C preprocessor defines in non-function form are assumed by C2Rust to be
constant if they are an expression and not a statement; the LED_PORT was
the only place in the code where that was wrong, and led to compiler
errors due to the value not being constant.
Altering the internal macro to use function form sidesteps that issue.
The generally preferred alternative of using a `const` is unavailable in
this case because the dereferencing operator is already part of the
vendor header file cpu/stellaris_common/include/vendor/cortex-m4-def.h.
The changed macro is documented as required by doccheck. The doccheck
rule that grandfathered in the LED_PORT macro as allowed undocumented is
not removed because it is also used in other board.h files.
Running
USEMODULE=stdio_rtt make BOARD=weact-g030f6 -C ... flash term
failed due the programmer not being specified by the time the serial
is configured:
makefiles/tools/serial.inc.mk:44: "Warning: No RIOT_TERMINAL set,
but using stdio_rtt: The default terminal is likely not to work."
This reorders the setup of the board to first configure the programmer
and then the serial, so that using stdio_rtt now works out of the box.
This fixes a race in `LED<NUM>_TOGGLE`, which is a read-copy-write
operation. Any access to a GPIO pin on the same GPIO port that
happens concurrently could result in data corruption. Using the
GPIO LL API, which is thread-safe, fixes the issue.
Note: The used GPIO LL functions will work even in when the GPIO LL
module is not used.
The API was based on the assumption that GPIO ports are mapped in memory
sanely, so that a `GPIO_PORT(num)` macro would work allow for constant
folding when `num` is known and still be efficient when it is not.
Some MCUs, however, will need a look up tables to efficiently translate
GPIO port numbers to the port's base address. This will prevent the use
of such a `GPIO_PORT(num)` macro in constant initializers.
As a result, we rather provide `GPIO_PORT_0`, `GPIO_PORT_1`, etc. macros
for each GPIO port present (regardless of MCU naming scheme), as well as
`GPIO_PORT_A`, `GPIO_PORT_B`, etc. macros if (and only if) the MCU port
naming scheme uses letters rather than numbers.
These can be defined as macros to the peripheral base address even when
those are randomly mapped into the address space. In addition, a C
function `gpio_port()` replaces the role of the `GPIO_PORT()` and
`gpio_port_num()` the `GPIO_PORT_NUM()` macro. Those functions will
still be implemented as efficient as possible and will allow constant
folding where it was formerly possible. Hence, there is no downside for
MCUs with sane peripheral memory mapping, but it is highly beneficial
for the crazy ones.
There are also two benefits for the non-crazy MCUs:
1. We can now test for valid port numbers with `#ifdef GPIO_PORT_<NUM>`
- This directly benefits the test in `tests/periph/gpio_ll`, which
can now provide a valid GPIO port for each and every board
- Writing to invalid memory mapped I/O addresses was treated as
triggering undefined behavior by the compiler and used as a
optimization opportunity
2. We can now detect at compile time if the naming scheme of the MCU
uses letters or numbers, and produce more user friendly output.
- This is directly applied in the test app
If there is an Ethernet peripheral (periph_eth feature provided), we
can conclude that an Ethernet network interface can be provided.
Co-authored-by: mguetschow <mikolai.guetschow@tu-dresden.de>
The `ethernet` feature has not yet been used, so renaming it should not
cause any issue.
The goal is to eventually have a number of `netif_<type>` features that
would allow filtering boards by the time of connectivity the have.
The ESP32 Ethernet Kit has only a single GPIO pin exposed to SAUL via
`saul_gpio`, but that GPIO doubles as PHY clock input when `esp32_eth`
is used. Hence, the `saul_gpio` dep should be optional and only kick
in when `esp32_eth` is not used.
This gets rid of a long list of boards with network interfaces and
instead let's boards (or MCUs with peripheral network interfaces)
provide the netif feature.
The apps that before used the long list are not depending on the
feature instead (in case of the default example, this is an
optional dependency).
Co-authored-by: mguetschow <mikolai.guetschow@tu-dresden.de>
Co-authored-by: mewen.berthelot <mewen.berthelot@orange.com>
The board has history of bugs that (such as broken timer configuration)
that would directly bite any users, but still have gone unnoticed for
significant time periods due to lack of testing. It appears, that none
of the core contributors has access to the boards and the number of
users is limited.
