public inbox for rust-for-linux@vger.kernel.org
 help / color / mirror / Atom feed
From: "Gary Guo" <gary@garyguo.net>
To: "Danilo Krummrich" <dakr@kernel.org>, "Gary Guo" <gary@garyguo.net>
Cc: "Zhi Wang" <zhiw@nvidia.com>, <rust-for-linux@vger.kernel.org>,
	<linux-pci@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
	<aliceryhl@google.com>, <bhelgaas@google.com>,
	<kwilczynski@kernel.org>, <ojeda@kernel.org>,
	<alex.gaynor@gmail.com>, <boqun.feng@gmail.com>,
	<bjorn3_gh@protonmail.com>, <lossin@kernel.org>,
	<a.hindborg@kernel.org>, <tmgross@umich.edu>,
	<markus.probst@posteo.de>, <helgaas@kernel.org>,
	<cjia@nvidia.com>, <smitra@nvidia.com>, <ankita@nvidia.com>,
	<aniketa@nvidia.com>, <kwankhede@nvidia.com>,
	<targupta@nvidia.com>, <acourbot@nvidia.com>,
	<joelagnelf@nvidia.com>, <jhubbard@nvidia.com>,
	<zhiwang@kernel.org>, <daniel.almeida@collabora.com>
Subject: Re: [PATCH v12 4/5] rust: pci: add config space read/write support
Date: Thu, 22 Jan 2026 14:26:35 +0000	[thread overview]
Message-ID: <DFV6RV4QOS48.25MHV8QDDCHNV@garyguo.net> (raw)
In-Reply-To: <DFV4IJDQC2J6.1Q91JOAL6CJSG@kernel.org>

On Thu Jan 22, 2026 at 12:40 PM GMT, Danilo Krummrich wrote:
> On Thu Jan 22, 2026 at 12:59 PM CET, Gary Guo wrote:
>> On Wed Jan 21, 2026 at 8:22 PM GMT, Zhi Wang wrote:
>>> +    /// Returns the size of configuration space in bytes.
>>> +    fn cfg_size(&self) -> Result<usize> {
>>> +        // SAFETY: `self.as_raw` is a valid pointer to a `struct pci_dev`.
>>> +        let size = unsafe { (*self.as_raw()).cfg_size };
>>> +        match size {
>>> +            256 | 4096 => Ok(size as usize),
>>> +            _ => {
>>> +                debug_assert!(false);
>>> +                Err(EINVAL)
>>> +            }
>>> +        }
>>> +    }
>>
>> This method is only invoked from maxsize, which turns error into `0`. Do apart
>> from the debug assertion, the error code is pointless. I think this function
>> should just return `usize` as it's specified in the device (we should trust the
>> C side that the value is sensible).
>
> That seems reasonable, but I also think we should keep the enum ConfigSpaceSize
> we had before and call the new trait ConfigSpaceKind instead, such that this
> method becomes:
>
> 	fn cfg_size(&self) -> ConfigSpaceSize;
>
>> The check, as Alex mentioned, need to be done when ConfigSpace is created in
>> the first place and is too late when you already hand out `Ok(ConfigSpace)`.
>
> We need the check for config_space_extended(), but not for config_space(), as it
> represents the minimum size, i.e. it's always valid.
>
> Here's a diff of what I think this should look like on top of this series.

The proposal looks good to me. Some comments below.

Reviewed-by: Gary Guo <gary@garyguo.net>

>
> (@Zhi: If we all agree on the diff and nothing else comes up you don't need to
> resend. :)
>
> diff --git a/rust/kernel/pci.rs b/rust/kernel/pci.rs
> index 9020959ce0c7..1d1a253e5d5d 100644
> --- a/rust/kernel/pci.rs
> +++ b/rust/kernel/pci.rs
> @@ -42,6 +42,7 @@
>  };
>  pub use self::io::{
>      Bar,
> +    ConfigSpaceKind,
>      ConfigSpaceSize,
>      Extended,
>      Normal, //
> diff --git a/rust/kernel/pci/io.rs b/rust/kernel/pci/io.rs
> index 39df41d0eaab..5dbdfe516418 100644
> --- a/rust/kernel/pci/io.rs
> +++ b/rust/kernel/pci/io.rs
> @@ -24,6 +24,31 @@
>      ops::Deref, //
>  };
>
> +/// Represents the size of a PCI configuration space.
> +///
> +/// PCI devices can have either a *normal* (legacy) configuration space of 256 bytes,
> +/// or an *extended* configuration space of 4096 bytes as defined in the PCI Express
> +/// specification.
> +#[repr(usize)]
> +#[derive(PartialEq)]

When `PartialEq` is derived, I would also derive `Eq` unless there's no
reflexivity in comparison.

> +pub enum ConfigSpaceSize {
> +    /// 256-byte legacy PCI configuration space.
> +    Normal = 256,
> +
> +    /// 4096-byte PCIe extended configuration space.
> +    Extended = 4096,
> +}
> +
> +impl ConfigSpaceSize {
> +    /// Get the raw value of this enum.
> +    #[inline(always)]
> +    pub const fn into_raw(self) -> usize {
> +        // CAST: PCI configuration space size is at most 4096 bytes, so the value always fits
> +        // within `usize` without truncation or sign change.
> +        self as usize
> +    }
> +}
> +
>  /// Marker type for normal (256-byte) PCI configuration space.
>  pub struct Normal;
>
> @@ -34,16 +59,16 @@
>  ///
>  /// This trait is implemented by [`Normal`] and [`Extended`] to provide
>  /// compile-time knowledge of the configuration space size.
> -pub trait ConfigSpaceSize {
> +pub trait ConfigSpaceKind {
>      /// The size of this configuration space in bytes.
>      const SIZE: usize;
>  }
>
> -impl ConfigSpaceSize for Normal {
> +impl ConfigSpaceKind for Normal {
>      const SIZE: usize = 256;
>  }
>
> -impl ConfigSpaceSize for Extended {
> +impl ConfigSpaceKind for Extended {
>      const SIZE: usize = 4096;
>  }
>
> @@ -55,7 +80,7 @@ impl ConfigSpaceSize for Extended {
>  /// The generic parameter `S` indicates the maximum size of the configuration space.
>  /// Use [`Normal`] for 256-byte legacy configuration space or [`Extended`] for
>  /// 4096-byte PCIe extended configuration space (default).
> -pub struct ConfigSpace<'a, S: ConfigSpaceSize = Extended> {
> +pub struct ConfigSpace<'a, S: ConfigSpaceKind = Extended> {
>      pub(crate) pdev: &'a Device<device::Bound>,
>      _marker: PhantomData<S>,
>  }
> @@ -118,11 +143,11 @@ macro_rules! call_config_write {
>  }
>
>  // PCI configuration space supports 8, 16, and 32-bit accesses.
> -impl<'a, S: ConfigSpaceSize> IoCapable<u8> for ConfigSpace<'a, S> {}
> -impl<'a, S: ConfigSpaceSize> IoCapable<u16> for ConfigSpace<'a, S> {}
> -impl<'a, S: ConfigSpaceSize> IoCapable<u32> for ConfigSpace<'a, S> {}
> +impl<'a, S: ConfigSpaceKind> IoCapable<u8> for ConfigSpace<'a, S> {}
> +impl<'a, S: ConfigSpaceKind> IoCapable<u16> for ConfigSpace<'a, S> {}
> +impl<'a, S: ConfigSpaceKind> IoCapable<u32> for ConfigSpace<'a, S> {}
>
> -impl<'a, S: ConfigSpaceSize> Io for ConfigSpace<'a, S> {
> +impl<'a, S: ConfigSpaceKind> Io for ConfigSpace<'a, S> {
>      const MIN_SIZE: usize = S::SIZE;
>
>      /// Returns the base address of the I/O region. It is always 0 for configuration space.
> @@ -134,7 +159,7 @@ fn addr(&self) -> usize {
>      /// Returns the maximum size of the configuration space.
>      #[inline]
>      fn maxsize(&self) -> usize {
> -        self.pdev.cfg_size().map_or(0, |v| v)
> +        self.pdev.cfg_size().into_raw()
>      }
>
>      // PCI configuration space does not support fallible operations.
> @@ -150,7 +175,7 @@ fn maxsize(&self) -> usize {
>  }
>
>  /// Marker trait indicating ConfigSpace has a known size at compile time.
> -impl<'a, S: ConfigSpaceSize> IoKnownSize for ConfigSpace<'a, S> {}
> +impl<'a, S: ConfigSpaceKind> IoKnownSize for ConfigSpace<'a, S> {}
>
>  /// A PCI BAR to perform I/O-Operations on.
>  ///
> @@ -281,29 +306,35 @@ pub fn iomap_region<'a>(
>          self.iomap_region_sized::<0>(bar, name)
>      }
>
> -    /// Returns the size of configuration space in bytes.
> -    fn cfg_size(&self) -> Result<usize> {
> +    /// Returns the size of configuration space.
> +    fn cfg_size(&self) -> ConfigSpaceSize {

If you keep this `fn` instead of `pub fn`, then the `ConfigSpaceSize` type being
`pub` is not very useful as it cannot be invoked by the user.

>          // SAFETY: `self.as_raw` is a valid pointer to a `struct pci_dev`.
>          let size = unsafe { (*self.as_raw()).cfg_size };
>          match size {
> -            256 | 4096 => Ok(size as usize),
> +            256 => ConfigSpaceSize::Normal,
> +            4096 => ConfigSpaceSize::Extended,
>              _ => {
> -                debug_assert!(false);
> -                Err(EINVAL)
> +                // PANIC: The PCI subsystem only ever reports the configuration space size as either
> +                // `ConfigSpaceSize::Normal` or `ConfigSpaceSize::Extended`.
> +                unreachable!();
>              }
>          }
>      }
>
>      /// Return an initialized normal (256-byte) config space object.
> -    pub fn config_space<'a>(&'a self) -> Result<ConfigSpace<'a, Normal>> {
> -        Ok(ConfigSpace {
> +    pub fn config_space<'a>(&'a self) -> ConfigSpace<'a, Normal> {

Nice. Less failing path = good.

> +        ConfigSpace {
>              pdev: self,
>              _marker: PhantomData,
> -        })
> +        }
>      }
>
>      /// Return an initialized extended (4096-byte) config space object.
>      pub fn config_space_extended<'a>(&'a self) -> Result<ConfigSpace<'a, Extended>> {
> +        if self.cfg_size() != ConfigSpaceSize::Extended {
> +            return Err(EINVAL);
> +        }
> +
>          Ok(ConfigSpace {
>              pdev: self,
>              _marker: PhantomData,
> diff --git a/samples/rust/rust_driver_pci.rs b/samples/rust/rust_driver_pci.rs
> index 1bc5bd1a8df5..8eea79e858a2 100644
> --- a/samples/rust/rust_driver_pci.rs
> +++ b/samples/rust/rust_driver_pci.rs
> @@ -67,8 +67,8 @@ fn testdev(index: &TestIndex, bar: &Bar0) -> Result<u32> {
>          Ok(bar.read32(Regs::COUNT))
>      }
>
> -    fn config_space(pdev: &pci::Device<Bound>) -> Result {
> -        let config = pdev.config_space()?;
> +    fn config_space(pdev: &pci::Device<Bound>) {
> +        let config = pdev.config_space();
>
>          // TODO: use the register!() macro for defining PCI configuration space registers once it
>          // has been move out of nova-core.
> @@ -89,8 +89,6 @@ fn config_space(pdev: &pci::Device<Bound>) -> Result {
>              "pci-testdev config space read32 BAR 0: {:x}\n",
>              config.read32(0x10)
>          );
> -
> -        Ok(())
>      }
>  }
>
> @@ -123,7 +121,7 @@ fn probe(pdev: &pci::Device<Core>, info: &Self::IdInfo) -> impl PinInit<Self, Er
>                          "pci-testdev data-match count: {}\n",
>                          Self::testdev(info, bar)?
>                      );
> -                    Self::config_space(pdev)?;
> +                    Self::config_space(pdev);
>                  },
>                  pdev: pdev.into(),
>              }))


  parent reply	other threads:[~2026-01-22 14:26 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-01-21 20:22 [PATCH v12 0/5] rust: pci: add config space read/write support Zhi Wang
2026-01-21 20:22 ` [PATCH v12 1/5] rust: devres: style for imports Zhi Wang
2026-01-22  4:25   ` Alexandre Courbot
2026-01-21 20:22 ` [PATCH v12 2/5] rust: io: separate generic I/O helpers from MMIO implementation Zhi Wang
2026-01-22  4:28   ` Alexandre Courbot
2026-01-22 11:52   ` Gary Guo
2026-01-22 12:52   ` Danilo Krummrich
2026-01-23  9:13   ` Alice Ryhl
2026-01-21 20:22 ` [PATCH v12 3/5] rust: io: factor out MMIO read/write macros Zhi Wang
2026-01-22  4:30   ` Alexandre Courbot
2026-01-22 11:54     ` Gary Guo
2026-01-21 20:22 ` [PATCH v12 4/5] rust: pci: add config space read/write support Zhi Wang
2026-01-22  4:25   ` Alexandre Courbot
2026-01-22 11:59   ` Gary Guo
2026-01-22 12:40     ` Danilo Krummrich
2026-01-22 12:52       ` Zhi Wang
2026-01-22 14:26       ` Gary Guo [this message]
2026-01-21 20:22 ` [PATCH v12 5/5] sample: rust: pci: add tests for config space routines Zhi Wang
2026-01-22  4:31   ` Alexandre Courbot
2026-01-26  5:08     ` Alexandre Courbot
2026-01-26  9:05       ` Zhi Wang
2026-01-26 10:21         ` Zhi Wang
2026-01-26 10:35         ` Alexandre Courbot
2026-01-22 12:00   ` Gary Guo
2026-01-24  0:12 ` [PATCH v12 0/5] rust: pci: add config space read/write support Danilo Krummrich

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=DFV6RV4QOS48.25MHV8QDDCHNV@garyguo.net \
    --to=gary@garyguo.net \
    --cc=a.hindborg@kernel.org \
    --cc=acourbot@nvidia.com \
    --cc=alex.gaynor@gmail.com \
    --cc=aliceryhl@google.com \
    --cc=aniketa@nvidia.com \
    --cc=ankita@nvidia.com \
    --cc=bhelgaas@google.com \
    --cc=bjorn3_gh@protonmail.com \
    --cc=boqun.feng@gmail.com \
    --cc=cjia@nvidia.com \
    --cc=dakr@kernel.org \
    --cc=daniel.almeida@collabora.com \
    --cc=helgaas@kernel.org \
    --cc=jhubbard@nvidia.com \
    --cc=joelagnelf@nvidia.com \
    --cc=kwankhede@nvidia.com \
    --cc=kwilczynski@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=lossin@kernel.org \
    --cc=markus.probst@posteo.de \
    --cc=ojeda@kernel.org \
    --cc=rust-for-linux@vger.kernel.org \
    --cc=smitra@nvidia.com \
    --cc=targupta@nvidia.com \
    --cc=tmgross@umich.edu \
    --cc=zhiw@nvidia.com \
    --cc=zhiwang@kernel.org \
    /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