xref: /linux/Documentation/bpf/kfuncs.rst (revision 4b99990cdf9560e8a071640baf19f312e6ae02f4)
1.. SPDX-License-Identifier: GPL-2.0
2
3.. _kfuncs-header-label:
4
5=============================
6BPF Kernel Functions (kfuncs)
7=============================
8
91. Introduction
10===============
11
12BPF Kernel Functions or more commonly known as kfuncs are functions in the Linux
13kernel which are exposed for use by BPF programs. Unlike normal BPF helpers,
14kfuncs do not have a stable interface and can change from one kernel release to
15another. Hence, BPF programs need to be updated in response to changes in the
16kernel. See :ref:`BPF_kfunc_lifecycle_expectations` for more information.
17
182. Defining a kfunc
19===================
20
21There are two ways to expose a kernel function to BPF programs, either make an
22existing function in the kernel visible, or add a new wrapper for BPF. In both
23cases, care must be taken that BPF program can only call such function in a
24valid context. To enforce this, visibility of a kfunc can be per program type.
25
26If you are not creating a BPF wrapper for existing kernel function, skip ahead
27to :ref:`BPF_kfunc_nodef`.
28
292.1 Creating a wrapper kfunc
30----------------------------
31
32When defining a wrapper kfunc, the wrapper function should have extern linkage.
33This prevents the compiler from optimizing away dead code, as this wrapper kfunc
34is not invoked anywhere in the kernel itself. It is not necessary to provide a
35prototype in a header for the wrapper kfunc.
36
37An example is given below::
38
39        /* Disables missing prototype warnings */
40        __bpf_kfunc_start_defs();
41
42        __bpf_kfunc struct task_struct *bpf_find_get_task_by_vpid(pid_t nr)
43        {
44                return find_get_task_by_vpid(nr);
45        }
46
47        __bpf_kfunc_end_defs();
48
49A wrapper kfunc is often needed when we need to annotate parameters of the
50kfunc. Otherwise one may directly make the kfunc visible to the BPF program by
51registering it with the BPF subsystem. See :ref:`BPF_kfunc_nodef`.
52
532.2 kfunc Parameters
54--------------------
55
56All kfuncs now require trusted arguments by default. This means that all
57pointer arguments must be valid, and all pointers to BTF objects must be
58passed in their unmodified form (at a zero offset, and without having been
59obtained from walking another pointer, with exceptions described below).
60
61There are two types of pointers to kernel objects which are considered "trusted":
62
631. Pointers which are passed as tracepoint or struct_ops callback arguments.
642. Pointers which were returned from a KF_ACQUIRE kfunc.
65
66Pointers to non-BTF objects (e.g. scalar pointers) may also be passed to
67kfuncs, and may have a non-zero offset.
68
69The definition of "valid" pointers is subject to change at any time, and has
70absolutely no ABI stability guarantees.
71
72As mentioned above, a nested pointer obtained from walking a trusted pointer is
73no longer trusted, with one exception. If a struct type has a field that is
74guaranteed to be valid (trusted or rcu, as in KF_RCU description below) as long
75as its parent pointer is valid, the following macros can be used to express
76that to the verifier:
77
78* ``BTF_TYPE_SAFE_TRUSTED``
79* ``BTF_TYPE_SAFE_RCU``
80* ``BTF_TYPE_SAFE_RCU_OR_NULL``
81
82For example,
83
84.. code-block:: c
85
86	BTF_TYPE_SAFE_TRUSTED(struct socket) {
87		struct sock *sk;
88	};
89
90or
91
92.. code-block:: c
93
94	BTF_TYPE_SAFE_RCU(struct task_struct) {
95		const cpumask_t *cpus_ptr;
96		struct css_set __rcu *cgroups;
97		struct task_struct __rcu *real_parent;
98		struct task_struct *group_leader;
99	};
100
101In other words, you must:
102
1031. Wrap the valid pointer type in a ``BTF_TYPE_SAFE_*`` macro.
104
1052. Specify the type and name of the valid nested field. This field must match
106   the field in the original type definition exactly.
107
108A new type declared by a ``BTF_TYPE_SAFE_*`` macro also needs to be emitted so
109that it appears in BTF. For example, ``BTF_TYPE_SAFE_TRUSTED(struct socket)``
110is emitted in the ``type_is_trusted()`` function as follows:
111
112.. code-block:: c
113
114	BTF_TYPE_EMIT(BTF_TYPE_SAFE_TRUSTED(struct socket));
115
1162.3 Annotating kfunc parameters
117-------------------------------
118
119Similar to BPF helpers, there is sometime need for additional context required
120by the verifier to make the usage of kernel functions safer and more useful.
121Hence, we can annotate a parameter by suffixing the name of the argument of the
122kfunc with a __tag, where tag may be one of the supported annotations.
123
1242.3.1 __sz Annotation
125---------------------
126
127This annotation is used to indicate a memory and size pair in the argument list.
128An example is given below::
129
130        __bpf_kfunc void bpf_memzero(void *mem, int mem__sz)
131        {
132        ...
133        }
134
135Here, the verifier will treat first argument as a PTR_TO_MEM, and second
136argument as its size. By default, without __sz annotation, the size of the type
137of the pointer is used. Without __sz annotation, a kfunc cannot accept a void
138pointer.
139
1402.3.2 __k Annotation
141--------------------
142
143This annotation is only understood for scalar arguments, where it indicates that
144the verifier must check the scalar argument to be a known constant, which does
145not indicate a size parameter, and the value of the constant is relevant to the
146safety of the program.
147
148An example is given below::
149
150        __bpf_kfunc void *bpf_obj_new(u32 local_type_id__k, ...)
151        {
152        ...
153        }
154
155Here, bpf_obj_new uses local_type_id argument to find out the size of that type
156ID in program's BTF and return a sized pointer to it. Each type ID will have a
157distinct size, hence it is crucial to treat each such call as distinct when
158values don't match during verifier state pruning checks.
159
160Hence, whenever a constant scalar argument is accepted by a kfunc which is not a
161size parameter, and the value of the constant matters for program safety, __k
162suffix should be used.
163
1642.3.3 __uninit Annotation
165-------------------------
166
167This annotation is used to indicate that the argument will be treated as
168uninitialized.
169
170An example is given below::
171
172        __bpf_kfunc int bpf_dynptr_from_skb(..., struct bpf_dynptr_kern *ptr__uninit)
173        {
174        ...
175        }
176
177Here, the dynptr will be treated as an uninitialized dynptr. Without this
178annotation, the verifier will reject the program if the dynptr passed in is
179not initialized.
180
1812.3.4 __nullable Annotation
182---------------------------
183
184This annotation is used to indicate that the pointer argument may be NULL.
185The verifier will allow passing NULL for such arguments.
186
187An example is given below::
188
189        __bpf_kfunc void bpf_task_release(struct task_struct *task__nullable)
190        {
191        ...
192        }
193
194Here, the task pointer may be NULL. The kfunc is responsible for checking if
195the pointer is NULL before dereferencing it.
196
197The __nullable annotation can be combined with other annotations. For example,
198when used with __sz or __szk annotations for memory and size pairs, the
199verifier will skip size validation when a NULL pointer is passed, but will
200still process the size argument to extract constant size information when
201needed::
202
203        __bpf_kfunc void *bpf_dynptr_slice(..., void *buffer__nullable,
204                                           u32 buffer__szk)
205
206Here, the buffer may be NULL. If the buffer is not NULL, it must be at least
207buffer__szk bytes in size. The kfunc is responsible for checking if the buffer
208is NULL before using it.
209
2102.3.5 __nonown_allowed Annotation
211---------------------------------
212
213This annotation is used to indicate that the parameter may be a non-owning reference.
214
215An example is given below::
216
217        __bpf_kfunc int bpf_list_add(..., struct bpf_list_node
218                                     *prev__nonown_allowed, ...)
219        {
220                ...
221        }
222
223For the ``prev__nonown_allowed`` parameter (resolved as ``KF_ARG_PTR_TO_LIST_NODE``),
224suffix ``__nonown_allowed`` retains the usual owning-pointer rules and also
225permits a non-owning reference with no ref_obj_id (e.g. the return value of
226bpf_list_front() / bpf_list_back()).
227
2282.3.6 __str Annotation
229----------------------
230This annotation is used to indicate that the argument is a constant string.
231
232An example is given below::
233
234        __bpf_kfunc bpf_get_file_xattr(..., const char *name__str, ...)
235        {
236        ...
237        }
238
239In this case, ``bpf_get_file_xattr()`` can be called as::
240
241        bpf_get_file_xattr(..., "xattr_name", ...);
242
243Or::
244
245        const char name[] = "xattr_name";  /* This need to be global */
246        int BPF_PROG(...)
247        {
248                ...
249                bpf_get_file_xattr(..., name, ...);
250                ...
251        }
252
253.. _BPF_kfunc_nodef:
254
2552.4 Using an existing kernel function
256-------------------------------------
257
258When an existing function in the kernel is fit for consumption by BPF programs,
259it can be directly registered with the BPF subsystem. However, care must still
260be taken to review the context in which it will be invoked by the BPF program
261and whether it is safe to do so.
262
2632.5 Annotating kfuncs
264---------------------
265
266In addition to kfuncs' arguments, verifier may need more information about the
267type of kfunc(s) being registered with the BPF subsystem. To do so, we define
268flags on a set of kfuncs as follows::
269
270        BTF_KFUNCS_START(bpf_task_set)
271        BTF_ID_FLAGS(func, bpf_get_task_pid, KF_ACQUIRE | KF_RET_NULL)
272        BTF_ID_FLAGS(func, bpf_put_pid, KF_RELEASE)
273        BTF_KFUNCS_END(bpf_task_set)
274
275This set encodes the BTF ID of each kfunc listed above, and encodes the flags
276along with it. Ofcourse, it is also allowed to specify no flags.
277
278kfunc definitions should also always be annotated with the ``__bpf_kfunc``
279macro. This prevents issues such as the compiler inlining the kfunc if it's a
280static kernel function, or the function being elided in an LTO build as it's
281not used in the rest of the kernel. Developers should not manually add
282annotations to their kfunc to prevent these issues. If an annotation is
283required to prevent such an issue with your kfunc, it is a bug and should be
284added to the definition of the macro so that other kfuncs are similarly
285protected. An example is given below::
286
287        __bpf_kfunc struct task_struct *bpf_get_task_pid(s32 pid)
288        {
289        ...
290        }
291
2922.5.1 KF_ACQUIRE flag
293---------------------
294
295The KF_ACQUIRE flag is used to indicate that the kfunc returns a pointer to a
296refcounted object. The verifier will then ensure that the pointer to the object
297is eventually released using a release kfunc, or transferred to a map using a
298referenced kptr (by invoking bpf_kptr_xchg). If not, the verifier fails the
299loading of the BPF program until no lingering references remain in all possible
300explored states of the program.
301
3022.5.2 KF_RET_NULL flag
303----------------------
304
305The KF_RET_NULL flag is used to indicate that the pointer returned by the kfunc
306may be NULL. Hence, it forces the user to do a NULL check on the pointer
307returned from the kfunc before making use of it (dereferencing or passing to
308another helper). This flag is often used in pairing with KF_ACQUIRE flag, but
309both are orthogonal to each other.
310
3112.5.3 KF_RELEASE flag
312---------------------
313
314The KF_RELEASE flag is used to indicate that the kfunc releases the pointer
315passed in to it. There can be only one referenced pointer that can be passed
316in. All copies of the pointer being released are invalidated as a result of
317invoking kfunc with this flag.
318
3192.5.4 KF_SLEEPABLE flag
320-----------------------
321
322The KF_SLEEPABLE flag is used for kfuncs that may sleep. Such kfuncs can only
323be called by sleepable BPF programs (BPF_F_SLEEPABLE).
324
3252.5.5 KF_DESTRUCTIVE flag
326--------------------------
327
328The KF_DESTRUCTIVE flag is used to indicate functions calling which is
329destructive to the system. For example such a call can result in system
330rebooting or panicking. Due to this additional restrictions apply to these
331calls. At the moment they only require CAP_SYS_BOOT capability, but more can be
332added later.
333
3342.5.6 KF_RCU flag
335-----------------
336
337The KF_RCU flag allows kfuncs to opt out of the default trusted args
338requirement and accept RCU pointers with weaker guarantees. The kfuncs marked
339with KF_RCU expect either PTR_TRUSTED or MEM_RCU arguments. The verifier
340guarantees that the objects are valid and there is no use-after-free. The
341pointers are not NULL, but the object's refcount could have reached zero. The
342kfuncs need to consider doing refcnt != 0 check, especially when returning a
343KF_ACQUIRE pointer. Note as well that a KF_ACQUIRE kfunc that is KF_RCU should
344very likely also be KF_RET_NULL.
345
3462.5.7 KF_RCU_PROTECTED flag
347---------------------------
348
349The KF_RCU_PROTECTED flag is used to indicate that the kfunc must be invoked in
350an RCU critical section. This is assumed by default in non-sleepable programs,
351and must be explicitly ensured by calling ``bpf_rcu_read_lock`` for sleepable
352ones.
353
354If the kfunc returns a pointer value, this flag also enforces that the returned
355pointer is RCU protected, and can only be used while the RCU critical section is
356active.
357
358The flag is distinct from the ``KF_RCU`` flag, which only ensures that its
359arguments are at least RCU protected pointers. This may transitively imply that
360RCU protection is ensured, but it does not work in cases of kfuncs which require
361RCU protection but do not take RCU protected arguments.
362
363.. _KF_deprecated_flag:
364
3652.5.8 KF_DEPRECATED flag
366------------------------
367
368The KF_DEPRECATED flag is used for kfuncs which are scheduled to be
369changed or removed in a subsequent kernel release. A kfunc that is
370marked with KF_DEPRECATED should also have any relevant information
371captured in its kernel doc. Such information typically includes the
372kfunc's expected remaining lifespan, a recommendation for new
373functionality that can replace it if any is available, and possibly a
374rationale for why it is being removed.
375
376Note that while on some occasions, a KF_DEPRECATED kfunc may continue to be
377supported and have its KF_DEPRECATED flag removed, it is likely to be far more
378difficult to remove a KF_DEPRECATED flag after it's been added than it is to
379prevent it from being added in the first place. As described in
380:ref:`BPF_kfunc_lifecycle_expectations`, users that rely on specific kfuncs are
381encouraged to make their use-cases known as early as possible, and participate
382in upstream discussions regarding whether to keep, change, deprecate, or remove
383those kfuncs if and when such discussions occur.
384
3852.5.9 KF_IMPLICIT_ARGS flag
386------------------------------------
387
388The KF_IMPLICIT_ARGS flag is used to indicate that the BPF signature
389of the kfunc is different from it's kernel signature, and the values
390for implicit arguments are provided at load time by the verifier.
391
392Only arguments of specific types are implicit.
393Currently only ``struct bpf_prog_aux *`` type is supported.
394
395A kfunc with KF_IMPLICIT_ARGS flag therefore has two types in BTF: one
396function matching the kernel declaration (with _impl suffix in the
397name by convention), and another matching the intended BPF API.
398
399Verifier only allows calls to the non-_impl version of a kfunc, that
400uses a signature without the implicit arguments.
401
402Example declaration:
403
404.. code-block:: c
405
406	__bpf_kfunc int bpf_task_work_schedule_signal(struct task_struct *task, struct bpf_task_work *tw,
407						      void *map__map, bpf_task_work_callback_t callback,
408						      struct bpf_prog_aux *aux) { ... }
409
410Example usage in BPF program:
411
412.. code-block:: c
413
414	/* note that the last argument is omitted */
415        bpf_task_work_schedule_signal(task, &work->tw, &arrmap, task_work_callback);
416
4172.6 Registering the kfuncs
418--------------------------
419
420Once the kfunc is prepared for use, the final step to making it visible is
421registering it with the BPF subsystem. Registration is done per BPF program
422type. An example is shown below::
423
424        BTF_KFUNCS_START(bpf_task_set)
425        BTF_ID_FLAGS(func, bpf_get_task_pid, KF_ACQUIRE | KF_RET_NULL)
426        BTF_ID_FLAGS(func, bpf_put_pid, KF_RELEASE)
427        BTF_KFUNCS_END(bpf_task_set)
428
429        static const struct btf_kfunc_id_set bpf_task_kfunc_set = {
430                .owner = THIS_MODULE,
431                .set   = &bpf_task_set,
432        };
433
434        static int init_subsystem(void)
435        {
436                return register_btf_kfunc_id_set(BPF_PROG_TYPE_TRACING, &bpf_task_kfunc_set);
437        }
438        late_initcall(init_subsystem);
439
4402.7  Specifying no-cast aliases with ___init
441--------------------------------------------
442
443The verifier will always enforce that the BTF type of a pointer passed to a
444kfunc by a BPF program, matches the type of pointer specified in the kfunc
445definition. The verifier, does, however, allow types that are equivalent
446according to the C standard to be passed to the same kfunc arg, even if their
447BTF_IDs differ.
448
449For example, for the following type definition:
450
451.. code-block:: c
452
453	struct bpf_cpumask {
454		cpumask_t cpumask;
455		refcount_t usage;
456	};
457
458The verifier would allow a ``struct bpf_cpumask *`` to be passed to a kfunc
459taking a ``cpumask_t *`` (which is a typedef of ``struct cpumask *``). For
460instance, both ``struct cpumask *`` and ``struct bpf_cpmuask *`` can be passed
461to bpf_cpumask_test_cpu().
462
463In some cases, this type-aliasing behavior is not desired. ``struct
464nf_conn___init`` is one such example:
465
466.. code-block:: c
467
468	struct nf_conn___init {
469		struct nf_conn ct;
470	};
471
472The C standard would consider these types to be equivalent, but it would not
473always be safe to pass either type to a trusted kfunc. ``struct
474nf_conn___init`` represents an allocated ``struct nf_conn`` object that has
475*not yet been initialized*, so it would therefore be unsafe to pass a ``struct
476nf_conn___init *`` to a kfunc that's expecting a fully initialized ``struct
477nf_conn *`` (e.g. ``bpf_ct_change_timeout()``).
478
479In order to accommodate such requirements, the verifier will enforce strict
480PTR_TO_BTF_ID type matching if two types have the exact same name, with one
481being suffixed with ``___init``.
482
4832.8 Accessing arena memory through kfunc arguments
484--------------------------------------------------
485
486A read or write at any address inside an arena does not oops the kernel.
487Unallocated arena pages are lazily backed by a scratch page and the
488access is reported through the program's BPF stream as an error. Only
489the BPF program's correctness is affected; the kernel itself remains
490intact.
491
492The arena is followed by a ``GUARD_SZ / 2`` (32 KiB) guard region that
493is also covered by this recovery. A kfunc handed an arena pointer may
494therefore access up to ``GUARD_SZ / 2`` past it without bounds-checking
495against the arena. Larger accesses must verify the range explicitly.
496
497.. _BPF_kfunc_lifecycle_expectations:
498
4993. kfunc lifecycle expectations
500===============================
501
502kfuncs provide a kernel <-> kernel API, and thus are not bound by any of the
503strict stability restrictions associated with kernel <-> user UAPIs. This means
504they can be thought of as similar to EXPORT_SYMBOL_GPL, and can therefore be
505modified or removed by a maintainer of the subsystem they're defined in when
506it's deemed necessary.
507
508Like any other change to the kernel, maintainers will not change or remove a
509kfunc without having a reasonable justification.  Whether or not they'll choose
510to change a kfunc will ultimately depend on a variety of factors, such as how
511widely used the kfunc is, how long the kfunc has been in the kernel, whether an
512alternative kfunc exists, what the norm is in terms of stability for the
513subsystem in question, and of course what the technical cost is of continuing
514to support the kfunc.
515
516There are several implications of this:
517
518a) kfuncs that are widely used or have been in the kernel for a long time will
519   be more difficult to justify being changed or removed by a maintainer. In
520   other words, kfuncs that are known to have a lot of users and provide
521   significant value provide stronger incentives for maintainers to invest the
522   time and complexity in supporting them. It is therefore important for
523   developers that are using kfuncs in their BPF programs to communicate and
524   explain how and why those kfuncs are being used, and to participate in
525   discussions regarding those kfuncs when they occur upstream.
526
527b) Unlike regular kernel symbols marked with EXPORT_SYMBOL_GPL, BPF programs
528   that call kfuncs are generally not part of the kernel tree. This means that
529   refactoring cannot typically change callers in-place when a kfunc changes,
530   as is done for e.g. an upstreamed driver being updated in place when a
531   kernel symbol is changed.
532
533   Unlike with regular kernel symbols, this is expected behavior for BPF
534   symbols, and out-of-tree BPF programs that use kfuncs should be considered
535   relevant to discussions and decisions around modifying and removing those
536   kfuncs. The BPF community will take an active role in participating in
537   upstream discussions when necessary to ensure that the perspectives of such
538   users are taken into account.
539
540c) A kfunc will never have any hard stability guarantees. BPF APIs cannot and
541   will not ever hard-block a change in the kernel purely for stability
542   reasons. That being said, kfuncs are features that are meant to solve
543   problems and provide value to users. The decision of whether to change or
544   remove a kfunc is a multivariate technical decision that is made on a
545   case-by-case basis, and which is informed by data points such as those
546   mentioned above. It is expected that a kfunc being removed or changed with
547   no warning will not be a common occurrence or take place without sound
548   justification, but it is a possibility that must be accepted if one is to
549   use kfuncs.
550
5513.1 kfunc deprecation
552---------------------
553
554As described above, while sometimes a maintainer may find that a kfunc must be
555changed or removed immediately to accommodate some changes in their subsystem,
556usually kfuncs will be able to accommodate a longer and more measured
557deprecation process. For example, if a new kfunc comes along which provides
558superior functionality to an existing kfunc, the existing kfunc may be
559deprecated for some period of time to allow users to migrate their BPF programs
560to use the new one. Or, if a kfunc has no known users, a decision may be made
561to remove the kfunc (without providing an alternative API) after some
562deprecation period so as to provide users with a window to notify the kfunc
563maintainer if it turns out that the kfunc is actually being used.
564
565It's expected that the common case will be that kfuncs will go through a
566deprecation period rather than being changed or removed without warning. As
567described in :ref:`KF_deprecated_flag`, the kfunc framework provides the
568KF_DEPRECATED flag to kfunc developers to signal to users that a kfunc has been
569deprecated. Once a kfunc has been marked with KF_DEPRECATED, the following
570procedure is followed for removal:
571
5721. Any relevant information for deprecated kfuncs is documented in the kfunc's
573   kernel docs. This documentation will typically include the kfunc's expected
574   remaining lifespan, a recommendation for new functionality that can replace
575   the usage of the deprecated function (or an explanation as to why no such
576   replacement exists), etc.
577
5782. The deprecated kfunc is kept in the kernel for some period of time after it
579   was first marked as deprecated. This time period will be chosen on a
580   case-by-case basis, and will typically depend on how widespread the use of
581   the kfunc is, how long it has been in the kernel, and how hard it is to move
582   to alternatives. This deprecation time period is "best effort", and as
583   described :ref:`above<BPF_kfunc_lifecycle_expectations>`, circumstances may
584   sometimes dictate that the kfunc be removed before the full intended
585   deprecation period has elapsed.
586
5873. After the deprecation period the kfunc will be removed. At this point, BPF
588   programs calling the kfunc will be rejected by the verifier.
589
5904. Core kfuncs
591==============
592
593The BPF subsystem provides a number of "core" kfuncs that are potentially
594applicable to a wide variety of different possible use cases and programs.
595Those kfuncs are documented here.
596
5974.1 struct task_struct * kfuncs
598-------------------------------
599
600There are a number of kfuncs that allow ``struct task_struct *`` objects to be
601used as kptrs:
602
603.. kernel-doc:: kernel/bpf/helpers.c
604   :identifiers: bpf_task_acquire bpf_task_release
605
606These kfuncs are useful when you want to acquire or release a reference to a
607``struct task_struct *`` that was passed as e.g. a tracepoint arg, or a
608struct_ops callback arg. For example:
609
610.. code-block:: c
611
612	/**
613	 * A trivial example tracepoint program that shows how to
614	 * acquire and release a struct task_struct * pointer.
615	 */
616	SEC("tp_btf/task_newtask")
617	int BPF_PROG(task_acquire_release_example, struct task_struct *task, u64 clone_flags)
618	{
619		struct task_struct *acquired;
620
621		acquired = bpf_task_acquire(task);
622		if (acquired)
623			/*
624			 * In a typical program you'd do something like store
625			 * the task in a map, and the map will automatically
626			 * release it later. Here, we release it manually.
627			 */
628			bpf_task_release(acquired);
629		return 0;
630	}
631
632
633References acquired on ``struct task_struct *`` objects are RCU protected.
634Therefore, when in an RCU read region, you can obtain a pointer to a task
635embedded in a map value without having to acquire a reference:
636
637.. code-block:: c
638
639	#define private(name) SEC(".data." #name) __hidden __attribute__((aligned(8)))
640	private(TASK) static struct task_struct *global;
641
642	/**
643	 * A trivial example showing how to access a task stored
644	 * in a map using RCU.
645	 */
646	SEC("tp_btf/task_newtask")
647	int BPF_PROG(task_rcu_read_example, struct task_struct *task, u64 clone_flags)
648	{
649		struct task_struct *local_copy;
650
651		bpf_rcu_read_lock();
652		local_copy = global;
653		if (local_copy)
654			/*
655			 * We could also pass local_copy to kfuncs or helper functions here,
656			 * as we're guaranteed that local_copy will be valid until we exit
657			 * the RCU read region below.
658			 */
659			bpf_printk("Global task %s is valid", local_copy->comm);
660		else
661			bpf_printk("No global task found");
662		bpf_rcu_read_unlock();
663
664		/* At this point we can no longer reference local_copy. */
665
666		return 0;
667	}
668
669----
670
671A BPF program can also look up a task from a pid. This can be useful if the
672caller doesn't have a trusted pointer to a ``struct task_struct *`` object that
673it can acquire a reference on with bpf_task_acquire().
674
675.. kernel-doc:: kernel/bpf/helpers.c
676   :identifiers: bpf_task_from_pid
677
678Here is an example of it being used:
679
680.. code-block:: c
681
682	SEC("tp_btf/task_newtask")
683	int BPF_PROG(task_get_pid_example, struct task_struct *task, u64 clone_flags)
684	{
685		struct task_struct *lookup;
686
687		lookup = bpf_task_from_pid(task->pid);
688		if (!lookup)
689			/* A task should always be found, as %task is a tracepoint arg. */
690			return -ENOENT;
691
692		if (lookup->pid != task->pid) {
693			/* bpf_task_from_pid() looks up the task via its
694			 * globally-unique pid from the init_pid_ns. Thus,
695			 * the pid of the lookup task should always be the
696			 * same as the input task.
697			 */
698			bpf_task_release(lookup);
699			return -EINVAL;
700		}
701
702		/* bpf_task_from_pid() returns an acquired reference,
703		 * so it must be dropped before returning from the
704		 * tracepoint handler.
705		 */
706		bpf_task_release(lookup);
707		return 0;
708	}
709
7104.2 struct cgroup * kfuncs
711--------------------------
712
713``struct cgroup *`` objects also have acquire and release functions:
714
715.. kernel-doc:: kernel/bpf/helpers.c
716   :identifiers: bpf_cgroup_acquire bpf_cgroup_release
717
718These kfuncs are used in exactly the same manner as bpf_task_acquire() and
719bpf_task_release() respectively, so we won't provide examples for them.
720
721----
722
723Other kfuncs available for interacting with ``struct cgroup *`` objects are
724bpf_cgroup_ancestor() and bpf_cgroup_from_id(), allowing callers to access
725the ancestor of a cgroup and find a cgroup by its ID, respectively. Both
726return a cgroup kptr.
727
728.. kernel-doc:: kernel/bpf/helpers.c
729   :identifiers: bpf_cgroup_ancestor
730
731.. kernel-doc:: kernel/bpf/helpers.c
732   :identifiers: bpf_cgroup_from_id
733
734Eventually, BPF should be updated to allow this to happen with a normal memory
735load in the program itself. This is currently not possible without more work in
736the verifier. bpf_cgroup_ancestor() can be used as follows:
737
738.. code-block:: c
739
740	/**
741	 * Simple tracepoint example that illustrates how a cgroup's
742	 * ancestor can be accessed using bpf_cgroup_ancestor().
743	 */
744	SEC("tp_btf/cgroup_mkdir")
745	int BPF_PROG(cgrp_ancestor_example, struct cgroup *cgrp, const char *path)
746	{
747		struct cgroup *parent;
748
749		/* The parent cgroup resides at the level before the current cgroup's level. */
750		parent = bpf_cgroup_ancestor(cgrp, cgrp->level - 1);
751		if (!parent)
752			return -ENOENT;
753
754		bpf_printk("Parent id is %d", parent->self.id);
755
756		/* Return the parent cgroup that was acquired above. */
757		bpf_cgroup_release(parent);
758		return 0;
759	}
760
7614.3 struct cpumask * kfuncs
762---------------------------
763
764BPF provides a set of kfuncs that can be used to query, allocate, mutate, and
765destroy struct cpumask * objects. Please refer to :ref:`cpumasks-header-label`
766for more details.
767