From: Danilo Krummrich <dakr@redhat.com>
To: Christian Schrefl <chrisi.schrefl@gmail.com>
Cc: mcgrof@kernel.org, russ.weight@linux.dev,
gregkh@linuxfoundation.org, linux-kernel@vger.kernel.org,
rust-for-linux@vger.kernel.org, Gary Guo <gary@garyguo.net>
Subject: Re: [PATCH 2/2] firmware_loader: fix soundness issue in `request_internal`
Date: Mon, 8 Jul 2024 13:28:42 +0200 [thread overview]
Message-ID: <ZovNap1NBusREw_-@pollux> (raw)
In-Reply-To: <dd56a9d9-08fe-4bd4-83e2-dfdc44f8c637@gmail.com>
On Sun, Jul 07, 2024 at 10:57:31PM +0200, Christian Schrefl wrote:
> Greetings.
>
> On 07.07.24 2:38 AM, Danilo Krummrich wrote:
> > `request_internal` must be called with one of the following function
> > pointers: request_firmware(), firmware_request_nowarn(),
> > firmware_request_platform() or request_firmware_direct().
> >
> > The previous `FwFunc` alias did not guarantee this, which is unsound.
> >
> > In order to fix this up, implement `FwFunc` as new type with a
> > corresponding type invariant and unsafe `new` function.
> >
> > Reported-by: Gary Guo <gary@garyguo.net>
> > Closes: https://lore.kernel.org/lkml/20240620143611.7995e0bb@eugeo/
> > Signed-off-by: Danilo Krummrich <dakr@redhat.com>
> > ---
> > rust/kernel/firmware.rs | 32 ++++++++++++++++++++++++++------
> > 1 file changed, 26 insertions(+), 6 deletions(-)
> >
> > diff --git a/rust/kernel/firmware.rs b/rust/kernel/firmware.rs
> > index 106a928a535e..d765ecc85d38 100644
> > --- a/rust/kernel/firmware.rs
> > +++ b/rust/kernel/firmware.rs
> > @@ -7,11 +7,24 @@
> > 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 =
> > +type RawFwFunc =
> > unsafe extern "C" fn(*mut *const bindings::firmware, *const i8, *mut bindings::device) -> i32;
> >
> > +/// # Invariants
> > +///
> > +/// One of the following: `bindings::request_firmware`, `bindings::firmware_request_nowarn`,
> > +/// `firmware_request_platform`, `bindings::request_firmware_direct`
> > +struct FwFunc(RawFwFunc);
> > +
> > +impl FwFunc {
> > + /// # Safety
> > + ///
> > + /// Must be one of the functions listed in the type invariants.
> > + unsafe fn from_raw(func: RawFwFunc) -> Self {
> > + Self(func)
> > + }
> Why not write methods that construct each possible FwFunc?
Thanks, this is a good idea indeed, will send a v2.
> That way the code that needs to know abut this invariant is only inside a single impl block.
> Something like:
> impl FwFunc {
> fn request_firmware() -> Self {
> // # Safety
> // As per the Type Invariant `bindings::request_firmware` is a valid vaule.
> FwFunc(bindings::request_firmware)
> }
> }
>
> > +}
> > +
> > /// Abstraction around a C `struct firmware`.
> > ///
> > /// This is a simple abstraction around the C firmware API. Just like with the C API, firmware can
> > @@ -48,7 +61,7 @@ fn request_internal(name: &CStr, dev: &Device, func: FwFunc) -> Result<Self> {
> >
> > // 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()) };
> > + let ret = unsafe { func.0(pfw as _, name.as_char_ptr(), dev.as_raw()) };
> > if ret != 0 {
> > return Err(Error::from_errno(ret));
> > }
> > @@ -60,13 +73,20 @@ fn request_internal(name: &CStr, dev: &Device, func: FwFunc) -> Result<Self> {
> >
> > /// 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)
> > + // SAFETY: `bindings::request_firmware` is valid by the safety requirement of `FwFunc`.
> > + let func = unsafe { FwFunc::from_raw(bindings::request_firmware) };
> > +
> > + Self::request_internal(name, dev, func)
> > }
> >
> > /// 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)
> > + // SAFETY: `bindings::firmware_request_nowarn` is valid by the safety requirement of
> > + // `FwFunc::new`.
> > + let func = unsafe { FwFunc::from_raw(bindings::firmware_request_nowarn) };
> > +
> > + Self::request_internal(name, dev, func)
> > }
> >
> > fn as_raw(&self) -> *mut bindings::firmware {
>
> Cheers,
> Christian
>
prev parent reply other threads:[~2024-07-08 11:28 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-07-07 0:38 [PATCH 1/2] firmware_loader: annotate doctests as `no_run` Danilo Krummrich
2024-07-07 0:38 ` [PATCH 2/2] firmware_loader: fix soundness issue in `request_internal` Danilo Krummrich
2024-07-07 20:57 ` Christian Schrefl
2024-07-08 11:28 ` Danilo Krummrich [this message]
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=ZovNap1NBusREw_-@pollux \
--to=dakr@redhat.com \
--cc=chrisi.schrefl@gmail.com \
--cc=gary@garyguo.net \
--cc=gregkh@linuxfoundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mcgrof@kernel.org \
--cc=russ.weight@linux.dev \
--cc=rust-for-linux@vger.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;
as well as URLs for NNTP newsgroup(s).