From: Zhi Wang <zhiw@nvidia.com>
To: <rust-for-linux@vger.kernel.org>, <linux-pci@vger.kernel.org>,
<linux-kernel@vger.kernel.org>
Cc: <dakr@kernel.org>, <aliceryhl@google.com>, <bhelgaas@google.com>,
<kwilczynski@kernel.org>, <ojeda@kernel.org>,
<alex.gaynor@gmail.com>, <boqun.feng@gmail.com>,
<gary@garyguo.net>, <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>,
Zhi Wang <zhiw@nvidia.com>
Subject: [PATCH v10 3/5] rust: io: factor out MMIO read/write macros
Date: Mon, 19 Jan 2026 22:22:45 +0200 [thread overview]
Message-ID: <20260119202250.870588-4-zhiw@nvidia.com> (raw)
In-Reply-To: <20260119202250.870588-1-zhiw@nvidia.com>
Refactor the existing MMIO accessors to use common call macros
instead of inlining the bindings calls in each `define_{read,write}!`
expansion.
This factoring separates the common offset/bounds checks from the
low-level call pattern, making it easier to add additional I/O accessor
families.
No functional change intended.
Cc: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Zhi Wang <zhiw@nvidia.com>
---
rust/kernel/io.rs | 147 ++++++++++++++++++++++++++++++++--------------
1 file changed, 102 insertions(+), 45 deletions(-)
diff --git a/rust/kernel/io.rs b/rust/kernel/io.rs
index fa1a81ba656b..45a6120c3af0 100644
--- a/rust/kernel/io.rs
+++ b/rust/kernel/io.rs
@@ -136,8 +136,65 @@ pub fn maxsize(&self) -> usize {
#[repr(transparent)]
pub struct Mmio<const SIZE: usize = 0>(MmioRaw<SIZE>);
+/// Internal helper macros used to invoke C MMIO read functions.
+///
+/// This macro is intended to be used by higher-level MMIO access macros (define_read) and provides
+/// a unified expansion for infallible vs. fallible read semantics. It emits a direct call into the
+/// corresponding C helper and performs the required cast to the Rust return type.
+///
+/// # Parameters
+///
+/// * `$c_fn` – The C function performing the MMIO write.
+/// * `$self` – The I/O backend object.
+/// * `$ty` – The type of the value to be read.
+/// * `$addr` – The MMIO address to read.
+///
+/// This macro does not perform any validation; all invariants must be upheld by the higher-level
+/// abstraction invoking it.
+macro_rules! call_mmio_read {
+ (infallible, $c_fn:ident, $self:ident, $type:ty, $addr:expr) => {
+ // SAFETY: By the type invariant `addr` is a valid address for MMIO operations.
+ unsafe { bindings::$c_fn($addr as *const c_void) as $type }
+ };
+
+ (fallible, $c_fn:ident, $self:ident, $type:ty, $addr:expr) => {{
+ // SAFETY: By the type invariant `addr` is a valid address for MMIO operations.
+ Ok(unsafe { bindings::$c_fn($addr as *const c_void) as $type })
+ }};
+}
+
+/// Internal helper macros used to invoke C MMIO write functions.
+///
+/// This macro is intended to be used by higher-level MMIO access macros (define_write) and provides
+/// a unified expansion for infallible vs. fallible read semantics. It emits a direct call into the
+/// corresponding C helper and performs the required cast to the Rust return type.
+///
+/// # Parameters
+///
+/// * `$c_fn` – The C function performing the MMIO write.
+/// * `$self` – The I/O backend object.
+/// * `$ty` – The type of the written value.
+/// * `$addr` – The MMIO address to write.
+/// * `$value` – The value to write.
+///
+/// This macro does not perform any validation; all invariants must be upheld by the higher-level
+/// abstraction invoking it.
+macro_rules! call_mmio_write {
+ (infallible, $c_fn:ident, $self:ident, $ty:ty, $addr:expr, $value:expr) => {
+ // SAFETY: By the type invariant `addr` is a valid address for MMIO operations.
+ unsafe { bindings::$c_fn($value, $addr as *mut c_void) }
+ };
+
+ (fallible, $c_fn:ident, $self:ident, $ty:ty, $addr:expr, $value:expr) => {{
+ // SAFETY: By the type invariant `addr` is a valid address for MMIO operations.
+ unsafe { bindings::$c_fn($value, $addr as *mut c_void) };
+ Ok(())
+ }};
+}
+
macro_rules! define_read {
- (infallible, $(#[$attr:meta])* $vis:vis $name:ident, $c_fn:ident -> $type_name:ty) => {
+ (infallible, $(#[$attr:meta])* $vis:vis $name:ident, $call_macro:ident($c_fn:ident) ->
+ $type_name:ty) => {
/// Read IO data from a given offset known at compile time.
///
/// Bound checks are performed on compile time, hence if the offset is not known at compile
@@ -147,12 +204,13 @@ macro_rules! define_read {
$vis fn $name(&self, offset: usize) -> $type_name {
let addr = self.io_addr_assert::<$type_name>(offset);
- // SAFETY: By the type invariant `addr` is a valid address for MMIO operations.
- unsafe { bindings::$c_fn(addr as *const c_void) }
+ // SAFETY: By the type invariant `addr` is a valid address for IO operations.
+ $call_macro!(infallible, $c_fn, self, $type_name, addr)
}
};
- (fallible, $(#[$attr:meta])* $vis:vis $try_name:ident, $c_fn:ident -> $type_name:ty) => {
+ (fallible, $(#[$attr:meta])* $vis:vis $try_name:ident, $call_macro:ident($c_fn:ident) ->
+ $type_name:ty) => {
/// Read IO data from a given offset.
///
/// Bound checks are performed on runtime, it fails if the offset (plus the type size) is
@@ -161,15 +219,16 @@ macro_rules! define_read {
$vis fn $try_name(&self, offset: usize) -> Result<$type_name> {
let addr = self.io_addr::<$type_name>(offset)?;
- // SAFETY: By the type invariant `addr` is a valid address for MMIO operations.
- Ok(unsafe { bindings::$c_fn(addr as *const c_void) })
+ // SAFETY: By the type invariant `addr` is a valid address for IO operations.
+ $call_macro!(fallible, $c_fn, self, $type_name, addr)
}
};
}
pub(crate) use define_read;
macro_rules! define_write {
- (infallible, $(#[$attr:meta])* $vis:vis $name:ident, $c_fn:ident <- $type_name:ty) => {
+ (infallible, $(#[$attr:meta])* $vis:vis $name:ident, $call_macro:ident($c_fn:ident) <-
+ $type_name:ty) => {
/// Write IO data from a given offset known at compile time.
///
/// Bound checks are performed on compile time, hence if the offset is not known at compile
@@ -179,12 +238,12 @@ macro_rules! define_write {
$vis fn $name(&self, value: $type_name, offset: usize) {
let addr = self.io_addr_assert::<$type_name>(offset);
- // SAFETY: By the type invariant `addr` is a valid address for MMIO operations.
- unsafe { bindings::$c_fn(value, addr as *mut c_void) }
+ $call_macro!(infallible, $c_fn, self, $type_name, addr, value);
}
};
- (fallible, $(#[$attr:meta])* $vis:vis $try_name:ident, $c_fn:ident <- $type_name:ty) => {
+ (fallible, $(#[$attr:meta])* $vis:vis $try_name:ident, $call_macro:ident($c_fn:ident) <-
+ $type_name:ty) => {
/// Write IO data from a given offset.
///
/// Bound checks are performed on runtime, it fails if the offset (plus the type size) is
@@ -193,9 +252,7 @@ macro_rules! define_write {
$vis fn $try_name(&self, value: $type_name, offset: usize) -> Result {
let addr = self.io_addr::<$type_name>(offset)?;
- // SAFETY: By the type invariant `addr` is a valid address for MMIO operations.
- unsafe { bindings::$c_fn(value, addr as *mut c_void) };
- Ok(())
+ $call_macro!(fallible, $c_fn, self, $type_name, addr, value)
}
};
}
@@ -428,46 +485,46 @@ fn try_write64(&self, value: u64, offset: usize) -> Result
// MMIO regions support 8, 16, and 32-bit accesses.
impl<const SIZE: usize> IoCapable<u8> for Mmio<SIZE> {
- define_read!(infallible, read, readb -> u8);
- define_write!(infallible, write, writeb <- u8);
+ define_read!(infallible, read, call_mmio_read(readb) -> u8);
+ define_write!(infallible, write, call_mmio_write(writeb) <- u8);
}
impl<const SIZE: usize> IoCapable<u16> for Mmio<SIZE> {
- define_read!(infallible, read, readw -> u16);
- define_write!(infallible, write, writew <- u16);
+ define_read!(infallible, read, call_mmio_read(readw) -> u16);
+ define_write!(infallible, write, call_mmio_write(writew) <- u16);
}
impl<const SIZE: usize> IoCapable<u32> for Mmio<SIZE> {
- define_read!(infallible, read, readl -> u32);
- define_write!(infallible, write, writel <- u32);
+ define_read!(infallible, read, call_mmio_read(readl) -> u32);
+ define_write!(infallible, write, call_mmio_write(writel) <- u32);
}
// MMIO regions on 64-bit systems also support 64-bit accesses.
#[cfg(CONFIG_64BIT)]
impl<const SIZE: usize> IoCapable<u64> for Mmio<SIZE> {
- define_read!(infallible, read, readq -> u64);
- define_write!(infallible, write, writeq <- u64);
+ define_read!(infallible, read, call_mmio_read(readq) -> u64);
+ define_write!(infallible, write, call_mmio_write(writeq) <- u64);
}
impl<const SIZE: usize> IoTryCapable<u8> for Mmio<SIZE> {
- define_read!(fallible, try_read, readb -> u8);
- define_write!(fallible, try_write, writeb <- u8);
+ define_read!(fallible, try_read, call_mmio_read(readb) -> u8);
+ define_write!(fallible, try_write, call_mmio_write(writeb) <- u8);
}
impl<const SIZE: usize> IoTryCapable<u16> for Mmio<SIZE> {
- define_read!(fallible, try_read, readw -> u16);
- define_write!(fallible, try_write, writew <- u16);
+ define_read!(fallible, try_read, call_mmio_read(readw) -> u16);
+ define_write!(fallible, try_write, call_mmio_write(writew) <- u16);
}
impl<const SIZE: usize> IoTryCapable<u32> for Mmio<SIZE> {
- define_read!(fallible, try_read, readl -> u32);
- define_write!(fallible, try_write, writel <- u32);
+ define_read!(fallible, try_read, call_mmio_read(readl) -> u32);
+ define_write!(fallible, try_write, call_mmio_write(writel) <- u32);
}
#[cfg(CONFIG_64BIT)]
impl<const SIZE: usize> IoTryCapable<u64> for Mmio<SIZE> {
- define_read!(fallible, try_read, readq -> u64);
- define_write!(fallible, try_write, writeq <- u64);
+ define_read!(fallible, try_read, call_mmio_read(readq) -> u64);
+ define_write!(fallible, try_write, call_mmio_write(writeq) <- u64);
}
impl<const SIZE: usize> Io for Mmio<SIZE> {
@@ -498,43 +555,43 @@ pub unsafe fn from_raw(raw: &MmioRaw<SIZE>) -> &Self {
unsafe { &*core::ptr::from_ref(raw).cast() }
}
- define_read!(infallible, pub read8_relaxed, readb_relaxed -> u8);
- define_read!(infallible, pub read16_relaxed, readw_relaxed -> u16);
- define_read!(infallible, pub read32_relaxed, readl_relaxed -> u32);
+ define_read!(infallible, pub read8_relaxed, call_mmio_read(readb_relaxed) -> u8);
+ define_read!(infallible, pub read16_relaxed, call_mmio_read(readw_relaxed) -> u16);
+ define_read!(infallible, pub read32_relaxed, call_mmio_read(readl_relaxed) -> u32);
define_read!(
infallible,
#[cfg(CONFIG_64BIT)]
pub read64_relaxed,
- readq_relaxed -> u64
+ call_mmio_read(readq_relaxed) -> u64
);
- define_read!(fallible, pub try_read8_relaxed, readb_relaxed -> u8);
- define_read!(fallible, pub try_read16_relaxed, readw_relaxed -> u16);
- define_read!(fallible, pub try_read32_relaxed, readl_relaxed -> u32);
+ define_read!(fallible, pub try_read8_relaxed, call_mmio_read(readb_relaxed) -> u8);
+ define_read!(fallible, pub try_read16_relaxed, call_mmio_read(readw_relaxed) -> u16);
+ define_read!(fallible, pub try_read32_relaxed, call_mmio_read(readl_relaxed) -> u32);
define_read!(
fallible,
#[cfg(CONFIG_64BIT)]
pub try_read64_relaxed,
- readq_relaxed -> u64
+ call_mmio_read(readq_relaxed) -> u64
);
- define_write!(infallible, pub write8_relaxed, writeb_relaxed <- u8);
- define_write!(infallible, pub write16_relaxed, writew_relaxed <- u16);
- define_write!(infallible, pub write32_relaxed, writel_relaxed <- u32);
+ define_write!(infallible, pub write8_relaxed, call_mmio_write(writeb_relaxed) <- u8);
+ define_write!(infallible, pub write16_relaxed, call_mmio_write(writew_relaxed) <- u16);
+ define_write!(infallible, pub write32_relaxed, call_mmio_write(writel_relaxed) <- u32);
define_write!(
infallible,
#[cfg(CONFIG_64BIT)]
pub write64_relaxed,
- writeq_relaxed <- u64
+ call_mmio_write(writeq_relaxed) <- u64
);
- define_write!(fallible, pub try_write8_relaxed, writeb_relaxed <- u8);
- define_write!(fallible, pub try_write16_relaxed, writew_relaxed <- u16);
- define_write!(fallible, pub try_write32_relaxed, writel_relaxed <- u32);
+ define_write!(fallible, pub try_write8_relaxed, call_mmio_write(writeb_relaxed) <- u8);
+ define_write!(fallible, pub try_write16_relaxed, call_mmio_write(writew_relaxed) <- u16);
+ define_write!(fallible, pub try_write32_relaxed, call_mmio_write(writel_relaxed) <- u32);
define_write!(
fallible,
#[cfg(CONFIG_64BIT)]
pub try_write64_relaxed,
- writeq_relaxed <- u64
+ call_mmio_write(writeq_relaxed) <- u64
);
}
--
2.51.0
next prev parent reply other threads:[~2026-01-19 20:23 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-01-19 20:22 [PATCH v10 0/5] rust: pci: add config space read/write support Zhi Wang
2026-01-19 20:22 ` [PATCH v10 1/5] rust: devres: style for imports Zhi Wang
2026-01-20 15:26 ` Gary Guo
2026-01-20 16:11 ` Miguel Ojeda
2026-01-21 9:29 ` Zhi Wang
2026-01-22 1:55 ` Miguel Ojeda
2026-01-22 2:24 ` John Hubbard
2026-01-22 11:43 ` Danilo Krummrich
2026-01-19 20:22 ` [PATCH v10 2/5] rust: io: separate generic I/O helpers from MMIO implementation Zhi Wang
2026-01-20 8:04 ` Alice Ryhl
2026-01-20 10:12 ` Danilo Krummrich
2026-01-20 11:26 ` Alice Ryhl
2026-01-20 12:38 ` Alice Ryhl
2026-01-20 15:59 ` Gary Guo
2026-01-20 16:09 ` Danilo Krummrich
2026-01-20 17:26 ` Zhi Wang
2026-01-20 17:29 ` Danilo Krummrich
2026-01-20 17:35 ` Gary Guo
2026-01-20 13:37 ` kernel test robot
2026-01-20 15:54 ` Gary Guo
2026-01-19 20:22 ` Zhi Wang [this message]
2026-01-19 20:22 ` [PATCH v10 4/5] rust: pci: add config space read/write support Zhi Wang
2026-01-19 20:22 ` [PATCH v10 5/5] sample: rust: pci: add tests for config space routines Zhi Wang
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=20260119202250.870588-4-zhiw@nvidia.com \
--to=zhiw@nvidia.com \
--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=gary@garyguo.net \
--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=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