Add a python script that allows having a single YAML file as source of
truth for what features are available, and also add documentation such
as help texts and grouping to them.
This utility converts such a YAML file to a Makefile with the contents
```
FEATURES_EXISTING := \
feature_a \
feature_b \
... \
features_z \
#
```
This allows the Makefile based build system to easily check if all
provided features and all requested features do actually exist.
In addition, this also converts the YAML file to a markdown file that
documents the features. This file is then intended to be used by
Doxygen to document the provided features.
Co-authored-by: mguetschow <mikolai.guetschow@tu-dresden.de>
Co-authored-by: chrysn <chrysn@fsfe.org>
This drops support for the legacy riscv-none-embed target triple. That
value has been incorrect since the beginning and the toolchain that
used that has been long declared obsolete and is fairly outdated.
With our CI updating the toolchain, we no longer need to check for
that.
On RISC-V a RAM overlow is reported as:
/opt/xpack-riscv-none-elf-gcc-13.2.0-2/bin/../lib/gcc/riscv-none-elf/13.2.0/../../../../riscv-none-elf/bin/ld: section .stack VMA [80003f00,80003fff] overlaps section .bss VMA [800000a8,80003fb7]
This extends the list of patters detecting RAM overflow.
Co-authored-by: benpicco <benpicco@googlemail.com>
This adds mspdebug as package, similar to EDBG, so that the
programmer/debugger is build from source.
This has the advantage that we can indeed provide patches of our own.
The first patch fixes a bug with the CPU detection of `mspdebug` in
combination with the Olimex MSP430-JTAG-TINY-V2. The second adds the
`--expect-id <CPU_NAME>` argument.
The RIOT integration is updated to directly make use of the
`--expect-id` parameter. No more spending time debugging why firmware
the firmware for the `olimex-msp430-h2618` doesn't run when flashed on
the `olimex-msp430h1611` hardware :D
- Test native64 like native in murdock
- Add native64 to "Platform: native" in github labeler
- Add "BUILDTEST_MCU_GROUP == x86_64" to `dist/tools/ci/build_and_test.sh`
This will use the make test-input-hash-changed feature to save
the test hashes with the results and optionally skip running
the test if nothing has changed.
Murdock already has this feature but it is not easily accessible.
This should prevent unneeded flash cycles as well as speeding up
constant rerunning of tests for boards.
Also complete and explain `PROGRAMMER=`, `OPENOCD_DEBUG_ADAPTER=`, and
`OPENOCD_RESET_USE_CONNECT_ASSERT_SRST=` when using zsh.
Co-authored-by: benpicco <benpicco@googlemail.com>
When using OpenOCD RTT (real-time transfer) for stdio with
`USEMODULE=stdio_rtt make ...`, it is no longer possible to debug
while being connected to stdio. By also opening gdb at RIOT's default
GDB port, `make debug-client` can be used to connect from GDB to the
OpenOCD instance providing stdio via RTT.
Sometimes we want to roll-back to a previous firmware version.
To do so we need a higher riotbot version numbers still.
Add an update command to the riotboot_gen_hdr tool so that an existing
firmware image can be re-rolled out with a new version number.
This adds command completion that, when manually installed, provides
RIOT-aware command completion for `make` in zsh. The completion
supports:
- most commonly used `make` goals including a help text
- most commonly used `make` variables (such as `BOARD=`, `LTO=`,
`BUILD_IN_DOCKER=`, `TOOLCHAIN=`)
- possible values for these variables, including a list of boards
detected at runtime to complete `BOARD=`
Co-authored-by: mguetschow <mikolai.guetschow@tu-dresden.de>
Since the native process is a child process of pyterm when called via
make we need to dig deeper for riotctrl. This also means that we require
pyserial to execute it.
Likely a change in the command line interface of EDBG broke the
`make reset` integration. This adds the missing flag.
Note: Since the binary is build on demand with our build system, we
can in fact rely on every using the same version of EDBG. Hence,
we don't need to be backwards compatible with the flag.
Previously the `add_insufficient_memory_board.sh` script would only
add the tested board to `Makefile.ci` when resource requirements grew,
but never remove it when resource requirements got lower.
This updates the script to also remove boards from `Makefile.ci` if
it is no longer needed. It is also rename (`add` --> `update`) to
match the change in behavior.
20009: cpu/native: fix bug in periph_timer r=MrKevinWeiss a=maribu
### Contribution description
While debugging https://github.com/RIOT-OS/RIOT/pull/18977#issuecomment-1764258356 it became obvious that the `periph_timer` in `native` is broken and issues early IRQs. This replaces the use of `setitimer` that cannot use a monotonic clock source with `timer_settime()`.
### Testing procedure
I have some non-publishable code that tests if the time an ISR fires in terms of `timer_read()` is no earlier than the time expected. This occasionally triggered with `master`, but I didn't see any of these issues anymore with this PR. I guess I should revive my PR to spice up the periph timer tests and add a polished version of this and let this run for an hour or two.
The tests ins `tests/periph/timer*` should still succeed on `native`. (They do for me in a container running `riot/riotbuild`).
### Issues/PRs references
Found while debugging https://github.com/RIOT-OS/RIOT/pull/18977#issuecomment-1764258356
20042: dist/tools/uf2: add target to also copy families.json file r=MrKevinWeiss a=MichelRottleuthner
### Contribution description
The updated UF2 pkg (#20035) stores the family ID in an external .json file. I overlooked that and flashing fails if this file is not present. This PR fixes it by also copying the json into the tool folder.
### Testing procedure
Check if the `feather-nrf52840-sense` can be flashed when the new UF2 pkg is cloned freshly.
### Issues/PRs references
Fixes a regression introduced with #20035
Co-authored-by: Marian Buschsieweke <marian.buschsieweke@posteo.net>
Co-authored-by: Michel Rottleuthner <michel.rottleuthner@haw-hamburg.de>
20035: Bump version of UF2 package r=MrKevinWeiss a=MichelRottleuthner
### Contribution description
Bump UF2 package version. The new version, besides many other things, also includes an improvement to search more potential mount paths, which gives better compatibility for different distros.
### Testing procedure
Test if flashing still works for boards that use UF2 (e.g., `feather-nrf52840`, `feather-nrf52840-sense`, `nrf52840-mdk-dongle`):
`BUILD_IN_DOCKER=1 BOARD=feather-nrf52840-sense make -C examples/hello-world all flash term`
With master you should get this error, during flashing: `No drive to deploy.` if you run it on a distro that mounts your device under something like `/run/media/<username>/FTHRSNSBOOT`, or other paths not supported by the old version. To me this happened on Arch.
**Note: you will still get this error if your system doesn't auto mount the device.** But now it will be found when mounted ;)
With this PR try again with the UF2 util:
`rm -rf build/pkg/UF2`
`BUILD_IN_DOCKER=1 BOARD=feather-nrf52840-sense make -C examples/hello-world all flash term`
It should use the new pkg version and flash successfully:
```
...
[INFO] uf2conv.py not found - fetching it from GitHub now
CC= CFLAGS= make -C /home/michel/devel/riot/dist/tools/uf2
[INFO] uf2conv.py successfully fetched!
...
Flashing /run/media/michel/FTHRSNSBOOT (nRF52840-Feather-Sense)
Wrote 63488 bytes to /run/media/michel/FTHRSNSBOOT/NEW.UF2
...
```
### Issues/PRs references
Fixes the mount path problem described in #20027.
Co-authored-by: Michel Rottleuthner <michel.rottleuthner@haw-hamburg.de>
