Re: [PATCH 04/10] rust: xarray: add `XArrayState`

[Andreas Hindborg <a.hindborg@kernel.org>]

Tamir Duberstein <tamird@gmail.com> writes:

> On Wed, Dec 3, 2025 at 5:27 PM Andreas Hindborg <a.hindborg@kernel.org> wrote:
>>
>> Add `XArrayState` as internal state for XArray iteration and entry
>> operations. This struct wraps the C `xa_state` structure and holds a
>> reference to a `Guard` to ensure exclusive access to the XArray for the
>> lifetime of the state object.
>>
>> The `XAS_RESTART` constant is also exposed through the bindings helper
>> to properly initialize the `xa_node` field.
>>
>> The struct and its constructor are marked with `#[expect(dead_code)]` as
>> there are no users yet. We will remove this annotation in a later patch.
>>
>> Signed-off-by: Andreas Hindborg <a.hindborg@kernel.org>
>> ---
>>  rust/bindings/bindings_helper.h |  1 +
>>  rust/kernel/xarray.rs           | 41 ++++++++++++++++++++++++++++++++++++++++-
>>  2 files changed, 41 insertions(+), 1 deletion(-)
>>
>> diff --git a/rust/bindings/bindings_helper.h b/rust/bindings/bindings_helper.h
>> index 2e43c66635a2c..86bca946faff0 100644
>> --- a/rust/bindings/bindings_helper.h
>> +++ b/rust/bindings/bindings_helper.h
>> @@ -109,6 +109,7 @@ const xa_mark_t RUST_CONST_HELPER_XA_PRESENT = XA_PRESENT;
>>  const gfp_t RUST_CONST_HELPER_XA_FLAGS_ALLOC = XA_FLAGS_ALLOC;
>>  const gfp_t RUST_CONST_HELPER_XA_FLAGS_ALLOC1 = XA_FLAGS_ALLOC1;
>>  const vm_flags_t RUST_CONST_HELPER_VM_MERGEABLE = VM_MERGEABLE;
>> +const size_t RUST_CONST_HELPER_XAS_RESTART = (size_t)XAS_RESTART;
>
> Can we avoid casting it here? It's quite distant from the site of use
> which makes it difficult to understand why it's necessary. It also
> suffers from C casting rules being quite lax, which means a change in
> the type on the C side would likely go unnoticed.

Yes, we should probably use `*mut bindings::xa_node` for this type, then
the casts at the use side in Rust will go away as well. I'll try that.

>
>>
>>  #if IS_ENABLED(CONFIG_ANDROID_BINDER_IPC_RUST)
>>  #include "../../drivers/android/binder/rust_binder.h"
>> diff --git a/rust/kernel/xarray.rs b/rust/kernel/xarray.rs
>> index ede48b5e1dba3..f6e610b059625 100644
>> --- a/rust/kernel/xarray.rs
>> +++ b/rust/kernel/xarray.rs
>> @@ -8,7 +8,10 @@
>>      iter,
>>      marker::PhantomData,
>>      pin::Pin,
>> -    ptr::NonNull, //
>> +    ptr::{
>> +        null_mut,
>> +        NonNull, //
>> +    },
>>  };
>>  use kernel::{
>>      alloc,
>> @@ -319,6 +322,42 @@ pub fn store(
>>      }
>>  }
>>
>> +/// Internal state for XArray iteration and entry operations.
>> +///
>> +/// # Invariants
>> +///
>> +/// - `state` is always a valid `bindings::xa_state`.
>> +#[expect(dead_code)]
>> +pub(crate) struct XArrayState<'a, 'b, T: ForeignOwnable> {
>> +    /// Holds a reference to the lock guard to ensure the lock is not dropped
>> +    /// while `Self` is live.
>> +    _access: &'b Guard<'a, T>,
>
> Can this comment explain why this isn't a PhantomData? IIUC we just
> need the lifetime, rather than the actual guard.

This can be a PhantomData. The use of the reference is leftover from
earlier prototypes. I'll change it.

>
>> +    state: bindings::xa_state,
>> +}
>> +
>> +impl<'a, 'b, T: ForeignOwnable> XArrayState<'a, 'b, T> {
>> +    #[expect(dead_code)]
>> +    fn new(access: &'b Guard<'a, T>, index: usize) -> Self {
>> +        let ptr = access.xa.xa.get();
>> +        // INVARIANT: We initialize `self.state` to a valid value below.
>> +        Self {
>> +            _access: access,
>> +            state: bindings::xa_state {
>> +                xa: ptr,
>> +                xa_index: index,
>> +                xa_shift: 0,
>> +                xa_sibs: 0,
>> +                xa_offset: 0,
>> +                xa_pad: 0,
>> +                xa_node: bindings::XAS_RESTART as *mut bindings::xa_node,
>
> Odd that we're casting it back to a pointer here.

Yes, full cycle. I'll try to use the pointer type for the constant.

Thanks for the comments.

Best regards,
Andreas Hindborg