Check for COMPILER_ISA_THUMB2 rather than CPU_CORTEX_M when determining
if special handling is required for thumb instructions.
This prevents false negative results on Cortex-R devices that may
generate a fault (test pass) even if the MMU/MPU is not configured
correctly.
Signed-off-by: Grant Ramsay <gramsay@enphaseenergy.com>
This commit updates the definition of z_shared_isr_table_entry
to use _isr_table_entry instead of specially created z_shared_isr_client.
Signed-off-by: Radosław Koppel <radoslaw.koppel@nordicsemi.no>
Adds a configurable delay to the test threads. It exists to keep
thread0 from hogging the scheduler's spin lock and giving threads
on another core a better chance to obtain that lock and not starve.
Signed-off-by: Peter Mitsis <peter.mitsis@intel.com>
Reduces the number of seconds that the test runs to 15 seconds
(down from 30 seconds). This should still be sufficient time
to determine if there is anything amiss with the synchonous
thread suspend/resume. Furthermore, the test itself has been
tagged as being slow and its timeout increased to 10 minutes
(600 seconds) as some simulators are rather slow.
Signed-off-by: Peter Mitsis <peter.mitsis@intel.com>
Instead of using the system heap, we define our own heap in incoherent
memory because stacks should be in cached memory.
Signed-off-by: Flavio Ceolin <flavio.ceolin@intel.com>
The CLIC has different software-triggerable irqs than the PLIC. I
adjusted them.
Additionally, in a clic the the irq flag 0 would mean triggering on a
positive level. To work with software-triggering, the irqs have to
trigger on a positive edge, i.e. 1.
Signed-off-by: Greter Raffael <rgreter@baumer.com>
This extends the smp_boot_delay test to test the newly
introduced function k_smp_cpu_custom_start().
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
This renames z_smp_cpu_start() to k_smp_cpu_start().
This effectively promotes z_smp_cpu_start() into a public API
which allows out of tree usage. Since this is a new API,
we can afford to change it signature, where it allows
an additional initialization steps to be done before a newly
powered up CPU starts participating in scheduling threads
to run.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
This makes the test a bit more generic so it can run on
other SMP enabled platforms. Though, we don't need CI
running this on every SMP platforms as this is not
testing the mechanism of bringing up a CPU. That is
being tested on the other SMP test. Therefore, only
enable qemu_x86_64 to be testing in CI so that this
feature will still be tested on CI.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
This test assumes that nothing will wake the CPU apart from
the timer set by the test.
But that is not necessarily the case.
Some other platform thread started at boot
may be waking the CPU every now and then.
Let's allow for some spurious wakes while we are testing k_cpu_idle,
while at the same time ensuring we are not just busy waiting
all the way.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
Because it was exclusively used by the "common" RISC-V privileged code
to build CPU idle routines that are now handled by arch level code.
Also, all platforms defaulted to "y", making it pointless in practice.
Signed-off-by: Gerard Marull-Paretas <gerard@teslabs.com>
This test assumes nothing else is running in the background
but when the logger is on, this board will by default
start the RTC sync mechanism, which will awake the CPU
every now and then.
Let's disable this mechanism to avoid disturbing the test.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
Adds a test to stress k_thread_suspend() and k_thread_resume() on
an SMP system. The test should take about 30 seconds.
Signed-off-by: Peter Mitsis <peter.mitsis@intel.com>
- Support for listing both testcases and testsuites
- Support for running single suites or single test cases
- Support shuffling tests and repeating execution based on command line
arguments.
For example, build with
west build -p -b qemu_cortex_m3 tests/kernel/sleep -t run -- \
-DCONFIG_ZTEST_SHUFFLE=y -DCONFIG_ZTEST_SHELL=y
Following commands are available:
uart:~$ ztest
ztest - Ztest commands
Subcommands:
run-all :Run all tests
shuffle :Shuffle tests
list-testsuites :List all test suites
list-testcases :List all test cases
run-testsuite :Run test suite
run-testcase :Run testcase
shuffle accepts two arguments --suite_iter and --case_iter which allows
repeated exercution of testcases or suites.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Introduce a new arch level Kconfig option to signal the implementation
of the RISCV Privileged ISA spec. This replaces
SOC_FAMILY_RISCV_PRIVILEGED, because this is not a SoC specific
property, nor a SoC family.
Note that the SoC family naming scheme will be fixed in upcoming
commits.
Signed-off-by: Gerard Marull-Paretas <gerard@teslabs.com>
tag nano libc tests with nano so we can exclude them if needed. Now they
cause issues when collecting coverage information from tests.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Reworked and combined the multi-level interrupt bit masks tests
to work on any configuration of bits in every level.
Updated the bits configuration in the testcase to use
non-symetric numbers so that shifting a level with the wrong
number of bits will certainly cause the test to fail.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
With MMU enabled on Xtensa, user_buffer is not marked as
executable. Invalidating the i-cache by region will cause
an instruction fetch prohibited exception. So skip all
i-cache tests, instead of just the range ones to avoid
confusions of only running the test partially.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
As the pointer to the data buffer passed to k_pipe_put() is now a
'const', that buffer must be initialized before calling k_pipe_put()
to avoid a compiler warning.
Signed-off-by: Peter Mitsis <peter.mitsis@intel.com>
This follows the idea to remove any z_ prefix. Since MMU has
a large number of these, separate out these changes into one
commit to ease review effort.
Since these are no longer have z_, these need proper doxygen
doc. So add them too.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
How cache test could be ran without CONFIG_CACHE_MANAGEMENT?
Adding qemu_x86/x86_64 as cache related API in x86 has proven to be
very brittle against changes recently: no test was at least verifying
it builds properly, not to mention running it.
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
Updated the tests to use info from the devicetree or Kconfig as
ground truth and compare that with the output from the
functions under test.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
Assert that the `local_irq` of each levels should only ranges
from `0` to `CONFIG_MAX_IRQ_PER_AGGREGATOR`, so that it doesn't
overflow the other aggregators.
Also, assert that the output of `z_get_sw_isr_table_idx` shouldn't
overflow the ISR table.
Update the `sw_isr_table` tests to test the range of
`CONFIG_MAX_IRQ_PER_AGGREGATOR` instead of the entire range of
level bits.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
Split the `sw_isr_irq_parent_table` test into one that tests
the IRQ table index function which is generic and should
always be tested, and one that tests multi-instance -related
functions which only work in interrupt controller drivers that
has multi-instance implementation.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
This option (CONFIG_HEAP_MEM_POOL_SIZE) is not used for the tests.
Moreover it overrides the option, specified in the board's Kconfigs.
Signed-off-by: Andrei Emeltchenko <andrei.emeltchenko@intel.com>
Test consistently fails with timeout for board `hifive1`, so exclude
this board until fixed.
Signed-off-by: Manuel Argüelles <manuel.arguelles@nxp.com>
This board have an NXP MPU with 8 configurable regions, of which five
of them are already used by the static MPU configuration. Many of
these tests are failing due to lack of MPU regions or free partition
slots available, even when HW stack protection is disabled when
building with userspace support.
Signed-off-by: Manuel Argüelles <manuel.arguelles@nxp.com>
When testing k_busy_wait(100), don't require that at least 100µs of CPU
time be consumed by the thread itself; some of that time may be used
elsewhere in the system. Instead, just make sure that at least 80µs is
accounted to the thread and allow 20µs to get used elsewhere.
Signed-off-by: Keith Packard <keithp@keithp.com>
Swith integration_platforms from native_posix(_64)
to native_sim(_64).
And replace it also in a comment.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
This test needs to include zephyr/kernel/mm/demand_paging.h, not
zephyr/kernel/mm.h, due to its use of k_mem_pin().
Signed-off-by: Henrik Brix Andersen <hebad@vestas.com>
This moves including of demand_paging.h out of kernel/mm.h,
so that users of demand paging APIs must include the header
explicitly. Since the main user is kernel itself, we can be
more discipline about header inclusion.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
qemu_x86_tiny has very limited memory resources; if too much text is
included in this test, it will not have enough remaining memory to run
it.
When using picolibc before 1.8.5, the only way to get 'long long' support
was to use the full version, including floating point support. This is too
large for this testcase.
Reduce the size of the printf code by switching to the version without
64-bit integer support. This allows the test to pass when using older
picolibc versions, such as that included with SDK version 0.16.3.
Signed-off-by: Keith Packard <keithp@keithp.com>
This enables tests/kernel/mem_protect/mem_protect to be tested
on Xtensa.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Signed-off-by: Flavio Ceolin <flavio.ceolin@intel.com>
This test requires architecture specific code to work.
Signed-off-by: Flavio Ceolin <flavio.ceolin@intel.com>
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
In minimal mode, format modifiers are not supported, leading to a lack
of width and precision support.
long long values are presented correctly in either long long or floating
mode.
Signed-off-by: Keith Packard <keithp@keithp.com>
