Linux PCI subsystem development
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: "Gary Guo" <gary@garyguo.net>
Cc: dri-devel@lists.freedesktop.org,
	"Alexandre Courbot" <acourbot@nvidia.com>,
	ojeda@kernel.org, linux-pci@vger.kernel.org,
	"Danilo Krummrich" <dakr@kernel.org>
Subject: Re: [PATCH v6 07/20] rust: io: implement `Mmio` as view type
Date: Mon, 06 Jul 2026 12:59:27 +0000	[thread overview]
Message-ID: <20260706125927.DEF411F00A3A@smtp.kernel.org> (raw)
In-Reply-To: <20260706-io_projection-v6-7-72cd5d055d54@garyguo.net>

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Infallible 64-bit I/O accessors (`read64`, `write64`) and typed 64-bit registers cannot be used on MMIO regions, and custom view types are unsupported by standard offsets.
--

commit 09225b49c257719774237bd1b25407607dd9df87
Author: Gary Guo <gary@garyguo.net>

rust: io: implement Mmio as view type

Implement Mmio as view type and convert RelaxedMmio to view type as
well. I/O implementations of MmioOwned are changed to delegate to the
Mmio view type.

> diff --git a/rust/kernel/io.rs b/rust/kernel/io.rs
> index f93be7f780694..8110b49aa4307 100644
> --- a/rust/kernel/io.rs
> +++ b/rust/kernel/io.rs
[ ... ]
> @@ -847,17 +915,12 @@ unsafe fn io_write(self, value: $ty, address: usize) {
>  }
>  
>  // MMIO regions support 8, 16, and 32-bit accesses.
> -impl_mmio_io_capable!(MmioOwned, u8, readb, writeb);
> -impl_mmio_io_capable!(MmioOwned, u16, readw, writew);
> -impl_mmio_io_capable!(MmioOwned, u32, readl, writel);
> +impl_mmio_io_capable!(Mmio, u8, readb, writeb);
> +impl_mmio_io_capable!(Mmio, u16, readw, writew);
> +impl_mmio_io_capable!(Mmio, u32, readl, writel);
>  // MMIO regions on 64-bit systems also support 64-bit accesses.
> -impl_mmio_io_capable!(
> -    MmioOwned,
> -    #[cfg(CONFIG_64BIT)]
> -    u64,
> -    readq,
> -    writeq
> -);
> +#[cfg(CONFIG_64BIT)]
> +impl_mmio_io_capable!(Mmio, u64, readq, writeq);

[Severity: High]
Does this actually allow infallible 64-bit I/O accesses when the target is
Region<SIZE>?

Region<SIZE> has a MIN_ALIGN of 4, so infallible 64-bit accessors like
read64() and write64() will fail the compile-time assertion in
io_view_assert():

    const_assert!(
        Alignment::of::<U>().as_usize() <= IO::Target::MIN_ALIGN.as_usize()
    );

For U = u64, the alignment is 8 on 64-bit architectures, which fails the
assertion since 8 <= 4 is false.

Additionally, while this patch introduces Mmio as a generic view type, can
users create a custom target type (e.g. with align(8)) to work around this?

Looking at impl_usize_ioloc in rust/kernel/io.rs and the IoLoc implementation
in rust/kernel/io/register.rs, both usize offsets and typed registers are
strictly implemented only for IoLoc<Region<SIZE>, T>.

Because standard offsets and registers are restricted to Region<SIZE>, a custom
view type cannot be used with them. Does this leave 64-bit infallible accesses
and 64-bit typed registers unusable on MMIO regions?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260706-io_projection-v6-0-72cd5d055d54@garyguo.net?part=7

  reply	other threads:[~2026-07-06 12:59 UTC|newest]

Thread overview: 41+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-07-06 12:44 [PATCH v6 00/20] rust: I/O type generalization and projection Gary Guo
2026-07-06 12:44 ` [PATCH v6 01/20] rust: io: add dynamically-sized `Region` type Gary Guo
2026-07-06 12:54   ` sashiko-bot
2026-07-06 12:44 ` [PATCH v6 02/20] rust: io: add missing safety requirement in `IoCapable` methods Gary Guo
2026-07-06 13:00   ` sashiko-bot
2026-07-06 12:44 ` [PATCH v6 03/20] rust: io: restrict untyped IO access and `register!` to `Region` Gary Guo
2026-07-06 13:03   ` sashiko-bot
2026-07-06 12:44 ` [PATCH v6 04/20] rust: io: implement `Io` on reference types instead Gary Guo
2026-07-06 12:53   ` sashiko-bot
2026-07-06 12:44 ` [PATCH v6 05/20] rust: io: generalize `MmioRaw` to pointer to arbitrary type Gary Guo
2026-07-06 12:56   ` sashiko-bot
2026-07-06 12:44 ` [PATCH v6 06/20] rust: io: rename `Mmio` to `MmioOwned` Gary Guo
2026-07-06 12:57   ` sashiko-bot
2026-07-06 12:44 ` [PATCH v6 07/20] rust: io: implement `Mmio` as view type Gary Guo
2026-07-06 12:59   ` sashiko-bot [this message]
2026-07-06 12:44 ` [PATCH v6 08/20] rust: pci: io: make `ConfigSpace` a view Gary Guo
2026-07-06 12:51   ` sashiko-bot
2026-07-06 12:44 ` [PATCH v6 09/20] rust: io: use view types instead of addresses for `Io` Gary Guo
2026-07-06 12:57   ` sashiko-bot
2026-07-06 12:44 ` [PATCH v6 10/20] pwm: th1520: remove unnecessary `deref` Gary Guo
2026-07-06 12:48   ` sashiko-bot
2026-07-06 12:44 ` [PATCH v6 11/20] rust: io: remove `MmioOwned` Gary Guo
2026-07-06 12:54   ` sashiko-bot
2026-07-06 12:44 ` [PATCH v6 12/20] rust: io: move `Io` methods to extension trait Gary Guo
2026-07-06 12:51   ` sashiko-bot
2026-07-06 12:44 ` [PATCH v6 13/20] rust: io: add projection macro and methods Gary Guo
2026-07-06 12:58   ` sashiko-bot
2026-07-06 12:44 ` [PATCH v6 14/20] rust: io: add I/O backend for system memory with volatile access Gary Guo
2026-07-06 13:01   ` sashiko-bot
2026-07-06 12:44 ` [PATCH v6 15/20] rust: io: implement a view type for `Coherent` Gary Guo
2026-07-06 13:03   ` sashiko-bot
2026-07-06 12:44 ` [PATCH v6 16/20] rust: io: add `read_val` and `write_val` functions on `Io` Gary Guo
2026-07-06 12:59   ` sashiko-bot
2026-07-06 12:44 ` [PATCH v6 17/20] gpu: nova-core: use I/O projection for cleaner encapsulation Gary Guo
2026-07-06 13:07   ` sashiko-bot
2026-07-06 12:44 ` [PATCH v6 18/20] rust: dma: drop `dma_read!` and `dma_write!` API Gary Guo
2026-07-06 13:10   ` sashiko-bot
2026-07-06 12:44 ` [PATCH v6 19/20] rust: io: add copying methods Gary Guo
2026-07-06 13:09   ` sashiko-bot
2026-07-06 12:44 ` [PATCH v6 20/20] rust: io: implement `IoSysMap` Gary Guo
2026-07-06 12:59   ` sashiko-bot

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=20260706125927.DEF411F00A3A@smtp.kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=acourbot@nvidia.com \
    --cc=dakr@kernel.org \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=gary@garyguo.net \
    --cc=linux-pci@vger.kernel.org \
    --cc=ojeda@kernel.org \
    --cc=sashiko-reviews@lists.linux.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