From: Rob Herring <robh@kernel.org>
To: Remo Senekowitsch <remo@buenzli.dev>
Cc: "Andy Shevchenko" <andriy.shevchenko@linux.intel.com>,
"Daniel Scally" <djrscally@gmail.com>,
"Heikki Krogerus" <heikki.krogerus@linux.intel.com>,
"Sakari Ailus" <sakari.ailus@linux.intel.com>,
"Dirk Behme" <dirk.behme@de.bosch.com>,
"Greg Kroah-Hartman" <gregkh@linuxfoundation.org>,
"Rafael J. Wysocki" <rafael@kernel.org>,
"Danilo Krummrich" <dakr@kernel.org>,
"Saravana Kannan" <saravanak@google.com>,
"Miguel Ojeda" <ojeda@kernel.org>,
"Alex Gaynor" <alex.gaynor@gmail.com>,
"Boqun Feng" <boqun.feng@gmail.com>,
"Gary Guo" <gary@garyguo.net>,
"Björn Roy Baron" <bjorn3_gh@protonmail.com>,
"Benno Lossin" <benno.lossin@proton.me>,
"Andreas Hindborg" <a.hindborg@kernel.org>,
"Alice Ryhl" <aliceryhl@google.com>,
"Trevor Gross" <tmgross@umich.edu>,
linux-kernel@vger.kernel.org, linux-acpi@vger.kernel.org,
devicetree@vger.kernel.org, rust-for-linux@vger.kernel.org
Subject: Re: [PATCH 09/10] rust: property: Add PropertyGuard
Date: Wed, 26 Mar 2025 17:25:29 -0500 [thread overview]
Message-ID: <20250326222529.GE2844851-robh@kernel.org> (raw)
In-Reply-To: <20250326171411.590681-10-remo@buenzli.dev>
On Wed, Mar 26, 2025 at 06:13:48PM +0100, Remo Senekowitsch wrote:
> Forcing users to specify whether a property is supposed to be required
> or not allows us to move error logging of missing required properties
> into core, preventing a lot of boilerplate in drivers.
We'll want to squash this one too.
>
> Signed-off-by: Remo Senekowitsch <remo@buenzli.dev>
> ---
> rust/kernel/property.rs | 146 ++++++++++++++++++++++++++++++++++------
> 1 file changed, 127 insertions(+), 19 deletions(-)
>
> diff --git a/rust/kernel/property.rs b/rust/kernel/property.rs
> index f1d0a33ba..e4f0e5f97 100644
> --- a/rust/kernel/property.rs
> +++ b/rust/kernel/property.rs
> @@ -42,7 +42,11 @@ pub fn property_match_string(&self, name: &CStr, match_str: &CStr) -> Result<usi
> }
>
> /// Returns firmware property `name` integer array values in a KVec
> - pub fn property_read_array_vec<T: Integer>(&self, name: &CStr, len: usize) -> Result<KVec<T>> {
> + pub fn property_read_array_vec<'fwnode, 'name, T: Integer>(
> + &'fwnode self,
> + name: &'name CStr,
> + len: usize,
> + ) -> Result<PropertyGuard<'fwnode, 'name, KVec<T>>> {
> self.fwnode().property_read_array_vec(name, len)
> }
>
> @@ -52,12 +56,18 @@ pub fn property_count_elem<T: Integer>(&self, name: &CStr) -> Result<usize> {
> }
>
> /// Returns firmware property `name` integer scalar value
> - pub fn property_read<T: Property>(&self, name: &CStr) -> Result<T> {
> + pub fn property_read<'fwnode, 'name, T: Property>(
> + &'fwnode self,
> + name: &'name CStr,
> + ) -> PropertyGuard<'fwnode, 'name, T> {
> self.fwnode().property_read(name)
> }
>
> /// Returns first matching named child node handle.
> - pub fn get_child_by_name(&self, name: &CStr) -> Option<ARef<FwNode>> {
> + pub fn get_child_by_name<'fwnode, 'name>(
> + &'fwnode self,
> + name: &'name CStr,
> + ) -> PropertyGuard<'fwnode, 'name, ARef<FwNode>> {
> self.fwnode().get_child_by_name(name)
> }
While child nodes can be required or optional, it's a bit less common to
be optional. There error message on missing required child nodes is
going to be a bit misleading saying 'property'. So maybe don't do the
PropertyGuard on child node handling?
>
> @@ -132,12 +142,16 @@ pub fn property_match_string(&self, name: &CStr, match_str: &CStr) -> Result<usi
> }
>
> /// Returns firmware property `name` integer array values in a KVec
> - pub fn property_read_array_vec<T: Integer>(&self, name: &CStr, len: usize) -> Result<KVec<T>> {
> + pub fn property_read_array_vec<'fwnode, 'name, T: Integer>(
> + &'fwnode self,
> + name: &'name CStr,
> + len: usize,
> + ) -> Result<PropertyGuard<'fwnode, 'name, KVec<T>>> {
> let mut val: KVec<T> = KVec::with_capacity(len, GFP_KERNEL)?;
>
> // SAFETY: `name` is non-null and null-terminated. `self.as_raw` is valid
> // because `self` is valid. `val.as_ptr` is valid because `val` is valid.
> - to_result(unsafe {
> + let err = unsafe {
> bindings::fwnode_property_read_int_array(
> self.as_raw(),
> name.as_char_ptr(),
> @@ -145,11 +159,19 @@ pub fn property_read_array_vec<T: Integer>(&self, name: &CStr, len: usize) -> Re
> val.as_ptr() as *mut c_void,
> len,
> )
> - })?;
> -
> - // SAFETY: fwnode_property_read_int_array() writes exactly `len` entries on success
> - unsafe { val.set_len(len) }
> - Ok(val)
> + };
> + let res = if err < 0 {
> + Err(Error::from_errno(err))
> + } else {
> + // SAFETY: fwnode_property_read_int_array() writes exactly `len` entries on success
> + unsafe { val.set_len(len) }
> + Ok(val)
> + };
> + Ok(PropertyGuard {
> + inner: res,
> + fwnode: self,
> + name,
> + })
> }
>
> /// Returns integer array length for firmware property `name`
> @@ -194,24 +216,42 @@ pub fn property_count_elem<T: Integer>(&self, name: &CStr) -> Result<usize> {
> /// # use crate::{device::Device, types::CString};
> /// fn examples(dev: &Device) -> Result {
> /// let fwnode = dev.fwnode();
> - /// let b: bool = fwnode.property_read("some-bool")?;
> - /// let s: CString = fwnode.property_read("some-str")?;
> + /// let b: bool = fwnode.property_read("some-bool").required()?;
> + /// if let Some(s) = fwnode.property_read::<CString>("some-str").optional() {
> + /// // ...
> + /// }
> /// }
> /// ```
> - pub fn property_read<T: Property>(&self, name: &CStr) -> Result<T> {
> - T::read(&self, name)
> + pub fn property_read<'fwnode, 'name, T: Property>(
> + &'fwnode self,
> + name: &'name CStr,
> + ) -> PropertyGuard<'fwnode, 'name, T> {
> + PropertyGuard {
> + inner: T::read(&self, name),
> + fwnode: self,
> + name,
> + }
> }
>
> /// Returns first matching named child node handle.
> - pub fn get_child_by_name(&self, name: &CStr) -> Option<ARef<Self>> {
> + pub fn get_child_by_name<'fwnode, 'name>(
> + &'fwnode self,
> + name: &'name CStr,
> + ) -> PropertyGuard<'fwnode, 'name, ARef<Self>> {
> // SAFETY: `self` and `name` are valid.
> let child =
> unsafe { bindings::fwnode_get_named_child_node(self.as_raw(), name.as_char_ptr()) };
> - if child.is_null() {
> - return None;
> + let res = if child.is_null() {
> + Err(ENOENT)
> + } else {
> + // SAFETY: `fwnode_get_named_child_node` returns a pointer with refcount incremented.
> + Ok(unsafe { Self::from_raw(child) })
> + };
> + PropertyGuard {
> + inner: res,
> + fwnode: self,
> + name,
> }
> - // SAFETY: `fwnode_get_named_child_node` returns a pointer with refcount incremented.
> - Some(unsafe { Self::from_raw(child) })
> }
>
> /// Returns an iterator over a node's children.
> @@ -365,3 +405,71 @@ pub enum NArgs<'a> {
> /// The known number of arguments.
> N(u32),
> }
> +
> +/// A helper for reading device properties.
> +///
> +/// Use [Self::required] if a missing property is considered a bug and
> +/// [Self::optional] otherwise.
> +///
> +/// For convenience, [Self::or] and [Self::or_default] are provided.
> +pub struct PropertyGuard<'fwnode, 'name, T> {
> + /// The result of reading the property.
> + inner: Result<T>,
> + /// The fwnode of the property, used for logging in the "required" case.
> + fwnode: &'fwnode FwNode,
> + /// The name of the property, used for logging in the "required" case.
> + name: &'name CStr,
> +}
> +
> +impl<T> PropertyGuard<'_, '_, T> {
> + /// Access the property, indicating it is required.
> + ///
> + /// If the property is not present, the error is automatically logged. If a
> + /// missing property is not an error, use [Self::optional] instead.
> + pub fn required(self) -> Result<T> {
> + if self.inner.is_err() {
> + // Get the device associated with the fwnode for device-associated
> + // logging.
> + // TODO: Are we allowed to do this? The field `fwnode_handle.dev`
> + // has a somewhat vague comment, which could mean we're not
> + // supposed to access it:
> + // https://elixir.bootlin.com/linux/v6.13.6/source/include/linux/fwnode.h#L51
> + // SAFETY: According to the invariant of FwNode, it is valid.
> + let dev = unsafe { (*self.fwnode.as_raw()).dev };
> +
> + if dev.is_null() {
> + pr_err!("property '{}' is missing\n", self.name);
We should print the node path in this case (and perhaps both cases).
We have "%pfw" printf format specifier for C. We're going to need the
same for rust.
> + } else {
> + // SAFETY: If dev is not null, it points to a valid device.
> + let dev: &Device = unsafe { &*dev.cast() };
> + dev_err!(dev, "property '{}' is missing\n", self.name);
> + };
> + }
> + self.inner
> + }
> +
> + /// Access the property, indicating it is optional.
> + ///
> + /// In contrast to [Self::required], no error message is logged if the
> + /// property is not present.
> + pub fn optional(self) -> Option<T> {
> + self.inner.ok()
> + }
> +
> + /// Access the property or the specified default value.
> + ///
> + /// Do not pass a sentinel value as default to detect a missing property.
> + /// Use [Self::required] or [Self::optional] instead.
> + pub fn or(self, default: T) -> T {
> + self.inner.unwrap_or(default)
> + }
> +}
> +
> +impl<T: Default> PropertyGuard<'_, '_, T> {
> + /// Access the property or a default value.
> + ///
> + /// Use [Self::or] to specify a custom default value.
> + pub fn or_default(self) -> T {
> + self.inner.unwrap_or_default()
> + }
> +}
> --
> 2.49.0
>
next prev parent reply other threads:[~2025-03-26 22:25 UTC|newest]
Thread overview: 83+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-03-26 17:13 [PATCH 0/10] More Rust bindings for device property reads Remo Senekowitsch
2025-03-26 17:13 ` [PATCH 01/10] rust: Move property_present to property.rs Remo Senekowitsch
2025-03-26 20:51 ` Rob Herring
2025-03-26 22:41 ` Rob Herring
2025-04-04 12:48 ` Remo Senekowitsch
2025-03-26 20:58 ` Andrew Ballance
2025-03-27 8:37 ` Andy Shevchenko
2025-03-27 13:55 ` Rob Herring
2025-03-27 17:49 ` Andy Shevchenko
2025-03-26 17:13 ` [PATCH 02/10] rust: Add an Integer trait Remo Senekowitsch
2025-03-26 20:00 ` Rob Herring
2025-03-26 17:13 ` [PATCH 03/10] device property: Add fwnode_property_read_int_array() Remo Senekowitsch
2025-03-27 8:41 ` Andy Shevchenko
2025-04-02 16:04 ` Remo Senekowitsch
2025-04-03 13:28 ` Andy Shevchenko
2025-04-03 16:08 ` Rob Herring
2025-04-03 16:15 ` Remo Senekowitsch
2025-04-03 17:04 ` Remo Senekowitsch
2025-04-03 17:22 ` Rob Herring
2025-04-04 12:29 ` Remo Senekowitsch
2025-04-03 16:04 ` Rob Herring
2025-04-03 16:15 ` Andy Shevchenko
2025-04-03 16:36 ` Rob Herring
2025-04-03 17:54 ` Andy Shevchenko
2025-04-03 18:48 ` Rob Herring
2025-04-03 20:36 ` Miguel Ojeda
2025-04-04 11:00 ` Andy Shevchenko
2025-04-04 14:12 ` Rob Herring
2025-04-04 16:35 ` Andy Shevchenko
2025-03-26 17:13 ` [PATCH 04/10] rust: Add bindings for reading device properties Remo Senekowitsch
2025-03-26 21:27 ` Rob Herring
2025-04-02 16:28 ` Remo Senekowitsch
2025-03-26 17:13 ` [PATCH 05/10] rust: Read properties via single generic method Remo Senekowitsch
2025-03-26 21:33 ` Rob Herring
2025-03-26 17:13 ` [PATCH 06/10] rust: property: Add child accessor and iterator Remo Senekowitsch
2025-03-26 21:04 ` Andrew Ballance
2025-03-26 21:40 ` Rob Herring
2025-03-26 17:13 ` [PATCH 07/10] rust: Add arrayvec Remo Senekowitsch
2025-03-26 21:06 ` Andrew Ballance
2025-03-27 14:40 ` Danilo Krummrich
2025-03-26 17:13 ` [PATCH 08/10] rust: property: Add property_get_reference_args Remo Senekowitsch
2025-03-26 21:07 ` Andrew Ballance
2025-03-26 21:25 ` Miguel Ojeda
2025-03-26 21:45 ` Remo Senekowitsch
2025-03-27 14:32 ` Danilo Krummrich
2025-03-26 17:13 ` [PATCH 09/10] rust: property: Add PropertyGuard Remo Senekowitsch
2025-03-26 21:10 ` Andrew Ballance
2025-03-26 22:25 ` Rob Herring [this message]
2025-03-26 17:13 ` [PATCH 10/10] samples: rust: platform: Add property read examples Remo Senekowitsch
2025-03-26 22:01 ` Rob Herring
2025-03-26 22:23 ` Remo Senekowitsch
2025-03-27 0:02 ` Rob Herring
2025-03-27 10:28 ` Danilo Krummrich
2025-03-26 20:54 ` [PATCH 0/10] More Rust bindings for device property reads Andrew Ballance
2025-03-27 8:42 ` Andy Shevchenko
2025-04-14 15:26 ` [PATCH v2 0/5] " Remo Senekowitsch
2025-04-14 15:26 ` [PATCH v2 1/5] rust: Move property_present to separate file Remo Senekowitsch
2025-04-14 16:00 ` Danilo Krummrich
2025-04-14 16:40 ` Remo Senekowitsch
2025-04-14 18:00 ` Danilo Krummrich
2025-04-15 11:17 ` Remo Senekowitsch
2025-04-14 15:26 ` [PATCH v2 2/5] rust: Add bindings for reading device properties Remo Senekowitsch
2025-04-14 17:44 ` Danilo Krummrich
2025-04-14 18:05 ` Danilo Krummrich
2025-04-14 23:55 ` Remo Senekowitsch
2025-04-15 9:48 ` Danilo Krummrich
2025-04-15 11:11 ` Remo Senekowitsch
2025-04-15 12:46 ` Rob Herring
2025-04-15 13:11 ` Danilo Krummrich
2025-04-15 14:47 ` Remo Senekowitsch
2025-04-16 18:28 ` Gary Guo
2025-04-23 12:29 ` Dirk Behme
2025-04-24 11:25 ` Remo Senekowitsch
2025-04-23 12:34 ` Dirk Behme
2025-04-14 15:26 ` [PATCH v2 3/5] rust: property: Add child accessor and iterator Remo Senekowitsch
2025-04-14 16:09 ` Danilo Krummrich
2025-04-14 15:26 ` [PATCH v2 4/5] rust: property: Add property_get_reference_args Remo Senekowitsch
2025-04-14 15:26 ` [PATCH v2 5/5] samples: rust: platform: Add property read examples Remo Senekowitsch
2025-04-23 12:39 ` Dirk Behme
2025-04-24 5:47 ` Dirk Behme
2025-04-14 15:38 ` [PATCH v2 0/5] More Rust bindings for device property reads Miguel Ojeda
2025-04-14 16:07 ` Remo Senekowitsch
2025-04-14 16:44 ` Miguel Ojeda
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=20250326222529.GE2844851-robh@kernel.org \
--to=robh@kernel.org \
--cc=a.hindborg@kernel.org \
--cc=alex.gaynor@gmail.com \
--cc=aliceryhl@google.com \
--cc=andriy.shevchenko@linux.intel.com \
--cc=benno.lossin@proton.me \
--cc=bjorn3_gh@protonmail.com \
--cc=boqun.feng@gmail.com \
--cc=dakr@kernel.org \
--cc=devicetree@vger.kernel.org \
--cc=dirk.behme@de.bosch.com \
--cc=djrscally@gmail.com \
--cc=gary@garyguo.net \
--cc=gregkh@linuxfoundation.org \
--cc=heikki.krogerus@linux.intel.com \
--cc=linux-acpi@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=ojeda@kernel.org \
--cc=rafael@kernel.org \
--cc=remo@buenzli.dev \
--cc=rust-for-linux@vger.kernel.org \
--cc=sakari.ailus@linux.intel.com \
--cc=saravanak@google.com \
--cc=tmgross@umich.edu \
/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;
as well as URLs for NNTP newsgroup(s).