function output invert feature. GPIO interrupt detection
of an output is inverted if the invert polarity is enabled.
Signed-off-by: Jay Vasanth <jay.vasanth@microchip.com>
pca9420_i2c.h has neem remamed to pca9420.h (device only has I2C, so
it's kind of redundant to have i2c in the header name).
pmic_i2c.h was again not generic but designed specifically for PCA9420.
All its macros have been renamed, s/PMIC/PCA9420, and moved to the
pca9420.h header.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The driver had a conceptual issue regarding current limitation. PCA9420
is able to limit the current flowing through VIN, ie input current. This
is a global setting, not individual to each regulator. This patch
creates a new DT property: nxp,vin-ilim-microamp to specify such limit.
It is applied when the device is initialized.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Voltage ranges were hardcoded in Devicetree, however, things can be
significantly simplified by using the recently introduce linear ranges
API. All values are now computed using information stored in the driver,
so there is no need to store any lookup table in ROM. Code should now
both be faster in average and consume less ROM.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Regulator registers were set for each BUCK/LDO in DT, likely because of
the way the devices were instantiated. When using a generic iterator,
ie, DT_INST_FOREACH_CHILD, there's no way to differentiate the child
being _parsed_. Since instantiation happens now based on child node
names, we are able to know which registers each devices gets assigned at
the driver level. This greatly simplifies Devicetree, and it actually
removes information that is not strictly hardware description from it.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The PCA9420 dt-bindings include file defines all device registers, but
many of them are not used (e.g. battery charger). It is also not the way
to use register definitions. Devicetree is about hardware description
after all. File will be cleaned up in subsequent commits.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Rather than hard-coding calls to each individual CRC function, it
may be more convenient to specify the crc algorithm to use
parametrically.
Signed-off-by: Chris Friedt <cfriedt@meta.com>
Added 3 helper methods for async calls as they exist for
the blocking implementation
Signed-off-by: Hernan I. Gutierrez-Vazquez <quic_hernang@quicinc.com>
Make ARC_MP_PRIMARY_CPU_ID definition public so it can be used in
other ARC code.
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>
Signed-off-by: Evgeniy Paltsev <PaltsevEvgeniy@gmail.com>
Remove the struct bt_vcp pointer from the volume
renderer API, as there is only ever a single
Volume Renderer instance on a device.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Split the `bt_vcp_cb` struct into separate structure for
the volume controller and the volume renderer.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
The VCP server, known as the volume renderer, has a
more explicity bt_vcp_vol_rend API naming scheme now.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Remove the bt_vcp_vocs API as it did the same as the
bt_vocs API, and did not really any functionality.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
The AICS functionality can be exercised using the
bt_aics API instead. These functions did very little
besides cluttering up the bt_vcp API.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Renames almost all occurences of VCS to VCP, as we want to
use the profile as the namespace for volume control.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
