This adds the bits to test the displaying the usage of
privileged stacks for user threads. Note that this only
adds the printk part as this is to test the code path to
fetch usage data rather than the display part.
Also add qemu_xtensa/dc233c/mmu to platform allow list so
the privileged stack code can be tested on that platform.
This replaces mps2/an385 with qemu_x86 as an integration
platform due to the privileged stack test. Since ARM does
not currently support getting privileged stack info, test
would result in twister error due to kconfig warnings.
The board cannot be excluded since it's an integration
platform. So for now, use qemu_x86 as an integration for
32-bit testing. Once ARM has proper support, we can go
back to having mps2/an385 as an integration platform.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
This disables QEMU icount for all the simulated platforms
in the allowed list. With icount enabled, the printout from
the thread analyzer would be continuous and non-stop instead
of the default once every 60 seconds. Besides, the console
harness only needs one, and these is no need to waste CPU
cycles and memory for processing the extra ones.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Update `shell_backend_uart` to take into account the new line print by
`z_cursor_next_line_move`.
Signed-off-by: Théo Battrel <theo.battrel@nordicsemi.no>
Instead of using filter (which is expensive in terms of CI as
it needs to run cmake on everything to get the full kconfigs
before filtering can be done), change that to a curated list
of allowed platforms. This avoids the expensive filtering
operation, yet can still run on some emulated platforms to
verify the dictionary output.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Wipe partition header prior to fatfs testsuite starting. This insures
that the disk will be in an unformatted state at the start of the test,
so the first fs_mount() calls will fail as expected.
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
Add support for the ST nucleo h755zi_q board which
uses an stm32h755 in the st nucleo-144 board
format.
Signed-off-by: Mike Banducci <michael.banducci@sandc.com>
This adds platform_allow to limit the scope of platforms to
build and run. This avoids the need to constantly nudge
the platform filter lists to exclude platforms which fail
to build or run due to incompatible kconfigs.
Fixes#76945
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
It's usual to show debug logs as blue.
Showing dbg and info with different colors are good,
but it's preferable to keep info logs with default
color.
Signed-off-by: Félix Piédallu <felix@piedallu.me>
Move samples config files from 'boards' to 'socs' in order to
remove multiple files with the same configuration and render
available samples for new boards. Only changed sample files
which are not board or hardware specific.
Signed-off-by: Raffael Rostagno <raffael.rostagno@espressif.com>
Today the thread analyzer depends on !ARCH_POSIX,
so this filter needs to filter any board for this architecture
not just native_sim.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
Packet buffer is designed to run writer and reader code on two CPUs.
The write function cannot be preempted by the read function in a CPU,
or the cache management could result in data corruption. In the tests,
the reader and the writer are executed on a single CPU. Disable data
cache management to prevent potential data corruption.
Signed-off-by: Krzysztof Chruściński <krzysztof.chruscinski@nordicsemi.no>
The test `storage.flash_map.mpu` for `nrf52840dk/nrf52840` fails
althought the output is correct. Time to test execution is not enough.
This change extend test execution timeout.
Signed-off-by: Katarzyna Giądła <katarzyna.giadla@nordicsemi.no>
Add z_test which uses new configs to capture multiple
threads in a core dump and with all of the context
necessary to debug the threads.
Signed-off-by: Mark Holden <mholden@meta.com>
Use an explicit callback name so that multiple instances of this do not
result in a:
redefinition of '_input_callback__longpress_cb'
error. This used to work when it was using unique generated wrappers,
but now it needs an index in the callback name.
Use it in one of the API tests as well, just in case.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
Add a void *user_data pointer to the input callback structure. This is
useful for driver to get back the driver data structure and avoid
defining wrapper functions.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
Add the missing timeout parameter to `modem_pipe_open()` and
`modem_pipe_close()` calls.
10 seconds is the default value used in the Zephyr tree.
Fixes a regression introduced in
https://github.com/zephyrproject-rtos/zephyr/pull/74325.
Signed-off-by: Tomi Fontanilles <tomi.fontanilles@nordicsemi.no>
The commit removes one-line overlay files, when copyrights excluded,
with direct Kconfig options selection inside testcase.yaml.
The overlays have been removed.
Signed-off-by: Dominik Ermel <dominik.ermel@nordicsemi.no>
The modem pipe APIs include synchronous calls to open/close,
which internally use a fixed timeout of 10 seconds. The timeout
should be configurable through the APIs, anywhere from K_NO_WAIT
to K_FOREVER.
This commit adds timeout parameters to the open/close APIs, and
updates in-tree usage of the open/close APIs to explicitly
provide the previously implicit timeout of 10 seconds.
Signed-off-by: Bjarki Arge Andreasen <bjarki@arge-andreasen.me>
Raise the descriptor table span to 201 due to the inclusion of the USB
configuration descriptor.
Tested with:
$ west build -p -b it82xx2_evb tests/subsys/usb/desc_sections
-T usb.desc_sections
Signed-off-by: Ren Chen <Ren.Chen@ite.com.tw>
Disable the flash disk driver in the RAM driver test to prevent the
test using the wrong disk in the tests.
Signed-off-by: Jordan Yates <jordan@embeint.com>
Change USBD_CONFIGURATION_DEFINE macro to take the address of a string
descriptor node as an argument. This is a breaking change for macro
users, but quite convenient and easy to implement.
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
Change the `platform_allow` to multi-line format so that newly
supported archs/boards can be added as a new line.
Refactor the tags out to `common`, and add
`CONFIG_ARCH_HAS_GDBSTUB` filter
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
Signed-off-by: Yong Cong Sin <yongcong.sin@gmail.com>
There are instances where on hardware tests would result in
multiple dictionary logging blocks. This is usually due to
flashing and running via twister in separate steps. Once
flashing is done, the device starts running and sending
logging string to output, but twister is not ready to read
them, thus leaving the strings in the host's UART buffer.
Then twister starts its testing run by resetting the device
which results in another dictionary logging block to show
up. So the pytest script has to be updated to only look at
the last dictionay logging block.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Since only the UART backend supports output in hexidecimal form
with dictionary logging, so filter for that or else the pytest
would fail due to none, or wrong captured output.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
When NO_BACKENDS was not defined then log backends were
initialized (even though not used). During the initialization
filters for that backends were configured and because of that
messages were not filtered out as expected.
Signed-off-by: Krzysztof Chruściński <krzysztof.chruscinski@nordicsemi.no>
Currently we get the build error when building
'tests/subsys/logging/log_backend_uart':
error: static_assert expression is not an integral constant expression
BUILD_ASSERT(strlen(TEST_DATA) < SAMPLE_DATA_SIZE);
~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
It can be easily workarounded by using sizeof instead of strlen
to calculate string size.
Signed-off-by: Evgeniy Paltsev <PaltsevEvgeniy@gmail.com>
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>
In `tests/subsys/logging/dictionary` we don't include stdio.h
header but use standard stream definitions (i.e. `stdout`)
which cause build issues (if the minimal libc is used).
Fix that.
Signed-off-by: Evgeniy Paltsev <PaltsevEvgeniy@gmail.com>
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>
Some AMP native simulator targets (like the nrf5340bsim net core)
will not produce a final executable if only one of the core images
is built, when IPC is enabled, as those images require the other
core image to be able to link.
So this test cannot be run when targetting that core alone
(without using sysbuild to build the other).
Let's filter out these targets to avoid twister building
but finding itself with no executable and error out badly
as it cannot run the test.
The filter condition is effectively:
If building for a native target, there must be an .exe being
produced.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
Some hardware takes longer to boot before any output is visible
from the app. So lengthen the initial timeout on waiting for
any output.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
When doing device testing, there may be some other characters
being printed before the dictionary logging header string.
Also the delay boot banner may also be there. So change
the regex to ignore those them.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
With sysbuild, the test app is actually built under another
layer of directory where the dictionary JSON file will reside.
Instead of simply using dut.device_config.build_dir, use
dut.devive_config.app_build_dir instead, which takes into
account of that.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Add pytest to test the output of dictionary logging to make sure
the encoded logs can be decoded back into strings, and to also
make sure the decoded logs have the expected strings.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
This is in preparation for introducing pytest to test the output
of dictionary logging. Code is the same, but the testcase.yaml
is trimmed to only those that can be tested.
Note that a copy is made instead of moving the whole sample over
is simply due to various documentation having references to
the sample. Since RST files under tests/ are not parsed for
zephyr:code-sample:<name>, the linking would have been vanished.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Add test cases for testing basic CFB functions.
These tests cannot run in CI because the CI environment has no display.
Mark these tests as `build_only`.
Signed-off-by: TOKITA Hiroshi <tokita.hiroshi@gmail.com>
`checkpatch.pl` requires that dts sources are indented with tabs,
fix all the spaces that slipped in while checkpatch wasn't watching.
Signed-off-by: Jordan Yates <jordan@embeint.com>
