This is meant as a substitute for sys_clock_timeout_end_calc()
Current sys_clock_timeout_end_calc() usage opens up many bug
possibilities due to the actual timeout evaluation's open-coded nature.
Issue ##50611 is one example.
- Some users store the returned value in a signed variable, others in
an unsigned one, making the comparison with UINT64_MAX (corresponding
to K_FOREVER) wrong in the signed case.
- Some users compute the difference and store that in a signed variable
to compare against 0 which still doesn't work with K_FOREVER. And when
this difference is used as a timeout argument then the K_FOREVER
nature of the timeout is lost.
- Some users complexify their code by special-casing K_NO_WAIT and
K_FOREVER inline which is bad for both code readability and binary
size.
Let's introduce a better abstraction to deal with absolute timepoints
with an opaque type to be used with a well-defined API.
The word "timeout" was avoided in the naming on purpose as the timeout
namespace is quite crowded already and it is preferable to make a
distinction between relative time periods (timeouts) and absolute time
values (timepoints).
A few stacks are also adjusted as they were too tight on X86.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
When building with the native simulator instead of attempting to
call directly to the host libC, use the trampolines provided
by the runner.
In this way we can build this code even if we are building
Zephyr with an embedded C library.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
So they depend or select on the right NATIVE_BUILD
instead of NATIVE_APPLICATION.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
When writing a test suite, it's more common to want the 'before'
hook to run in privilaged mode, even when the test is run in userspace.
Reconfigure ztest to first run the test thread callback in privilaged
mode and only enter userspace after the test rule and suite's 'before'
functions ran.
Signed-off-by: Yuval Peress <peress@google.com>
The extra ztest functionality for the posix arch
requires the host libC. Disable it if we are
building with an embedded libC.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
This adds a few line use zephyr_syscall_header() to include
headers containing syscall function prototypes.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Before this change we had 2 top-level Doxygen entries for Testing APIs:
- Zephyr tests
- Zephyr testing suite
This patch creates a single top-level group to contain ZTest and FFF
extensions.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
All posix arch boards are required to provide the same
tracing/print API.
So, instead of having a different driver for native_posix and the
bsim boards, let's have a common one which uses this API.
This in turn results in the printk strings being printed in
the same underlaying backend as before with individual drivers.
A part from this, the native_posix console driver was a full
backend for the now long gone Zephyr console shell
(named legacy_shell from 527256501f
until it was retired in fd0b7f7767).
The whole input handling in this driver was dead code
(since 140a8d0c8a)
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
The current approach of failing the build on ztest with no optimization
broke coverage builds, and generally raised some concerns about being
too aggressive.
Downgrade the error to a warning and rework the option to inhibit the
warning, while also dropping it automatically for POSIX (that are not
really affected by stack size) and coverage run (that always runs with
no optimization).
Will reconsider this down the road if we still see issues filed for the
tests broken with no optimization and no further tuning.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
Many tests are known to fail when built and no compiler optimizations.
Add a CMake check to error out when building a ztest based test with no
optimization, ask not file issues about it but also adds an opt-out
option to bypass the error for tests are actually designed to work in
this setup.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
Until now iterable sections APIs have been part of the toolchain
(common) headers. They are not strictly related to a toolchain, they
just rely on linker providing support for sections. Most files relied on
indirect includes to access the API, now, it is included as needed.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Fix ztress_execute() out-of-bounds in case it is called with timer
and for maximum number of threads allowed (CONFIG_ZTRESS_MAX_THREADS).
Signed-off-by: Dmitrii Golovanov <dmitrii.golovanov@intel.com>
Corrected printed message on failing test with expected or received
NULL pointer in ztest_check_expected_data.
Signed-off-by: Stine Åkredalen <stine.akredalen@nordicsemi.no>
MISRA Rule 5.7 requires uniqueness of tag identifiers. Shell is
frequently problematic because many code uses `const struct shell
*shell`. This causes CI noise every time one of these shell files is
edited, so let's update all of them with `const struct shell *sh`
instead.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
As both C and C++ standards require applications running under an OS to
return 'int', adapt that for Zephyr to align with those standard. This also
eliminates errors when building with clang when not using -ffreestanding,
and reduces the need for compiler flags to silence warnings for both clang
and gcc.
Most of these changes were automated using coccinelle with the following
script:
@@
@@
- void
+ int
main(...) {
...
- return;
+ return 0;
...
}
Approximately 40 files had to be edited by hand as coccinelle was unable to
fix them.
Signed-off-by: Keith Packard <keithp@keithp.com>
Fix checkpatch issue: UNNECESSARY_INT: Prefer 'unsigned long' over
'unsigned long int' as the int is unnecessary
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Many areas of Zephyr divide and round up without using the DIV_ROUND_UP
macro. Make use of it, so that we make use of a tested system macro and
at the same time we make code more readable.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The init infrastructure, found in `init.h`, is currently used by:
- `SYS_INIT`: to call functions before `main`
- `DEVICE_*`: to initialize devices
They are all sorted according to an initialization level + a priority.
`SYS_INIT` calls are really orthogonal to devices, however, the required
function signature requires a `const struct device *dev` as a first
argument. The only reason for that is because the same init machinery is
used by devices, so we have something like:
```c
struct init_entry {
int (*init)(const struct device *dev);
/* only set by DEVICE_*, otherwise NULL */
const struct device *dev;
}
```
As a result, we end up with such weird/ugly pattern:
```c
static int my_init(const struct device *dev)
{
/* always NULL! add ARG_UNUSED to avoid compiler warning */
ARG_UNUSED(dev);
...
}
```
This is really a result of poor internals isolation. This patch proposes
a to make init entries more flexible so that they can accept sytem
initialization calls like this:
```c
static int my_init(void)
{
...
}
```
This is achieved using a union:
```c
union init_function {
/* for SYS_INIT, used when init_entry.dev == NULL */
int (*sys)(void);
/* for DEVICE*, used when init_entry.dev != NULL */
int (*dev)(const struct device *dev);
};
struct init_entry {
/* stores init function (either for SYS_INIT or DEVICE*)
union init_function init_fn;
/* stores device pointer for DEVICE*, NULL for SYS_INIT. Allows
* to know which union entry to call.
*/
const struct device *dev;
}
```
This solution **does not increase ROM usage**, and allows to offer clean
public APIs for both SYS_INIT and DEVICE*. Note that however, init
machinery keeps a coupling with devices.
**NOTE**: This is a breaking change! All `SYS_INIT` functions will need
to be converted to the new signature. See the script offered in the
following commit.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
init: convert SYS_INIT functions to the new signature
Conversion scripted using scripts/utils/migrate_sys_init.py.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
manifest: update projects for SYS_INIT changes
Update modules with updated SYS_INIT calls:
- hal_ti
- lvgl
- sof
- TraceRecorderSource
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
tests: devicetree: devices: adjust test
Adjust test according to the recently introduced SYS_INIT
infrastructure.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
tests: kernel: threads: adjust SYS_INIT call
Adjust to the new signature: int (*init_fn)(void);
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
For ztest twister would parse the symbol information that was
generated as part of the build (zephyr.symbols). However the format
of the zephyr.symbols files is highly dependant on the toolchain.
Move to using pyelf to parse the symbol information directly from
zephyr.elf instead so that this works regardless of toolchain.
Signed-off-by: Kumar Gala <kumar.gala@intel.com>
Some cases, especially those dealing with faults need to inject output
manually, so provied a macro for special case.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Add supplementary header <zephyr/fff_extensions.h>. Add macro
to fff_extensions.h for simplifying definition of custom fake functions
needing call-unique information for producing desired output data.
When an array of custom fake context structures is defined and
the return field within the first structure instance is registered
with the standard SET_RETURN_SEQ() macro of FFF, the
RETURN_HANDLED_CONTEXT() macro provides the inverse logic to
recover the context structure for this called instance. The body of
the custom fake handler is provided to the RETURN_HANDLED_CONTEXT()
macro for appropriate execution and access to the custom fake
parameters.
A test suite is also provided to verify macro implementation and
illustrate usage. It is at:
zephyr/tests/subsys/testsuite/fff_fake_contexts/
This code was verified by:
1. (Pass) west build -p always \
-b unit_testing tests/subsys/testsuite/fff_fake_contexts/ && \
./build/testbinary
2. (Pass) west build -p always \
-b native_posix tests/subsys/testsuite/fff_fake_contexts/ && \
./build/zephyr/zephyr.exe
3. (Pass) ./scripts/twister -p unit_testing \
-T tests/subsys/testsuite/fff_fake_contexts/
4. (Pass) ./scripts/twister -p native_posix \
-T tests/subsys/testsuite/fff_fake_contexts/
5. (Pass) cd doc && build html-fast
Fix#55246
Signed-off-by: Gregory Shue <gregory.shue@legrand.com>
Add the zexpect API, inspired by GoogleTest's EXPECT API. This API reports
test failures while allowing test execution to continue. This enables test
reports to show more than a singule failing property on a failing test.
Signed-off-by: Aaron Massey <aaronmassey@google.com>
LEGACY_INCLUDE_PATH has been defaulting to "n" and marked as deprecated
in both v3.2 and v3.3. Drop the option entirely for v3.4.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
While porting the coverage.c file from RIOT to Zephyr, which employs
different compiler flags, I noticed several -Wcast-align GCC warnings on
arm. I think, as is, the current implementation may perform unaligned
memory accesses which may not be supported on certain platforms. To
workaround that, I have rewritten the code for RIOT using
bytewise-writes with `memcpy`.
Signed-off-by: Sören Tempel <tempel@uni-bremen.de>
This commit updates all in-tree code to use `CONFIG_CPP` instead of
`CONFIG_CPLUSPLUS`, which is now deprecated.
Signed-off-by: Stephanos Ioannidis <stephanos.ioannidis@nordicsemi.no>
Accurate timekeeping is something that is often taken for granted.
However, reliability of timekeeping code is critical for most core
and subsystem code. Furthermore, Many higher-level timekeeping
utilities in Zephyr work off of ticks but there is no way to modify
ticks directly which would require either unnecessary delays in
test code or non-ideal compromises in test coverage.
Since timekeeping is so critical, there should be as few barriers
to testing timekeeping code as possible, while preserving
integrity of the kernel's public interface.
With this, we expose `sys_clock_tick_set()` as a system call only
when `CONFIG_ZTEST` is set, declared within the ztest framework.
Signed-off-by: Chris Friedt <cfriedt@meta.com>
Test summary can add a lot of noise to the logs when debugging
a specific test using `-test` argument.
Add control to turn this off.
Signed-off-by: Al Semjonovs <asemjonovs@google.com>
Some missing features for getting coverage data for unit tests:
- Setting the unit_testing board to have coverage support and native
application.
- Fixing the CONFIG_COVERAGE check
Signed-off-by: Yuval Peress <peress@google.com>
The heap size for serial dump of gcov data is currently
defined by MALLOC_MAX_HEAP_SIZE, which cannot be
adjusted by kconfig. This commit adds a new kconfig
option, CONFIG_COVERAGE_GCOV_HEAP_SIZE, which retains
the behavior or MALLOC_MAX_HEAP_SIZE by default.
Signed-off-by: Jason Wright <jwright@synchron.com>
We get compiler warnings on testcases that set CONFIG_MP_NUM_CPUS=1
and platforms that have CONFIG_SMP=y. Qualify the code so its only
built if CONFIG_SMP && (CONFIG_MP_NUM_CPUS > 1).
Signed-off-by: Kumar Gala <kumar.gala@intel.com>
Move the function in the `subsys/testsuite/ztest/src/ztest_mock.c` files.
This is motivated by the fact that there is others re-implementation of
`*printk` functions using libc counterparts in the `ztest_mock.c` file.
Signed-off-by: Théo Battrel <theo.battrel@nordicsemi.no>
For test_1cpu_start/test_1cpu_stop make the code only build if
CONFIG_SMP and move to using arch_num_cpus() for runtime loops
and CONFIG_MP_MAX_NUM_CPUS for array decleration.
Signed-off-by: Kumar Gala <kumar.gala@intel.com>
3000 milliseconds may not always be enough time for all 1cpu type
tests to finish on all platforms. Making the CPU hold time
configurable allows for additional flexibility.
Signed-off-by: Peter Mitsis <peter.mitsis@intel.com>
Change for loops of the form:
for (i = 0; i < CONFIG_MP_NUM_CPUS; i++)
...
to
unsigned int num_cpus = arch_num_cpus();
for (i = 0; i < num_cpus; i++)
...
We do the call outside of the for loop so that it only happens once,
rather than on every iteration.
Signed-off-by: Kumar Gala <kumar.gala@intel.com>
Adding the new Kconfig (enabled by default) to make a failed assumption
mark the final result as failed. This change has the following benefits
which have been asked for by the Zephyr community:
1. A failed assumption does not go silent. In this example, the failed
assumption will still mark the test as skipped, but the final result
will be to mark the full test run as failed. This would allow
blocking the CI when an assume fails.
2. Normal test skipping via the ztest_test_skip() is unaffected by this
change. Those tests will be marked as skipped, but the binary will
still pass.
Signed-off-by: Yuval Peress <peress@google.com>
This fixes the compiler warning "comparison of integer expressions of
different signedness: 'int' and 'unsigned int' [-Werror=sign-compare]"
Signed-off-by: Benjamin Gwin <bgwin@google.com>
Update fff.h to the pending PR upstream which allows for using a custom
function signature. This enables the use of C++ std::function as the
mock.
Signed-off-by: Yuval Peress <peress@google.com>
Due to the diverse coding styles and lack of preprocessing when
scanning for test cases, there were many unintended combinations
of yaml test scenarios and C test functions with the regex-based
test case discovery, which caused an inaccurate test plan and test
result stats.
As the new ztest fx is used, the test cases of a test instance can
be accurately determined via zephyr.symbols file.
Because the zephyr.symbols file is not available until after build,
test cases determination is placed right after the build stage of
the pipeline and before the runtime test collection.
For those test instances that don't go through the build stage,
such as those with "Skip filtered" (statically filtered) reason
before entering the pipeline, they are not affected.
This patch also adjust the stats logic because the zephyr.symbols
file is generated after build. So ExecutionCounter update is split
and some must be postponed until test execution pipeline is completed.
Some concepts:
A test instance = a yaml scenario + a platform
"Test instance" and "test config" are synonyms in twister, unfortunately
excessive IMHO...
A test instance can be filtered or skipped in 3 ways.
Let's define "runtime" as "after entering the execution pipeline".
1) statically filtered (before runtime)
Such test instance is filtered by analyzing the yaml and never
enters the execution pipeline.
2) cmake filtered (runtime)
Such test instance enters pipeline and is filtered at cmake stage.
3) build skipped (also runtime)
Such test instance enters pipeline and is skipped at build stage.
All the test instances that enter the execution pipeline will go
through the report stage, where ExecutionCounter is further updated.
The meaning of the fields of ExecutionCounter are:
.skipped_configs = static filtered + cmake filtered + build skipped
.skipped_runtime = cmake filtered + build skipped
.skipped_filter = static filtered
.done = instances that enter the execution pipeline
.passed = instances that actually executed and passed
Definition of the overall stats:
total_complete = .done + .skipped_filter
total = yaml test scenario * applicable platforms
complete percentage = total_complete / total
pass rate = passed / (total - skipped_configs)
Signed-off-by: Ming Shao <ming.shao@intel.com>
Some testsuites dump lots of output very fast where some systems are not
able to capture the complete output from the tests. Add a slight delay
between each test in the suite.
New kconfig CONFIG_ZTEST_TEST_DELAY_MS is added to ztest.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
The GCC/GCOV version over 12 has slight format change of the gcno
and gcda. Make some adaption in the gcov dump function to fix the
code coverage report.
Mainly two places change:
1. Added the checksum in the struct gcov_info. This fix the crash
in qemu_x86, and mps2_an385 when run with --coverage.
2. Adjust the GCOV_TAG_FUNCTION_LENGTH accroding to gcov-io.h. It's
length unit is caculated by bytes now.
Fixes#50255.
Fixes#50257.
Signed-off-by: Enjia Mai <enjia.mai@intel.com>
When CONFIG_ZTEST_SHUFFLE is enabled and ASSERTS are disabled
`start_pos` becomes an unused variable leading to a compile error.
Cleaned-up shuffling algorithm to not need a `start_pos` check.
Signed-off-by: Al Semjonovs <asemjonovs@google.com>
