From: Benno Lossin <lossin@kernel.org>
To: gary@garyguo.net
Cc: a.hindborg@kernel.org, acourbot@nvidia.com, aliceryhl@google.com,
bjorn3_gh@protonmail.com, boqun@kernel.org, dakr@kernel.org,
daniel.almeida@collabora.com, linux-kernel@vger.kernel.org,
lossin@kernel.org, ojeda@kernel.org,
rust-for-linux@vger.kernel.org, tamird@kernel.org,
tmgross@umich.edu, work@onurozkan.dev
Subject: Re: [PATCH 00/20] rust: pin-init: create self references safely
Date: Thu, 8 Oct 2026 17:20:31 +0100 [thread overview]
Message-ID: <20261008162031.158761-1-lossin@kernel.org> (raw)
In-Reply-To: <20261008-dev-selfref-v1-0-6c1eb269fe57@garyguo.net>
On Thu, 08 Oct 2026 14:23:47 +0200, Gary Guo <gary@garyguo.net> wrote:
> This is a big series that add support for one field to reference a sibling
> field in a safe and ergnonomic way. Unlike many other crates in the
> userspace Rust ecosystem, no additional allocation is required, thus it
> requires the struct to be pinned, which is exactly what pin-init provides.
>
> This is a very powerful feature, and thus can raise concerns about whether
> it is sound. I spent a lot of time studying the rules to make this sound,
> and presented durirng Kangrejos [1].
>
> The simple use case looks like this:
>
> #[pin_data]
> struct MyDriver<'bound> {
> dev: &'bound Device,
> #[pin]
> irq: irq::Registration<'bound, MyIrqHandler<'bound, 'bar>>,
> bar: Bar<'bound, BAR_SIZE>,
> }
>
> You simply need to mention a lifetime that shares the name with the field.
> There are more advanced use cases which require annotations like
>
> #[uses('bar: invariant)]
>
> to explicitly declare that the lifetime is used in invariant manner, and
> also
>
> where
> exists<'dev>: 'a
>
> to create new existential lifetimes as a way to erase invariant lifetimes
> that would otherwise bubble up to users. More info can be seen in my
> Plumbers slide [2]. Note that the syntax for explicit variance annotation
> has changed following discussions during Plumbers.
>
> The initial plan was to upstream the simple use case only and iron out the
> advanced features subsequently; however it turns out that DRM jobqueue
> would need the variance annotation feature, and Nova `Cmdq` would need to
> use the existential lifetime feature.
>
> During Plumbers the upstream schedule is discussed, and instead it was
> agreed that all the features should be upstreamed at once, but the usage of
> the advanced feature usage limited to the pre-agreed users only to limit
> the blast radius in case the feature needs to be reworked.
>
> Detailed documentation about the pin-init self-reference feature would be
> added the following cycle, when the usage of them become more clear.
>
> The pull request on GitHub [3] has a bunch of test suites, which are not
> synchronized to kernel tree.
>
> Link: https://kangrejos.com/2026/Self%20referential%20pin-init.pdf [1]
> Link: https://lpc.events/event/20/contributions/2498/attachments/2200/4855/presentation.pdf [2]
> Link: https://github.com/Rust-for-Linux/pin-init/pull/181 [3]
> Signed-off-by: Gary Guo <gary@garyguo.net>
> ---
> Gary Guo (20):
> kbuild: rust: allow `clippy::comparison_chain` globally
> rust: pin-init: internal: pin_data: infer self-referential struct
> rust: pin-init: internal: pin_data: rewrite fields that borrow others
> rust: pin-init: internal: pin_data: pin borrowed fields with wrapper
> rust: pin-init: internal: pin_data: teach drop check about generics that cannot dangle
> rust: pin-init: internal: pin_data: self-referential drop order checks
> rust: pin-init: internal: pin_data: check covariance of self-referential fields
> rust: pin-init: internal: pin_data: implement initialization of borrowed structs
> rust: pin-init: internal: pin_data: project self-referential fields
> rust: pin-init: internal: pin_data: add `with_project` method
> rust: pin-init: internal: pin_data: enable self-referential support
> rust: pin-init: internal: pin_data: allow lifetime to be shortened per field drop order
> rust: pin-init: internal: pin_data: parse explicit `#[borrowed]` annotation
> rust: pin-init: internal: pin_data: support mutable borrows
> rust: pin-init: internal: pin_data: parse explicit `#[uses]` annotation
> rust: pin-init: internal: pin_data: make field lifetime invariance imply type invariance
> rust: pin-init: internal: pin_data: complete invariant borrow support
> rust: pin-init: internal: pin_data: perform AST lifetime replacement if possible
> rust: pin-init: internal: pin_data: support shared projection
> rust: pin-init: internal: pin_data: support existential lifetimes
>
> Makefile | 2 +
> rust/pin-init/examples/selfref.rs | 60 ++
> rust/pin-init/internal/src/init.rs | 42 +-
> rust/pin-init/internal/src/pin_data.rs | 1556 +++++++++++++++++++++++++++++++-
> rust/pin-init/internal/src/util.rs | 268 +++++-
> rust/pin-init/src/__internal.rs | 357 +++++++-
> rust/pin-init/src/lib.rs | 1 +
> 7 files changed, 2238 insertions(+), 48 deletions(-)
incredible work, Gary! when you first told me about this, I didn't
imagine that it would look this clean from the user's side. I sadly am
unable to take a detailed look at all of the patches. but we've talked
extensively about the design; so I'm sure we'll be able to fix all bugs
that pop up:
Acked-by: Benno Lossin <lossin@kernel.org>
just for the record: I think that users of this feature should come
gradually. that way, fixing any issues (if there are any :) will be much
less stressful, since there are fewer places to fix. we've discussed
this and also mentioned it to Miguel. additionally, it would be great if
users would take a careful stance when using this feature. I'm not
worried, just cautious, since an unsoundness hole here could end up
being extremely ugly to fix and require re-architecting the users.
once again, great work!
best,
Benno
> ---
> base-commit: 0d9aa4b994379c6e0099737221ed421deb7e1a31
> change-id: 20261006-dev-selfref-27c37abfa849
>
> Best regards,
> --
> Gary Guo <gary@garyguo.net>
prev parent reply other threads:[~2026-10-08 16:20 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-10-08 12:23 Gary Guo
2026-10-08 12:23 ` [PATCH 01/20] kbuild: rust: allow `clippy::comparison_chain` globally Gary Guo
2026-10-08 12:23 ` [PATCH 02/20] rust: pin-init: internal: pin_data: infer self-referential struct Gary Guo
2026-10-08 12:23 ` [PATCH 03/20] rust: pin-init: internal: pin_data: rewrite fields that borrow others Gary Guo
2026-10-08 12:23 ` [PATCH 04/20] rust: pin-init: internal: pin_data: pin borrowed fields with wrapper Gary Guo
2026-10-08 12:23 ` [PATCH 05/20] rust: pin-init: internal: pin_data: teach drop check about generics that cannot dangle Gary Guo
2026-10-08 12:23 ` [PATCH 06/20] rust: pin-init: internal: pin_data: self-referential drop order checks Gary Guo
2026-10-08 12:23 ` [PATCH 07/20] rust: pin-init: internal: pin_data: check covariance of self-referential fields Gary Guo
2026-10-08 12:23 ` [PATCH 08/20] rust: pin-init: internal: pin_data: implement initialization of borrowed structs Gary Guo
2026-10-08 12:23 ` [PATCH 09/20] rust: pin-init: internal: pin_data: project self-referential fields Gary Guo
2026-10-08 12:23 ` [PATCH 10/20] rust: pin-init: internal: pin_data: add `with_project` method Gary Guo
2026-10-08 12:23 ` [PATCH 11/20] rust: pin-init: internal: pin_data: enable self-referential support Gary Guo
2026-10-08 12:23 ` [PATCH 12/20] rust: pin-init: internal: pin_data: allow lifetime to be shortened per field drop order Gary Guo
2026-10-08 12:24 ` [PATCH 13/20] rust: pin-init: internal: pin_data: parse explicit `#[borrowed]` annotation Gary Guo
2026-10-08 12:24 ` [PATCH 14/20] rust: pin-init: internal: pin_data: support mutable borrows Gary Guo
2026-10-08 12:24 ` [PATCH 15/20] rust: pin-init: internal: pin_data: parse explicit `#[uses]` annotation Gary Guo
2026-10-08 12:24 ` [PATCH 16/20] rust: pin-init: internal: pin_data: make field lifetime invariance imply type invariance Gary Guo
2026-10-08 12:24 ` [PATCH 17/20] rust: pin-init: internal: pin_data: complete invariant borrow support Gary Guo
2026-10-08 12:24 ` [PATCH 18/20] rust: pin-init: internal: pin_data: perform AST lifetime replacement if possible Gary Guo
2026-10-08 12:24 ` [PATCH 19/20] rust: pin-init: internal: pin_data: support shared projection Gary Guo
2026-10-08 12:24 ` [PATCH 20/20] rust: pin-init: internal: pin_data: support existential lifetimes Gary Guo
2026-10-08 16:20 ` Benno Lossin [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=20261008162031.158761-1-lossin@kernel.org \
--to=lossin@kernel.org \
--cc=a.hindborg@kernel.org \
--cc=acourbot@nvidia.com \
--cc=aliceryhl@google.com \
--cc=bjorn3_gh@protonmail.com \
--cc=boqun@kernel.org \
--cc=dakr@kernel.org \
--cc=daniel.almeida@collabora.com \
--cc=gary@garyguo.net \
--cc=linux-kernel@vger.kernel.org \
--cc=ojeda@kernel.org \
--cc=rust-for-linux@vger.kernel.org \
--cc=tamird@kernel.org \
--cc=tmgross@umich.edu \
--cc=work@onurozkan.dev \
/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
all inboxes | Powered by JetHome®