projects/linux·blob v6.13 · refs

380 lines · 15.0 KiB

1// SPDX-License-Identifier: GPL-2.0
2
3//! Tasks (threads and processes).
4//!
5//! C header: [`include/linux/sched.h`](srctree/include/linux/sched.h).
6
7use crate::{
8 bindings,
9 ffi::{c_int, c_long, c_uint},
10 pid_namespace::PidNamespace,
11 types::{ARef, NotThreadSafe, Opaque},
12};
13use core::{
14 cmp::{Eq, PartialEq},
15 ops::Deref,
16 ptr,
17};
18
19/// A sentinel value used for infinite timeouts.
20pub const MAX_SCHEDULE_TIMEOUT: c_long = c_long::MAX;
21
22/// Bitmask for tasks that are sleeping in an interruptible state.
23pub const TASK_INTERRUPTIBLE: c_int = bindings::TASK_INTERRUPTIBLE as c_int;
24/// Bitmask for tasks that are sleeping in an uninterruptible state.
25pub const TASK_UNINTERRUPTIBLE: c_int = bindings::TASK_UNINTERRUPTIBLE as c_int;
26/// Convenience constant for waking up tasks regardless of whether they are in interruptible or
27/// uninterruptible sleep.
28pub const TASK_NORMAL: c_uint = bindings::TASK_NORMAL as c_uint;
29
30/// Returns the currently running task.
31#[macro_export]
32macro_rules! current {
33 () => {
34 // SAFETY: Deref + addr-of below create a temporary `TaskRef` that cannot outlive the
35 // caller.
36 unsafe { &*$crate::task::Task::current() }
37 };
38}
39
40/// Returns the currently running task's pid namespace.
41#[macro_export]
42macro_rules! current_pid_ns {
43 () => {
44 // SAFETY: Deref + addr-of below create a temporary `PidNamespaceRef` that cannot outlive
45 // the caller.
46 unsafe { &*$crate::task::Task::current_pid_ns() }
47 };
48}
49
50/// Wraps the kernel's `struct task_struct`.
51///
52/// # Invariants
53///
54/// All instances are valid tasks created by the C portion of the kernel.
55///
56/// Instances of this type are always refcounted, that is, a call to `get_task_struct` ensures
57/// that the allocation remains valid at least until the matching call to `put_task_struct`.
58///
59/// # Examples
60///
61/// The following is an example of getting the PID of the current thread with zero additional cost
62/// when compared to the C version:
63///
64/// ```
65/// let pid = current!().pid();
66/// ```
67///
68/// Getting the PID of the current process, also zero additional cost:
69///
70/// ```
71/// let pid = current!().group_leader().pid();
72/// ```
73///
74/// Getting the current task and storing it in some struct. The reference count is automatically
75/// incremented when creating `State` and decremented when it is dropped:
76///
77/// ```
78/// use kernel::{task::Task, types::ARef};
79///
80/// struct State {
81/// creator: ARef<Task>,
82/// index: u32,
83/// }
84///
85/// impl State {
86/// fn new() -> Self {
87/// Self {
88/// creator: current!().into(),
89/// index: 0,
90/// }
91/// }
92/// }
93/// ```
94#[repr(transparent)]
95pub struct Task(pub(crate) Opaque<bindings::task_struct>);
96
97// SAFETY: By design, the only way to access a `Task` is via the `current` function or via an
98// `ARef<Task>` obtained through the `AlwaysRefCounted` impl. This means that the only situation in
99// which a `Task` can be accessed mutably is when the refcount drops to zero and the destructor
100// runs. It is safe for that to happen on any thread, so it is ok for this type to be `Send`.
101unsafe impl Send for Task {}
102
103// SAFETY: It's OK to access `Task` through shared references from other threads because we're
104// either accessing properties that don't change (e.g., `pid`, `group_leader`) or that are properly
105// synchronised by C code (e.g., `signal_pending`).
106unsafe impl Sync for Task {}
107
108/// The type of process identifiers (PIDs).
109type Pid = bindings::pid_t;
110
111/// The type of user identifiers (UIDs).
112#[derive(Copy, Clone)]
113pub struct Kuid {
114 kuid: bindings::kuid_t,
115}
116
117impl Task {
118 /// Returns a raw pointer to the current task.
119 ///
120 /// It is up to the user to use the pointer correctly.
121 #[inline]
122 pub fn current_raw() -> *mut bindings::task_struct {
123 // SAFETY: Getting the current pointer is always safe.
124 unsafe { bindings::get_current() }
125 }
126
127 /// Returns a task reference for the currently executing task/thread.
128 ///
129 /// The recommended way to get the current task/thread is to use the
130 /// [`current`] macro because it is safe.
131 ///
132 /// # Safety
133 ///
134 /// Callers must ensure that the returned object doesn't outlive the current task/thread.
135 pub unsafe fn current() -> impl Deref<Target = Task> {
136 struct TaskRef<'a> {
137 task: &'a Task,
138 _not_send: NotThreadSafe,
139 }
140
141 impl Deref for TaskRef<'_> {
142 type Target = Task;
143
144 fn deref(&self) -> &Self::Target {
145 self.task
146 }
147 }
148
149 let current = Task::current_raw();
150 TaskRef {
151 // SAFETY: If the current thread is still running, the current task is valid. Given
152 // that `TaskRef` is not `Send`, we know it cannot be transferred to another thread
153 // (where it could potentially outlive the caller).
154 task: unsafe { &*current.cast() },
155 _not_send: NotThreadSafe,
156 }
157 }
158
159 /// Returns a PidNamespace reference for the currently executing task's/thread's pid namespace.
160 ///
161 /// This function can be used to create an unbounded lifetime by e.g., storing the returned
162 /// PidNamespace in a global variable which would be a bug. So the recommended way to get the
163 /// current task's/thread's pid namespace is to use the [`current_pid_ns`] macro because it is
164 /// safe.
165 ///
166 /// # Safety
167 ///
168 /// Callers must ensure that the returned object doesn't outlive the current task/thread.
169 pub unsafe fn current_pid_ns() -> impl Deref<Target = PidNamespace> {
170 struct PidNamespaceRef<'a> {
171 task: &'a PidNamespace,
172 _not_send: NotThreadSafe,
173 }
174
175 impl Deref for PidNamespaceRef<'_> {
176 type Target = PidNamespace;
177
178 fn deref(&self) -> &Self::Target {
179 self.task
180 }
181 }
182
183 // The lifetime of `PidNamespace` is bound to `Task` and `struct pid`.
184 //
185 // The `PidNamespace` of a `Task` doesn't ever change once the `Task` is alive. A
186 // `unshare(CLONE_NEWPID)` or `setns(fd_pidns/pidfd, CLONE_NEWPID)` will not have an effect
187 // on the calling `Task`'s pid namespace. It will only effect the pid namespace of children
188 // created by the calling `Task`. This invariant guarantees that after having acquired a
189 // reference to a `Task`'s pid namespace it will remain unchanged.
190 //
191 // When a task has exited and been reaped `release_task()` will be called. This will set
192 // the `PidNamespace` of the task to `NULL`. So retrieving the `PidNamespace` of a task
193 // that is dead will return `NULL`. Note, that neither holding the RCU lock nor holding a
194 // referencing count to
195 // the `Task` will prevent `release_task()` being called.
196 //
197 // In order to retrieve the `PidNamespace` of a `Task` the `task_active_pid_ns()` function
198 // can be used. There are two cases to consider:
199 //
200 // (1) retrieving the `PidNamespace` of the `current` task
201 // (2) retrieving the `PidNamespace` of a non-`current` task
202 //
203 // From system call context retrieving the `PidNamespace` for case (1) is always safe and
204 // requires neither RCU locking nor a reference count to be held. Retrieving the
205 // `PidNamespace` after `release_task()` for current will return `NULL` but no codepath
206 // like that is exposed to Rust.
207 //
208 // Retrieving the `PidNamespace` from system call context for (2) requires RCU protection.
209 // Accessing `PidNamespace` outside of RCU protection requires a reference count that
210 // must've been acquired while holding the RCU lock. Note that accessing a non-`current`
211 // task means `NULL` can be returned as the non-`current` task could have already passed
212 // through `release_task()`.
213 //
214 // To retrieve (1) the `current_pid_ns!()` macro should be used which ensure that the
215 // returned `PidNamespace` cannot outlive the calling scope. The associated
216 // `current_pid_ns()` function should not be called directly as it could be abused to
217 // created an unbounded lifetime for `PidNamespace`. The `current_pid_ns!()` macro allows
218 // Rust to handle the common case of accessing `current`'s `PidNamespace` without RCU
219 // protection and without having to acquire a reference count.
220 //
221 // For (2) the `task_get_pid_ns()` method must be used. This will always acquire a
222 // reference on `PidNamespace` and will return an `Option` to force the caller to
223 // explicitly handle the case where `PidNamespace` is `None`, something that tends to be
224 // forgotten when doing the equivalent operation in `C`. Missing RCU primitives make it
225 // difficult to perform operations that are otherwise safe without holding a reference
226 // count as long as RCU protection is guaranteed. But it is not important currently. But we
227 // do want it in the future.
228 //
229 // Note for (2) the required RCU protection around calling `task_active_pid_ns()`
230 // synchronizes against putting the last reference of the associated `struct pid` of
231 // `task->thread_pid`. The `struct pid` stored in that field is used to retrieve the
232 // `PidNamespace` of the caller. When `release_task()` is called `task->thread_pid` will be
233 // `NULL`ed and `put_pid()` on said `struct pid` will be delayed in `free_pid()` via
234 // `call_rcu()` allowing everyone with an RCU protected access to the `struct pid` acquired
235 // from `task->thread_pid` to finish.
236 //
237 // SAFETY: The current task's pid namespace is valid as long as the current task is running.
238 let pidns = unsafe { bindings::task_active_pid_ns(Task::current_raw()) };
239 PidNamespaceRef {
240 // SAFETY: If the current thread is still running, the current task and its associated
241 // pid namespace are valid. `PidNamespaceRef` is not `Send`, so we know it cannot be
242 // transferred to another thread (where it could potentially outlive the current
243 // `Task`). The caller needs to ensure that the PidNamespaceRef doesn't outlive the
244 // current task/thread.
245 task: unsafe { PidNamespace::from_ptr(pidns) },
246 _not_send: NotThreadSafe,
247 }
248 }
249
250 /// Returns a raw pointer to the task.
251 #[inline]
252 pub fn as_ptr(&self) -> *mut bindings::task_struct {
253 self.0.get()
254 }
255
256 /// Returns the group leader of the given task.
257 pub fn group_leader(&self) -> &Task {
258 // SAFETY: The group leader of a task never changes after initialization, so reading this
259 // field is not a data race.
260 let ptr = unsafe { *ptr::addr_of!((*self.as_ptr()).group_leader) };
261
262 // SAFETY: The lifetime of the returned task reference is tied to the lifetime of `self`,
263 // and given that a task has a reference to its group leader, we know it must be valid for
264 // the lifetime of the returned task reference.
265 unsafe { &*ptr.cast() }
266 }
267
268 /// Returns the PID of the given task.
269 pub fn pid(&self) -> Pid {
270 // SAFETY: The pid of a task never changes after initialization, so reading this field is
271 // not a data race.
272 unsafe { *ptr::addr_of!((*self.as_ptr()).pid) }
273 }
274
275 /// Returns the UID of the given task.
276 pub fn uid(&self) -> Kuid {
277 // SAFETY: It's always safe to call `task_uid` on a valid task.
278 Kuid::from_raw(unsafe { bindings::task_uid(self.as_ptr()) })
279 }
280
281 /// Returns the effective UID of the given task.
282 pub fn euid(&self) -> Kuid {
283 // SAFETY: It's always safe to call `task_euid` on a valid task.
284 Kuid::from_raw(unsafe { bindings::task_euid(self.as_ptr()) })
285 }
286
287 /// Determines whether the given task has pending signals.
288 pub fn signal_pending(&self) -> bool {
289 // SAFETY: It's always safe to call `signal_pending` on a valid task.
290 unsafe { bindings::signal_pending(self.as_ptr()) != 0 }
291 }
292
293 /// Returns task's pid namespace with elevated reference count
294 pub fn get_pid_ns(&self) -> Option<ARef<PidNamespace>> {
295 // SAFETY: By the type invariant, we know that `self.0` is valid.
296 let ptr = unsafe { bindings::task_get_pid_ns(self.as_ptr()) };
297 if ptr.is_null() {
298 None
299 } else {
300 // SAFETY: `ptr` is valid by the safety requirements of this function. And we own a
301 // reference count via `task_get_pid_ns()`.
302 // CAST: `Self` is a `repr(transparent)` wrapper around `bindings::pid_namespace`.
303 Some(unsafe { ARef::from_raw(ptr::NonNull::new_unchecked(ptr.cast::<PidNamespace>())) })
304 }
305 }
306
307 /// Returns the given task's pid in the provided pid namespace.
308 #[doc(alias = "task_tgid_nr_ns")]
309 pub fn tgid_nr_ns(&self, pidns: Option<&PidNamespace>) -> Pid {
310 let pidns = match pidns {
311 Some(pidns) => pidns.as_ptr(),
312 None => core::ptr::null_mut(),
313 };
314 // SAFETY: By the type invariant, we know that `self.0` is valid. We received a valid
315 // PidNamespace that we can use as a pointer or we received an empty PidNamespace and
316 // thus pass a null pointer. The underlying C function is safe to be used with NULL
317 // pointers.
318 unsafe { bindings::task_tgid_nr_ns(self.as_ptr(), pidns) }
319 }
320
321 /// Wakes up the task.
322 pub fn wake_up(&self) {
323 // SAFETY: It's always safe to call `signal_pending` on a valid task, even if the task
324 // running.
325 unsafe { bindings::wake_up_process(self.as_ptr()) };
326 }
327}
328
329// SAFETY: The type invariants guarantee that `Task` is always refcounted.
330unsafe impl crate::types::AlwaysRefCounted for Task {
331 fn inc_ref(&self) {
332 // SAFETY: The existence of a shared reference means that the refcount is nonzero.
333 unsafe { bindings::get_task_struct(self.as_ptr()) };
334 }
335
336 unsafe fn dec_ref(obj: ptr::NonNull<Self>) {
337 // SAFETY: The safety requirements guarantee that the refcount is nonzero.
338 unsafe { bindings::put_task_struct(obj.cast().as_ptr()) }
339 }
340}
341
342impl Kuid {
343 /// Get the current euid.
344 #[inline]
345 pub fn current_euid() -> Kuid {
346 // SAFETY: Just an FFI call.
347 Self::from_raw(unsafe { bindings::current_euid() })
348 }
349
350 /// Create a `Kuid` given the raw C type.
351 #[inline]
352 pub fn from_raw(kuid: bindings::kuid_t) -> Self {
353 Self { kuid }
354 }
355
356 /// Turn this kuid into the raw C type.
357 #[inline]
358 pub fn into_raw(self) -> bindings::kuid_t {
359 self.kuid
360 }
361
362 /// Converts this kernel UID into a userspace UID.
363 ///
364 /// Uses the namespace of the current task.
365 #[inline]
366 pub fn into_uid_in_current_ns(self) -> bindings::uid_t {
367 // SAFETY: Just an FFI call.
368 unsafe { bindings::from_kuid(bindings::current_user_ns(), self.kuid) }
369 }
370}
371
372impl PartialEq for Kuid {
373 #[inline]
374 fn eq(&self, other: &Kuid) -> bool {
375 // SAFETY: Just an FFI call.
376 unsafe { bindings::uid_eq(self.kuid, other.kuid) }
377 }
378}
379
380impl Eq for Kuid {}