imxrt-rs / imxrt-iomuxc

Pin definitions, configurations, and multiplexing API
Apache License 2.0
1 stars 14 forks source link

imxrt-iomuxc

i.MX RT pad definitions and pin configuration. Use this library to

imxrt-iomuxc targets the latest stable Rust compiler.

Getting started

If you're already using imxrt-hal as you're hardware abstraction layer, there's nothing more to do. imxrt-hal re-exports this API and ensures compatibility with your i.MX RT processor. See the imxrt-hal documentation for more information on acquiring pads and instantiating drivers.

If you're defining a new i.MX RT hardware driver, consider using imxrt-iomuxc to constrain the pads that are compatible with your driver. Depend on the latest imxrt-iomuxc package, and do not enable any package features. Then, implement your APIs with the various imxrt-iomuxc pin traits. See the imxrt-iomuxc API documentation for examples. For even more examples, study the imxrt-hal package.

If you're defining a board support package, consider using imxrt-iomuxc to rename the pads that are supported by your board. Depend on the latest imxrt-iomuxc package, and enable the feature for your i.MX RT processor. Then, expose type aliases and / or objects that rename i.MX RT processor pads according to your board. If you're looking for an example of this pattern, consult the teensy4-pins package.

Supported chip

The table below shows the chips and their corresponding crate feature. Note that chips are denoted by reference manuals and may support different variants. For instance, the 1060 feature supports 1061 and 1062 chip variants.

Chip Feature
1010 "imxrt1010"
1020 "imxrt1020"
1060 "imxrt1060"
1170 "imxrt1170"

Read on if you're interested in adding support for another i.MX RT microcontroller, or if you want to expand existing support.

Contributing

We appreciate your contributions. After discussing the development basics, this section describes how you can

Generally, open an issue if you're interested in any type of contribution, or if you find a defect.

Development basics

Build the package with cargo. The first command only builds the pin traits and configuration API; it does not include any chip support. The second command includes 1060 pad definitions and pin implementations. The third command includes all supported pad definitions and pin implementations for an ARM target.

cargo build
cargo build --features=imxrt1060
cargo build --all-features --target=thumbv7em-none-eabihf

All code must build for your host and for an embedded target. Notice how a target like thumbv7em-none-eabihf is optional.

Run tests with

cargo test --all-features

In particular, documementation tests may that the crate is built with the 1060 feature. Please follow this precedent when adding documentation examples.

Generate documentation with cargo doc. You can optionally enable (all) chip features.

cargo doc
cargo doc --features=imxrt1060
cargo doc --all-features

If you're interested in generating the docs.rs documentation, use a nightly toolchain, and supply the docsrs configuration. This will include the documentation tags that highlight build configurations.

cargo +nightly rustdoc --all-features -- --cfg=docsrs

Adding support for new chips

Use the iomuxc.py script to generate all pads for a new i.MX RT chip. The script extracts pad definitions from a system view description (SVD) file. Consider using the SVDs maintained in the imxrt-ral repository.

Once you have an SVD and can generate the pads module, integrate the pads module into the package. Use the existing 1010, 1060, and 1170 support as your guide. As of now, the process roughly follows

  1. Conditionally include the module in lib.rs.
  2. Create a new directory for your chip, and add a mod.rs file. Include and re-export your script-generated pads module.
  3. Add a Cargo.toml feature for your chip.

By the end of this process, you have definitions for your i.MX RT chip's pads. You should also have GPIO pin implementations. However, you do not have pin implementations for other peripherals. Read on to learn about that contribution process.

Extending pin implementations for existing pads

This section assumes that a pin trait exists for your peripheral. If this doesn't exist, read the next section for guidance.

When you implement a pin trait for a processor pad, you signal to users that this pad supports a particular peripheral function. As of this writing, we write these implementations by hand using the reference manual (RM) or SVD as a source of truth. The general imxrt-rs project documentation has tips for obtaining a RM. Once you have an RM, the table in the "External Signals and Pin Multiplexing" chapter should reveal the pads that support peripheral functions.

To extend pin implementations, either append to the existing pin module for your chip, or create a new pin module for your chip. If your pin implementation requires daisy registers, consider using the daisy.py script to extract them from a SVD file. (If you don't have SVD files, see the previous section.) Again, use the existing 1010, 1060, and 1170 support as your guide.

We welcome any contribution to automate this process. See the tracking issue for more information.

Define new pin traits

When you define a new pin trait, you specify all of the information needed for a pad to be configured for a particular peripheral function. For example, an LPUART pin trait needs to be concerned with specifing

The trait describes these parameters, and functions designed to those traits prepare the pads for the chosen configuration.

The trait complexity may vary depending on the peripheral. If you're not sure what a trait might look like, start by studying the pad multiplexing options, then translate that into a trait. Consult the existing pin traits for different design techniques.

License

Licensed under either of

at your option.

Unless you explicitly state otherwise, any contribution intentionally submitted for inclusion in the work by you, as defined in the Apache-2.0 license, shall be dual licensed as above, without any additional terms or conditions.