From: Markus Probst <markus.probst@posteo.de>
To: Igor Korotin <igor.korotin.linux@gmail.com>,
Miguel Ojeda <ojeda@kernel.org>,
Alex Gaynor <alex.gaynor@gmail.com>,
Wolfram Sang <wsa+renesas@sang-engineering.com>
Cc: "Boqun Feng" <boqun.feng@gmail.com>,
"Gary Guo" <gary@garyguo.net>,
"Björn Roy Baron" <bjorn3_gh@protonmail.com>,
"Benno Lossin" <lossin@kernel.org>,
"Andreas Hindborg" <a.hindborg@kernel.org>,
"Alice Ryhl" <aliceryhl@google.com>,
"Trevor Gross" <tmgross@umich.edu>,
"Danilo Krummrich" <dakr@kernel.org>,
"Greg Kroah-Hartman" <gregkh@linuxfoundation.org>,
"Viresh Kumar" <viresh.kumar@linaro.org>,
"Asahi Lina" <lina+kernel@asahilina.net>,
"Wedson Almeida Filho" <wedsonaf@gmail.com>,
"Alex Hung" <alex.hung@amd.com>,
"Tamir Duberstein" <tamird@gmail.com>,
"Xiangfei Ding" <dingxiangfei2009@gmail.com>,
linux-kernel@vger.kernel.org, rust-for-linux@vger.kernel.org,
linux-i2c@vger.kernel.org
Subject: Re: [PATCH v8 0/4] rust: i2c: Add basic I2C driver abstractions
Date: Mon, 17 Nov 2025 16:15:42 +0000 [thread overview]
Message-ID: <25c70010f67e9f0760840e35cc722d3bb89fd703.camel@posteo.de> (raw)
In-Reply-To: <20251116162104.171420-1-igor.korotin.linux@gmail.com>
add me to cc please.
My led driver currently relies on the i2c bindings.
Thanks
- Markus Probst
[1]
https://lore.kernel.org/rust-for-linux/20251008181027.662616-1-markus.probst@posteo.de/
On Sun, 2025-11-16 at 16:21 +0000, Igor Korotin wrote:
> This patch series lays the groundwork for writing Linux I2C drivers in
> Rust by:
>
> 1. Core abstractions
> Introduce `i2c::I2cClient`, `i2c::I2cAdapter`, `i2c::Driver` and
> built on the existing `struct i2c_client`, `struct i2c_adapter`
> and `struct i2c_driver`, with safe Rust wrappers around probe,
> transfer, and teardown logic.
>
> 2. Manual device creation
> Provide an API to register an I2C device at runtime from Rust using
> `I2cBoardInfo` and `I2cAdapter`, including automatic cleanup when
> the driver unloads.
>
> 3. Sample I2C driver (legacy table, OF & ACPI)
> Add `rust_driver_i2c`, a sample that:
> - binds to an I2C client via:
> - legacy I2C-ID table,
> - Open Firmware (device-tree) compatible strings, or
> - ACPI IDs.
> 4. Sample I2C client registration (OF & ACPI)
> Add `rust_i2c_client`, a sample that:
> - binds to a platform parent device using OF/ACPI table
> - registers new I2C client bound to the parent platform device
> using `i2c::Registration::new()`
> - destroyes the I2C client device on exit.
>
> Together, these four patches:
>
> - Establish the essential Rust traits and types for I2C drivers.
> - Enable driver binding via legacy ID table, device-tree (OF), or ACPI
> - Enable manual device creation at runtime.
> - Ship a samples showing typical usage
>
> Igor Korotin (4):
> rust: i2c: add basic I2C device and driver abstractions
> rust: i2c: add manual I2C device creation abstractions
> samples: rust: add Rust I2C sample driver
> samples: rust: add Rust I2C client registration sample
>
> Changelog
> ---------
> v8:
> - Rebase commits on top of [1]
> - Applied kernel vertical style [2]
> - Replace core::mem::zeroed() with pin_init::zeroed()
> - Rename `I2cAdapter::get_nr` to `I2cAdapter::index`
> - fixed some minor format issues
> - Link to v7: [3]
>
> [1] https://git.kernel.org/pub/scm/linux/kernel/git/driver-core/driver-core.git/log/?h=driver-core-next
> [2] https://docs.kernel.org/rust/coding-guidelines.html#imports
> [3] https://lore.kernel.org/rust-for-linux/20251110112437.50405-1-igor.korotin.linux@gmail.com/
>
> v7:
> - Rebase and update `i2c::Driver::probe` function based on patch
> series [1].
> - use `i2c::Driver::unbind` instead of `drop` in
> `i2c::Adapter::remove_callback`
> - Implement and use `i2c::Driver::unbind` for I2C SampleDriver
> instead of Drop.
> - `i2c::Registration::new` requires a parent device to which a new
> I2C client will be bound.
> - `i2c::Registration::new` returns `impl PinInit<Devres<Self>, Error>`
> - The existing rust_driver_i2c sample has been split into two:
> `rust_driver_i2c`, which demonstrates the implementation of an I2C
> driver, and `rust_i2c_client`, which demonstrates registering a new
> I2C client bound to a parent platform device.
> - Link to v6: [2]
>
> [1] https://lore.kernel.org/all/20251016125544.15559-1-dakr@kernel.org/
> [2] https://lore.kernel.org/rust-for-linux/20251005102226.41876-1-igor.korotin.linux@gmail.com/
>
> v6:
> - Add implementation of unbind for `i2c::Driver` trait;
> - Add argument `Pin<&Self>` to `i2c::Driver::shutdown` method;
> - Adjust usage of `i2c::Driver::shutdown` in
> `i2c::Adapter::shutdown_callback` in `i2c::Driver` trait code
> example and in rust_driver_i2c code;
> - Remove dummy AsRef implementation for I2cAdapter. Adjust code
> in rust_driver_i2c;
> - Add `i2c::I2cAdapter::get_nr` method that returns I2cAdapter index;
> - Optimize unsafe sections in inc_ref/dec_ref in AlwaysRefCounted
> for I2cAdapter implementation;
> - Remove unnecessary Drop implementation for I2cAdapter, because
> I2cAdapter is always a reference;
> - Remove unnecessary type definition `Ops<T>` in rust_driver_i2c
> - Simplify call of `i2c::I2cAdapter::get` in `try_pin_init!` macro
> for rust_driver_i2c
> - Link to v5: https://lore.kernel.org/rust-for-linux/20250911154717.96637-1-igor.korotin.linux@gmail.com/
> v5:
> - Rename missed pdev variables to idev (thanks to Daniel).
> - Import `crate::device_id::RawDeviceIdIndex` and
> `crate::types::AlwaysRefCounted` in i2c.rs.
> - Switch dev_dbg to dev_info in the sample I2C driver messages.
> - Make `I2cAdapter::get()` return `ARef<I2cAdapter>` instead of
> `&I2cAdapter`.
> - Remove `TryFrom<device::Device<Ctx>> for I2cAdapter<Ctx>` (unused;
> to be reintroduced in a later I2C series).
> - Remove `AsRef<device::Device<Ctx>> for I2cAdapter<Ctx>` (unused;
> to be reintroduced in a later I2C series).
> - Add `AsRef<I2cAdapter> for I2cAdapter<Ctx>`.
> - Use i2c_get/put_adapter instead of get/put_device for
> `AlwaysRefCounted<I2cAdapter>`.
> - Update safety comment for `unsafe impl Sync for Registration {}`.
> - Tweak comment for `I2cBoardInfo::new`.
> - Adjust build-time assertion message in `Adapter::register`.
> - Link to v4: https://lore.kernel.org/rust-for-linux/20250820151427.1812482-1-igor.korotin.linux@gmail.com/
> v4:
> - Renamed `i2c::I2cAdapterRef` to `i2c::I2cAdapter`.
> - Renamed `i2c::Device` to `i2c::I2cClient` for consistency with
> `i2c::I2cAdapter` and to avoid confusion with `i2c::Adapter`
> - Reworked `i2c::I2cAdapter` to be an Opaque around `i2c_adapter` struct
> - Implemented AlwaysRefCounted trait for `i2c::I2cAdapter`.
> - Fixed numerous comment mistakes and typos all over the code, thanks
> to Danilo and Daniel
> - Got rid of all unwrap() use-cases in i2c.rs and rust_driver_i2c.rs.
> This covers 0-day kernel panic <202508071027.8981cbd4-lkp@intel.com>
> - Removed unnecessary casts.
> - Replaced all addr_of_mut! macros to &raw mut.
> - In `i2c::Adapter::register` method build assert if all ID tables are
> None.
> - Renamed all pdrv and pdev instances to idrv and idev respectivly
> - Implemented an ealry return in `i2c::Adapter::i2c_id_info`
> - Added all missing Safety comments.
> - Removed `unsafe impl<Ctx: device::DeviceContext> crate::types::AlwaysRefCounted for Device<Ctx>`
> implementation which came to v3 from v2 by mistake.
> - Added more details regarding i2c-stub driver usage in rust_driver_i2c
> comment.
> - Changed `i2c::I2cAdapter::get` return type from `Option<Self>` to
> `Result<&'static Self>`.
> - Added Daniel Almeida as a reviewer to the "I2C Subsystem [RUST]" entry
> in MAINTAINERS, per his offer.
> - Link to v3: https://lore.kernel.org/rust-for-linux/20250801153742.13472-1-igor.korotin.linux@gmail.com/
> v3:
> - removed unnecessary i2c_get_clientdata and i2c_set_clientdata rust
> helpers. Using generic accessors implemented in [1] instead.
> - Reimplemented i2c::DeviceId based on changes in [2].
> - Using from_result in i2c::Adapter::probe_callback
> - Using explicit drop() for i2c client private data in
> `i2c::Adapter::remove_callback`
> - replaced device::Device::as_ref() with device::Device::from_raw in
> `i2c::Device::as_ref()`. It is renamed in device::Device.
> - Build Rust I2C only if I2C is built-in
> - Reimplement overcomplicated trait i2c::DeviceOwned the same way it is
> implemented in auxiliary [3].
> - Merge rust_device_i2c and rust_driver_i2c samples. Resulting
> rust_driver_i2c creates pined i2c_client using i2c::Registration::new
> and probes newly created i2c_client.
> - Created a new entry in MAINTAINERS file containing i2c.rs and
> rust_driver_i2c.rs in it.
> - Link to v2: [4]
>
> [1] https://lore.kernel.org/lkml/20250621195118.124245-3-dakr@kernel.org/
> [2] https://lore.kernel.org/rust-for-linux/20250711040947.1252162-1-fujita.tomonori@gmail.com/
> [3] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/rust/kernel/auxiliary.rs?h=v6.16-rc4#n299
> [4] https://lore.kernel.org/rust-for-linux/20250704153332.1193214-1-igor.korotin.linux@gmail.com/
>
> v2:
> - Merged separated ACPI support patches since ACPI-table support is
> merged into driver-core-next.
> - Added I2cAdapterRef and I2cBoardInfo abstractions
> - Added DeviceState generic parameter which is used for `i2c::Device`
> as a sign if the device is created manually
> - Added `DeviceOwned` abstraction which is a safe reference to a
> manually created `i2c::Device<Ctx, state::Owned>`.
> - Added Rust manual I2C device creation sample
> - Link to v1: https://lore.kernel.org/rust-for-linux/20250626174623.904917-1-igor.korotin.linux@gmail.com/
>
> MAINTAINERS | 9 +
> rust/bindings/bindings_helper.h | 1 +
> rust/kernel/i2c.rs | 589 ++++++++++++++++++++++++++++++++
> rust/kernel/lib.rs | 2 +
> samples/rust/Kconfig | 23 ++
> samples/rust/Makefile | 2 +
> samples/rust/rust_driver_i2c.rs | 74 ++++
> samples/rust/rust_i2c_client.rs | 157 +++++++++
> 8 files changed, 857 insertions(+)
> create mode 100644 rust/kernel/i2c.rs
> create mode 100644 samples/rust/rust_driver_i2c.rs
> create mode 100644 samples/rust/rust_i2c_client.rs
>
>
> base-commit: ededb7bcdfdbcfbb7af93e3a543165a9553e1683
next prev parent reply other threads:[~2025-11-17 16:15 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-11-16 16:21 [PATCH v8 0/4] rust: i2c: Add basic I2C driver abstractions Igor Korotin
2025-11-16 16:21 ` [PATCH v8 1/4] rust: i2c: add basic I2C device and " Igor Korotin
2025-11-17 16:12 ` Markus Probst
2025-11-16 16:21 ` [PATCH v8 2/4] rust: i2c: add manual I2C device creation abstractions Igor Korotin
2025-11-17 16:13 ` Markus Probst
2025-11-16 16:22 ` [PATCH v8 3/4] samples: rust: add Rust I2C sample driver Igor Korotin
2025-11-16 16:22 ` [PATCH v8 4/4] samples: rust: add Rust I2C client registration sample Igor Korotin
2025-11-17 16:15 ` Markus Probst [this message]
2025-11-22 14:00 ` [PATCH v8 0/4] rust: i2c: Add basic I2C driver abstractions Igor Korotin
2025-11-17 21:40 ` Danilo Krummrich
2025-11-22 13:59 ` Igor Korotin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=25c70010f67e9f0760840e35cc722d3bb89fd703.camel@posteo.de \
--to=markus.probst@posteo.de \
--cc=a.hindborg@kernel.org \
--cc=alex.gaynor@gmail.com \
--cc=alex.hung@amd.com \
--cc=aliceryhl@google.com \
--cc=bjorn3_gh@protonmail.com \
--cc=boqun.feng@gmail.com \
--cc=dakr@kernel.org \
--cc=dingxiangfei2009@gmail.com \
--cc=gary@garyguo.net \
--cc=gregkh@linuxfoundation.org \
--cc=igor.korotin.linux@gmail.com \
--cc=lina+kernel@asahilina.net \
--cc=linux-i2c@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=lossin@kernel.org \
--cc=ojeda@kernel.org \
--cc=rust-for-linux@vger.kernel.org \
--cc=tamird@gmail.com \
--cc=tmgross@umich.edu \
--cc=viresh.kumar@linaro.org \
--cc=wedsonaf@gmail.com \
--cc=wsa+renesas@sang-engineering.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox