Re: [PATCH v4 7/9] rust: pci: add SR-IOV enable and disable tokens

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] rust: missing #[inline] on small abstraction functions
--

commit e7af204882b55dcbfacee7ce650e2197a9830e0f
Author: Zhi Wang <zhiw@nvidia.com>

rust: pci: add SR-IOV enable and disable tokens

This commit introduces SR-IOV enable and disable tokens for PCI device
drivers in Rust. It utilizes callback-scoped guards to ensure SR-IOV is
disabled on error and coordinates final VF teardown with registration
tokens.

> diff --git a/rust/kernel/pci/iov.rs b/rust/kernel/pci/iov.rs
> index c61462595141f..449f1e9b12fda 100644
> --- a/rust/kernel/pci/iov.rs
> +++ b/rust/kernel/pci/iov.rs

[ ... ]

> +pub struct SriovEnable<'a> {
> +    pdev: &'a Device<device::CoreInternal<'a>>,
> +    num_vfs: u16,
> +}
> +
> +impl<'a> SriovEnable<'a> {
> +    /// Returns the number of VFs requested for this callback.
> +    pub fn num_vfs(&self) -> u16 {
> +        self.num_vfs
> +    }

[Severity: Low]
Should this trivial accessor be annotated with #[inline]? Subsystem
guidelines suggest small functions in the kernel crate abstraction should be
explicitly annotated to prevent suboptimal code generation.

[ ... ]

> +/// Enabled VFs awaiting successful completion of `sriov_enable()`.
> +pub struct SriovEnabled<'a> {
> +    pdev: &'a Device<device::CoreInternal<'a>>,
> +    num_vfs: u16,
> +}
> +
> +impl SriovEnabled<'_> {
> +    #[expect(dead_code)]
> +    fn disarm(self) -> u16 {
> +        ManuallyDrop::new(self).num_vfs
> +    }

[Severity: Low]
Does this small wrapper manipulating object destruction need an #[inline]
annotation to ensure optimal inlining across crate boundaries?

[ ... ]

> +pub struct SriovDisable<'a> {
> +    pdev: &'a Device<device::CoreInternal<'a>>,
> +}
> +
> +impl SriovDisable<'_> {
> +    /// Disables all VFs and waits for their drivers to unbind.
> +    ///
> +    /// The PF resources used by VF drivers must remain available until this method returns.
> +    pub fn disable(self) {
> +        self.pdev.disable_sriov();
> +    }

[Severity: Low]
Could this small forwarding method benefit from an #[inline] annotation?
The Rust subsystem guidelines recommend #[inline] for functions forwarding
to other methods in abstractions.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261004120732.1045629-1-zhiw@nvidia.com?part=7