Rust for Linux List
 help / color / mirror / Atom feed
From: Mike Lothian <mike@fireburn.co.uk>
To: linux-usb@vger.kernel.org
Cc: "Mike Lothian" <mike@fireburn.co.uk>,
	"Miguel Ojeda" <ojeda@kernel.org>,
	"Boqun Feng" <boqun@kernel.org>, "Gary Guo" <gary@garyguo.net>,
	"Björn Roy Baron" <bjorn3_gh@protonmail.com>,
	"Benno Lossin" <lossin@kernel.org>,
	"Andreas Hindborg" <a.hindborg@kernel.org>,
	"Alice Ryhl" <aliceryhl@google.com>,
	"Trevor Gross" <tmgross@umich.edu>,
	"Danilo Krummrich" <dakr@kernel.org>,
	"Daniel Almeida" <daniel.almeida@collabora.com>,
	"Tamir Duberstein" <tamird@kernel.org>,
	"Alexandre Courbot" <acourbot@nvidia.com>,
	"Onur Özkan" <work@onurozkan.dev>,
	"Greg Kroah-Hartman" <gregkh@linuxfoundation.org>,
	"Colin Braun" <colinbrauncl@gmail.com>,
	rust-for-linux@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: [PATCH v3 3/5] rust: usb: expose device descriptor fields and queue readiness
Date: Wed, 26 Aug 2026 17:30:39 +0100	[thread overview]
Message-ID: <20260826163101.4168-4-mike@fireburn.co.uk> (raw)
In-Reply-To: <20260826163101.4168-1-mike@fireburn.co.uk>

A driver that identifies hardware before it decides to drive it needs the
device descriptor: idVendor, idProduct, bcdDevice, bcdUSB, the enumerated
speed, and the cached iManufacturer, iProduct and iSerialNumber strings.
bcdDevice in particular is the only revision a driver can read without
already speaking the device's own protocol.

Add can_send_n() alongside them, which reports whether the next count
queue slots can be submitted without waiting and reaps completed slots on
the way. A protocol that must not block halfway through a multi-URB record
uses it to defer the whole record and service its control plane first.

Assisted-by: Claude:claude-opus-5
Signed-off-by: Mike Lothian <mike@fireburn.co.uk>
---
 rust/kernel/usb.rs | 96 ++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 96 insertions(+)

diff --git a/rust/kernel/usb.rs b/rust/kernel/usb.rs
index ad40c814616a..782bae53584d 100644
--- a/rust/kernel/usb.rs
+++ b/rust/kernel/usb.rs
@@ -1430,6 +1430,7 @@ fn is_active(&self) -> bool {
         matches!(self.urb, Some(QueueUrb::Active(_)))
     }
 
+    #[inline]
     fn wait(&self, timeout: Delta) -> bool {
         let millis = timeout.as_millis();
         let millis = if millis <= 0 {
@@ -1441,6 +1442,7 @@ fn wait(&self, timeout: Delta) -> bool {
             .wait_for_completion_timeout(crate::time::msecs_to_jiffies(millis))
     }
 
+    #[inline]
     fn finish(&mut self) -> Result<(i32, usize)> {
         let state = self.urb.take().ok_or(EIO)?;
         let active = match state {
@@ -1700,6 +1702,34 @@ fn reap(&mut self, i: usize, timeout: Delta) -> Result<bool> {
         Ok(true)
     }
 
+    /// Reports whether the next `count` queue slots can be submitted without waiting.
+    ///
+    /// Completed slots are reaped and any transfer error is returned. This is useful when a
+    /// higher-level protocol must not block halfway through a multi-URB record while waiting for
+    /// endpoint progress; callers can defer the whole record and service its control plane first.
+    #[inline]
+    pub fn can_send_n(&mut self, io: &Io<'_>, count: usize) -> Result<bool> {
+        self.inner.check(io)?;
+        if count > self.slots.len() {
+            return Ok(false);
+        }
+        for off in 0..count {
+            let i = (self.cursor + off) % self.slots.len();
+            if self.slots[i].is_active() {
+                if !self.slots[i].wait(Delta::ZERO) {
+                    return Ok(false);
+                }
+                // `wait_for_completion_timeout()` consumes the completion signal. Reap the URB
+                // now rather than leaving `send()` to wait for the signal a second time.
+                let (status, _) = self.slots[i].finish()?;
+                if status != 0 {
+                    return Err(Error::from_errno(status));
+                }
+            }
+        }
+        Ok(true)
+    }
+
     /// Submits `data` as a bulk OUT transfer without waiting for it to complete.
     ///
     /// If the slot about to be reused still has a transfer outstanding, this blocks up to
@@ -2626,6 +2656,72 @@ fn inner(&self) -> &bindings::usb_device {
     fn devnum(&self) -> u32 {
         self.inner().devnum as u32
     }
+
+    /// Returns the `idVendor` of the device descriptor.
+    pub fn vendor_id(&self) -> u16 {
+        self.inner().descriptor.idVendor
+    }
+
+    /// Returns the `idProduct` of the device descriptor.
+    pub fn product_id(&self) -> u16 {
+        self.inner().descriptor.idProduct
+    }
+
+    /// Returns the `bcdDevice` of the device descriptor.
+    ///
+    /// Vendors conventionally use this as the device revision, and it is the only version a driver
+    /// can read without speaking the device's own protocol.
+    pub fn bcd_device(&self) -> u16 {
+        self.inner().descriptor.bcdDevice
+    }
+
+    /// Returns the `bcdUSB` of the device descriptor.
+    pub fn bcd_usb(&self) -> u16 {
+        self.inner().descriptor.bcdUSB
+    }
+
+    /// Returns the enumerated bus speed as a human-readable string.
+    pub fn speed_str(&self) -> &'static str {
+        match self.inner().speed {
+            bindings::usb_device_speed_USB_SPEED_LOW => "low (1.5 Mbps)",
+            bindings::usb_device_speed_USB_SPEED_FULL => "full (12 Mbps)",
+            bindings::usb_device_speed_USB_SPEED_HIGH => "high (480 Mbps)",
+            bindings::usb_device_speed_USB_SPEED_WIRELESS => "wireless",
+            bindings::usb_device_speed_USB_SPEED_SUPER => "super (5 Gbps)",
+            bindings::usb_device_speed_USB_SPEED_SUPER_PLUS => "super-plus (10+ Gbps)",
+            _ => "unknown",
+        }
+    }
+
+    /// Returns the device's `iManufacturer` string, if the core cached one.
+    pub fn manufacturer(&self) -> Option<&CStr> {
+        // SAFETY: `manufacturer` is either null or a NUL-terminated string owned by the USB core
+        // for as long as the device exists, which outlives the borrow of `self`.
+        unsafe { Self::opt_cstr(self.inner().manufacturer) }
+    }
+
+    /// Returns the device's `iProduct` string, if the core cached one.
+    pub fn product(&self) -> Option<&CStr> {
+        // SAFETY: As for `manufacturer`.
+        unsafe { Self::opt_cstr(self.inner().product) }
+    }
+
+    /// Returns the device's `iSerialNumber` string, if the core cached one.
+    pub fn serial(&self) -> Option<&CStr> {
+        // SAFETY: As for `manufacturer`.
+        unsafe { Self::opt_cstr(self.inner().serial) }
+    }
+
+    /// # Safety
+    ///
+    /// `p` must be null or point to a NUL-terminated string that outlives `'a`.
+    unsafe fn opt_cstr<'a>(p: *mut crate::ffi::c_char) -> Option<&'a CStr> {
+        if p.is_null() {
+            return None;
+        }
+        // SAFETY: The caller guarantees `p` is a NUL-terminated string valid for `'a`.
+        Some(unsafe { CStr::from_char_ptr(p) })
+    }
 }
 
 impl Device<device::Bound> {

  parent reply	other threads:[~2026-08-26 16:31 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-08-26 16:30 [PATCH v3 0/5] rust: usb: host-side abstractions for a bulk-endpoint driver Mike Lothian
2026-08-26 16:30 ` [PATCH v3 1/5] rust: usb: add revocable typed interface I/O Mike Lothian
2026-08-26 18:59   ` Danilo Krummrich
2026-08-26 16:30 ` [PATCH v3 2/5] rust: usb: add reusable URBs and persistent bulk queues Mike Lothian
2026-08-26 16:30 ` Mike Lothian [this message]
2026-08-26 16:30 ` [PATCH v3 4/5] rust: usb: add a vendor-and-interface-info device id constructor Mike Lothian
2026-08-26 16:30 ` [PATCH v3 5/5] rust: usb: let a driver keep its interface usable while unbinding Mike Lothian
2026-08-26 18:25 ` [PATCH v3 0/5] rust: usb: host-side abstractions for a bulk-endpoint driver 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=20260826163101.4168-4-mike@fireburn.co.uk \
    --to=mike@fireburn.co.uk \
    --cc=a.hindborg@kernel.org \
    --cc=acourbot@nvidia.com \
    --cc=aliceryhl@google.com \
    --cc=bjorn3_gh@protonmail.com \
    --cc=boqun@kernel.org \
    --cc=colinbrauncl@gmail.com \
    --cc=dakr@kernel.org \
    --cc=daniel.almeida@collabora.com \
    --cc=gary@garyguo.net \
    --cc=gregkh@linuxfoundation.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-usb@vger.kernel.org \
    --cc=lossin@kernel.org \
    --cc=ojeda@kernel.org \
    --cc=rust-for-linux@vger.kernel.org \
    --cc=tamird@kernel.org \
    --cc=tmgross@umich.edu \
    --cc=work@onurozkan.dev \
    /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