Add devicetree support to Rust on Zephyr #76074
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Add the `CONFIG_RUST` Kconfig. This can be set to indicate that an application wishes to use Rust support. Signed-off-by: David Brown <david.brown@linaro.org>
Until further variants are needed, this provides a `main()` function that simply calls into the `rust_main()` function. This is adequate for applications where main can be directly written in Rust. Signed-off-by: David Brown <david.brown@linaro.org>
The initial support crate for zephyr use within a Rust application. This crate does two simple things: - Processes the .config file generated for this build, and ensures that CONFIG_RUST is enabled. - Provide a hanging Rust panic handler. Signed-off-by: David Brown <david.brown@linaro.org>
This provides the function `rust_cargo_application()` within Cmake to give applications written in Rust a simple way to start. This implements the basic functionality needed to build a rust application on Arm, including target mapping, and invoking cargo during the build. Signed-off-by: David Brown <david.brown@linaro.org>
This implements a simple hello world application in Rust. Since there are no bindings yet, this just directly calls `printk` from Rust code. Signed-off-by: David Brown <david.brown@linaro.org>
Indicate the new code is maintained. Signed-off-by: David Brown <david.brown@linaro.org>
Create a crate `zephyr-build` that contains the code that will run at build time. Move the bool kconfig handing from this. This will make it easier for an application crate that needs access to config entries to do access them by: - Adding a dependency to zephyr-build to `[build-dependencies]` in their Cargo.toml. - Creating a build.rs, or adding a call to: zephyr_build::export_bool_kconfig() in it. Signed-off-by: David Brown <david.brown@linaro.org>
As part of the build of the `zephyr` crate, convert string and numeric entries from the .config file of the current build into constants. This mimics how these are available as defines in C code. The defines are available under `zephyr::kconfig::CONFIG_...`. Signed-off-by: David Brown <david.brown@linaro.org>
This is a bit awkward, as we don't have a clean interface between Rust and the Zephyr C code, but copy and null-terminate the current board, and include that in the printed message. This demonstrates that the kconfig module generation is working. Signed-off-by: David Brown <david.brown@linaro.org>
Add initial docs for Rust language support. Explains the support currently present, how to install the toolchain needed, and explains what is in the current sample. Signed-off-by: David Brown <david.brown@linaro.org>
The documentation system requires all of the samples to have unique names, no matter what directory they are in. Rename our sample to `hello_rust` to avoid a conflict with the existing `hello_world` sample. Signed-off-by: David Brown <david.brown@linaro.org>
Code-blocks for shell commands are of type 'shell', to avoid this warning. Signed-off-by: David Brown <david.brown@linaro.org>
Distinguish between 'int' and 'hex' Kconfig values. Use `usize` for the hex values, as these are unsigned and may have the high bit set. But, for 'int' values, use `isize`, and allow the constants to be negative. Signed-off-by: David Brown <david.brown@linaro.org>
This is still very experimental, mark it in the Kconfig as such. Signed-off-by: David Brown <david.brown@linaro.org>
For Arm targets, match the ABI selection defines from the cmake/compiler/gcc/target_arm.cmake file. What is important here is that the floating point selection is the same between gcc and rustc. Signed-off-by: David Brown <david.brown@linaro.org>
Because of a mismatched quote, shell quoting for this block fails sphinx. Removing the quite causes compliance to fail due to the mispelling. Fix this by just removing formatting entirely from the block. Signed-off-by: David Brown <david.brown@linaro.org>
Compliance fails if even an example refers to non-existent Kconfig options. Change these to real options, even though that is slightly distracting. Signed-off-by: David Brown <david.brown@linaro.org>
The Cortex-M4 should be the thumbv7em not thumbv7m. Signed-off-by: David Brown <david.brown@linaro.org>
Make this a proper C prototype function that takes no arguments. Signed-off-by: David Brown <david.brown@linaro.org>
zephyrbot
added
area: Documentation
area: Build System
area: Samples
zephyrbot
requested review from
57300,
carlescufi,
jeremybettis,
kartben,
MaureenHelm and
nashif
|
I've rebased this on top of #76337, as having generated bindings makes a lot of this quite a bit easier. |
Gcc automatically defines __SOFTFP__ on Cortex-M targets where soft floating point is used. Clang, which is used by bindgen, does not define this. Workaround this by detecting this situation, and defining this. Signed-off-by: David Brown <david.brown@linaro.org>
Implement a general critical section handler, using spinlocks, for Rust users to be able to use the `critical-section` crate. This crate is used by other crates as well. Signed-off-by: David Brown <david.brown@linaro.org>
Whether a device uses a gpio pin depends on the DT. Allow this symbol to be unused, for targets that doing have gpio pins defined. Signed-off-by: David Brown <david.brown@linaro.org>
pdgendt
reviewed
Jul 30, 2024
| unsafe extern "C" fn blink(_p1: *mut c_void, _p2: *mut c_void, _p3: *mut c_void) { | ||
| warn!("Inside of blinky"); | ||
|
|
||
| let mut led0 = zephyr::devicetree::aliases::led0::get_instance(); |
There was a problem hiding this comment.
Can we somehow make use of Option types, similar to how DEVICE_DT_GET_OR_NULL would work? This would allow to re-use code for different platforms where runtime checks are forced if a device tree node would be available or not.
There was a problem hiding this comment.
That seems reasonable, but I'll have to think about how that would work. I do like the idea of the DT being actually present as modules, but I'm not aware of a way to query that, and it doesn't really make sense to query the existence of modules at compile time.
This crate runs bindgen at build time to generate bindings to the Zephyr API surface. Signed-off-by: David Brown <david.brown@linaro.org>
This adds an atomic to each static Rust-defined kobject to ensure that they are properly initialized. It adds a small amount of overhead to each object, as well as to when they are first used. Signed-off-by: David Brown <david.brown@linaro.org>
The bindgen tool will generate bindings for #defines that resolve to numeric constants. This misses some otherwise constant bindings that have a type cast in them. We can, however, export some of these, by making a const in the C code that bindgen will export. Signed-off-by: David Brown <david.brown@linaro.org>
Implement a simple static wrapper to use with Zephyr kernel objects, and use this to implement static threads. This defines a static thread and a static thread stack that can be initialized with a simple rust function, with no arguments. Signed-off-by: David Brown <david.brown@linaro.org>
Start an implementation of the dining philosopher's problem, in Rust. There is just a child thread, and the mutex use is unsafe. But, this demonstrates the new Rust thread support. Signed-off-by: David Brown <david.brown@linaro.org>
When alloc is enabled for Rust code, provide an easier to use `spawn` function that boxes a closure and sends it to the child side. This requires allocations to hold the data passed to the child. Signed-off-by: David Brown <david.brown@linaro.org>
Change this to test the `spawn` function we have available with allocation. Signed-off-by: David Brown <david.brown@linaro.org>
Mark this as a TODO, disabling putting the k_thread in the noinit section. It is supposed to only do this when we aren't debugging initialization, but the conditional attribute doesn't seem to work. For now, put it in initialized memory, which will run correctly, just a little longer on init time, but the atomic will be initialized properly. Signed-off-by: David Brown <david.brown@linaro.org>
`zephyr::sys::Mutex` is a simple wrapper for the k_mutex that provides safe access to lock and unlock. Signed-off-by: David Brown <david.brown@linaro.org>
This eliminates the unsafe operations here. Signed-off-by: David Brown <david.brown@linaro.org>
This is a little complicated. We currently just jump directly to rust's panic, so we aren't likely to actually need these unwind tables. But, they would be needed if we ever reall paniced. As of the time of this change, the tables seem to just be 24 bytes of ROM space. Signed-off-by: David Brown <david.brown@linaro.org>
This is a Mutex object, similar to that in Rust's std::sync::Mutex. It doesn't implement poisoning yet, as we have no poisoning to speak of. This can be used as an `Arc<Mutex<T>>` to protect a block of data. The initializer requires a sys::Mutex to be passed in, as these are typically static in Rust applications. The wrapped mutex can live in an Arc. Signed-off-by: David Brown <david.brown@linaro.org>
Make these symbols available to Rust code. They are too complex for bindgen to inline, so represent the values as consts, which does depend on their specific values. Signed-off-by: David Brown <david.brown@linaro.org>
The low-level mutex can be safely cloned (and is send as well). Be sure to derive clone so that it can be used this way. Signed-off-by: David Brown <david.brown@linaro.org>
The philosopher demo now uses the "friendly" wrapped mutexes, and as part of the demo, places an incrementing counter inside of the mutex. Signed-off-by: David Brown <david.brown@linaro.org>
|
This pull request has been marked as stale because it has been open (more than) 60 days with no activity. Remove the stale label or add a comment saying that you would like to have the label removed otherwise this pull request will automatically be closed in 14 days. Note, that you can always re-open a closed pull request at any time. |
|
Rust on Zephyr work is happening on the zephyr-lang-rust module. This is no longer relevant. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This is the beginnings of adding support for the devicetree to Rust on Zephyr. This currently:
zephyr.dtsand thedevicetree_generated.hheader, and builds a representation of that DT in memory at build time.zephyr::devicetree::aliases::led0and get the same node as the actual led node this alias points to. Deeper nested phandles are not yet resolved._DEBUGappended to the symbol name, and the string a debug dump of the value data structure of that property.Also, this is built on top of #75904.