Updates the ipi_optimize test to use 4 CPUs for qemu_x86_64.
This provides enough CPUs for the test to verify the recently
added directed IPI support for x86_64.
Signed-off-by: Peter Mitsis <peter.mitsis@intel.com>
When disabling the __deprecated macros in tests/kernel/pipe/deprecated,
the macros was set to ''. For IAR tools this was expanded to '' which
caused a compilation error.
I have replaced them with /* deprecated */ which should work for all
toolchains.
Signed-off-by: Lars-Ove Karlsson <lars-ove.karlsson@iar.com>
This test was buggy. The first call to k_thread_abort() would
terminate the calling thread (because it handled a panic!) and so the
second case (an essential thread self-aborting) wasn't actually being
exercised and was silently "passing". Oops.
Fix by splitting the cases into two tests, as suggested by @fsammoura
But unmasking that shows that there are some arch buglets[1] to fix
before this case can pass for everyone. So skip on x86/riscv/sparc
for now.
[1] See the comment. It's not really a "bug" as we've never demanded
that arch layers do this. But it does work on many of them.
Signed-off-by: Andy Ross <andyross@google.com>
Improve robustness of test_all_stats_usage test. Test was relying on a
fact that system did not go to idle before this test is run. It means
that it was assuming that it will be the first test which might not be
true. Additionally, on some targets there might be some idle waits
during system boot. Test is adjusted to validate gathered statistics
with an assumption that there might have been idle period prior to the
test.
Signed-off-by: Krzysztof Chruściński <krzysztof.chruscinski@nordicsemi.no>
No files use MS_TO_US except tests/kernel/interrupt/src/nested_irq.c.
While the macro is convenient, we should put it in a more central
location and use it everywhere if we want to use it rather than just in
interrupt_util.h. For now, remove it and replace the two usages to match
how code is written in the rest of Zephyr.
Signed-off-by: Tom Hughes <tomhughes@chromium.org>
This revert the idea of 3fa7d78 from #78845.
The 3rd level IRQ APIs won't compile when
CONFIG_3RD_LEVEL_INTERRUPT_BITS=0.
Updated testcases accordingly.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
Signed-off-by: Yong Cong Sin <yongcong.sin@gmail.com>
First, SoC is not supported, but when I tried to add it, build results
in RAM overflow of ~10K. Excluding the platform for now, so CI is happy.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Since it was added, `arch.interrupt.gen_isr_table_local.arm_mainline`
filters on `CONFIG_ARMV6_M_ARMV8_M_BASELINE`. I'm not entirely sure whether
this was intentional, but this seems odd to me given a) the naming of the
test b) that the `platform_allow` is `qemu_cortex_m3` and c) that the
`arch.interrupt.gen_isr_table.arm_mainline` test it inherits part of its
config from filters on `CONFIG_ARMV7_M_ARMV8_M_MAINLINE`.
The current filtering setup also means that a command like
`west twister -c -T tests/kernel/gen_isr_table/` currently filters out all
of the `*gen_isr_table_local*` tests which was a surprise to me when
testing locally (no tests broke for a change I made as they were
all filtered out).
Assuming this is undesireable, change the filter to check against
`CONFIG_ARMV7_M_ARMV8_M_MAINLINE` to make it possible to run the tests
on the expected platform without needing to work around the filter.
Signed-off-by: Jonathon Penix <jpenix@quicinc.com>
For some target and optimization level combinations (aarch64 and -O2, for
example), GCC will over align _isr_list_sname instances. The local isr
declaration parser doesn't expect the extra padding that may be introduced
and will produce unexpected errors about bad IRQs (`error: IRQ 1903323438
(offset=0) exceeds the maximum of 42`, for example). Prevent this issue by
forcing the expected alignment of _isr_list_sname.
Fixes#81254
Signed-off-by: Jonathon Penix <jpenix@quicinc.com>
Thread abort of essential threads had a buglet where the panic would
be signaled but, if caught by k_sys_fatal_error_handler(), the thread
would not actually be aborted. This test wasn't validating that the
thread was actually dead (in fac, it RELIED on the thread to signal a
semaphore after its "abort"!).
Make sure it's dead and that the panic is caught. Also add a case for
aborting _current which is a separate code path that needs coverage.
Signed-off-by: Andy Ross <andyross@google.com>
In the current pipe_api test file we inadvertantly use VLA.
Toolchains are allowed by standard to allocate VLA on
heap for example. Therefore in my opinion we shouldn't
use VLA atleast in kernel+kernel tests.
Signed-off-by: Robin Kastberg <robin.kastberg@iar.com>
Building kernel.threads.apis with clang warns:
tests/kernel/threads/thread_apis/src/main.c:362:6: error: variable 'ret'
is used uninitialized whenever 'if' condition is true
[-Werror,-Wsometimes-uninitialized]
if (m == ISR_ALREADY_EXIT || m == ISR_RUNNING) {
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
tests/kernel/threads/thread_apis/src/main.c:380:6: note: uninitialized
use occurs here
if (ret != 0) {
^~~
tests/kernel/threads/thread_apis/src/main.c:362:2: note: remove the 'if'
if its condition is always false
if (m == ISR_ALREADY_EXIT || m == ISR_RUNNING) {
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
tests/kernel/threads/thread_apis/src/main.c:362:6: error: variable 'ret'
is used uninitialized whenever '||' condition is true
[-Werror,-Wsometimes-uninitialized]
if (m == ISR_ALREADY_EXIT || m == ISR_RUNNING) {
^~~~~~~~~~~~~~~~~~~~~
tests/kernel/threads/thread_apis/src/main.c:380:6: note: uninitialized
use occurs here
if (ret != 0) {
^~~
tests/kernel/threads/thread_apis/src/main.c:362:6: note: remove the '||'
if its condition is always false
if (m == ISR_ALREADY_EXIT || m == ISR_RUNNING) {
^~~~~~~~~~~~~~~~~~~~~~~~
tests/kernel/threads/thread_apis/src/main.c:329:9: note: initialize the
variable 'ret' to silence this warning
int ret;
^
= 0
Signed-off-by: Tom Hughes <tomhughes@chromium.org>
The intention of the test is to have a meta-IRQ preempt a cooperative
thread, then ensure the same cooperative thread is resumed back to instead
of a higher priority thread. Thus, the test needs to assert that the lower
priority cooperative thread is not yet complete when it makes the higher
priority thread ready to run.
This change reveals bugs like #80574, where the meta-IRQ is unable to
preempt a cooperative thread.
Signed-off-by: Kalle Kietäväinen <kalle.kietavainen@silabs.com>
Test test_timer_remaining is comparing time captured by k_busy_wait and
k_timer_remaining_ticks. Test was accepting 1 tick difference between
those two. If system clock frequency is low (e.g. 100 Hz) 1 tick is a
long time but if system clock is high then 1 tick may not cover for
processing latency. Instead of using fixed 1 tick test is now converting
100 us to ticks (ceiling) to cover for cases where system clock is high.
100 us processing latency time is an arbitrary value.
Signed-off-by: Krzysztof Chruściński <krzysztof.chruscinski@nordicsemi.no>
On SMP systems, threads going through k_thread_abort() may still
be running while the test moves on to the next one. This creates
some interferences and may result in the next test failing. So
after each test, we need to do k_thread_join() on those threads
to make sure they are no longer active.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Fix so the compiler does not throw a warning thinking that those
buffers may be used unitialized inside k_pipe_write/read().
Note clang does not the array size being of a variable size and
initialized as it throws a
"error: variable length array folded to constant array as an extension"
The issue was discovered with with gcc 14.2
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
This commit adds new test cases for the pipe API rework.
* basic.c: Sanity check for pipe operations.
* concurrency.c: Test pipe operations with multiple threads.
* stress.c: Test pipe operations under stress conditions.
And moves the old pipe test cases to the deprecated folder.
Signed-off-by: Måns Ansgariusson <Mansgariusson@gmail.com>
Increase the MAX_THREAD_BYTES from 5->6 to accomodate
for devices with an increased number of threads in the
tests schedule_api and dynamic_thread_stack.
Signed-off-by: Georgios Vasilakis <georgios.vasilakis@nordicsemi.no>
Test test_one_tick_timer_train was failing due not being able to
execute enough 1 tick timeouts. Decrease system ticks frequency
to make the test pass.
Signed-off-by: Krzysztof Chruściński <krzysztof.chruscinski@nordicsemi.no>
Although many platforms may have an unmapped address in common,
this address will not be the same for all platforms. This commit
makes such an address configurable to get better coverage.
Signed-off-by: Peter Mitsis <peter.mitsis@intel.com>
Adjust testcase.yaml files to changes in Twister schema which
now allows multiple recording patterns ('record: regex:').
Signed-off-by: Dmitrii Golovanov <dmitrii.golovanov@intel.com>
Test was using interrupt lines which does not exist on nRF VPR
targets. nRF54Lx FLPR has interrupts >= 16 and nRF54Hx PPR does
not have interrupt 17. Added configuration which works for
nrf54h20_cpuppr and nrf54lx_flpr.
Signed-off-by: Krzysztof Chruściński <krzysztof.chruscinski@nordicsemi.no>
The number of mpu regions that can be configured is less than four cases.
Therefore, only remove this case on cm33 cores, failed log show below:
"num_parts of 4 exceeds maximum allowable partitions (3)"
samples: kernel: mem_protect: userspace: remove mimxrt700_evk cm33 cores
The maximum number of partitions is less than the number of partitions
currently required, causing case to be unable to apply for or delete
excess regions.
Signed-off-by: Lucien Zhao <lucien.zhao@nxp.com>
Currently the timer_behavior test uses fabs.
Since the C library functions being used in the kernel
is restricted in Rule A.4 in the Coding Standard
We should probably not use these functions in tests either.
Signed-off-by: Robin Kastberg <robin.kastberg@iar.com>
Mostly a revert of commit b1def7145f ("arch: deprecate `_current`").
This commit was part of PR #80716 whose initial purpose was about providing
an architecture specific optimization for _current. The actual deprecation
was sneaked in later on without proper discussion.
The Zephyr core always used _current before and that was fine. It is quite
prevalent as well and the alternative is proving rather verbose.
Furthermore, as a concept, the "current thread" is not something that is
necessarily architecture specific. Therefore the primary abstraction
should not carry the arch_ prefix.
Hence this revert.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
Setting a deadline is not a schedule point. This makes it a
perfect place to verify the behavior of the new k_reschedule()
routine at both thread and ISR level.
Signed-off-by: Peter Mitsis <peter.mitsis@intel.com>
This adds data cache manipulations, flushing and invalidation,
to the tests where buffer content are being written and
compared. These tests map different virtual pages to the same
physical pages, and write to one of the mapped virtual
addresses. Some SoCs may cache the virtual address separately
and writes to one virtual address will not be reflected to
another virtual address, this failing the comparison. So we
need to manually flush the cache after writing to the buffer,
and invalidating cache before reading. Note that not all reads
and writes need this treatment as some of them only needs to
test for access permissions, and not the memory content.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
For CPUs with incoherent cache under SMP, the tests to read/write
buffer (... majority of tests here) may not work correctly if
the test thread jumps between CPUs. So use the test infrastructure
to limit the test to 1 CPU.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
At the present time, Zephyr does has overlap between sleeping and
suspending. Not only should sleeping and suspended be orthogonal
states, but we should ensure users always employ the correct API.
For example, to wake a sleeping thread, k_wakeup() should be used,
and to resume a suspended thread, k_thread_resume() should be used.
However, at the present time k_thread_resume() can be used on a
thread that called k_sleep(K_FOREVER). Sleeping should have nothing
to do with suspension.
This commit introduces the new _THREAD_SLEEPING thread state along
with some prep-work to facilitate the decoupling of the sleeping and
suspended thread states.
Signed-off-by: Peter Mitsis <peter.mitsis@intel.com>
Change to prevent build failure when CONFIG_SCHED_IPI_SUPPORTED is
defined and CONFIG_TRACE_SCHED_IPI is not.
Signed-off-by: William Tambe <williamt@cadence.com>
CONFIG_ENTROPY_GENERATOR is now enabled by default as soon as
"zephyr,entropy" is set in the device-tree. Since the sub-Kconfig
CONFIG_ENTROPY_NRF5_RNG is also enabled by default on nrf boards,
this means that the driver is pulled in the build and intialized
at boot. This might cause runtime issues during some IRQ/kernel
tests, so it's better to remove "zephyr,entropy" on these tests.
Signed-off-by: Valerio Setti <vsetti@baylibre.com>
Both tests set CONFIG_COMMON_LIBC_MALLOC_ARENA_SIZE=0 so extract
it to prj.conf. The tests require malloc arena size to be 0 to
run successful. By refactoring the config, it is now able to
with west build and still passing.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
