linux-stable.git/rust/kernel/acpi.rs, branch v7.3-rc2 Linux kernel stable tree rust: driver: store pointers in `DeviceId` 2026-07-30T20:50:15+00:00 Gary Guo gary@garyguo.net 2026-06-29T12:39:42+00:00 0b76335e8f321a59c32a5f52759f635d2e5db5c8 The common practice in C drivers is to store pointers into `driver_data` field of device IDs. The Rust code is however currently storing indices into the fields and then carry a side table that maps the index to pointers. It is much simpler to just have `DeviceId` carry the pointer like C code does. However, just doing so naively would cause a "pointers cannot be cast to integers during const eval" error, as kernel_ulong_t does not have provenance while pointers do, and Rust forbids `expose_provenance` during consteval. Work around this limitation by wrapping raw IDs in `MaybeUninit`. `MaybeUninit` is allowed to host arbitrary bytes with or without provenance, so we can just then use `unsafe` to store a pointer with provenance there. This has the same effect as changing the C-side definition to use `void*` instead of `kernel_ulong_t`, but without actually changing the C side. Signed-off-by: Gary Guo <gary@garyguo.net> Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Link: https://patch.msgid.link/20260629-id_info-v2-8-56fccbe9c5ef@garyguo.net Signed-off-by: Danilo Krummrich <dakr@kernel.org> 
The common practice in C drivers is to store pointers into `driver_data`
field of device IDs. The Rust code is however currently storing indices
into the fields and then carry a side table that maps the index to
pointers.

It is much simpler to just have `DeviceId` carry the pointer like C code
does. However, just doing so naively would cause a "pointers cannot be cast
to integers during const eval" error, as kernel_ulong_t does not have
provenance while pointers do, and Rust forbids `expose_provenance` during
consteval.

Work around this limitation by wrapping raw IDs in `MaybeUninit`.
`MaybeUninit` is allowed to host arbitrary bytes with or without
provenance, so we can just then use `unsafe` to store a pointer with
provenance there. This has the same effect as changing the C-side
definition to use `void*` instead of `kernel_ulong_t`, but without actually
changing the C side.

Signed-off-by: Gary Guo <gary@garyguo.net>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Link: https://patch.msgid.link/20260629-id_info-v2-8-56fccbe9c5ef@garyguo.net
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
rust: driver: centralize device ID handling 2026-07-30T20:50:15+00:00 Gary Guo gary@garyguo.net 2026-06-29T12:39:40+00:00 1d92f2e8b8810556d631cd5dffb536a66ff55553 Move the `IdArray` creation from individual buses to be handled by shared code in `device_id.rs`. Signed-off-by: Gary Guo <gary@garyguo.net> Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Link: https://patch.msgid.link/20260629-id_info-v2-6-56fccbe9c5ef@garyguo.net Signed-off-by: Danilo Krummrich <dakr@kernel.org> 
Move the `IdArray` creation from individual buses to be handled by shared
code in `device_id.rs`.

Signed-off-by: Gary Guo <gary@garyguo.net>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Link: https://patch.msgid.link/20260629-id_info-v2-6-56fccbe9c5ef@garyguo.net
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
rust: acpi: replace `core::mem::zeroed` with `pin_init::zeroed` 2025-10-27T19:27:05+00:00 Siyuan Huang huangsiyuan@kylinos.cn 2025-10-20T03:12:04+00:00 040beccb038b7a7dec60ef20383278347a2aa09e All types in `bindings` implement `Zeroable` if they can, so use `pin_init::zeroed` instead of relying on `unsafe` code. If this ends up not compiling in the future, something in bindgen or on the C side changed and is most likely incorrect. Link: https://github.com/Rust-for-Linux/linux/issues/1189 Suggested-by: Benno Lossin <lossin@kernel.org> Signed-off-by: Siyuan Huang <huangsiyuan@kylinos.cn> Reviewed-by: Alice Ryhl <aliceryhl@google.com> Reviewed-by: Benno Lossin <lossin@kernel.org> Acked-by: Danilo Krummrich <dakr@kernel.org> Reviewed-by: Kunwu Chan <chentao@kylinos.cn> Link: https://patch.msgid.link/20251020031204.78917-1-huangsiyuan@kylinos.cn Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> 
All types in `bindings` implement `Zeroable` if they can, so use
`pin_init::zeroed` instead of relying on `unsafe` code.

If this ends up not compiling in the future, something in bindgen or on
the C side changed and is most likely incorrect.

Link: https://github.com/Rust-for-Linux/linux/issues/1189
Suggested-by: Benno Lossin <lossin@kernel.org>
Signed-off-by: Siyuan Huang <huangsiyuan@kylinos.cn>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Reviewed-by: Benno Lossin <lossin@kernel.org>
Acked-by: Danilo Krummrich <dakr@kernel.org>
Reviewed-by: Kunwu Chan <chentao@kylinos.cn>
Link: https://patch.msgid.link/20251020031204.78917-1-huangsiyuan@kylinos.cn
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
rust: acpi: use `core::ffi::CStr` method names 2025-09-16T07:26:59+00:00 Tamir Duberstein tamird@gmail.com 2025-08-13T15:42:02+00:00 657403637f7d343352efb29b53d9f92dcf86aebb Prepare for `core::ffi::CStr` taking the place of `kernel::str::CStr` by avoid methods that only exist on the latter. Signed-off-by: Tamir Duberstein <tamird@gmail.com> Reviewed-by: Benno Lossin <lossin@kernel.org> Signed-off-by: Miguel Ojeda <ojeda@kernel.org> 
Prepare for `core::ffi::CStr` taking the place of `kernel::str::CStr` by
avoid methods that only exist on the latter.

Signed-off-by: Tamir Duberstein <tamird@gmail.com>
Reviewed-by: Benno Lossin <lossin@kernel.org>
Signed-off-by: Miguel Ojeda <ojeda@kernel.org>
rust: device_id: split out index support into a separate trait 2025-07-16T21:38:21+00:00 FUJITA Tomonori fujita.tomonori@gmail.com 2025-07-11T04:09:45+00:00 8d84b32075fb2d9bd95c7e47b165942411d74bba Introduce a new trait `RawDeviceIdIndex`, which extends `RawDeviceId` to provide support for device ID types that include an index or context field (e.g., `driver_data`). This separates the concerns of layout compatibility and index-based data embedding, and allows `RawDeviceId` to be implemented for types that do not contain a `driver_data` field. Several such structures are defined in include/linux/mod_devicetable.h. Refactor `IdArray::new()` into a generic `build()` function, which takes an optional offset. Based on the presence of `RawDeviceIdIndex`, index writing is conditionally enabled. A new `new_without_index()` constructor is also provided for use cases where no index should be written. This refactoring is a preparation for enabling the PHY abstractions to use the RawDeviceId trait. The changes to acpi.rs and driver.rs were made by Danilo. Reviewed-by: Trevor Gross <tmgross@umich.edu> Signed-off-by: FUJITA Tomonori <fujita.tomonori@gmail.com> Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Link: https://lore.kernel.org/r/20250711040947.1252162-2-fujita.tomonori@gmail.com Signed-off-by: Danilo Krummrich <dakr@kernel.org> 
Introduce a new trait `RawDeviceIdIndex`, which extends `RawDeviceId`
to provide support for device ID types that include an index or
context field (e.g., `driver_data`). This separates the concerns of
layout compatibility and index-based data embedding, and allows
`RawDeviceId` to be implemented for types that do not contain a
`driver_data` field. Several such structures are defined in
include/linux/mod_devicetable.h.

Refactor `IdArray::new()` into a generic `build()` function, which
takes an optional offset. Based on the presence of `RawDeviceIdIndex`,
index writing is conditionally enabled. A new `new_without_index()`
constructor is also provided for use cases where no index should be
written.

This refactoring is a preparation for enabling the PHY abstractions to
use the RawDeviceId trait.

The changes to acpi.rs and driver.rs were made by Danilo.

Reviewed-by: Trevor Gross <tmgross@umich.edu>
Signed-off-by: FUJITA Tomonori <fujita.tomonori@gmail.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Link: https://lore.kernel.org/r/20250711040947.1252162-2-fujita.tomonori@gmail.com
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
rust: acpi: remove unneeded cast to clean future Clippy warning 2025-07-03T17:18:59+00:00 Miguel Ojeda ojeda@kernel.org 2025-07-01T17:46:56+00:00 97ba207a994f7d0cc58f19d448e98db6735f5e77 A future Clippy warning, `clippy::as_underscore`, is getting enabled in parallel in the rust-next tree: error: using `as _` conversion --> rust/kernel/acpi.rs:25:9 | 25 | self.0.driver_data as _ | ^^^^^^^^^^^^^^^^^^^^^^- | | | help: consider giving the type explicitly: `usize` The type is already `ulong`, which nowadays is always `usize`, so the cast is unneeded. Thus remove it, which in turn will avoid the warning in the future. Other abstractions of device tables do not use a cast here either. Signed-off-by: Miguel Ojeda <ojeda@kernel.org> Reviewed-by: Trevor Gross <tmgross@umich.edu> Link: https://lore.kernel.org/r/20250701174656.62205-1-ojeda@kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
A future Clippy warning, `clippy::as_underscore`, is getting enabled in
parallel in the rust-next tree:

    error: using `as _` conversion
      --> rust/kernel/acpi.rs:25:9
       |
    25 |         self.0.driver_data as _
       |         ^^^^^^^^^^^^^^^^^^^^^^-
       |                               |
       |                               help: consider giving the type explicitly: `usize`

The type is already `ulong`, which nowadays is always `usize`, so the
cast is unneeded. Thus remove it, which in turn will avoid the warning
in the future.

Other abstractions of device tables do not use a cast here either.

Signed-off-by: Miguel Ojeda <ojeda@kernel.org>
Reviewed-by: Trevor Gross <tmgross@umich.edu>
Link: https://lore.kernel.org/r/20250701174656.62205-1-ojeda@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
rust: acpi: add `acpi::DeviceId` abstraction 2025-06-26T21:22:17+00:00 Igor Korotin igor.korotin.linux@gmail.com 2025-06-20T15:24:25+00:00 a74931eb59cf6b914e1b9471d9abd4429176f6af `acpi::DeviceId` is an abstraction around `struct acpi_device_id`. Enable drivers to build ACPI device ID tables, to be consumed by the corresponding bus abstractions, such as platform or I2C. Signed-off-by: Igor Korotin <igor.korotin.linux@gmail.com> Acked-by: Rafael J. Wysocki <rafael@kernel.org> Link: https://lore.kernel.org/r/20250620152425.285683-1-igor.korotin.linux@gmail.com [ Always inline DeviceId::new() and use &'static CStr; slightly reword commit message. - Danilo ] Signed-off-by: Danilo Krummrich <dakr@kernel.org> 
`acpi::DeviceId` is an abstraction around `struct acpi_device_id`.

Enable drivers to build ACPI device ID tables, to be consumed by the
corresponding bus abstractions, such as platform or I2C.

Signed-off-by: Igor Korotin <igor.korotin.linux@gmail.com>
Acked-by: Rafael J. Wysocki <rafael@kernel.org>
Link: https://lore.kernel.org/r/20250620152425.285683-1-igor.korotin.linux@gmail.com
[ Always inline DeviceId::new() and use &'static CStr; slightly reword
  commit message. - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>