Message-ID: <ZnC9xajuhN6nSQb-@cassiopeiae>
Date: Tue, 18 Jun 2024 00:50:45 +0200
From: Danilo Krummrich <dakr@...hat.com>
To: Boqun Feng <boqun.feng@...il.com>
Cc: gregkh@...uxfoundation.org, rafael@...nel.org, mcgrof@...nel.org,
	russ.weight@...ux.dev, ojeda@...nel.org, alex.gaynor@...il.com,
	wedsonaf@...il.com, gary@...yguo.net, bjorn3_gh@...tonmail.com,
	benno.lossin@...ton.me, a.hindborg@...sung.com,
	aliceryhl@...gle.com, airlied@...il.com, fujita.tomonori@...il.com,
	pstanner@...hat.com, ajanulgu@...hat.com, lyude@...hat.com,
	rust-for-linux@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3 2/2] rust: add firmware abstractions

On Mon, Jun 17, 2024 at 03:05:32PM -0700, Boqun Feng wrote:
> On Mon, Jun 17, 2024 at 10:29:41PM +0200, Danilo Krummrich wrote:
> > Add an abstraction around the kernels firmware API to request firmware
> > images. The abstraction provides functions to access the firmware's size
> > and backing buffer.
> > 
> > The firmware is released once the abstraction instance is dropped.
> > 
> > Signed-off-by: Danilo Krummrich <dakr@...hat.com>
> > ---
> >  drivers/base/firmware_loader/Kconfig |  7 ++
> >  rust/bindings/bindings_helper.h      |  1 +
> >  rust/kernel/firmware.rs              | 98 ++++++++++++++++++++++++++++
> >  rust/kernel/lib.rs                   |  2 +
> >  4 files changed, 108 insertions(+)
> >  create mode 100644 rust/kernel/firmware.rs
> > 
> > diff --git a/drivers/base/firmware_loader/Kconfig b/drivers/base/firmware_loader/Kconfig
> > index 5ca00e02fe82..a03701674265 100644
> > --- a/drivers/base/firmware_loader/Kconfig
> > +++ b/drivers/base/firmware_loader/Kconfig
> > @@ -37,6 +37,13 @@ config FW_LOADER_DEBUG
> >  	  SHA256 checksums to the kernel log for each firmware file that is
> >  	  loaded.
> >  
> > +config RUST_FW_LOADER_ABSTRACTIONS
> > +	bool "Rust Firmware Loader abstractions"
> > +	depends on RUST
> > +	depends on FW_LOADER=y
> > +	help
> > +	  This enables the Rust abstractions for the firmware loader API.
> > +
> >  if FW_LOADER
> >  
> >  config FW_LOADER_PAGED_BUF
> > diff --git a/rust/bindings/bindings_helper.h b/rust/bindings/bindings_helper.h
> > index ddb5644d4fd9..18a3f05115cb 100644
> > --- a/rust/bindings/bindings_helper.h
> > +++ b/rust/bindings/bindings_helper.h
> > @@ -9,6 +9,7 @@
> >  #include <kunit/test.h>
> >  #include <linux/errname.h>
> >  #include <linux/ethtool.h>
> > +#include <linux/firmware.h>
> >  #include <linux/jiffies.h>
> >  #include <linux/mdio.h>
> >  #include <linux/phy.h>
> > diff --git a/rust/kernel/firmware.rs b/rust/kernel/firmware.rs
> > new file mode 100644
> > index 000000000000..05a4f84cfd42
> > --- /dev/null
> > +++ b/rust/kernel/firmware.rs
> > @@ -0,0 +1,98 @@
> > +// SPDX-License-Identifier: GPL-2.0
> > +
> > +//! Firmware abstraction
> > +//!
> > +//! C header: [`include/linux/firmware.h`](srctree/include/linux/firmware.h")
> > +
> > +use crate::{bindings, device::Device, error::Error, error::Result, str::CStr};
> > +use core::ptr::NonNull;
> > +
> > +// One of the following: `bindings::request_firmware`, `bindings::firmware_request_nowarn`,
> > +// `firmware_request_platform`, `bindings::request_firmware_direct`
> > +type FwFunc =
> > +    unsafe extern "C" fn(*mut *const bindings::firmware, *const i8, *mut bindings::device) -> i32;
> > +
> > +/// Abstraction around a C `struct firmware`.
> > +///
> > +/// This is a simple abstraction around the C firmware API. Just like with the C API, firmware can
> > +/// be requested. Once requested the abstraction provides direct access to the firmware buffer as
> > +/// `&[u8]`. The firmware is released once [`Firmware`] is dropped.
> > +///
> > +/// # Invariants
> > +///
> > +/// The pointer is valid, and has ownership over the instance of `struct firmware`.
> > +///
> > +/// # Examples
> > +///
> > +/// ```
> > +/// # use kernel::{c_str, device::Device, firmware::Firmware};
> > +///
> > +/// # // SAFETY: *NOT* safe, just for the example to get an `ARef<Device>` instance
> > +/// # let dev = unsafe { Device::from_raw(core::ptr::null_mut()) };
> > +///
> > +/// let fw = Firmware::request(c_str!("path/to/firmware.bin"), &dev).unwrap();
> > +/// let blob = fw.data();
> > +/// ```
> > +pub struct Firmware(NonNull<bindings::firmware>);
> > +
> 
> I feel like eventually we need a very simple smart pointer type for
> these case, for example:
> 
>     /// A smart pointer owns the underlying data.
>     pub struct Owned<T: Ownable> {
>         ptr: NonNull<T>,
>     }
> 
>     impl<T: Ownable> Owned<T> {
>         /// # Safety
> 	/// `ptr` needs to be a valid pointer, and it should be the
> 	/// unique owner to the object, in other words, no one can touch
> 	/// or free the underlying data.
>         pub unsafe to_owned(ptr: *mut T) -> Self {
> 	    // SAFETY: Per function safety requirement.
> 	    Self { ptr: unsafe { NonNull::new_unchecked(ptr) } }
> 	}
> 
> 	/// other safe constructors are available if a initializer (impl
> 	/// Init) is provided
>     }
> 
>     /// A Ownable type is a type that can be put into `Owned<T>`, and
>     /// when `Owned<T>` drops, `ptr_drop` will be called.
>     pub unsafe trait Ownable {
>         /// # Safety
> 	/// This could only be called in the `Owned::drop` function.
>         unsafe fn ptr_drop(ptr: *mut Self);
>     }
> 
>     impl<T: Ownable> Drop for Owned<T> {
>         fn drop(&mut self) {
> 	    /// SAFETY: In Owned<T>::drop.
> 	    unsafe {
> 	        <T as Ownable>::ptr_drop(self.as_mut_ptr());
> 	    }
> 	}
>     }
> 
> we can implement Deref and DerefMut easily on `Owned<T>`. And then we
> could define Firmware as
> 
>     #[repr(transparent)]
>     pub struct Firmware(Opaque<bindings::firmware>);
> 
> and
> 
>     unsafe impl Ownable for Firmware {
>         unsafe fn ptr_drop(ptr: *mut Self) {
> 	    // SAFETY: Per function safety, this is called in
> 	    // Owned::drop(), so `ptr` is a unique pointer to object,
> 	    // it's safe to release the firmware.
>             unsafe { bindings::release_firmware(ptr.cast()); }
>         }
>     }
> 
> and the request_*() will return a `Result<Owned<Self>>`. 
> 
> Alice mentioned the need of this in page as well:
> 
> 	https://lore.kernel.org/rust-for-linux/CAH5fLgjrt0Ohj1qBv=GrqZumBTMQ1jbsKakChmxmG2JYDJEM8w@mail.gmail.com		

I think in the `Page` case this is useful to create `Page` references from
previously allocated memory.

In the case of `Firmware`, I agree it makes sense to use it once we have it,
but other than for consistency, is there any advantage?

> 
> Just bring it up while we are (maybe not? ;-)) at it. Also I would like
> to hear whether this would work for Firmware in the longer-term ;-) But
> yes, I'm not that worried about merging it as it is if others are all
> OK.

I think there's not too much to add here in the future, once we got an allocator
API (I should get back to that soon), I want to add a method that copies the
data to a new buffer allocated with a given allocator. And maybe we want to
support a few other request_firmware_* functions in the future, but none of that
should require the above abstraction.

> 
> > +impl Firmware {
> > +    fn request_internal(name: &CStr, dev: &Device, func: FwFunc) -> Result<Self> {
> > +        let mut fw: *mut bindings::firmware = core::ptr::null_mut();
> > +        let pfw: *mut *mut bindings::firmware = &mut fw;
> > +
> > +        // SAFETY: `pfw` is a valid pointer to a NULL initialized `bindings::firmware` pointer.
> > +        // `name` and `dev` are valid as by their type invariants.
> > +        let ret = unsafe { func(pfw as _, name.as_char_ptr(), dev.as_raw()) };
> > +        if ret != 0 {
> > +            return Err(Error::from_errno(ret));
> > +        }
> > +
> > +        // SAFETY: `func` not bailing out with a non-zero error code, guarantees that `fw` is a
> > +        // valid pointer to `bindings::firmware`.
> > +        Ok(Firmware(unsafe { NonNull::new_unchecked(fw) }))
> > +    }
> > +
> > +    /// Send a firmware request and wait for it. See also `bindings::request_firmware`.
> > +    pub fn request(name: &CStr, dev: &Device) -> Result<Self> {
> > +        Self::request_internal(name, dev, bindings::request_firmware)
> > +    }
> > +
> > +    /// Send a request for an optional firmware module. See also
> > +    /// `bindings::firmware_request_nowarn`.
> > +    pub fn request_nowarn(name: &CStr, dev: &Device) -> Result<Self> {
> > +        Self::request_internal(name, dev, bindings::firmware_request_nowarn)
> > +    }
> > +
> > +    fn as_raw(&self) -> *mut bindings::firmware {
> > +        self.0.as_ptr()
> > +    }
> > +
> > +    /// Returns the size of the requested firmware in bytes.
> > +    pub fn size(&self) -> usize {
> > +        // SAFETY: Safe by the type invariant.
> > +        unsafe { (*self.as_raw()).size }
> > +    }
> > +
> > +    /// Returns the requested firmware as `&[u8]`.
> > +    pub fn data(&self) -> &[u8] {
> > +        // SAFETY: Safe by the type invariant. Additionally, `bindings::firmware` guarantees, if
> 
> Does this "Safe by the type invariant" also covers the following safe
> requirement of `from_raw_parts`?
> 
> 	The memory referenced by the returned slice must not be mutated for the duration of lifetime 'a, except inside an UnsafeCell.
> 
> in that `&[u8]` has the same lifetime as `&self`, and as long as
> `&self` exists, no function can touch the inner `data`? If so, I
> probably want to call this out.

Yes, nothing should ever modify the firmware buffer after it has been requested
successfully. I can add this to the type invariant.

> 
> Regards,
> Boqun
> 
> > +        // successfully requested, that `bindings::firmware::data` has a size of
> > +        // `bindings::firmware::size` bytes.
> > +        unsafe { core::slice::from_raw_parts((*self.as_raw()).data, self.size()) }
> > +    }
> > +}
> > +
> > +impl Drop for Firmware {
> > +    fn drop(&mut self) {
> > +        // SAFETY: Safe by the type invariant.
> > +        unsafe { bindings::release_firmware(self.as_raw()) };
> > +    }
> > +}
> > +
> > +// SAFETY: `Firmware` only holds a pointer to a C `struct firmware`, which is safe to be used from
> > +// any thread.
> > +unsafe impl Send for Firmware {}
> > +
> > +// SAFETY: `Firmware` only holds a pointer to a C `struct firmware`, references to which are safe to
> > +// be used from any thread.
> > +unsafe impl Sync for Firmware {}
> > diff --git a/rust/kernel/lib.rs b/rust/kernel/lib.rs
> > index dd1207f1a873..7707cb013ce9 100644
> > --- a/rust/kernel/lib.rs
> > +++ b/rust/kernel/lib.rs
> > @@ -30,6 +30,8 @@
> >  mod build_assert;
> >  pub mod device;
> >  pub mod error;
> > +#[cfg(CONFIG_RUST_FW_LOADER_ABSTRACTIONS)]
> > +pub mod firmware;
> >  pub mod init;
> >  pub mod ioctl;
> >  #[cfg(CONFIG_KUNIT)]
> > -- 
> > 2.45.1
> > 
> 

Powered by blists - more mailing lists