Message-Id: <20260205-rust-pci-sriov-v2-9-ef9400c7767b@redhat.com>
Date: Thu, 05 Feb 2026 15:59:56 -0500
From: Peter Colberg <pcolberg@...hat.com>
To: Danilo Krummrich <dakr@...nel.org>, Bjorn Helgaas <bhelgaas@...gle.com>, 
 Krzysztof Wilczyński <kwilczynski@...nel.org>, 
 Miguel Ojeda <ojeda@...nel.org>, Alex Gaynor <alex.gaynor@...il.com>, 
 Boqun Feng <boqun.feng@...il.com>, Gary Guo <gary@...yguo.net>, 
 Björn Roy Baron <bjorn3_gh@...tonmail.com>, 
 Benno Lossin <lossin@...nel.org>, Andreas Hindborg <a.hindborg@...nel.org>, 
 Alice Ryhl <aliceryhl@...gle.com>, Trevor Gross <tmgross@...ch.edu>, 
 Abdiel Janulgue <abdiel.janulgue@...il.com>, 
 Daniel Almeida <daniel.almeida@...labora.com>, 
 Robin Murphy <robin.murphy@....com>, 
 Greg Kroah-Hartman <gregkh@...uxfoundation.org>, 
 Dave Ertman <david.m.ertman@...el.com>, Ira Weiny <ira.weiny@...el.com>, 
 Leon Romanovsky <leon@...nel.org>, David Airlie <airlied@...il.com>, 
 Simona Vetter <simona@...ll.ch>, Jonathan Corbet <corbet@....net>, 
 Xu Yilun <yilun.xu@...el.com>, Tom Rix <trix@...hat.com>, 
 Moritz Fischer <mdf@...nel.org>, "Rafael J. Wysocki" <rafael@...nel.org>
Cc: linux-pci@...r.kernel.org, rust-for-linux@...r.kernel.org, 
 linux-kernel@...r.kernel.org, Alexandre Courbot <acourbot@...dia.com>, 
 Alistair Popple <apopple@...dia.com>, 
 Joel Fernandes <joelagnelf@...dia.com>, John Hubbard <jhubbard@...dia.com>, 
 Zhi Wang <zhiw@...dia.com>, nouveau@...ts.freedesktop.org, 
 dri-devel@...ts.freedesktop.org, linux-doc@...r.kernel.org, 
 linux-fpga@...r.kernel.org, driver-core@...ts.linux.dev, 
 Peter Colberg <pcolberg@...hat.com>, Jason Gunthorpe <jgg@...pe.ca>
Subject: [PATCH v2 09/10] rust: pci: add physfn(), to return PF device for
 VF device

Add a method to return the Physical Function (PF) device for a Virtual
Function (VF) device in the bound device context.

Unlike for a PCI driver written in C, guarantee that when a VF device is
bound to a driver, the underlying PF device is bound to a driver, too,
by always setting the flag managed_sriov in the pci_driver structure.

In case SR-IOV has been enabled by a C driver that has not set the flag
managed_sriov in pci_driver, return an error from physfn().

This change depends on commit a995fe1a3aa7 ("rust: driver: drop device
private data post unbind") to also uphold the safety guarantee in case
a (broken) PF driver re-enables SR-IOV in its unbind() callback. That
commit extends the lifetime of the device private data beyond the
remove_callback() wrapper. In particular, that commit ensures that the
device private data for the PF device is still alive until after the
function pci_iov_remove() is called and forcibly re-disables SR_IOV,
which means the data can be safely accessed by VF drivers until then.

Suggested-by: Danilo Krummrich <dakr@...nel.org>
Signed-off-by: Peter Colberg <pcolberg@...hat.com>
---
Changes in v2:
- Uphold safety guarantee when PF driver is written in C.
- Let physfn() return error if driver flag managed_sriov is unset.
---
 rust/kernel/pci.rs | 53 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 53 insertions(+)

diff --git a/rust/kernel/pci.rs b/rust/kernel/pci.rs
index 581930d0afe98ccc29d729e4d9aab75b4144e46c..3b11f73a9f2b69a02fe003b8feadd61864adc8c0 100644
--- a/rust/kernel/pci.rs
+++ b/rust/kernel/pci.rs
@@ -525,6 +525,59 @@ pub fn pci_class(&self) -> Class {
     }
 }
 
+impl Device<device::Bound> {
+    /// Returns the Physical Function (PF) device for a Virtual Function (VF) device.
+    ///
+    /// # Examples
+    ///
+    /// The following example illustrates how to obtain the private driver data of the PF device,
+    /// where `vf_pdev` is the VF device of reference type `&Device<Core>` or `&Device<Bound>`.
+    ///
+    /// ```
+    /// # use kernel::{device::Core, pci};
+    /// /// A PCI driver that binds to both the PF and its VF devices.
+    /// struct MyDriver;
+    ///
+    /// impl MyDriver {
+    ///     fn connect(vf_pdev: &pci::Device<Core>) -> Result {
+    ///         let pf_pdev = vf_pdev.physfn()?;
+    ///         let pf_drvdata = pf_pdev.as_ref().drvdata::<Self>()?;
+    ///         Ok(())
+    ///     }
+    /// }
+    /// ```
+    #[cfg(CONFIG_PCI_IOV)]
+    pub fn physfn(&self) -> Result<&Device<device::Bound>> {
+        if !self.is_virtfn() {
+            return Err(EINVAL);
+        }
+
+        // SAFETY: `self.as_raw()` returns a valid pointer to a `struct pci_dev`.
+        // `physfn` is a valid pointer to a `struct pci_dev` since `is_virtfn()` is `true`.
+        let pf_dev = unsafe { (*self.as_raw()).__bindgen_anon_1.physfn };
+
+        // SAFETY: `pf_dev` is a valid pointer to a `struct pci_dev`.
+        // `driver` is either NULL or a valid pointer to a `struct pci_driver`.
+        let pf_drv = unsafe { (*pf_dev).driver };
+        if pf_drv.is_null() {
+            return Err(EINVAL);
+        }
+
+        // SAFETY: `pf_drv` is a valid pointer to a `struct pci_driver`.
+        if !unsafe { (*pf_drv).managed_sriov } {
+            return Err(EINVAL);
+        }
+
+        // SAFETY: `physfn` may be cast to a `Device<device::Bound>` since the
+        // driver flag `managed_sriov` forces SR-IOV to be disabled when the
+        // PF driver is unbound, i.e., all VF devices are destroyed. This
+        // guarantees that the underlying PF device is bound to a driver
+        // when the VF device is bound to a driver, which is the case since
+        // `Device::physfn()` requires a `&Device<Bound>` reference.
+        Ok(unsafe { &*pf_dev.cast() })
+    }
+}
+
 impl Device<device::Core> {
     /// Enable memory resources for this device.
     pub fn enable_device_mem(&self) -> Result {

-- 
2.52.0

Powered by blists - more mailing lists