From: Boqun Feng <boqun.feng@gmail.com>
To: Benno Lossin <benno.lossin@proton.me>
Cc: "Alice Ryhl" <alice@ryhl.io>, "Alice Ryhl" <aliceryhl@google.com>,
"Wedson Almeida Filho" <wedsonaf@gmail.com>,
rust-for-linux@vger.kernel.org, "Miguel Ojeda" <ojeda@kernel.org>,
"Alex Gaynor" <alex.gaynor@gmail.com>,
"Gary Guo" <gary@garyguo.net>,
"Björn Roy Baron" <bjorn3_gh@protonmail.com>,
"Andreas Hindborg" <a.hindborg@samsung.com>,
linux-kernel@vger.kernel.org,
"Wedson Almeida Filho" <walmeida@microsoft.com>
Subject: Re: [PATCH v2 2/2] rust: arc: remove `ArcBorrow` in favour of `WithRef`
Date: Mon, 25 Sep 2023 16:24:08 -0700 [thread overview]
Message-ID: <ZRIWmPa0mN6s7Yoj@boqun-archlinux> (raw)
In-Reply-To: <ZRILEIsKW6LsGdfb@boqun-archlinux>
On Mon, Sep 25, 2023 at 03:34:56PM -0700, Boqun Feng wrote:
[...]
> > >>>
> > >>> Hmm... but we do the similar thing in `Arc::from_raw()`, right?
> > >>>
> > >>> pub unsafe fn from_raw(ptr: *const T) -> Self {
> > >>> ..
> > >>> }
> > >>>
> > >>> , what we have is a pointer to T, and we construct a pointer to
> > >>> `ArcInner<T>/WithRef<T>`, in that function. Because the `sub` on pointer
> > >>> gets away from provenance? If so, we can also do a sub(0) in the above
> > >>> code.
> > >>
> > >> Not sure what you mean. Operations on raw pointers leave provenance
> > >> unchanged.
> > >
> > > Let's look at the function from_raw(), the input is a pointer to T,
> > > right? So you only have the provenance to T, but in that function, the
> > > pointer is casted to a pointer to WithRef<T>/ArcInner<T>, that means you
> > > have the provenance to the whole WithRef<T>/ArcInner<T>, right? My
> > > question is: why isn't that a UB?
> >
> > The pointer was originally derived by a call to `into_raw`:
> > ```
> > pub fn into_raw(self) -> *const T {
> > let ptr = self.ptr.as_ptr();
> > core::mem::forget(self);
> > // SAFETY: The pointer is valid.
> > unsafe { core::ptr::addr_of!((*ptr).data) }
> > }
> > ```
> > So in this function the origin (also the origin of the provenance)
> > of the pointer is `ptr` which is of type `NonNull<WithRef<T>>`.
> > Raw pointers do not lose this provenance information when you cast
> > it and when using `addr_of`/`addr_of_mut`. So provenance is something
> > that is not really represented in the type system for raw pointers.
>
> Ah, I see, that's the thing I was missing. Now it makes much sense to
> me, thank you both!
>
> >
> > When doing a round trip through a reference though, the provenance is
> > newly assigned and thus would only be valid for a `T`:
> > ```
> > let raw = arc.into_raw();
> > let reference = unsafe { &*raw };
> > let raw: *const T = reference;
> > let arc = unsafe { Arc::from_raw(raw) };
> > ```
>
> Agreed. This example demonstrates the key point: the provenances of raw
> pointers are decided at derive time.
>
So the original problem the Alice brought up is also because of the
provenance, right? To get a `&WithRef<T>`, we reborrow the pointer to
get a `&`, and any pointer derived from that reference will have a
different (and read-only) provenance, which causes the problem. Like:
```rust
let raw = Box::into_raw(arc);
let reference = unsafe { &*raw }; // as_with_ref()
let raw: *mut T = reference as *const _ as *mut _ ;
let arc = unsafe { Box::from_raw(raw) };
```
Regards,
Boqun
> Regards,
> Boqun
>
>
> > Miri would complain about the above code.
> >
> > --
> > Cheers,
> > Benno
> >
> >
next prev parent reply other threads:[~2023-09-25 23:24 UTC|newest]
Thread overview: 41+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-09-23 14:49 [PATCH v2 0/2] Remove `ArcBorrow` Wedson Almeida Filho
2023-09-23 14:49 ` [PATCH v2 1/2] rust: arc: rename `ArcInner` to `WithRef` Wedson Almeida Filho
2023-09-23 19:31 ` Martin Rodriguez Reboredo
2023-09-24 11:59 ` Benno Lossin
2023-09-24 13:41 ` Jianguo Bao
2023-09-25 6:21 ` Alice Ryhl
2023-09-23 14:49 ` [PATCH v2 2/2] rust: arc: remove `ArcBorrow` in favour of `WithRef` Wedson Almeida Filho
2023-09-23 19:32 ` Martin Rodriguez Reboredo
2023-09-24 11:59 ` Benno Lossin
2023-09-24 13:36 ` Jianguo Bao
2023-09-25 6:29 ` Alice Ryhl
2023-09-25 9:14 ` Benno Lossin
2023-09-25 14:49 ` Boqun Feng
2023-09-25 15:00 ` Alice Ryhl
2023-09-25 15:17 ` Boqun Feng
2023-09-25 15:30 ` Alice Ryhl
2023-09-25 16:02 ` Boqun Feng
2023-09-25 16:11 ` Benno Lossin
2023-09-25 15:07 ` Benno Lossin
2023-09-25 16:16 ` Boqun Feng
2023-09-25 17:00 ` Benno Lossin
2023-09-25 18:51 ` Boqun Feng
2023-09-25 21:03 ` Benno Lossin
2023-09-25 21:55 ` Boqun Feng
2023-09-25 21:58 ` Alice Ryhl
2023-09-25 22:02 ` Boqun Feng
2023-09-25 22:06 ` Boqun Feng
2023-09-25 22:26 ` Benno Lossin
2023-09-25 22:34 ` Boqun Feng
2023-09-25 23:24 ` Boqun Feng [this message]
2023-09-26 8:26 ` Gary Guo
2023-09-26 15:24 ` Boqun Feng
2023-09-26 15:41 ` Alice Ryhl
2023-09-26 16:35 ` Boqun Feng
2023-09-26 17:15 ` Benno Lossin
2023-09-26 17:43 ` Boqun Feng
2023-09-26 18:26 ` Benno Lossin
2023-09-26 21:31 ` Alice Ryhl
2023-09-26 18:20 ` Boqun Feng
2023-09-26 21:27 ` Alice Ryhl
2023-09-25 15:04 ` Alice Ryhl
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=ZRIWmPa0mN6s7Yoj@boqun-archlinux \
--to=boqun.feng@gmail.com \
--cc=a.hindborg@samsung.com \
--cc=alex.gaynor@gmail.com \
--cc=alice@ryhl.io \
--cc=aliceryhl@google.com \
--cc=benno.lossin@proton.me \
--cc=bjorn3_gh@protonmail.com \
--cc=gary@garyguo.net \
--cc=linux-kernel@vger.kernel.org \
--cc=ojeda@kernel.org \
--cc=rust-for-linux@vger.kernel.org \
--cc=walmeida@microsoft.com \
--cc=wedsonaf@gmail.com \
/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).