The SDK FSL DMA driver converts descriptor addresses to DMA's address
space when linking descriptors. The Zephyr dma_mcux_lpc driver is
missing the inverse conversion when dereferencing the linked next
descriptor pointer.
This isn't a problem when this driver is used on the M33 core
of the MIMXRT595S because the M33 can access the address space
of the DMA (0x20000000+). But when the Fusion F1 DSP core uses
this driver, the DSP cannot access the DMA's address space
so the inverse conversion is needed.
Signed-off-by: Yicheng Li <yichengli@google.com>
Support three different FIFO contents which are selectable through
a new DT property, fifo-mode-sel, which may be set to one of the
following values:
- 0x0 # 1x Accelerometer @12bit and 1x temperature @12bit samples
- 0x1 # 1x Accelerometer @16bit sample
- 0x2 # 2x Accelerometer @8bit samples (previous and current)
Signed-off-by: Armando Visconti <armando.visconti@st.com>
Updated the condition in the gpio_ra_interrupt_unset function to use a
logical OR instead of AND, since it should return if either port_num or
pin is not the same as the current callback.
Signed-off-by: Benjamin Cabé <benjamin@zephyrproject.org>
Update board document to remove dcache off command under U-Boot as
CONFIG_ARM64_BOOT_DISABLE_DCACHE has been enabled.
Signed-off-by: Jiafei Pan <Jiafei.Pan@nxp.com>
Remove logically dead if and else-if conditions as shown by the static
analysis, replacing with the else statement.
When data_opt is assigned to the MIN of fifo_wmark_cfg and fifo_full_cfg,
both those variables will be non-NULL as an earlier condition ensures the
function returns if either one of the variables is NULL before assigning
to data_opt.
Signed-off-by: Alexander Apostolu <apostolu240@gmail.com>
Add sample to read from up to 10 streaming devices, using the
sensor_stream() API and the RTIO framework:
https://docs.zephyrproject.org/latest/services/rtio/index.html
The devices has to be aliased as streamN (0 <= N <= 9) in DT, and
will be instantiated using SENSOR_DT_STREAM_IODEV() macro.
Currently the sample gets/prints data on SENSOR_TRIG_DATA_READY
trigger basis for the following sensor channels:
- SENSOR_CHAN_ACCEL_XYZ
Signed-off-by: Armando Visconti <armando.visconti@st.com>
Add SENSOR_TRIG_DATA_READY trigger support to RTIO streaming.
Currently it just handle XL drdy.
Signed-off-by: Armando Visconti <armando.visconti@st.com>
Enable/disable devices as necessary to ensure memc test will test the
attached HyperRAM and no other memory nodes.
Signed-off-by: Pete Johanson <pete.johanson@analog.com>
Fix build error when CONFIG_MCUX_COMPONENT_driver.fro_calib is set.
Error message:
modules/hal/nxp/mcux/mcux-sdk-ng/cmake/extension/function.cmake:2084 (if)
if given arguments:
"IN_LIST" "NEED_WRAP_TOOLCHAINS"
The error is because, mcux sdk ng cmake functions `mcux_add_xxx` support
add content based on toolchains. CMake should set current toolchain in
CMake Variable `CONFIG_TOOLCHAIN` to use this feature.
Signed-off-by: Jason Yu <zejiang.yu@nxp.com>
In MCUX SDK NG, the `core_id` and `core_id_suffix_name` are
different concepts, generally the values are the same. But
there are exeptions, such as RT595 F1.
Zephyr's `CONFIG_MCUX_CORE_SUFFIX` is actually MCUX SDK's
`core_id_suffix_name`. SDK NG CMake needs `core_id`, but
current integration layer uses `CONFIG_MCUX_CORE_SUFFIX`
as `core_id`, so there is build error with RT595 F1.
Fix the code that convert `CONFIG_MCUX_CORE_SUFFIX` to `core_id`,
handle the special case.
Signed-off-by: Jason Yu <zejiang.yu@nxp.com>
Adapt MAX32690 driver to use Wrap_MXC_CAN_Init to handle differences
in the MSDK API (see analogdevicesinc/msdk#1306) between the
MAX32690 and MAX32662.
can_driver_api.timing_min required phase_seg1 >= 3 and phase_seg2 >= 2
when configuring CAN bit timing. Both microcontrollers covered by this
driver (MAX32662, MAX32690) support values down to 1 for both of these
timing parameters.
Refer to the docs for registers CAN_BUSTIM1, CANn_BUSTIM1.
Add a can0 node to the MAX32662 dtsi.
Signed-off-by: Ioan Dragomir <ioan.dragomir@analog.com>
Currently, the `display_read_write` test execution can be suppressed
by `harness: display`, but since the display harness does not exist,
there is no way to run tests.
Using a fixture, it will be possible only to build if not specified,
and run tests if specified.
Signed-off-by: TOKITA Hiroshi <tokita.hiroshi@gmail.com>
Drop this redundant check that was likely meant to be checking that Z
field is zero but in the case of reception the RFC says it MUST be
ignored anyway.
Signed-off-by: Benjamin Cabé <benjamin@zephyrproject.org>
When CONFIG_NO_OPTIMIZATIONS the compiler will not inline
socket_offload_dns_is_enabled(), which means calls to
socket_offload* remain, and the linker will fail with
in function `zsock_getaddrinfo': undefined reference to
`socket_offload_getaddrinfo'
in function `zsock_freeaddrinfo': undefined reference to
`socket_offload_freeaddrinfo'
Instead of relaying on that function being inlined and the
if'ed code being removed, let's just use the preprocessor.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
Add external HSPI flash on STM32U5G9J-DK2
Use external loader to flash it with STM32CubeProgrammer
Signed-off-by: Mathias Markussen <mathias.markussen@st.com>
- with the updates made to the spi_ll_stm32 driver,
we no longer need to configure the 16-bit frame/word size
in the DTS overlay file. The test_spi_word_size_16 testsuite
helps us verify the 16-bit frame mode supported on
the ST platform.
- remove testcases that use thoses files in testcase.yaml
Signed-off-by: Fabrice DJIATSA <fabrice.djiatsa-ext@st.com>
Update the source_data and dest_data sizes with the frame size
in bytes before performing the transfer.
With these changes, we can handle cases where the buffers switch
between 8-bit and 16-bit modes.
Signed-off-by: Fabrice DJIATSA <fabrice.djiatsa-ext@st.com>
Workaround to avoid getting the -EFAULT (14) error when running
the test_spi_word_size_x (x = 7, 9, 24) test cases:
Since these testcase configurations aren't supported, update
the error type to ENOTSUP to allow skipping the test cases using ztest.
Signed-off-by: Fabrice DJIATSA <fabrice.djiatsa-ext@st.com>
Place transfer buffers in non-cacheable memory when
CONFIG_NOCACHE_MEMORY=y. This change ensures that DMA transfer
buffers are allocated in non-cacheable memory on platforms where
CONFIG_NOCACHE_MEMORY is enabled. This avoids potential cache
coherence issues that are not handled by the SPI driver.
Signed-off-by: Fabrice DJIATSA <fabrice.djiatsa-ext@st.com>
Set a convenient latency for each platform, since each platform
reacts differently to the driver.
Signed-off-by: Fabrice DJIATSA <fabrice.djiatsa-ext@st.com>
The driver treats the `source_data_size` and `dest_data_size` as a
width in bits and converts 8 bits to 1, 16 bits to 2, and 32 bits to 3.
This should be a width in bytes with 1 byte mapping to 0, 2 bytes to
1, and 4 bytes to 3.
Note that this preserves the current behaviour of silently accepting
invalid transfer bit widths.
Signed-off-by: Michael Hope <michaelh@juju.nz>
This adds `tests/drivers/dma/chan_blen_transfer` and
`tests/drivers/dma/loop_transfer` as build tests to any manual Twister
run.
Signed-off-by: Michael Hope <michaelh@juju.nz>
Sets LinkServer as the default runner for the mimxrt1060-evk board,
as the board is configured for CMSIS-DAP by default.
Signed-off-by: Andrej Butok <andrey.butok@nxp.com>
