* Re: [PATCH v6] rust: aref: make `AlwaysRefCounted::inc_ref` an associated function
From: Gary Guo @ 2026-06-29 14:37 UTC (permalink / raw)
To: Trevor Chan, gregkh, rafael, dakr, ojeda, a.hindborg, paul,
aliceryhl, airlied, simona, viro, brauner, igor.korotin, vireshk,
nm, sboyd, m.wilczynski, boqun, gary, axboe, daniel.almeida,
shankari.ak0208, lyude, j, lossin, acourbot, markus.probst,
driver-core, rust-for-linux, linux-kernel, linux-block,
linux-security-module, dri-devel, linux-fsdevel, linux-mm,
linux-pm, linux-pci, linux-pwm
Cc: david.m.ertman, iweiny, leon, bjorn3_gh, tmgross, tamird, work,
sergeh, matthew.brost, thomas.hellstrom, jack, ljs, liam,
bhelgaas, kwilczynski, ptikhomirov
In-Reply-To: <20260628095132.47753-1-trev@trevrosa.dev>
On Sun Jun 28, 2026 at 10:51 AM BST, Trevor Chan wrote:
> `AlwaysRefCounted::inc_ref` is a function that shouldn't be called lightly.
>
> To prevent accidentally calling it, change `inc_ref` to be an associated function.
>
> Modify all `AlwaysRefCounted` implementors to work with this change.
>
> Suggested-by: Benno Lossin <lossin@kernel.org>
> Link: https://github.com/Rust-for-Linux/linux/issues/1177
> Reviewed-by: Alice Ryhl <aliceryhl@google.com>
> Signed-off-by: Trevor Chan <trev@trevrosa.dev>
Reviewed-by: Gary Guo <gary@garyguo.net>
> ---
> Changes in v2:
> - Don't word wrap the patch
> Changes in v3:
> - Make argument name of `Empty::inc_ref` consistent with `Empty::dec_ref`
> Changes in v4:
> - Rebase to new rust-next, change new implementors
> - Reword explanation for change in `AlwaysRefCounted::inc_ref` doc comment
> Changes in v5:
> - Change commit message to be imperative
> Changes in v6:
> - Change all the implementors
> ---
> rust/kernel/auxiliary.rs | 4 ++--
> rust/kernel/block/mq/request.rs | 4 ++--
> rust/kernel/cred.rs | 4 ++--
> rust/kernel/device.rs | 4 ++--
> rust/kernel/device/property.rs | 4 ++--
> rust/kernel/drm/device.rs | 4 ++--
> rust/kernel/drm/gem/mod.rs | 4 ++--
> rust/kernel/drm/gpuvm/mod.rs | 4 ++--
> rust/kernel/drm/gpuvm/vm_bo.rs | 4 ++--
> rust/kernel/fs/file.rs | 8 ++++----
> rust/kernel/i2c.rs | 8 ++++----
> rust/kernel/mm.rs | 8 ++++----
> rust/kernel/mm/mmput_async.rs | 4 ++--
> rust/kernel/opp.rs | 4 ++--
> rust/kernel/pci.rs | 4 ++--
> rust/kernel/pid_namespace.rs | 4 ++--
> rust/kernel/platform.rs | 4 ++--
> rust/kernel/pwm.rs | 2 +-
> rust/kernel/sync/aref.rs | 11 +++++++----
> rust/kernel/task.rs | 4 ++--
> rust/kernel/usb.rs | 8 ++++----
> 21 files changed, 54 insertions(+), 51 deletions(-)
^ permalink raw reply
* Re: [PATCH stable/linux-5.10.y 0/7] Backport Fix incorrect overlayfs mmap() and mprotect() LSM access controls
From: Amir Goldstein @ 2026-06-29 17:31 UTC (permalink / raw)
To: Cai Xinchen
Cc: viro, brauner, jack, miklos, paul, jmorris, serge,
stephen.smalley.work, omosnace, gregkh, sashal, bboscaccy,
linux-fsdevel, linux-kernel, linux-unionfs, linux-security-module,
selinux, bpf, stable, lujialin4
In-Reply-To: <20260629070653.580879-1-caixinchen1@huawei.com>
On Mon, Jun 29, 2026 at 8:38 AM Cai Xinchen <caixinchen1@huawei.com> wrote:
>
> ackport the patch series
> "Fix incorrect overlayfs mmap() and mprotect() LSM access controls" [1]
> to 5.10 lts
Chai,
First of all, I don't think that stable maintainers are picking backports
to 5.10 that were not backported to 6.1 and 5.15.
Second, backporting backing_file as a dependency to LTS kernels is a pretty
intrusive change, so your description above is very much lacking.
Please do not backport backing_file to any of the LTS kernels without providing
detailed explanation to try and convince the vfs maintainers that you
verified this
bacport is safe for the LTS kernel, because honestly, this looks a bit
risky for me.
Thanks,
Amir.
^ permalink raw reply
page: | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox