Some devices do not need to perform any initialization, so allow the
init function to be NULL. In this case, the initialization code will
just mark the device as initialized, i.e. ready.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
If the timer driver only implements sys_clock_cycle_get_32() (meaning
CONFIG_TIMER_HAS_64BIT_CYCLE_COUNTER=n) and the hardware clock is high
enough then the reported cycle count may wrap an uint32_t during the
test. This makes validating the total test duration pointless as it
cannot be measured. Just print a warning instead of failing the test
in that case.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
Looks like switching the main return value to int means that stack
frame persists and increases stack usage by a few bytes. Increase the
main stack size to avoid overflows.
Signed-off-by: Keith Packard <keithp@keithp.com>
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>
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>
HiFive Unmatched is using size and address cells of length 2. It has to
use different overlays (with reg properties of correct length).
Signed-off-by: Franciszek Zdobylak <fzdobylak@antmicro.com>
When building this test with the armclang compiler we get the following
warning:
tests/kernel/interrupt/src/dynamic_isr.c
tests/kernel/interrupt/src/dynamic_isr.c:23:32: error: 'used' attribute
ignored on a non-definition declaration [-Werror,-Wignored-attributes]
extern struct _isr_table_entry __sw_isr_table _sw_isr_table[];
^
There is no need to add the __sw_isr_table on the extern, so remove it
to address the warning.
Signed-off-by: Kumar Gala <kumar.gala@intel.com>
Clang complains when an unsigned value is passed to abs, even though there
is an implicit cast to a signed type. Insert an explicit cast to make clang
happy.
Signed-off-by: Keith Packard <keithp@keithp.com>
TC_START is used to evaluate output of tests and is used internally by
ztest when a test starts, no need to call this manually here.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
TC_START is used to evaluate output of tests and is used internally by
ztest when a test starts, no need to call this manually here.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
TC_START is used to evaluate output of tests and is used internally by
ztest when a test starts, no need to call this manually here.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Nordic targets use 24-bit RTC peripheral for system clock. Nordic system
clock timeout implementation relies on RTC CC (capture compare) when
the timeout is in future. Nordic system clock driver allows setting
alarm only to 3 or more counts from current counter value due to silicon
limitation (to ensure that CC event triggers before counter overflow).
RTC CC limitation does not have much impact on normal applications where
there is no need to schedule such short timeouts, but is problematic in
a timer test that expects being able to repeatedly schedule timeouts on
subsequent ticks.
Reduce system tick rate to 8192 on nRF targets to allow setting CC to
the very next tick. With system tick rate being 4 times less than the
hardware tick rate, it is always possible to schedule timeout to happen
in the next tick because ticks are 4 counts apart, i.e. current timer
value + 3 never runs past the next tick.
Fixes: #54211
Signed-off-by: Tomasz Moń <tomasz.mon@nordicsemi.no>
This commit marks testcases that require working Power Managament with
the appropriate `pm` tag to allow proper testcase filtering in the board
YAML file.
Signed-off-by: Filip Kokosinski <fkokosinski@antmicro.com>
SMP tests `inc_concurrency` and `smp_switch_torture` use a 'stressing'
approach to verify their results: run something for some time (or some
number of repetitions). However, in some environments, current 'stress'
levels can be quite high, making tests take a long time - environments
like emulators/simulators.
This patch adds a Kconfig that allows one to define a percentage factor
to the 'stress' (time or repetitions) used by these tests.
Signed-off-by: Ederson de Souza <ederson.desouza@intel.com>
If there are any CPU exceptions, printing a failed message
would allow twister to stop early instead of waiting for
timeout.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
If there are any CPU exceptions, printing a failed message
would allow twister to stop early instead of waiting for
timeout.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
The variable need_recover_spinlock is always set to false so
the spinlock recovery code is effectively no-op. So remove
everything related to the variable.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
A new Z_SPIN_DELAY() macro has been added which
can be used to reduce a bit the amount of noise
due to the POSIX arch need to break busy loops with
k_busy_wait().
Use it.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
Don't sample the first entry outside the timer as this is a different
code path which produces a different offset from the clock tick.
Use sys_clock_hw_cycles_per_sec() to be compatible with systems that
read their hardware clock frequency at run time.
Perform cycle difference computations with uint64_t. If ever the
magnitude of the absolute clock cycle values is greater than 52 bits
then the cast to a double will actually lose accuracy.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
Adds REQUIRED to samples and tests for finding the zephyr package
to align all samples and tests with the same call and parameters.
Signed-off-by: Jamie McCrae <jamie.mccrae@nordicsemi.no>
An assertion statement was a bit too strict. Period drift may come about
not only from kernel ticks being large but also from time conversion being
inexact due to division truncation.
Fixes: #55136
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
For some kernel tests, faults and exceptions are expected.
They are caught and the test would continue if the reasons
for faults are as expected. However, when the unexpected
reasons are encountered, the code simply prints a message
and calls k_fatal_halt(). When running under twister,
these messages are not the expected failed messages so
twister will spin till timeout although the execution
has already been halted. This adds another printk() before
halt to signal twister that the test has failed and bails
early.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
...test_inherit_resource_pool. This waits for the newly created
threads to finish before moving on to the next test. This fixes
an issue on qemu_x86_tiny where there would be a double fault
after all tests have run.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
qemu_x86 seems to take an extra instruction after the sti instruction
(irq_unlock) happens before it posts the interrupts. This can issues
if the instruction after the sti ends up reading the state that is
suppose to be updated by the ISR handler.
We see this behavior when building with LLVM. To workaround this issue
we add an arch_nop() to provide an extra instruction to allow the
interrupts to post.
Opened zephyrproject-rtos/sdk-ng#629 to track qemu issue.
Signed-off-by: Kumar Gala <kumar.gala@intel.com>
LLVM doesn't support SSE + 387 math. As such if SSE is enabled we
have to utilize SSE floating point. To utilize 387 math, SSE has
to be disabled.
Update the floating point related tests to introduce 387 only variants
that will build on both GCC & LLVM based tools. Than we exclude llvm
based (llvm, oneApi) toolchains from the CONFIG_X86_SSE_FP_MATH=n and
CONFIG_X86_SSE=y test variants.
Signed-off-by: Kumar Gala <kumar.gala@intel.com>
Looks like some implementors decided not to implement the full set of
PMP range matching modes. Let's rearrange the code so that any of those
modes can be disabled.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
This test relies on one thread interrupting another to exercize the FPU
sharing. On SMP those threads get one CPU each with no sharing of their
FPU making the test rather pointless.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
Fix comments in board DTS files referring to AN521 tables defining
memory areas, and choose node label names that more accurately reflect
the entries of interest in those tables.
Adjust the one in-tree user of the affected node labels.
Signed-off-by: Martí Bolívar <marti.bolivar@nordicsemi.no>
When building with LLVM on qemu_x86 we see the compiler ends up
inlining the check_input function. This breaks the stack overflow
that the test is trying to generate, so mark the check_input()
function as noinline to fix the issue.
Signed-off-by: Kumar Gala <kumar.gala@intel.com>
Provide an estimate of the test duration.
Make the output nicer than a few overloaded and wrapped lines.
Provide more context in the presence of period time drift.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
Print the "perfect" reference period for easier evaluation.
Suggest a remedy to the missed ticks problem.
Still, that wasn't satisfactory. Implemented a count of missed ticks
to get to the bottom of this issue. Found that missed ticks always came
to a perfect count of 40.
Incidentally, the busy loop prints a line every 250 ms and the test spans
10 seconds. There are no such coincidences.
Turns out that CONFIG_PRINTK_SYNC was set by default. This disables IRQs
for the serial output duration, which can be quite long at 115200 bauds.
Given a 60-ish character line length, this represents more than 5 ms of
no IRQ servicing during a timer latency measurement test which is bad.
So make sure CONFIG_PRINTK_SYNC=n for proper statistics.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
Disable tests/kernel/mem_protect/syscalls for qemu_arc_em where
we trigger ARC QEMU bug which cause illegal instruction exception
on perfectly valid ARC code.
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>
Signed-off-by: Evgeniy Paltsev <PaltsevEvgeniy@gmail.com>
Change expected reason code for cpu exception to be generic and
in compliance with a3774fd51aFixes#54335
Signed-off-by: Aastha Grover <aastha.grover@intel.com>
This test has trouble on qemu_x86_tiny and randomly generates a Double
Fault error. I couldn't get it to reliably run with picolibc as a Double
Fault usually occured before the test completed.
I spent a couple of hours attempting to track this down and found that it
happens when code pages for the main thread get unmapped because the
qemu_x86_tiny intentionally offers very few available PTEs.
Work around this by just using the minimal libc for this test.
Signed-off-by: Keith Packard <keithp@keithp.com>
Tune quantum parameter for selected kernel tests
targetting the HiFive Unleashed platform.
Those tests require higher fidelity of the virtual
time flow which is achievable on multi-core platforms
in Renode by reducing the quantum.
Signed-off-by: Jan Malek <jmalek@internships.antmicro.com>
Signed-off-by: Mateusz Holenko <mholenko@antmicro.com>
With lazy FPU context switching, k_float_disable() is merely triggering
a synchronous FPU context save and k_float_enable() is a no-op.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
After the dbe3874079 - (tests: kernel/smp: wait for threads to exits
between tests) I've started seeing sporadic kernel.multiprocessing.smp
test failures on our platforms.
------------------------------->8---------------------------------
[*snip*]
===================================================================
START - test_fatal_on_smp
E: r0: 0x3 r1: 0x0 r2: 0x0 r3: 0x0
E: r4: 0x80000194 r5: 0x0 r6: 0x0 r7: 0x0
E: r8: 0x800079c4 r9: 0x82802 r10: 0x80008d8c r11: 0x8000dad8
E: r0: 0x3 r1: 0x2712 r2: 0x114 r3: 0x0
E: r4: 0xf4240000 r5: 0x0 r6: 0xf424 r7: 0xbe40
E: r8: 0x2540 r9: 0x0 r10: 0x80008d8c r11: 0x8000db8c
E: r12: 0x8000ddf0 r13: 0x0 pc: 0x80000aec
E: blink: 0x80000ae6 status32: 0x80082002
E: >>> ZEPHYR FATAL ERROR 3: Kernel oops on CPU 0
E: Current thread: 0x8000db8c (test_fatal_on_smp)
E: r12: 0x8000ddf0 r13: 0x0 pc: 0x8000019a
PASS - test_fatal_on_smp in 0.014 seconds
===================================================================
START - test_get_cpu
E: blink: 0x80001490 status32: 0x80082002
E: >>> ZEPHYR FATAL ERROR 3: Kernel oops on CPU 1
E: Current thread: 0x8000dad8 (unknown)
------------------------------->8---------------------------------
The rootcause if that we doesn't proper cleanup resources after
test_fatal_on_smp test case. So child thread we start test_fatal_on_smp
may continue running for some time after the test_fatal_on_smp
test case is finished.
As in the next test case (test_get_cpu) we use same thead structures
again to create new child thread we may actually rewrite some data of
thread which is still running (or vise versa).
As we trigger the crash in test_fatal_on_smp we can't simply join
child thread in the end of test case (as we never get here). We can't
simply use join child thread before we initiate crash in test_fatal_on_smp
either as we don't want to introduce reschedule point here which may break
the test logic.
So, to fix that, we'll just do k_busy_wait in test_fatal_on_smp
thread after we start child thread to wait for thread trigger
exception and being terminated.
To verify that we also assert that child thread is dead by the
time when we stop busy waiting.
Signed-off-by: Evgeniy Paltsev <PaltsevEvgeniy@gmail.com>
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>
Use namespacing with extra_configs in some tests and remove duplicated
scenarios the were made arch or platform specifc.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Adds a check for `K_ERR_ARM_USAGE_ILLEGAL_EPSR` as the reason code
when running this test for `CONFIG_ARMV7_M_ARMV8_M_MAINLINE`.
Signed-off-by: Kevin Townsend <kevin.townsend@linaro.org>
Fix all line-length errors detected by yamllint:
yamllint -f parsable -c .yamllint $( find -regex '.*\.y[a]*ml' ) | \
grep '(line-length)'
Using a limit is set to 100 columns, not touching the commandlines in
GitHub workflows (at least for now).
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
Fix all comments-indentation errors detected by yamllint:
yamllint -f parsable -c .yamllint $( find -regex '.*\.y[a]*ml' ) | \
grep '(comments-indentation)'
This checks that the comment is aligned with the content.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
Fix all thruthy errors detected by yamllint:
yamllint -f parsable -c .yamllint $( find -regex '.*\.y[a]*ml' ) | \
grep '(truthy)'
This only accepts true/false for boolean properties. Seems like python
takes all sort of formats:
https://github.com/yaml/pyyaml/blob/master/lib/yaml/constructor.py#L224-L235
But the current specs only mention "true" or "false"
https://yaml.org/spec/1.2.2/#10212-boolean
Which is the standard yamllint config.
Excluding codeconv and workflow files, as some are using yes/no instead
in the respective documentation.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
The test in its default configuration needs 3600 seconds to complete,
adjust timeout for twister to meet that.
Signed-off-by: Piotr Kosycarz <piotr.kosycarz@nordicsemi.no>
The spin loop to ensure time goes past the timeout is done in terms
of the core clock, while the spin lock is timed on the system clock.
This difference is exasperated on systems where the core clock is much
faster than the system clock and the test failed. Add a significant
multiplier so the test works even when the system clock is much slower.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
Re-enable several mem_protect tests which were disabled due to
issues in ARC QEMU (which are fixed and fixes were propagated to
Zephyr SDK)
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>
Signed-off-by: Evgeniy Paltsev <PaltsevEvgeniy@gmail.com>
... to avoid undesirable delays in waking up of the threads on
scheduled timeouts. In specific configurations such additional load
of the CPU can be even harmful to the test. This was observed for
nRF platforms where the UART_NRF_DK_SERIAL_WORKAROUND Kconfig option
was enabled.
All calls to TC_PRINT() in this test suite are replaced with calls to
LOG_DBG(), so the status messages are not printed by default, but they
can be enabled if needed by changing CONFIG_LOG_DEFAULT_LEVEL.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
In the message reporting the wrong order of woken up threads
(e.g. "thread 3 woke up, expected 2"), provide indexes of
the threads in the timeout order array, not their timeout
order values.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
integration_platforms help us control what get built/executed in CI and
for each PR submitted. They do not filter out platforms, instead they
just minimize the amount of builds/testing for a particular
tests/sample.
Tests still run on all supported platforms when not in integration mode.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
This adds a test to see if z_phys_unmap() can reclaim memory
correctly, so that the next z_phys_map() re-uses the same
address (with identical input arguments).
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
The requirement of being able to spend only 10% of processing time
on execution of timer handlers that are scheduled on every tick is
not really possible to fulfill on platforms like the nRF ones where
the tick period is quite short (~30 us in this case). Relax this
requirement and accept if at least one-third of the processing time
is available for other work while handling the timer tick train.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
The test in its default configuration needs 3600 seconds to complete,
so use such timeout value in testcase.yaml so that twister called with
--enable-slow option can successfully execute it.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
Test assumes that system clock is slow enough so 1 tick timeout
will not expire before k_timer_start function exists. That is
not the case when system clock is fast (relatively to the cpu
clock). Increase the timeout and add synchronization point.
Signed-off-by: Krzysztof Chruscinski <krzysztof.chruscinski@nordicsemi.no>
The mem_map test was skipped on all the phsical x86 boards when
running twister to test them. This error happens when migrating
the new ztest. Remove the incorrect platform allow to fix this
error.
Signed-off-by: Enjia Mai <enjia.mai@intel.com>
Use the picolibc definition of _WANT_IO_LONG_LONG as that can only be set
when building the library, and not selected by the consumer.
Signed-off-by: Keith Packard <keithp@keithp.com>
This avoids problems when using timers for random numbers; run too fast and
all the values are the same.
Signed-off-by: Keith Packard <keithp@keithp.com>
After writing to mapped_rw, we should also check if the backing
buffer has the correct data. Or we could have a situation where
on systems that need explicit cache controls, the newly updated
mapped_rw is cached but the backing buffer still contain old
data. Comparing the backing buffer to mapped_ro does not really
matter in this case as the content would certain match.
Also, this moves the mapping of mapped_ro earlier so that we
map both mapped_rw and mapped_ro because data manipulation.
And that we also need to verify the values of the backing and
mapped buffers.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
There is an assumption on test_page buffer that the MMU page
size is 4kb so that there is a 8kb buffer for read/write.
However, page size may not be 4kb on all architectures.
We need to make sure the test buffer is large enough for
the read/write test.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
The sys_sem.nouser test does not enable userspace which makes
k_thread_access_grant() no-op. However, XCC would still emit
LOOP instructions for the for-loop. Since there is nothing
to do, the XCC assembler complains about it being an empty
loop and errors out. So guard the k_thread_access_grant()
calls so they are only compiled if userspace is enabled.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Previously, this change was added to `mutex_error_case`.
That worked fine in `main`, but once the change was backported to
`v2.7-branch`, the test would fail because it *did not* cause a
failure. The reason for that, was that the `mutex_error_case`
suite has `CONFIG_ZTEST_FATAL_HOOK=y`.
With the newer ztest API, it allowed a separate suite to be used,
allowing the test to pass (although it did not really fit in with
the rest of the testsuite).
The solution is to simply merge it with the `mutex_api` suite
which uses non-inverted success logic.
This change will also have to be cherry-picked for the backport
in #49031.
Fixes#48056.
tests: kernel: mutex: move race timeout test to mutex_api
Previously, this change was added to `mutex_error_case`.
That worked fine in `main`, but once the change was backported to
`v2.7-branch`, the test would fail because it *did not* cause a
failure. The reason for that, was that the `mutex_error_case`
suite has `CONFIG_ZTEST_FATAL_HOOK=y`.
With the newer ztest API, it allowed a separate suite to be used,
allowing the test to pass (although it did not really fit in with
the rest of the testsuite).
The solution is to simply merge it with the `mutex_api` suite
which uses non-inverted success logic.
This change will also have to be cherry-picked for the backport
in #49031.
Fixes#48056.
Signed-off-by: Chris Friedt <cfriedt@meta.com>
Move runtime checks to use arch_num_cpus() and build checks
to use CONFIG_MP_MAX_NUM_CPUS. This is to allow runtime
determination of the number of CPUs in the future.
Signed-off-by: Kumar Gala <kumar.gala@intel.com>
Move runtime checks to use arch_num_cpus() to use
CONFIG_MP_MAX_NUM_CPUS. This is to allow runtime
determination of the number of CPUs in the future.
Signed-off-by: Kumar Gala <kumar.gala@intel.com>
Move runtime checks to use arch_num_cpus() and build checks
to use CONFIG_MP_MAX_NUM_CPUS. This is to allow runtime
determination of the number of CPUs in the future.
Signed-off-by: Kumar Gala <kumar.gala@intel.com>
Pick a platform that actual supports SMP - qemu_x86_64. Remove setting
CONFIG_TIMESLICING as that is already set.
Signed-off-by: Kumar Gala <kumar.gala@intel.com>
Enable all cbprintf / logging related tests which were previously
disabled for qemu_arc_hs6x.
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>
Signed-off-by: Evgeniy Paltsev <PaltsevEvgeniy@gmail.com>
Spin locks must be in coherent memory for cavs. Initially this variable
was at the compilation unit scope but warnings about it being unused
from a twister run lead me to move it to be in the ifdef scope in the
function.
Move it back into the compilation units scope and wrap it in an
ifdef to ensure its not labeled as unused.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
Set the time limit to be long enough not to trigger too early. Do
not unlock after assert when doing the time limit test.
Signed-off-by: Tom Burdick <thomas.burdick@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>
For tests that set CONFIG_MP_NUM_CPUS, switch to using
CONFIG_MP_MAX_NUM_CPUS instead as we work to phase out
CONFIG_MP_NUM_CPUS.
Signed-off-by: Kumar Gala <kumar.gala@intel.com>
Spin locks held for any lengthy duration prevent interrupts and
in a real time system where interrupts drive tasks this can be
problematic. Add an option to assert if a spin lock is held for
a duration longer than the configurable number of microseconds.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
Change automated searching for files using "IRQ_CONNECT()" API not
including <zephyr/irq.h>.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
This test case will call k_sem_give() twice and expect both to be
received by k_sem_take(), yet the semaphore is initialized with a
maximum count of one!
The reason this worked was an undocumented misfeature of k_sem: if
k_sem_take() was called on a semaphore with a pended thread, it would
wake up that thread synchronously instead of incrementing the count.
So you could call it once to wake up the thread and again to queue the
count and not overflow. The problem is that this is a priority bug (a
high priority runnable thread should have the chance to run and call
k_sem_take() before a low priority thread that got woken).
Zync corrects that, and so needs to have two slots if you want two
semaphore events.
Signed-off-by: Andy Ross <andyross@google.com>
Test timers with a train of one tick timers to test that
a configured SYS_CLOCK_TICKS_PER_SEC is sensible. If the TICKS_PER_SEC
is too high the timer train will take longer than expected to reach
the station. Worse, if the timer driver has too short of a minimum
delay for its processing power and the tick rate is too high its
possible the device will get caught in an interrupt loop
preventing any threads from running while processing timers.
This test validates that the tick rate configured is actually able to be
processed without delays while also having work done in threads ensuring
that no thread scheduling delays occur either from delayed timers or an
interrupt loop from preventing threads from running.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
Adds a custom test_main and renames the test suite for jitter_drift.
Runs the jitter_drift test suite.
The order of these tests matter on hardware as the counter is often
reset on loading the test program. This is useful as its far less likely
to encounter a clock counter rollover. On arm this is especially useful.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
Move the main.c timer behavior test code to jitter_drift.c
so that other tests may be added to the suite.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
The _SYS_INIT_LEVEL* definitions were used to indicate the index entry
into the levels array defined in init.c (z_sys_init_run_level). init.c
uses this information internally, so there is no point in exposing this
in a public header. It has been replaced with an enum inside init.c. The
device shell was re-using the same defines to index its own array. This
is a fragile design, the shell needs to be responsible of its own data
indexing. A similar situation happened with some unit tests.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
We have cases where some devices needs to be initialized very early and
before c_start is call, i.e. to setup very early console or to setup
memory. Traditionally this would be hardcoded as part of the soc layer
and not using device model or the init levels.
This patch adds a new level ARCH, which will be called in early
architecture code and before we jump to the kernel code.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
In test_busy_wait and test_k_sleep test cases of
tests/kernel/context test we measure not only execution time of the
primitives itself (k_busy_wait and k_msleep respectively) but also
the overall test thread execution time.
The issue here is that we do printing in test threads which
means that we do printing in time-critical sections. That breaks
test if we do printing via some device which isn't fast enough.
Fix that by removing print from time-critical section
Signed-off-by: Evgeniy Paltsev <PaltsevEvgeniy@gmail.com>
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>
This changes to compile the IRQ offload test only if
CONFIG_IRQ_OFFLOAD=y. For architectures that do not support
IRQ offload (or that developers with to disable IRQ offload
during bring-up), compiling the code would result in linking
errors.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Say threadA holds a mutex and threadB tries
to lock it with a timeout, a race would occur
if threadA unlock that mutex after threadB
got unpended by sys_clock and before it gets
scheduled and calls k_spin_lock.
This patch supplies the test that can be used
to reproduce the problem and the fix that was
provided in #48056Fixes#48056
Signed-off-by: Christopher Friedt <cfriedt@fb.com>
In the default configuration of the test, with 10000 timer samples,
the `periodic_data` array is too big to fit in SRAM on many targets.
Use lower counts of samples for those, depending on their SRAM size,
leaving at least 8 kB for other variables, buffers, stacks etc.
Exclude the test for targets with less than 16 kB.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
The SMP config for RISC-V on QEMU triggers this:
|START - test_sem_queue_mutual_exclusion
|
|Assertion failed at
| WEST_TOPDIR/zephyr/subsys/testsuite/ztest/src/ztest_new.c:155:
| cpu_hold: (dt < 3000 is false)
|1cpu test took too long (4090 ms)
|ERROR: cannot fail in test 'after()', bailing
Looping 10000 times is maybe a bit excessive.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
This reverts the commit 7d8a119213
because GCC is now configured to not emit ldp/stp Qn instructions for
consecutive 32-byte loads and stores, and the nested interrupt handling
failure due to the missing emulation of these instructions no longer
occurs.
For more details, refer to the GitHub issue #49491 and #49806.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
Zephyr timer is based on system ticks, there usually exists some time drift
due to round up/down errors between cycles, ticks and time delay, we
need to add those expected time drift into the bound calculation for
running this test.
Add a new config TIMER_TEST_PERIOD_MAX_DRIFT_PERCENT for users to set
expected maximum drift percentage for the timer period.
Signed-off-by: Chen Peng1 <peng1.chen@intel.com>
Tickless test enables PM which implies use of LPTIM as ticker on STM32
platforms.
Specifically on nucleo_l073rz, this configuration is fragile as only
LSI(37KHz) could be used as LPTIM tick source, whith a huge accuracy
tolerance (20%).
This works on most cases, when a specific tick freq (4000 ticks/sec),
but this tests explicitly requires tick frequency set to 100.
Excludes nucleo_l073rz for this test.
Signed-off-by: Erwan Gouriou <erwan.gouriou@linaro.org>
The test case was originally designed for the coverage of
nop()/arch_nop() function. It is not very meaningful for
testing something but causing lots of false alarms on the
kernel/common test so far. Suggest removing this test case.
Signed-off-by: Enjia Mai <enjia.mai@intel.com>
Make sure msg is initialized before being used, fixes compiler warning:
main.c:735:9: error: 'msg' may be used uninitialized
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
The k_sys_fatal_error_handler() function is declared in zephyr/fatal.h
with a name of "esf" for the second parameter, not "pEsf". For
unknown reasons, this is showing up in CI as a documentation
generation failure pointing at the (correct) header.
Still, no reason not to synchronize.
Signed-off-by: Andy Ross <andyross@google.com>
Obviously the test of the feature won't work if we don't have an IPI.
And there were two threads that spawned threads that enter busy loops,
expecting to be able to abort them from another CPU. That doesn't
work[1] without an IPI. Just skip these cases rather than trying to
kludge up some kind of abort signal.
[1] Rather, it does work, it just takes infinite time for these
particular test cases. Eventually the CPU would be expected to
receive some other interrupt like a timeout, which would work to
abort the running thread. But no such timer was registered.
Signed-off-by: Andy Ross <andyross@google.com>
This commit changes some tests from using zassert_equal to validate
the pointers to using the zassert_is_null and zassert_not_null.
Signed-off-by: Michał Barnaś <mb@semihalf.com>
As of today <zephyr/zephyr.h> is 100% equivalent to <zephyr/kernel.h>.
This patch proposes to then include <zephyr/kernel.h> instead of
<zephyr/zephyr.h> since it is more clear that you are including the
Kernel APIs and (probably) nothing else. <zephyr/zephyr.h> sounds like a
catch-all header that may be confusing. Most applications need to
include a bunch of other things to compile, e.g. driver headers or
subsystem headers like BT, logging, etc.
The idea of a catch-all header in Zephyr is probably not feasible
anyway. Reason is that Zephyr is not a library, like it could be for
example `libpython`. Zephyr provides many utilities nowadays: a kernel,
drivers, subsystems, etc and things will likely grow. A catch-all header
would be massive, difficult to keep up-to-date. It is also likely that
an application will only build a small subset. Note that subsystem-level
headers may use a catch-all approach to make things easier, though.
NOTE: This patch is **NOT** removing the header, just removing its usage
in-tree. I'd advocate for its deprecation (add a #warning on it), but I
understand many people will have concerns.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
This commit disables the RISC-V direct ISR test due to the broken
`vectors` section placement with the IRQ vector table enabled,
introduced by the commit d2f8ec70235208f4a70e371ccb1ed8dfe0f573c5.
For more details, refer to the GitHub issue #49903 tracking this bug.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
SYS_CLOCK_TICKS_PER_SEC of it8xxx2 is 4096 (244us).
Running test_sleep_abs item on it8xxx2 and we get
k_us_to_ticks_ceil32(250) = 2 and late = 2, so it failed.
After we enable the CONFIG_PM, it needs more time to resume
from low power mode, so I modify the logic to <= for passing
the test.
fixes#49605
Signed-off-by: Ruibin Chang <Ruibin.Chang@ite.com.tw>
The test will always fail on emulated/simulated environments. Exclude
this one which was failing.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
NSEC_PER_MSEC should be defined along with the rest of the
per-sec macros in sys_clock.h. Currently, it's defined
multiply in a few separate locations.
Signed-off-by: Christopher Friedt <cfriedt@fb.com>
Inside test_get_cpu, the current CPU ID is stored in the test
thread's stack. Another thread is spawned with a pointer to
the variable holding this CPU ID, where this thread is supposed
to run on another CPU. On a cache incoherent platform, this
value of this variable may not have been updated on other CPU's
internal cache. Therefore when checking CPU IDs inside the newly
spawned thread, it is not checking the passed in CPU ID, but
actually whatever is on the another CPU's cache. This results in
random failure on the test_get_cpu test. Since for cache
incoherence architectures, CONFIG_KERNEL_COHERENCE is enabled by
default on SMP where shared data is placed in multiprocessor
coherent (generally "uncached") memory. The fix to this is to
simply make this variable as a global variable, as global
variable are consided shared data and will be placed in
multiprocessor coherent memory, and thus the correct value will
be referenced inside the newly spawned thread.
Fixes#49442
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
This adds a bunch of k_thread_join() to make sure threads spawned
for a test are no longer running between exiting that test. This
prevents interference between tests if some threads are still
running when assumed not.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
This commit disables the `kernel.poll` test on `qemu_arc_hs6x` because
it fails at run-time when compiled with GCC 12.
Revert this commit when the GitHub issue #49492, which tracks this bug,
is fixed.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
This commit disables the `arch.interrupt` test on the ARM64 QEMU
targets (`qemu_cortex_a53` and `qemu_cortex_a53_smp`) because the
nested interrupt test fails when compiled with GCC 12.
Revert this commit when the GitHub issue #49491, which tracks this bug,
is fixed.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
Test and validate the behavior of a timer driver.
Takes a number of absolute timer cycle samples of a periodic timer then
calculates statistical mean, variance, stddev along with total drift over
the entire test time. Ensures standard deviation and drift are within
a given configurable bound.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
The kernel.scheduler.dumb_no_timeslicing is duplicated.
Should be kernel.scheduler.dumb_timeslicing.
Otherwise, there'll be only 7 test configurations while
actually there should be 8.
Signed-off-by: Ming Shao <ming.shao@intel.com>
The global variable thread_idx should be properly initialized
for test_slice_scheduling. This issue is found when run the
test repeatedly with the new ztest fx.
Signed-off-by: Ming Shao <ming.shao@intel.com>
Old implementation of tests/kernel/sched/metairq used
static threads. The new ztest fx doesn't support static
threads for repeated test execution.(See issue: #48018)
So change to use dynamic threads to embrace the new
ztest fx.
Signed-off-by: Ming Shao <ming.shao@intel.com>
Old tests/kernel/sched/preempt cannot be run repeatedly.
If you register the test_preempt() method twice, the
second run will fail. It is because the participating
threads don't exit properly when the main thread exits.
The new ztest fx implies that a unit test should be able
to run repeatedly. This commit enables that for the preempt
test by adding explicit epilogue.
Signed-off-by: Ming Shao <ming.shao@intel.com>
The deadline test should initialize the n_exec so
that it can be executed repeatedly and in a shuffled
way with other tests, which is a paradigm offered
by the new ztest fx.
Signed-off-by: Ming Shao <ming.shao@intel.com>
This adds the bits so that we can use qemu_x86_tiny for
coverage, as this is currently the only board that can do
demand paging.
This uses the board revision as a way to specify the RAM
size as coverage requires more memory available to store
the coverage data. By piggybacking onto board revision,
this avoids adding another board config just for coverage.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Use of pipes is now configurable. All tests that use pipes must enable
that feature. (Note: no sample projects currently use pipes.)
Signed-off-by: Peter Mitsis <peter.mitsis@intel.com>
with CONFIG_ZTEST_NEW_API, test cases can be run in any order, every
test case need to do necessary cleanup.
Signed-off-by: Meng xianglin <xianglinx.meng@intel.com>
Partially revert commit 0028e9733295316d152eba07bf56677d83f4b1b5.
Timeout for tests/posix/common must be still increased for slow
platforms (previously was 120 sec).
Signed-off-by: Manuel Arguelles <manuel.arguelles@nxp.com>
After moving the mbox_api test to new ztest API, one test failed
due to stack overflow on qemu_x86_lakemont. Add a little 64 extra
stack size for adapting it.
Signed-off-by: Enjia Mai <enjia.mai@intel.com>
Loosen some timing constraints in order to hack arond spurious
failures in upstream Zephyr's CI while working on an unrelated task.
Signed-off-by: Martí Bolívar <marti.bolivar@nordicsemi.no>
Current thread_competition() case cannot verify the order of mutex
taking. This commit enhances on that.
Signed-off-by: Ming Shao <ming.shao@intel.com>
As of now, both the old and new ztest framework don't have a logic to
launch static threads. Static threads can only be launched by Zephyr
itself during boot. If a test interact with static threads, when it
is executed multiple times without reboot, only the first few runs can
have static threads to interact with. After the static threads exit,
later runs will have nothing to interact with, which can lead to failure.
One unfortunate example is the sys_mutex test.
This commit replaces static threads with dynamic thread to make the test
repeatable. In future, maybe static threads launching logic should be
covered by the ztest framework so richer scenarios can be tested.
Signed-off-by: Ming Shao <ming.shao@intel.com>
This workaround is no longer needed as sparc stack footprint
is reduced in this commit:
5cf2083e8b87f918a522423a1659ae58137ceea0
Signed-off-by: Julius Barendt <julius.barendt@gaisler.com>
This commit disables the infinite recursion warning
(`-Winfinite-recursion`), which may be reported by the GCC 12 and
above, for the `stack_smasher` function because that is the intended
behaviour.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
Add a bunch of missing "zephyr/" prefixes to #include statements in
various test and test framework files.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
This commit sets an initial value of 0 for the `elapsed_ms` variable,
which may be used uninitialised when the while loop below does not
execute.
This fixes the "‘elapsed_ms’ may be used uninitialized" warning
generated by the GCC 12.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
There is no easy way to clear event bits without
the potential for a race to exist between producer(s)
and consumer(s). The result of this race is that events
can be lost through the various resetting mechanisms
available (flag to k_event_wait(), or k_event_set()).
Add k_event_set_masked() which permits bits to be set or cleared.
This allows consumers to clear just the bits that they have read
without (accidentally) discarding any new bits.
Update unit tests to verify the functionality.
Partly Fixes#46117.
Signed-off-by: Andrew Jackson <andrew.jackson@amd.com>
When running tests/kernel/fpu_sharing/generic on qemu_leon3 with the
new ztest API, a fatal error is raised while test is reported success.
So workaround this by changing the main stack size to 2048.
Signed-off-by: Guo Lixin <lixinx.guo@intel.com>
Any project with Kconfig option CONFIG_LEGACY_INCLUDE_PATH set to n
couldn't be built because some files were missing zephyr/ prefix in
includes
Re-run the migrate_includes.py script to fix all legacy include paths
Signed-off-by: Tomislav Milkovic <milkovic@byte-lab.com>
irq_lock() returns an unsigned integer key.
Generated by spatch using semantic patch
scripts/coccinelle/irq_lock.cocci
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
Move the direct interrupt test to tests/arch/x86/direct_isr. Two
reasons:
1. The direct interrupt is only for x86. It's arch-specific.
2. And it need extra gcc option to pass the build, that will
include testsuite number. Although it seems like we add a
extra testsuite for it, actually we can reduce whole tests
configuration in tests/kernel/interrupt. And also make this
test more generic as it used to be.
Signed-off-by: Enjia Mai <enjia.mai@intel.com>
Timeout must be increased for fvp_baser_aemv8r_aarch32 board. Enabling
MPU on this board makes simulation slower.
Signed-off-by: Manuel Arguelles <manuel.arguelles@nxp.com>
Every architecture must export the z_irq_spurious definition. Just unify
that in one single header file.
Signed-off-by: Carlo Caione <ccaione@baylibre.com>
Fix this test adding a proper support for RISCV when in vectored mode
with IRQ vector table generation.
Signed-off-by: Carlo Caione <ccaione@baylibre.com>
Following zephyr's style guideline, all if statements, including single
line statements shall have braces.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Check that `SYS_INIT_NAMED` allows multiple instances of the same
initialisation function, as long as names are unique.
Signed-off-by: Jordan Yates <jordan.yates@data61.csiro.au>
This test was written to idle for exactly 1ms and wake up with zero
error, which is just too tight for some platforms (and worked on
emulators where the tick rate is 10x coarser only because 0 == 0!).
And it's not clear that it's testing anything we promise in
documentation, regardless. Early wakeups are not an error and
absolutely not disallowed, yet the test is treating the wakeup like a
sleep.
Clean it up a bit and relax the tolerance to what we can compute
reliably: do all the math in ticks, idle for 10ms (i.e. longer than a
host quantum for emulators), and allow 1 tick of slop on either side to
permit slightly early wakeups while still verifying that "yes, the idle
did idle".
Fixes#46641
Signed-off-by: Andy Ross <andyross@google.com>
Now that picolibc's malloc arena configuration always allocates
some space, we don't need explicit allocations for tests.
Signed-off-by: Keith Packard <keithp@keithp.com>
When running with picolibc, we need more MPU resources for these
tests. Get rid of picolibc malloc arena too.
Signed-off-by: Keith Packard <keithp@keithp.com>
As with previous commit, make the timer irq a simple integer variable
exported by the timer driver for the benefit of this one test
(tests/kernel/context).
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
This test has gotten out of control. It has a giant #if cascade
enumerating every timer driver in the Zephyr tree and extracting its
interrupt number. Which means that every driver needs to somehow
expose that interrupt in its platform headers or some other API.
Make it a simple integer variable exported by the timer driver for the
benefit of this one test.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Picolibc has subtly different output from the minimal libc as a result
of different handling for code built without real long long support.
Signed-off-by: Keith Packard <keithp@keithp.com>
This runs the existing kernel common tests using picolibc as some of those
depend on libc functionality.
Signed-off-by: Keith Packard <keithp@keithp.com>
change the test interrupt line number to a bigger one, because
the low number usually will be used by other devices.
Signed-off-by: Chen Peng1 <peng1.chen@intel.com>
When SMP feature is enabled on board, this test suite will skip.
Enable the test suite by removing the filter.
Note that these testings only run SMP platform on single core
which are handled by setting MP_NUM_CPUS=1 in prj.conf.
Signed-off-by: Yinfang Wang <yinfang.wang@intel.com>
This commit updates all deprecated `K_THREAD_STACK_ARRAY_EXTERN` macro
usages to use the `K_THREAD_STACK_ARRAY_DECLARE` macro instead.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
This commit updates all deprecated `K_THREAD_STACK_EXTERN` macro usages
to use the `K_THREAD_STACK_DECLARE` macro instead.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
This commit updates all deprecated `K_THREAD_STACK_EXTERN` macro usages
to use the `K_THREAD_STACK_DECLARE` macro instead.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
This commit updates all deprecated `K_THREAD_STACK_EXTERN` macro usages
to use the `K_THREAD_STACK_DECLARE` macro instead.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
This commit selectively disables the dangling pointer warning
(`-Wdangling-pointer`) for the compilation of the `alternate_thread`
function because it deliberately makes use of a dangling pointer in
order to test stack randomisation.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
Currently the device MMIO APIs is only able to map single DT-defined
regions and also the _NAMED variant is assuming that each DT-defined
device has only one single region to map.
This is a limitation and a problem when in the DT are defined devices
with multiple regions that need to be mapped.
This patch is trying to overcome this limitation by introducing the
DEVICE_MMIO_NAMED_ROM_INIT_BY_NAME macro that leveraged the 'reg-names'
DT property to map multiple regions defined by a single device.
So for example in the DT we can have a device like:
driver@c4000000 {
reg = <0xc4000000 0x1000>, <0xc4001000 0x1000>;
reg-names = "region0", "region1";
};
and then we can use DEVICE_MMIO_NAMED_ROM_INIT_BY_NAME doing:
struct driver_config config = {
DEVICE_MMIO_NAMED_ROM_INIT_BY_NAME(region0, DT_DRV_INST(0)),
DEVICE_MMIO_NAMED_ROM_INIT_BY_NAME(region1, DT_DRV_INST(0)),
};
Signed-off-by: Carlo Caione <ccaione@baylibre.com>
Avoid potentially calling __builtin_clz() twice with non-constant
values. Also add a test for it.
Clang produces false positive vla warnings so disable them. GCC will
spot real vla's already.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
Files including <zephyr/kernel.h> do not have to include
<zephyr/zephyr.h>, a shim to <zephyr/kernel.h>.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
This commit relaxes the precision requirements for idle event statistic
test when RISC-V machine timer driver is used. This is needed for some
platforms (e.g. hifive1), for which the cycle count is too low to pass
the checks where a percent deviation of peak cycles count is allowed.
Signed-off-by: Filip Kokosinski <fkokosinski@antmicro.com>
Subtracting with a uint64_t operand yields a uint64_t result, for which
the absolute value is not terribly interesting. Cast the operand to
int64_t.
Use llabs instead of abs as abs takes an int parameter and not an
int64_t. This appears to work even with the minimal C library.
Signed-off-by: Keith Packard <keithp@keithp.com>
Due to qemu_x86_tiny having very small defined SRAM area,
enabling userspace results in not having enough free physical
pages to run the tests. So make the memory a bit larger so
we can actually test memory mapping with userspace.
Fixes#46398
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
As the type of A(n) is integer, and A3 and A5 are close to
each other. Sometimes A3 is equal to A5. So change the ">" to
">="
Signed-off-by: Hu Zhenyu <zhenyu.hu@intel.com>
The tests is currently testing all the memory mapping parameters but
K_MEM_PERM_USER. Add a test case to test that as well.
Signed-off-by: Carlo Caione <ccaione@baylibre.com>
ehl_crb supports HPET timer by default. Add test suite to test APIC
timer's TSC deadline mode on ehl_crb.
Signed-off-by: Yinfang Wang <yinfang.wang@intel.com>
When test tests/kernel/sched/schedule_api, it shows "ASSERTION FAIL:
timeslice in ticks much be divisible by two", then break and fail
the test.
Fixes#44887.
On board it8xxx2_evb, it can pass schedule_api test without
the assertion, so I add the floating part back to half_slice_cyc
when the timeslice in ticks can't be divisible by two.
After change, the time slice will be:
1.slice_ticks = (200x8192+999)/1000 = 1639 (not changed)
2.before add the deviation (not handle the floating part):
half_slice_cyc = (1639/2) * (32768/8192) = (819) * (32768/8192) = 3276,
after add the deviation:
half_slice_cyc = 3276 + (32768/8192/2) = 3278,
and it's equal (1639/2) * (32768/8192) = 3278.
Verified by test pattern:
west build -p always -b it8xxx2_evb tests/kernel/sched/schedule_api
Signed-off-by: Ruibin Chang <Ruibin.Chang@ite.com.tw>
Bug #45779 discovered an edge case with nested interrupts on Xtensa
where they might select an incorrect thread context to return to
instead of the (mandatory!) return to the outer interrupt context.
Cleverly adjust the nested_irq_offload to exercise this. It now
creates a thread that it knows it will interrupt, then suspends that
thread from within the inner/nested interrupt. This guarantees that
_current will be different on exit from the second interrupt, which is
the case that tripped up Xtensa.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
Test was setting up timer for 1 system tick and then work was
cancelled. It was assumed that work will be cancelled before
timer expires. This is the case for low frequency system clock
(e.g. qemu targets using 100Hz) but there are cases when system
clock has higher frequency (32kHz on nRF). In that case, timer
was occasionally expiring before cancellation and test was
randomly failing.
Signed-off-by: Krzysztof Chruscinski <krzysztof.chruscinski@nordicsemi.no>
This test was written to do a k_oops() in the main thread. That's an
essential thread, and aborting it is actually a system panic now. The
test was written contra the docs on this, but it worked fine for
years.
Just reflag the thread as a simple workaround rather than trying
anything fancy. This is a very simple test.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
Documentation specifies that aborting/terminating/exiting essential
threads is a system panic condition, but we didn't actually implement
that and allowed it as for other threads. At least one app wants to
exploit this documented behavior as a "watchdog" kind of condition,
and that seems reasonable. Do what we say we're supposed to do.
This also includes a small fix to a test, which seemed like it was
written to exercise exactly this condition. Except that it failed to
detect whether or not a system fatal error was actually signaled and
was (incorrectly) indicating "success". Check that we actually enter
the handler.
Fixes#45545
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
A separate privileged stack is used when CONFIG_GEN_PRIV_STACKS=y. The
main stack guard area is no longer needed and can be made available to
the application upon transitioning to user mode. And that's actually
required if we want a naturally aligned power-of-two buffer to let the
PMP map a NAPOT entry on it which is the whole point of having this
CONFIG_PMP_POWER_OF_TWO_ALIGNMENT option in the first place.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
The StackGuard area is used to save the esf and run the exception code
resulting from a StackGuard trap. Size it appropriately.
Remove redundancy, clarify documentation, etc.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
Some test suites have different test case lists in test_main(), that
conforms to different test scenarios defined in testcase.yaml. We
use if statement to decide which test case list should run under
specific config.
But for thoses boards who do not support those configs, we will run test
cases on the other side of the if statement even if it has deviated from
the original test scenario.
So add filter to avoid test scenario running under mismatch config.
Signed-off-by: Guo Lixin <lixinx.guo@intel.com>
Mutual exclusion test assume that the excution order of two threads like
this:
mutual_exclusion1 -> mutual_exclusion2 -> mutual_exclusion1 ...
but some times the excution order of two threads would be this:
mutual_exclusion1 -> mutual_exclusion2 -> mutual_exclusion2 ...
This patch increase the loop cycle, add a variable 'tmp' to store the
value of 'critical_var' before operating it.
Signed-off-by: Huifeng Zhang <Huifeng.Zhang@arm.com>
The issue is caused by multiple threads which have taken the semaphore
to increase or decrease the normal count variable. Change its type with
atomic_t.
Signed-off-by: Jaxson Han <jaxson.han@arm.com>
This commit adds the `volatile` qualifier to the timing variables, in
order to ensure that the compiler does not try to optimise the test in
a way that can affect the execution time measurements.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
When an object availability event triggers a k_work_poll
item, the object lock should not be held anymore
during the execution of the work callback.
Signed-off-by: Lucas Dietrich <ld.adecy@gmail.com>
Linker files were not migrated with the new <zephyr/...> prefix. Note
that the conversion has been scripted, refer to #45388 for more details.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
In order to bring consistency in-tree, migrate all tests to the new
prefix <zephyr/...>. Note that the conversion has been scripted, refer
to #45388 for more details.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Validate the behaviour of `k_can_yield` in pre-kernel, ISR, and idle
thread and standard thread contexts.
Signed-off-by: Jordan Yates <jordan.yates@data61.csiro.au>
For testing, assume that the Cortex-A/R platforms are using a GIC
interrupt controller. Use the last GIC SGI to trigger an interrupt for
the test.
Signed-off-by: Bradley Bolen <bbolen@lexmark.com>
Add the appropriate hooks effectively replacing the old implementation
with the new one.
Also the stackguard wasn't properly enforced especially with the
usermode combination. This is now fixed.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
When a memory domain is initialized, the z_libc_partition must be
included so that critical libc-related data can be accessed.
On ARM processors without TPIDRURO when THREAD_LOCAL_STORAGE is enabled,
this includes the TLS base pointer, which is used for several
thread-local variables in the kernel.
Signed-off-by: Keith Packard <keithp@keithp.com>
When active, z_libc_partition consumes an MPU region which leaves too
few for some MPU tests. Free up one by disabling HW stack protection.
Signed-off-by: Keith Packard <keithp@keithp.com>
When using THREAD_LOCAL_STORAGE the thread_userspace_local_data stuff
isn't used, so these tests wouldn't build.
Signed-off-by: Keith Packard <keithp@keithp.com>
There are tests failing due to timeout for a few seconds in simulators,
slightly increase the timeout for these cases.
Signed-off-by: Guo Lixin <lixinx.guo@intel.com>
When threads are in more than one state at a time, k_thread_state_str()
returns a string that lists each of its states delimited by a '+'.
This in turn necessitates a change to the API that includes both a
pointer to the buffer to use for the string and the size of the buffer.
Signed-off-by: Peter Mitsis <peter.mitsis@intel.com>
This commit excludes the kernel monotonic timer test for the
`qemu_arc_hs` platform because this test may fail with the ARC QEMU
6.2 on certain host systems.
Refer to the following issues for more details:
* foss-for-synopsys-dwc-arc-processors/qemu#67
* zephyrproject-rtos/zephyr#44862
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
This commit excludes the kernel syscall test for the `qemu_arc_em`
platform because this test may fail with the ARC QEMU 6.2 on certain
host systems.
Refer to the following issues for more details:
* foss-for-synopsys-dwc-arc-processors/qemu#66
* zephyrproject-rtos/zephyr#44862
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
The autoconf.h header is not required because the definitions present in
the file are exposed using the compiler `-imacros` flag.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Some names of the test cases are duplicated within the project.
This commit contains the proposed names of the test scenarios.
Signed-off-by: Katarzyna Giadla <katarzyna.giadla@nordicsemi.no>
There are tests failing sometimes due timeouts in simulators, slightly
increase the default timeout for these cases.
Signed-off-by: Flavio Ceolin <flavio.ceolin@intel.com>
The residency policy, is in reality, influences by other parameters for
example constraints. It has been renamed to "DEFAULT" policy to make it
more general. The "APP" policy has been renamed to "CUSTOM" to better
represent its purpose.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
This test was working by accident onarm64 and riscv64. Those
architectures have large register files, even more so considering
their 64-bit nature.
This test works by calling k_object_alloc(K_OBJ_THREAD) until thread
index exhaustion. However here it exhausted heap memory before running
out of thread indexes. There was a test to make sure that wasn't the
case by attempting a k_malloc(256). But here that succeeded just
because 256 is far smaller than a struct k_thread on the above
architectures.
Fix this by:
- attempting an additional allocation with the actual object size
instead of an arbitrary 256 bites
- increasing the heap size as 8192 was clearly insufficient for the
above platforms.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
When operating on different kinds of heaps sometimes there's a need to
perform special operations on heap, poweroff memory bank when releasing
memory etc. Therefore some additional data may be required.
Metadata is a point to keep such data.
Signed-off-by: Marcin Szkudlinski <marcin.szkudlinski@intel.com>
Simple coverage exerciser for the per-thread timeslice feature. Added
as a(nother) new variant of the schedule_api test.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
This test uses k_yield() to "sync" between threads, so it's implicitly
supposed to run on a single CPU. Make it explicit, to avoid issues on
platforms with more cores.
Signed-off-by: Ederson de Souza <ederson.desouza@intel.com>
FIXKFLOATDISABLE
The thread context test has insufficient checkings for failing
so add them:
() The variable rv for pass/fail is set but never checked. So
add a check to fail the test if such indicated.
() Each thread's pass variable is set to TC_FAIL (== 1) and
the check for successful thread execution simply checks
if pass variable is not zero, which is always true. So
change it so the check for failing condition is reasonable.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Change the CPU_CORTEX_R kconfig option to CPU_AARCH32_CORTEX_R to
distinguish the armv7 version from the armv8 version of Cortex-R.
Signed-off-by: Bradley Bolen <bbolen@lexmark.com>
Fix priorities for the test threads to allow execution when test thread
yields.
Also cleanup some strings.
Fixes#42723
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Skip kernel_timer_interrupts when CONFIG_TICKLESS_KERNEL
is disabled, because timer won't generate interrupts
anymore after invoking irq_disable and irq_enable
to enable timer interrupt again in TICK mode.
Signed-off-by: Chen Peng1 <peng1.chen@intel.com>
Add a very simple test of the CONFIG_IRQ_OFFSET_NESTED feature that
exercises nested interrupts in a portable way. It calls irq_offset()
from within a k_timer callback and validates that the return lands
back in the original interrupt successfully.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
No reason to exclude this platform, we only have been using wrong
logging defaults (which was set in the soc) and not printing log
messages which is required by the test.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
ARM does not guarantee the timing effects of NOP
instruction. Hence skip the test_nop test.
Fix for Issue#42666
Signed-off-by: Mahesh Mahadevan <mahesh.mahadevan@nxp.com>
K_OBJ_MSGQ, K_OBJ_PIPE, and K_OBJ_STACK objects have pointers
to additional memory that can be allocated. The k_obj_alloc()
returns these objects as uninitialized so when they are freed
there are random opportunities for freeing invalid memory
and causing random faults.
Signed-off-by: David Leach <david.leach@nxp.com>
The K_OBJ_MSGQ object is unitialized so when the thread cleanup occurs
after an expected fault for invalid access the test case can randomly
fault again because the cleanup of the thread will sometimes attempt
to free invalid buffer_start pointer in the msgq object.
Fixes#42705
Signed-off-by: David Leach <david.leach@nxp.com>
On SMP, and especially using qemu on a busy system, it is possible for
a thread with a later timeout to get ahead of another one with an
earlier timeout. The tight timeout value difference (10ms) makes it
possible albeit difficult to reproduce. The result is something like:
|START - test_timeout_threads_pend_on_lifo
| thread (q order: 2, t/o: 0, lifo 0x4001d350)
|
| Assertion failed at main.c:140:
|test_multiple_threads_pending: (data->timeout_order not equal to ii)
| *** thread 2 woke up, expected 1
Let's make timeout values 10 times larger to make this unlikely race
even less likely.
While at it... The timeout field in struct timeout_order_data is some ms
value and not a number of ticks, so change the type accordingly.
And leverage k_cyc_to_ms_floor32() to simplify computation in
is_timeout_in_range().
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
Both alloc_pending tests requires the main thread to utilize
the heap so that child threads must pend on memory allocations.
However, the previous implementation was not SMP friendly where
the child threads could run and succeeded in memory allocation
on other CPUs while the main thread continued to allocate
memory. The main thread would fail to allocate memory if
the child thread (on other CPU) has not freed the memory yet.
Not to mention that, in this scenario, the child thread was not
pending on memory allocation which defeated the purpose of
the test. So to fix this, make sure the main thread allocates
enough memory so future allocations must go into pending.
Also, check that the child thread cannot allocation memory
when first entered so it is actually going into pending.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
This test is already running with CONFIG_MP_NUM_CPUS=1. All those
1cpu declarations are needless. And some of them (like the init
tests) have side effects that make it difficult to do things like
"filter for only MP cases".
(Indeed, this is a heavily MP-unsafe test; almost all cases written to
rely on ordering between a parent thread and its child. And that's
doubly so for COHERENE platforms because lots and lots of the test
objects are on stacks. MP_NUM_CPUS=1 is definitely the right thing
here.)
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
There's no reason to wait a whole second here just to know if a tick
should have fired (though, yes, on some older/legacy/non-tickless
configurations, 128 ticks is actually more than a second).
Some simulators are very slow; busy waiting is expensive.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
This reverts commit ae8745df6f.
Patch "kernel/init.c: Initialise logging subsystem after arch" should
fix this, so no more need to filter this test out.
Signed-off-by: Ederson de Souza <ederson.desouza@intel.com>
When building with XCC, k_therad_access_grant() expands to
a loop but does nothing if no building for userspace. XCC
does not like this and emits error:
main.s: Assembler messages:
main.s:4563: Error: invalid empty loop
So add #ifdef to only enable the loop if userspace is enabled.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
When building with XCC, the for loop to call k_therad_access_grant()
is an empty loop because k_thread_access_grant() does nothing
if no building for userspace. XCC does not like this and emits
error:
main.s: Assembler messages:
main.s:1951: Error: invalid empty loop
So add #ifdef to only enable the loop if userspace is enabled.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
This test is triggering some kind of bug that will reliably cause a
full host crash/hang of the x86 host environment on TGL/cAVS 2.5.
It's interfering with CI testing, so filter it out for now while we
figure it out.
Interestingly it doesn't have any trouble on older cavs15. And even
more so, it seems to be some kind of build interaction. If I disable
LOG=y, it passes. But when it fails, it actually fails BEFORE the boot
entry and core 0 initialization code is reached (i.e. LONG before any
logging initialization). Something is wrong with the generated file;
maybe a linker or rimage bug? The signature is reported OK by the
ROM, but that's the last we hear from the device before it blows up.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
The test_queue_multithread_competition case wants to be sure that an
inserted item is recevied by the highest priority thread of several
waiting, but that only works if the threads aren't racing against each
other on different CPUs.
Also, the test_queue_loop case would produce a LOT of console output
very quickly. On a few occasions, I saw this overflow the 8k output
buffer of the intel_adsp devices at exactly the wrong time (with
respect to the polling loop in the host python script), cause a flush
of the stream, and then miss the SUCCESSFUL message. Quiet things
down a bit, there's not a lot of value of verbosity in a CI test.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>