Re: [PATCH v4 3/5] rust: driver: Consolidate `Adapter` methods using `#[cfg]`

[Igor Korotin]

On Tue, Jun 10, 2025 at 4:15 PM Danilo Krummrich <dakr@xxxxxxxxxx> wrote:
>
> On Tue, Jun 10, 2025 at 04:05:38PM +0100, Igor Korotin wrote:
> > Refactor the `acpi_id_info` and `of_id_info` methods in the `Adapter` trait
> > to reduce duplication. Previously, each method had two versions selected
> > via `#[cfg(...)]` and `#[cfg(not(...))]`. This change merges them into a
> > single method per case by using `#[cfg]` blocks within the method bodies.
> >
> > Suggested-by: Benno Lossin <lossin@xxxxxxxxxx>
> > Signed-off-by: Igor Korotin <igor.korotin.linux@xxxxxxxxx>
> > ---
> >  rust/kernel/driver.rs | 76 +++++++++++++++++++++----------------------
> >  1 file changed, 38 insertions(+), 38 deletions(-)
> >
> > diff --git a/rust/kernel/driver.rs b/rust/kernel/driver.rs
> > index 638f1d270af8..805f61bd4a50 100644
> > --- a/rust/kernel/driver.rs
> > +++ b/rust/kernel/driver.rs
> > @@ -147,30 +147,30 @@ pub trait Adapter {
> >      /// Returns the driver's private data from the matching entry in the [`acpi::IdTable`], if any.
> >      ///
> >      /// If this returns `None`, it means there is no match with an entry in the [`acpi::IdTable`].
> > -    #[cfg(CONFIG_ACPI)]
> >      fn acpi_id_info(dev: &device::Device) -> Option<&'static Self::IdInfo> {
> > -        let table = Self::acpi_id_table()?;
> > +        #[cfg(not(CONFIG_ACPI))] {
> > +            let _ = dev;
> > +            return None;
> > +        }
> >
> > -        // SAFETY:
> > -        // - `table` has static lifetime, hence it's valid for read,
> > -        // - `dev` is guaranteed to be valid while it's alive, and so is `pdev.as_ref().as_raw()`.
> > -        let raw_id = unsafe { bindings::acpi_match_device(table.as_ptr(), dev.as_raw()) };
> > +        #[cfg(CONFIG_ACPI)] {
> > +            let table = Self::acpi_id_table()?;
> >
> > -        if raw_id.is_null() {
> > -            None
> > -        } else {
> > -            // SAFETY: `DeviceId` is a `#[repr(transparent)` wrapper of `struct of_device_id` and
> > -            // does not add additional invariants, so it's safe to transmute.
> > -            let id = unsafe { &*raw_id.cast::<acpi::DeviceId>() };
> > +            // SAFETY:
> > +            // - `table` has static lifetime, hence it's valid for read,
> > +            // - `dev` is guaranteed to be valid while it's alive, and so is `pdev.as_ref().as_raw()`.
> > +            let raw_id = unsafe { bindings::acpi_match_device(table.as_ptr(), dev.as_raw()) };
> >
> > -            Some(table.info(<acpi::DeviceId as crate::device_id::RawDeviceId>::index(id)))
> > -        }
> > -    }
> > +            if raw_id.is_null() {
> > +                None
> > +            } else {
> > +                // SAFETY: `DeviceId` is a `#[repr(transparent)` wrapper of `struct of_device_id` and
> > +                // does not add additional invariants, so it's safe to transmute.
> > +                let id = unsafe { &*raw_id.cast::<acpi::DeviceId>() };
> >
> > -    #[cfg(not(CONFIG_ACPI))]
> > -    #[allow(missing_docs)]
> > -    fn acpi_id_info(_dev: &device::Device) -> Option<&'static Self::IdInfo> {
> > -        None
> > +                Some(table.info(<acpi::DeviceId as crate::device_id::RawDeviceId>::index(id)))
> > +            }
> > +        }
> >      }
>
> It seems there has been a misunderstanding, the idea was to refactor
> Adapter::of_id_info() in one patch and introduce Adapter::acpi_id_info() in a
> subsequent patch already being the suggested implementation by Benno.
>
> We try to avoid introducing code, just to change it again in a subsequent patch.

That does make perfect sense. My bad. I'll reorder changes in the v5.

Sorry for the mess. Thanks for the review.

Best Regards
Igor