The test script pinvalidate.py requires the following which are
missing in the unit test stub PinName.h:
* A comment "MBED TARGET LIST"
* `CONSOLE_TX` and `CONSOLE_RX` in the `PinName` enum
This commit adds them.
Individual libraries' `target_h` stub headers have now all been moved
from `mbed-headers-base` to `mbed-headers-<library>`.
Note: Even though headers previously in `target_h` are technically
stubs/fakes too, they are used by not only unit tests but also regular
libraries when compiled for unit tests, because no target-specific HAL
implementation exists in this case. In order for regular library
sources to pick up `target_h` headers, those headers must
* have the same names as regular headers
* appear first in include paths
This is why those headers are part of `mbed-headers-<library>` and not
`mbed-stubs-<library>`. Before this refactoring, `mbed-headers-base`
was the first in unit tests' include paths.
Stubs previously in UNITTESTS/target_h/ have the same names as
regular Mbed OS headers, intending to override the latter directly.
We move hal target_h stubs into hal/tests/UNITTESTS/doubles/.
Note: In Mbed OS, the standard include format requires each header to
be prefixed with its module name, for example "hal/gpio_api.h". This
requires headers to be organized in a module directory. But unit tests
stubs for hal correspond to what a real Mbed target would have
implemented (in a non-test scenario), and targets do not currently put
headers in hal/, so we similary put stub headers directly in
hal/tests/UNITTESTS/doubles/ instead of add a hal/ subdirectory there.
The stub header randLIB.h overrides the header of the same name in
platform/randlib/ which is an external repository vendored into
the mbed-os codebase. As the repository is synchronized regularly,
it is better not to put overrides there, so we put the randLIB.h
stub into the regular platform doubles directory.
The header `mbed.h` is a convenient wrapper that pre-includes some
platform headers, for use by user applications. Libraries and tests
internal to Mbed OS should not use it, and they should explicitly
include headers they need. So a stub is not needed.
Stubs previously in UNITTESTS/target_h/ have the same names as
regular Mbed OS headers, intending to override the latter directly.
We move platform target_h stubs into
platform/tests/UNITTESTS/doubles/platform/.
Note: nvic_wrapper.h is normally implemented and used by Mbed targets
as needed. But as unit tests do not have a real target, we treat it
as a stub for the platform.
For FPU, use armasm to select fpu selection. This will be fixed upstream in the next version of CMSIS. Meanwhile, we use our local patch.
Taken from --cpu selection for armasm:
--cpu | {FPU}
Cortex-M4.fp.sp | VFPv4_SP_D16
Cortex-M7.fp.sp | FPv5-SP
Cortex-M7.fp.dp | FPv5_D16
Tracking issue: https://github.com/ARM-software/CMSIS_5/issues/1266
We use preprocessor for asm files even for Armcc. If symbol is defined it's replaced by preprocessor,
asembler would just see 1 or 0 in this case and errors:
TARGET_M33\\irq_armv8mml.S", line 31: Error: A1185E: Symbol missing
Use preprocessor instead.
rtx_def includes two CMSIS headers that pull in C/C++ headers in our case. As I found out,
they should only define macros. We can fix it but it will require some refactoring as our targets
use mbed rtx headers to define heap using stdin header, plus some other offenders.
Workaround is to exclude the headers we do not need in irq assembly files.
Tracking issue https://github.com/ARMmbed/mbed-os/issues/14962
ARMCC provides __semihost via compat header. As CMSIS 5.8.0 removed this compat header,
we need to explicitly include it to fix definition missing error.
Reference: CMSIS 5.8.0 known issues and https://github.com/ARM-software/CMSIS_5/issues/1211
This fixes the error about redefinition of enable/disable irq. we need compat header because of
semihosting (not yet provided in CMSIS).
References to time should do so using std::chrono. We reworked tests in
connectivity and drivers to use std::chrono and new APIs in order to
remove deprecation warnings resulting from deprecated API calls.
This required addition of a macro for test assertions using std::chrono
values.
As host test "timing_drift_auto" expects time values represented as an
integral number of microseconds, we explicitly provide this in place
using "microseconds{TICKER_TIME}.count()" in the relevant ticker tests.
We recognise this is ugly, but thought it best to descriptively convert
from std::chrono to the host test's required representation.
Co-authored-by: Hari Limaye <hari.limaye@arm.com>
We failed to set `LoRaMac_stub::mlme_ind_ptr` to a valid pointer. When
we tried to dereference it in `LoRaWANStack::process_reception` we hit a
SEGFAULT.
This test was relying on the state of `LoRaMac_stub::bool_true_counter`
to be set to 1 from a previous test. When the test was ran in isolation
the `LoRaWANStack::rx_timeout_interrupt_handler` callback would fail
because we weren't returning true from the stub implementation of
`LoRaMac::nwk_joined`. Due to this
`LoRaWANStack::post_process_tx_no_reception` was never called. This
caused the LoRaWANStack object to think the "tx_metadata" was "stale"
(i.e the data hadn't changed since any previous read). So, when we
attempted to call `LoRaWANStack::acquire_tx_metadata` it returned
`LORAWAN_STATUS_METADATA_NOT_AVAILABLE` as it thought we had no new
metadata to report, causing the test assertion to fail.
We had commented out a line where we reset LoRaPHY_stub::uint16_value to
0. This was causing an invalid array access in
LoRaMac::handle_data_frame, when trying to extract the
_mps_indication.channel from list of channel_params_t returned by
lora_phy->get_phy_channels.
Some tests in athandlertest.cpp relied on the state of stubs set in
previous tests. This caused test failures if the tests cases were ran in
isolation. Remove the interdependencies between tests by setting stub
values in the tests that rely on them.
Previously a test executable was recognised as a single test by CTest.
However, test executables usually contain multiple test cases, the
results of the test cases should be individually reported. With our
previous setup we could miss test case failures that didn't cause the
executable to return an error code.
This commit uses gtest_discover_test to discover all test cases in a
test executable. This enables CTest to match test passes and failures
from the googletest binary output.
The CircularBuffer doesn't allow pushing zero elements; you must push at
least one. Update the CircularBuffer unit test to avoid invalid use of
the CircularBuffer.
We previously made up an invalid message type, which was roughly
`MCPS_MULTICAST | 0x40`. This causes an MBED_ASSERT failure in the
handle_rx function, as it uses an assert to validate the message type
passed to it (in convert_to_msg_flag()). Use the valid message type
MCPS_MULTICAST instead in the handle_rx unit test.
Lingkai Dong
Martin Kojtal
Jaeden Amero
Martin Kojtal
Mohammed Mubeen
harmut01
Charles
Robert Walton