[PATCH v5 0/2] strncpy_from_user for Rust

[PATCH v5 0/2] strncpy_from_user for Rust

[Alice Ryhl]

There is currently no easy way to read NUL-terminated strings from
userspace. Trying to use the ordinary read function on an array of the
maximum length doesn't work because it could fail with EFAULT when the C
string is shorter than the maximum length. In this case,
strncpy_from_user is better because it doesn't return EFAULT even if it
encounters a page fault on bytes that are after the NUL-terminator but
before the maximum length.

I don't see a clear owner of uaccess in MAINTAINERS, so I suggest that
it lands through Miguel's tree.

Signed-off-by: Alice Ryhl <aliceryhl@google.com>
---
Changes in v5:
- Add comment about not updating self.length.
- Add `` in "This implies that len == buf.len()."
- Update guarantees for raw_strncpy_from_user.
- Add Reviewed-by for Benno.
- Link to v4: https://lore.kernel.org/r/20250527-strncpy-from-user-v4-0-82168470d472@google.com

Changes in v4:
- Swap order of arguments to raw_strncpy_from_user, and rename buf to
  dst.
- Update safety comment on CStr::from_bytes_with_nul_unchecked.
- Add `` in "This implies that len == dst.len() < buf.len()."
- Pick up Reviewed-by tags.
- Link to v3: https://lore.kernel.org/r/20250505-strncpy-from-user-v3-0-85c677fd4f91@google.com

Changes in v3:
- Remove pub from raw_strncpy_from_user.
- Mention that some data may have been copied on EFAULT.
- Add more comments to strcpy_into_buf about tricky cases.
- Rewrite documentation of strcpy_into_buf.
- Add documentation alias.
- Pick up Reviewed-by tags.
- Link to v2: https://lore.kernel.org/r/20250429-strncpy-from-user-v2-0-7e6facac0bf0@google.com

Changes in v2:
- Rename the raw wrapper around strncpy_from_user to raw_strncpy_from_user.
- Add a more convenient helper on top that adds the missing
  NUL-terminator when necessary.
- Link to v1: https://lore.kernel.org/r/20250424-strncpy-from-user-v1-1-f983fe21685a@google.com

---
Alice Ryhl (2):
      uaccess: rust: add strncpy_from_user
      uaccess: rust: add UserSliceReader::strcpy_into_buf

 rust/kernel/uaccess.rs | 94 +++++++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 93 insertions(+), 1 deletion(-)
---
base-commit: 19272b37aa4f83ca52bdf9c16d5d81bdd1354494
change-id: 20250424-strncpy-from-user-1f2d06b0cdde

Best regards,
-- 
Alice Ryhl <aliceryhl@google.com>

[PATCH v5 1/2] uaccess: rust: add strncpy_from_user

[Alice Ryhl]

This patch adds a direct wrapper around the C function of the same name.
It's not really intended for direct use by Rust code since
strncpy_from_user has a somewhat unfortunate API where it only
nul-terminates the buffer if there's space for the nul-terminator. This
means that a direct Rust wrapper around it could not return a &CStr
since the buffer may not be a cstring. However, we still add the method
to build more convenient APIs on top of it, which will happen in
subsequent patches.

Reviewed-by: Danilo Krummrich <dakr@kernel.org>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Boqun Feng <boqun.feng@gmail.com>
Reviewed-by: Benno Lossin <lossin@kernel.org>
Signed-off-by: Alice Ryhl <aliceryhl@google.com>
---
 rust/kernel/uaccess.rs | 36 +++++++++++++++++++++++++++++++++++-
 1 file changed, 35 insertions(+), 1 deletion(-)

diff --git a/rust/kernel/uaccess.rs b/rust/kernel/uaccess.rs
index 6d70edd8086a9f1ce38f78d312d6a7cfc03141a1..635a03e0989f3fe99be80987aa47763782de1d3f 100644
--- a/rust/kernel/uaccess.rs
+++ b/rust/kernel/uaccess.rs
@@ -8,7 +8,7 @@
     alloc::{Allocator, Flags},
     bindings,
     error::Result,
-    ffi::c_void,
+    ffi::{c_char, c_void},
     prelude::*,
     transmute::{AsBytes, FromBytes},
 };
@@ -367,3 +367,37 @@ pub fn write<T: AsBytes>(&mut self, value: &T) -> Result {
         Ok(())
     }
 }
+
+/// Reads a nul-terminated string into `dst` and returns the length.
+///
+/// This reads from userspace until a NUL byte is encountered, or until `dst.len()` bytes have been
+/// read. Fails with [`EFAULT`] if a read happens on a bad address (some data may have been
+/// copied). When the end of the buffer is encountered, no NUL byte is added, so the string is
+/// *not* guaranteed to be NUL-terminated when `Ok(dst.len())` is returned.
+///
+/// # Guarantees
+///
+/// When this function returns `Ok(len)`, it is guaranteed that the first `len` bytes of `dst` are
+/// initialized and non-zero. Furthermore, if `len < dst.len()`, then `dst[len]` is a NUL byte.
+#[inline]
+#[expect(dead_code)]
+fn raw_strncpy_from_user(dst: &mut [MaybeUninit<u8>], src: UserPtr) -> Result<usize> {
+    // CAST: Slice lengths are guaranteed to be `<= isize::MAX`.
+    let len = dst.len() as isize;
+
+    // SAFETY: `dst` is valid for writing `dst.len()` bytes.
+    let res = unsafe {
+        bindings::strncpy_from_user(dst.as_mut_ptr().cast::<c_char>(), src as *const c_char, len)
+    };
+
+    if res < 0 {
+        return Err(Error::from_errno(res as i32));
+    }
+
+    #[cfg(CONFIG_RUST_OVERFLOW_CHECKS)]
+    assert!(res <= len);
+
+    // GUARANTEES: `strncpy_from_user` was successful, so `dst` has contents in accordance with the
+    // guarantees of this function.
+    Ok(res as usize)
+}

-- 
2.50.0.rc2.692.g299adb8693-goog

[PATCH v5 2/2] uaccess: rust: add UserSliceReader::strcpy_into_buf

[Alice Ryhl]

This patch adds a more convenient method for reading C strings from
userspace. Logic is added to NUL-terminate the buffer when necessary so
that a &CStr can be returned.

Note that we treat attempts to read past `self.length` as a fault, so
this returns EFAULT if that limit is exceeded before `buf.len()` is
reached.

Reviewed-by: Danilo Krummrich <dakr@kernel.org>
Signed-off-by: Alice Ryhl <aliceryhl@google.com>
---
 rust/kernel/uaccess.rs | 60 +++++++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 59 insertions(+), 1 deletion(-)

diff --git a/rust/kernel/uaccess.rs b/rust/kernel/uaccess.rs
index 635a03e0989f3fe99be80987aa47763782de1d3f..106aa05ea1b88868fe48f64ca9c86b20ad7db68e 100644
--- a/rust/kernel/uaccess.rs
+++ b/rust/kernel/uaccess.rs
@@ -291,6 +291,65 @@ pub fn read_all<A: Allocator>(mut self, buf: &mut Vec<u8, A>, flags: Flags) -> R
         unsafe { buf.inc_len(len) };
         Ok(())
     }
+
+    /// Read a NUL-terminated string from userspace and return it.
+    ///
+    /// The string is read into `buf` and a NUL-terminator is added if the end of `buf` is reached.
+    /// Since there must be space to add a NUL-terminator, the buffer must not be empty. The
+    /// returned `&CStr` points into `buf`.
+    ///
+    /// Fails with [`EFAULT`] if the read happens on a bad address (some data may have been
+    /// copied).
+    #[doc(alias = "strncpy_from_user")]
+    pub fn strcpy_into_buf<'buf>(self, buf: &'buf mut [u8]) -> Result<&'buf CStr> {
+        if buf.is_empty() {
+            return Err(EINVAL);
+        }
+
+        // SAFETY: The types are compatible and `strncpy_from_user` doesn't write uninitialized
+        // bytes to `buf`.
+        let mut dst = unsafe { &mut *(buf as *mut [u8] as *mut [MaybeUninit<u8>]) };
+
+        // We never read more than `self.length` bytes.
+        if dst.len() > self.length {
+            dst = &mut dst[..self.length];
+        }
+
+        let mut len = raw_strncpy_from_user(dst, self.ptr)?;
+        if len < dst.len() {
+            // Add one to include the NUL-terminator.
+            len += 1;
+        } else if len < buf.len() {
+            // This implies that `len == dst.len() < buf.len()`.
+            //
+            // This means that we could not fill the entire buffer, but we had to stop reading
+            // because we hit the `self.length` limit of this `UserSliceReader`. Since we did not
+            // fill the buffer, we treat this case as if we tried to read past the `self.length`
+            // limit and received a page fault, which is consistent with other `UserSliceReader`
+            // methods that also return page faults when you exceed `self.length`.
+            return Err(EFAULT);
+        } else {
+            // This implies that `len == buf.len()`.
+            //
+            // This means that we filled the buffer exactly. In this case, we add a NUL-terminator
+            // and return it. Unlike the `len < dst.len()` branch, don't modify `len` because it
+            // already represents the length including the NUL-terminator.
+            //
+            // SAFETY: Due to the check at the beginning, the buffer is not empty.
+            unsafe { *buf.last_mut().unwrap_unchecked() = 0 };
+        }
+
+        // This method consumes `self`, so it can only be called once, thus we do not need to
+        // update `self.length`. This sidesteps concerns such as whether `self.length` should be
+        // incremented by `len` or `len-1` in the `len == buf.len()` case.
+
+        // SAFETY: There are two cases:
+        // * If we hit the `len < dst.len()` case, then `raw_strncpy_from_user` guarantees that
+        //   this slice contains exactly one NUL byte at the end of the string.
+        // * Otherwise, `raw_strncpy_from_user` guarantees that the string contained no NUL bytes,
+        //   and we have since added a NUL byte at the end.
+        Ok(unsafe { CStr::from_bytes_with_nul_unchecked(&buf[..len]) })
+    }
 }
 
 /// A writer for [`UserSlice`].
@@ -380,7 +439,6 @@ pub fn write<T: AsBytes>(&mut self, value: &T) -> Result {
 /// When this function returns `Ok(len)`, it is guaranteed that the first `len` bytes of `dst` are
 /// initialized and non-zero. Furthermore, if `len < dst.len()`, then `dst[len]` is a NUL byte.
 #[inline]
-#[expect(dead_code)]
 fn raw_strncpy_from_user(dst: &mut [MaybeUninit<u8>], src: UserPtr) -> Result<usize> {
     // CAST: Slice lengths are guaranteed to be `<= isize::MAX`.
     let len = dst.len() as isize;

-- 
2.50.0.rc2.692.g299adb8693-goog

Re: [PATCH v5 2/2] uaccess: rust: add UserSliceReader::strcpy_into_buf

[Benno Lossin]

On Mon Jun 16, 2025 at 2:41 PM CEST, Alice Ryhl wrote:
> This patch adds a more convenient method for reading C strings from
> userspace. Logic is added to NUL-terminate the buffer when necessary so
> that a &CStr can be returned.
>
> Note that we treat attempts to read past `self.length` as a fault, so
> this returns EFAULT if that limit is exceeded before `buf.len()` is
> reached.
>
> Reviewed-by: Danilo Krummrich <dakr@kernel.org>
> Signed-off-by: Alice Ryhl <aliceryhl@google.com>

I have one concern left below, when we fix or resolve that:

Reviewed-by: Benno Lossin <lossin@kernel.org>

> ---
>  rust/kernel/uaccess.rs | 60 +++++++++++++++++++++++++++++++++++++++++++++++++-
>  1 file changed, 59 insertions(+), 1 deletion(-)
>
> diff --git a/rust/kernel/uaccess.rs b/rust/kernel/uaccess.rs
> index 635a03e0989f3fe99be80987aa47763782de1d3f..106aa05ea1b88868fe48f64ca9c86b20ad7db68e 100644
> --- a/rust/kernel/uaccess.rs
> +++ b/rust/kernel/uaccess.rs
> @@ -291,6 +291,65 @@ pub fn read_all<A: Allocator>(mut self, buf: &mut Vec<u8, A>, flags: Flags) -> R
>          unsafe { buf.inc_len(len) };
>          Ok(())
>      }
> +
> +    /// Read a NUL-terminated string from userspace and return it.
> +    ///
> +    /// The string is read into `buf` and a NUL-terminator is added if the end of `buf` is reached.
> +    /// Since there must be space to add a NUL-terminator, the buffer must not be empty. The
> +    /// returned `&CStr` points into `buf`.
> +    ///
> +    /// Fails with [`EFAULT`] if the read happens on a bad address (some data may have been
> +    /// copied).
> +    #[doc(alias = "strncpy_from_user")]
> +    pub fn strcpy_into_buf<'buf>(self, buf: &'buf mut [u8]) -> Result<&'buf CStr> {
> +        if buf.is_empty() {
> +            return Err(EINVAL);
> +        }
> +
> +        // SAFETY: The types are compatible and `strncpy_from_user` doesn't write uninitialized
> +        // bytes to `buf`.
> +        let mut dst = unsafe { &mut *(buf as *mut [u8] as *mut [MaybeUninit<u8>]) };
> +
> +        // We never read more than `self.length` bytes.
> +        if dst.len() > self.length {
> +            dst = &mut dst[..self.length];
> +        }
> +
> +        let mut len = raw_strncpy_from_user(dst, self.ptr)?;
> +        if len < dst.len() {
> +            // Add one to include the NUL-terminator.
> +            len += 1;
> +        } else if len < buf.len() {
> +            // This implies that `len == dst.len() < buf.len()`.
> +            //
> +            // This means that we could not fill the entire buffer, but we had to stop reading
> +            // because we hit the `self.length` limit of this `UserSliceReader`. Since we did not
> +            // fill the buffer, we treat this case as if we tried to read past the `self.length`
> +            // limit and received a page fault, which is consistent with other `UserSliceReader`
> +            // methods that also return page faults when you exceed `self.length`.
> +            return Err(EFAULT);
> +        } else {
> +            // This implies that `len == buf.len()`.
> +            //
> +            // This means that we filled the buffer exactly. In this case, we add a NUL-terminator
> +            // and return it. Unlike the `len < dst.len()` branch, don't modify `len` because it
> +            // already represents the length including the NUL-terminator.
> +            //
> +            // SAFETY: Due to the check at the beginning, the buffer is not empty.
> +            unsafe { *buf.last_mut().unwrap_unchecked() = 0 };

What about the case `self.length == 0`? Will `raw_strncpy_from_user`
return early with a page fault, or will it return with `len == 0`?
Because if it is the latter, then this will result in UB.

---
Cheers,
Benno

> +        }
> +
> +        // This method consumes `self`, so it can only be called once, thus we do not need to
> +        // update `self.length`. This sidesteps concerns such as whether `self.length` should be
> +        // incremented by `len` or `len-1` in the `len == buf.len()` case.
> +
> +        // SAFETY: There are two cases:
> +        // * If we hit the `len < dst.len()` case, then `raw_strncpy_from_user` guarantees that
> +        //   this slice contains exactly one NUL byte at the end of the string.
> +        // * Otherwise, `raw_strncpy_from_user` guarantees that the string contained no NUL bytes,
> +        //   and we have since added a NUL byte at the end.
> +        Ok(unsafe { CStr::from_bytes_with_nul_unchecked(&buf[..len]) })
> +    }
>  }
>  
>  /// A writer for [`UserSlice`].
> @@ -380,7 +439,6 @@ pub fn write<T: AsBytes>(&mut self, value: &T) -> Result {
>  /// When this function returns `Ok(len)`, it is guaranteed that the first `len` bytes of `dst` are
>  /// initialized and non-zero. Furthermore, if `len < dst.len()`, then `dst[len]` is a NUL byte.
>  #[inline]
> -#[expect(dead_code)]
>  fn raw_strncpy_from_user(dst: &mut [MaybeUninit<u8>], src: UserPtr) -> Result<usize> {
>      // CAST: Slice lengths are guaranteed to be `<= isize::MAX`.
>      let len = dst.len() as isize;

Re: [PATCH v5 2/2] uaccess: rust: add UserSliceReader::strcpy_into_buf

[Alice Ryhl]

On Tue, Jun 17, 2025 at 9:38 AM Benno Lossin <lossin@kernel.org> wrote:
>
> On Mon Jun 16, 2025 at 2:41 PM CEST, Alice Ryhl wrote:
> > This patch adds a more convenient method for reading C strings from
> > userspace. Logic is added to NUL-terminate the buffer when necessary so
> > that a &CStr can be returned.
> >
> > Note that we treat attempts to read past `self.length` as a fault, so
> > this returns EFAULT if that limit is exceeded before `buf.len()` is
> > reached.
> >
> > Reviewed-by: Danilo Krummrich <dakr@kernel.org>
> > Signed-off-by: Alice Ryhl <aliceryhl@google.com>
>
> I have one concern left below, when we fix or resolve that:
>
> Reviewed-by: Benno Lossin <lossin@kernel.org>
>
> > ---
> >  rust/kernel/uaccess.rs | 60 +++++++++++++++++++++++++++++++++++++++++++++++++-
> >  1 file changed, 59 insertions(+), 1 deletion(-)
> >
> > diff --git a/rust/kernel/uaccess.rs b/rust/kernel/uaccess.rs
> > index 635a03e0989f3fe99be80987aa47763782de1d3f..106aa05ea1b88868fe48f64ca9c86b20ad7db68e 100644
> > --- a/rust/kernel/uaccess.rs
> > +++ b/rust/kernel/uaccess.rs
> > @@ -291,6 +291,65 @@ pub fn read_all<A: Allocator>(mut self, buf: &mut Vec<u8, A>, flags: Flags) -> R
> >          unsafe { buf.inc_len(len) };
> >          Ok(())
> >      }
> > +
> > +    /// Read a NUL-terminated string from userspace and return it.
> > +    ///
> > +    /// The string is read into `buf` and a NUL-terminator is added if the end of `buf` is reached.
> > +    /// Since there must be space to add a NUL-terminator, the buffer must not be empty. The
> > +    /// returned `&CStr` points into `buf`.
> > +    ///
> > +    /// Fails with [`EFAULT`] if the read happens on a bad address (some data may have been
> > +    /// copied).
> > +    #[doc(alias = "strncpy_from_user")]
> > +    pub fn strcpy_into_buf<'buf>(self, buf: &'buf mut [u8]) -> Result<&'buf CStr> {
> > +        if buf.is_empty() {
> > +            return Err(EINVAL);
> > +        }
> > +
> > +        // SAFETY: The types are compatible and `strncpy_from_user` doesn't write uninitialized
> > +        // bytes to `buf`.
> > +        let mut dst = unsafe { &mut *(buf as *mut [u8] as *mut [MaybeUninit<u8>]) };
> > +
> > +        // We never read more than `self.length` bytes.
> > +        if dst.len() > self.length {
> > +            dst = &mut dst[..self.length];
> > +        }
> > +
> > +        let mut len = raw_strncpy_from_user(dst, self.ptr)?;
> > +        if len < dst.len() {
> > +            // Add one to include the NUL-terminator.
> > +            len += 1;
> > +        } else if len < buf.len() {
> > +            // This implies that `len == dst.len() < buf.len()`.
> > +            //
> > +            // This means that we could not fill the entire buffer, but we had to stop reading
> > +            // because we hit the `self.length` limit of this `UserSliceReader`. Since we did not
> > +            // fill the buffer, we treat this case as if we tried to read past the `self.length`
> > +            // limit and received a page fault, which is consistent with other `UserSliceReader`
> > +            // methods that also return page faults when you exceed `self.length`.
> > +            return Err(EFAULT);
> > +        } else {
> > +            // This implies that `len == buf.len()`.
> > +            //
> > +            // This means that we filled the buffer exactly. In this case, we add a NUL-terminator
> > +            // and return it. Unlike the `len < dst.len()` branch, don't modify `len` because it
> > +            // already represents the length including the NUL-terminator.
> > +            //
> > +            // SAFETY: Due to the check at the beginning, the buffer is not empty.
> > +            unsafe { *buf.last_mut().unwrap_unchecked() = 0 };
>
> What about the case `self.length == 0`? Will `raw_strncpy_from_user`
> return early with a page fault, or will it return with `len == 0`?
> Because if it is the latter, then this will result in UB.

If `self.length == 0`, then you will either:
1. If buf.is_empty() then you return EINVAL at the top.
2. Otherwise, you return EFAULT from the `len < buf.len()` case.

Alice

Re: [PATCH v5 2/2] uaccess: rust: add UserSliceReader::strcpy_into_buf

[Benno Lossin]

On Tue Jun 17, 2025 at 10:55 AM CEST, Alice Ryhl wrote:
> On Tue, Jun 17, 2025 at 9:38 AM Benno Lossin <lossin@kernel.org> wrote:
>> What about the case `self.length == 0`? Will `raw_strncpy_from_user`
>> return early with a page fault, or will it return with `len == 0`?
>> Because if it is the latter, then this will result in UB.
>
> If `self.length == 0`, then you will either:
> 1. If buf.is_empty() then you return EINVAL at the top.
> 2. Otherwise, you return EFAULT from the `len < buf.len()` case.

Ah that makes sense. Thanks for explaining.

---
Cheers,
Benno

Re: [PATCH v5 0/2] strncpy_from_user for Rust

[Miguel Ojeda]

On Mon, Jun 16, 2025 at 2:42 PM Alice Ryhl <aliceryhl@google.com> wrote:
>
> There is currently no easy way to read NUL-terminated strings from
> userspace. Trying to use the ordinary read function on an array of the
> maximum length doesn't work because it could fail with EFAULT when the C
> string is shorter than the maximum length. In this case,
> strncpy_from_user is better because it doesn't return EFAULT even if it
> encounters a page fault on bytes that are after the NUL-terminator but
> before the maximum length.
>
> I don't see a clear owner of uaccess in MAINTAINERS, so I suggest that
> it lands through Miguel's tree.
>
> Signed-off-by: Alice Ryhl <aliceryhl@google.com>

Applied to `rust-next` -- thanks everyone!

    [ Reworded title. - Miguel ]

    [ Use `from_mut` to clean `clippy::ref_as_ptr` lint. Reworded
      title. - Miguel ]

Cheers,
Miguel