kernel/rust, branch linux-6.11.y

rust: block: fix formatting of `kernel::block::mq::request` module

2024-12-05T12:53:48Z

[ Upstream commit 28e848386b92645f93b9f2fdba5882c3ca7fb3e2 ] Fix several issues with rustdoc formatting for the `kernel::block::mq::Request` module, in particular: - An ordered list not rendering correctly, fixed by using numbers prefixes instead of letters. - Code snippets formatted as regular text, fixed by wrapping the code with `back-ticks`. - References to types missing intra-doc links, fixed by wrapping the types with [square brackets]. Reported-by: Miguel Ojeda Closes: https://github.com/Rust-for-Linux/linux/issues/1108 Signed-off-by: Francesco Zardi Acked-by: Andreas Hindborg Fixes: 3253aba3408a ("rust: block: introduce `kernel::block::mq` module") Link: https://lore.kernel.org/r/20240903173027.16732-3-frazar00@gmail.com [ Added an extra intra-doc link. Took the chance to add some periods for consistency. Reworded slightly. - Miguel ] Signed-off-by: Miguel Ojeda Signed-off-by: Sasha Levin

rust: macros: fix documentation of the paste! macro

2024-12-05T12:53:47Z

[ Upstream commit 15541c9263ce34ff95a06bc68f45d9bc5c990bcd ] One of the example in this section uses a curious mix of the constant and function declaration syntaxes; fix it. Signed-off-by: Paolo Bonzini Reviewed-by: Alice Ryhl Fixes: 823d4737d4c2 ("rust: macros: add `paste!` proc macro") Link: https://lore.kernel.org/r/20241019072208.1016707-1-pbonzini@redhat.com Signed-off-by: Miguel Ojeda Signed-off-by: Sasha Levin

rust: kernel: fix THIS_MODULE header path in ThisModule doc comment

2024-12-05T12:53:47Z

[ Upstream commit 8b55dc8610acf816a66373be53ca6e3bbe2d313a ] The doc comment for `ThisModule` incorrectly states the C header file for `THIS_MODULE` as `include/linux/export.h`, while the correct path is `include/linux/init.h`. This is because `THIS_MODULE` was moved in commit 5b20755b7780 ("init: move THIS_MODULE from to "). Update the doc comment for `ThisModule` to reflect the correct header file path for `THIS_MODULE`. Fixes: 5b20755b7780 ("init: move THIS_MODULE from to ") Signed-off-by: Yutaro Ohno Reviewed-by: Greg Kroah-Hartman Link: https://lore.kernel.org/r/ZxXDZwxWgoEiIYkj@ohnotp Signed-off-by: Miguel Ojeda Signed-off-by: Sasha Levin

rust: device: change the from_raw() function

2024-11-08T15:30:51Z

[ Upstream commit cc4332afb5631b0e9d2ce5699b7f4b7caf743526 ] The function Device::from_raw() increments a refcount by a call to bindings::get_device(ptr). This can be confused because usually from_raw() functions don't increment a refcount. Hence, rename Device::from_raw() to avoid confuion with other "from_raw" semantics. The new name of function should be "get_device" to be consistent with the function get_device() already exist in .c files. This function body also changed, because the `into()` will convert the `&'a Device` into `ARef` and also call `inc_ref` from the `AlwaysRefCounted` trait implemented for Device. Signed-off-by: Guilherme Giacomo Simoes Acked-by: Danilo Krummrich Closes: https://github.com/Rust-for-Linux/linux/issues/1088 Reviewed-by: Boqun Feng Link: https://lore.kernel.org/r/20241001205603.106278-1-trintaeoitogc@gmail.com Signed-off-by: Greg Kroah-Hartman Signed-off-by: Sasha Levin

rust: sync: require `T: Sync` for `LockedBy::access`

2024-10-10T10:03:47Z

commit a8ee30f45d5d57467ddb7877ed6914d0eba0af7f upstream. The `LockedBy::access` method only requires a shared reference to the owner, so if we have shared access to the `LockedBy` from several threads at once, then two threads could call `access` in parallel and both obtain a shared reference to the inner value. Thus, require that `T: Sync` when calling the `access` method. An alternative is to require `T: Sync` in the `impl Sync for LockedBy`. This patch does not choose that approach as it gives up the ability to use `LockedBy` with `!Sync` types, which is okay as long as you only use `access_mut`. Cc: stable@vger.kernel.org Fixes: 7b1f55e3a984 ("rust: sync: introduce `LockedBy`") Signed-off-by: Alice Ryhl Suggested-by: Boqun Feng Reviewed-by: Gary Guo Link: https://lore.kernel.org/r/20240915-locked-by-sync-fix-v2-1-1a8d89710392@google.com Signed-off-by: Miguel Ojeda Signed-off-by: Greg Kroah-Hartman

rust: mutex: fix __mutex_init() usage in case of PREEMPT_RT

2024-10-10T10:03:03Z

[ Upstream commit d065cc76054d21e48a839a2a19ba99dbc51a4d11 ] In case CONFIG_PREEMPT_RT is enabled __mutex_init() becomes a macro instead of an extern function (simplified from include/linux/mutex.h): #ifndef CONFIG_PREEMPT_RT extern void __mutex_init(struct mutex *lock, const char *name, struct lock_class_key *key); #else #define __mutex_init(mutex, name, key) \ do { \ rt_mutex_base_init(&(mutex)->rtmutex); \ __mutex_rt_init((mutex), name, key); \ } while (0) #endif The macro isn't resolved by bindgen, then. What results in a build error: error[E0425]: cannot find function `__mutex_init` in crate `bindings` --> rust/kernel/sync/lock/mutex.rs:104:28 | 104 | unsafe { bindings::__mutex_init(ptr, name, key) } | ^^^^^^^^^^^^ help: a function with a similar name exists: `__mutex_rt_init` | ::: rust/bindings/bindings_generated.rs:23722:5 | 23722 | / pub fn __mutex_rt_init( 23723 | | lock: *mut mutex, 23724 | | name: *const core::ffi::c_char, 23725 | | key: *mut lock_class_key, 23726 | | ); | |_____- similarly named function `__mutex_rt_init` defined here Fix this by adding a helper. As explained by Gary Guo in [1] no #ifdef CONFIG_PREEMPT_RT is needed here as rust/bindings/lib.rs prefers externed function to helpers if an externed function exists. Reported-by: Conor Dooley Link: https://lore.kernel.org/rust-for-linux/20240913-shack-estate-b376a65921b1@spud/ Link: https://lore.kernel.org/rust-for-linux/20240915123626.1a170103.gary@garyguo.net/ [1] Fixes: 6d20d629c6d8 ("rust: lock: introduce `Mutex`") Signed-off-by: Dirk Behme Tested-by: Conor Dooley Reviewed-by: Gary Guo Link: https://lore.kernel.org/r/20240916073752.3123484-1-dirk.behme@de.bosch.com [ Reworded to include the proper example by Dirk. - Miguel ] Signed-off-by: Miguel Ojeda Signed-off-by: Sasha Levin

rust: kbuild: auto generate helper exports

2024-10-10T10:03:02Z

[ Upstream commit e26fa546042add70944d018b930530d16b3cf626 ] This removes the need to explicitly export all symbols. Generate helper exports similarly to what's currently done for Rust crates. These helpers are exclusively called from within Rust code and therefore can be treated similar as other Rust symbols. Signed-off-by: Gary Guo Reviewed-by: Boqun Feng Tested-by: Boqun Feng Link: https://lore.kernel.org/r/20240817165302.3852499-1-gary@garyguo.net [ Fixed dependency path, reworded slightly, edited comment a bit and rebased on top of the changes made when applying Andreas' patch (e.g. no `README.md` anymore, so moved the edits). - Miguel ] Signed-off-by: Miguel Ojeda Stable-dep-of: d065cc76054d ("rust: mutex: fix __mutex_init() usage in case of PREEMPT_RT") Signed-off-by: Sasha Levin

rust: kbuild: split up helpers.c

2024-10-10T10:03:02Z

[ Upstream commit 876346536c1b59a5b1b5e44477b1b3ece77647fd ] This patch splits up the rust helpers C file. When rebasing patch sets on upstream linux, merge conflicts in helpers.c is common and time consuming [1]. Thus, split the file so that each kernel component can live in a separate file. This patch lists helper files explicitly and thus conflicts in the file list is still likely. However, they should be more simple to resolve than the conflicts usually seen in helpers.c. [ Removed `README.md` and undeleted the original comment since now, in v3 of the series, we have a `helpers.c` again; which also allows us to keep the "Sorted alphabetically" line and makes the diff easier. In addition, updated the Documentation/ mentions of the file, reworded title and removed blank lines at the end of `page.c`. - Miguel ] Link: https://rust-for-linux.zulipchat.com/#narrow/stream/288089-General/topic/Splitting.20up.20helpers.2Ec/near/426694012 [1] Signed-off-by: Andreas Hindborg Reviewed-by: Gary Guo Acked-by: Dirk Behme Reviewed-by: Alice Ryhl Reviewed-by: Benno Lossin Link: https://lore.kernel.org/r/20240815103016.2771842-1-nmi@metaspace.dk Signed-off-by: Miguel Ojeda Stable-dep-of: d065cc76054d ("rust: mutex: fix __mutex_init() usage in case of PREEMPT_RT") Signed-off-by: Sasha Levin

rust: macros: provide correct provenance when constructing THIS_MODULE

2024-09-02T07:14:28Z

Currently while defining `THIS_MODULE` symbol in `module!()`, the pointer used to construct `ThisModule` is derived from an immutable reference of `__this_module`, which means the pointer doesn't have the provenance for writing, and that means any write to that pointer is UB regardless of data races or not. However, the usage of `THIS_MODULE` includes passing this pointer to functions that may write to it (probably in unsafe code), and this will create soundness issues. One way to fix this is using `addr_of_mut!()` but that requires the unstable feature "const_mut_refs". So instead of `addr_of_mut()!`, an extern static `Opaque` is used here: since `Opaque` is transparent to `T`, an extern static `Opaque` will just wrap the C symbol (defined in a C compile unit) in an `Opaque`, which provides a pointer with writable provenance via `Opaque::get()`. This fix the potential UBs because of pointer provenance unmatched. Reported-by: Alice Ryhl Signed-off-by: Boqun Feng Reviewed-by: Alice Ryhl Reviewed-by: Trevor Gross Reviewed-by: Benno Lossin Reviewed-by: Gary Guo Closes: https://rust-for-linux.zulipchat.com/#narrow/stream/x/topic/x/near/465412664 Fixes: 1fbde52bde73 ("rust: add `macros` crate") Cc: stable@vger.kernel.org # 6.6.x: be2ca1e03965: ("rust: types: Make Opaque::get const") Link: https://lore.kernel.org/r/20240828180129.4046355-1-boqun.feng@gmail.com [ Fixed two typos, reworded title. - Miguel ] Signed-off-by: Miguel Ojeda

rust: alloc: eschew `Box>::write`

2024-08-26T22:07:05Z

Upstream Rust's libs-api team has consensus for stabilizing some of `feature(new_uninit)`, but not for `Box>::write`. Instead, we can use `MaybeUninit::write`, so Rust for Linux can drop the feature after stabilization. That will happen after merging, as the FCP has completed [1]. This is required before stabilization because remaining-unstable API will be divided into new features. This code doesn't know about those yet. It can't: they haven't landed, as the relevant PR is blocked on rustc's CI testing Rust-for-Linux without this patch. [ The PR has landed [2] and will be released in Rust 1.82.0 (expected on 2024-10-17), so we could conditionally enable the new unstable feature (`box_uninit_write` [3]) instead, but just for a single `unsafe` block it is probably not worth it. For the time being, I added it to the "nice to have" section of our unstable features list. - Miguel ] Link: https://github.com/rust-lang/rust/issues/63291#issuecomment-2183022955 [1] Link: https://github.com/rust-lang/rust/pull/129416 [2] Link: https://github.com/rust-lang/rust/issues/129397 [3] Signed-off-by: Jubilee Young Reviewed-by: Alice Ryhl Reviewed-by: Trevor Gross [ Reworded slightly. - Miguel ] Signed-off-by: Miguel Ojeda