Re: [PATCH v14 6/8] mm: rust: add VmaNew for f_ops->mmap()

[Alice Ryhl <aliceryhl@google.com>]

On Tue, Feb 25, 2025 at 5:24 PM Gary Guo <gary@garyguo.net> wrote:
>
> On Thu, 13 Feb 2025 11:04:05 +0000
> Alice Ryhl <aliceryhl@google.com> wrote:
>
> > This type will be used when setting up a new vma in an f_ops->mmap()
> > hook. Using a separate type from VmaRef allows us to have a separate set
> > of operations that you are only able to use during the mmap() hook. For
> > example, the VM_MIXEDMAP flag must not be changed after the initial
> > setup that happens during the f_ops->mmap() hook.
> >
> > To avoid setting invalid flag values, the methods for clearing
> > VM_MAYWRITE and similar involve a check of VM_WRITE, and return an error
> > if VM_WRITE is set. Trying to use `try_clear_maywrite` without checking
> > the return value results in a compilation error because the `Result`
> > type is marked #[must_use].
> >
> > For now, there's only a method for VM_MIXEDMAP and not VM_PFNMAP. When
> > we add a VM_PFNMAP method, we will need some way to prevent you from
> > setting both VM_MIXEDMAP and VM_PFNMAP on the same vma.
> >
> > Acked-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
> > Reviewed-by: Jann Horn <jannh@google.com>
> > Reviewed-by: Andreas Hindborg <a.hindborg@kernel.org>
> > Signed-off-by: Alice Ryhl <aliceryhl@google.com>
> > ---
> >  rust/kernel/mm/virt.rs | 186 ++++++++++++++++++++++++++++++++++++++++++++++++-
> >  1 file changed, 185 insertions(+), 1 deletion(-)
> >
> > diff --git a/rust/kernel/mm/virt.rs b/rust/kernel/mm/virt.rs
> > index 3e2eabcc2145..31803674aecc 100644
> > --- a/rust/kernel/mm/virt.rs
> > +++ b/rust/kernel/mm/virt.rs
> > @@ -16,7 +16,7 @@
> >
> >  use crate::{
> >      bindings,
> > -    error::{to_result, Result},
> > +    error::{code::EINVAL, to_result, Result},
> >      mm::MmWithUser,
> >      page::Page,
> >      types::Opaque,
> > @@ -198,6 +198,190 @@ pub fn vm_insert_page(&self, address: usize, page: &Page) -> Result {
> >      }
> >  }
> >
> > +/// A configuration object for setting up a VMA in an `f_ops->mmap()` hook.
> > +///
> > +/// The `f_ops->mmap()` hook is called when a new VMA is being created, and the hook is able to
> > +/// configure the VMA in various ways to fit the driver that owns it. Using `VmaNew` indicates that
> > +/// you are allowed to perform operations on the VMA that can only be performed before the VMA is
> > +/// fully initialized.
> > +///
> > +/// # Invariants
> > +///
> > +/// For the duration of 'a, the referenced vma must be undergoing initialization in an
> > +/// `f_ops->mmap()` hook.
> > +pub struct VmaNew {
> > +    vma: VmaRef,
> > +}
> > +
> > +// Make all `VmaRef` methods available on `VmaNew`.
>
> Are there operations that can't be performed when VMA is still being
> setup? If so, using typestate might be more preferrable to Deref.

I don't think so.

> > +impl Deref for VmaNew {
> > +    type Target = VmaRef;
> > +
> > +    #[inline]
> > +    fn deref(&self) -> &VmaRef {
> > +        &self.vma
> > +    }
> > +}
> > +
> > +impl VmaNew {
> > +    /// Access a virtual memory area given a raw pointer.
> > +    ///
> > +    /// # Safety
> > +    ///
> > +    /// Callers must ensure that `vma` is undergoing initial vma setup for the duration of 'a.
> > +    #[inline]
> > +    pub unsafe fn from_raw<'a>(vma: *mut bindings::vm_area_struct) -> &'a Self {
> > +        // SAFETY: The caller ensures that the invariants are satisfied for the duration of 'a.
> > +        unsafe { &*vma.cast() }
> > +    }
> > +
> > +    /// Internal method for updating the vma flags.
> > +    ///
> > +    /// # Safety
> > +    ///
> > +    /// This must not be used to set the flags to an invalid value.
> > +    #[inline]
> > +    unsafe fn update_flags(&self, set: vm_flags_t, unset: vm_flags_t) {
> > +        let mut flags = self.flags();
> > +        flags |= set;
> > +        flags &= !unset;
> > +
> > +        // SAFETY: This is not a data race: the vma is undergoing initial setup, so it's not yet
>
> It is possible to make this API `&mut self` then?

No because of pinning. And taking `self: Pin<&mut Self>` is obnoxious.

> > +        // shared. Additionally, `VmaNew` is `!Sync`, so it cannot be used to write in parallel.
> > +        // The caller promises that this does not set the flags to an invalid value.
>
> Does `VmaRef` has to be `!Sync`? I wonder if we should explicitly mark
> `VmaNew` as `!Sync` so this is more explicitly correct.

See above.

Alice