xref: /openbmc/linux/Documentation/bpf/kfuncs.rst (revision 83b975b5)
1=============================
2BPF Kernel Functions (kfuncs)
3=============================
4
51. Introduction
6===============
7
8BPF Kernel Functions or more commonly known as kfuncs are functions in the Linux
9kernel which are exposed for use by BPF programs. Unlike normal BPF helpers,
10kfuncs do not have a stable interface and can change from one kernel release to
11another. Hence, BPF programs need to be updated in response to changes in the
12kernel.
13
142. Defining a kfunc
15===================
16
17There are two ways to expose a kernel function to BPF programs, either make an
18existing function in the kernel visible, or add a new wrapper for BPF. In both
19cases, care must be taken that BPF program can only call such function in a
20valid context. To enforce this, visibility of a kfunc can be per program type.
21
22If you are not creating a BPF wrapper for existing kernel function, skip ahead
23to :ref:`BPF_kfunc_nodef`.
24
252.1 Creating a wrapper kfunc
26----------------------------
27
28When defining a wrapper kfunc, the wrapper function should have extern linkage.
29This prevents the compiler from optimizing away dead code, as this wrapper kfunc
30is not invoked anywhere in the kernel itself. It is not necessary to provide a
31prototype in a header for the wrapper kfunc.
32
33An example is given below::
34
35        /* Disables missing prototype warnings */
36        __diag_push();
37        __diag_ignore_all("-Wmissing-prototypes",
38                          "Global kfuncs as their definitions will be in BTF");
39
40        struct task_struct *bpf_find_get_task_by_vpid(pid_t nr)
41        {
42                return find_get_task_by_vpid(nr);
43        }
44
45        __diag_pop();
46
47A wrapper kfunc is often needed when we need to annotate parameters of the
48kfunc. Otherwise one may directly make the kfunc visible to the BPF program by
49registering it with the BPF subsystem. See :ref:`BPF_kfunc_nodef`.
50
512.2 Annotating kfunc parameters
52-------------------------------
53
54Similar to BPF helpers, there is sometime need for additional context required
55by the verifier to make the usage of kernel functions safer and more useful.
56Hence, we can annotate a parameter by suffixing the name of the argument of the
57kfunc with a __tag, where tag may be one of the supported annotations.
58
592.2.1 __sz Annotation
60---------------------
61
62This annotation is used to indicate a memory and size pair in the argument list.
63An example is given below::
64
65        void bpf_memzero(void *mem, int mem__sz)
66        {
67        ...
68        }
69
70Here, the verifier will treat first argument as a PTR_TO_MEM, and second
71argument as its size. By default, without __sz annotation, the size of the type
72of the pointer is used. Without __sz annotation, a kfunc cannot accept a void
73pointer.
74
75.. _BPF_kfunc_nodef:
76
772.3 Using an existing kernel function
78-------------------------------------
79
80When an existing function in the kernel is fit for consumption by BPF programs,
81it can be directly registered with the BPF subsystem. However, care must still
82be taken to review the context in which it will be invoked by the BPF program
83and whether it is safe to do so.
84
852.4 Annotating kfuncs
86---------------------
87
88In addition to kfuncs' arguments, verifier may need more information about the
89type of kfunc(s) being registered with the BPF subsystem. To do so, we define
90flags on a set of kfuncs as follows::
91
92        BTF_SET8_START(bpf_task_set)
93        BTF_ID_FLAGS(func, bpf_get_task_pid, KF_ACQUIRE | KF_RET_NULL)
94        BTF_ID_FLAGS(func, bpf_put_pid, KF_RELEASE)
95        BTF_SET8_END(bpf_task_set)
96
97This set encodes the BTF ID of each kfunc listed above, and encodes the flags
98along with it. Ofcourse, it is also allowed to specify no flags.
99
1002.4.1 KF_ACQUIRE flag
101---------------------
102
103The KF_ACQUIRE flag is used to indicate that the kfunc returns a pointer to a
104refcounted object. The verifier will then ensure that the pointer to the object
105is eventually released using a release kfunc, or transferred to a map using a
106referenced kptr (by invoking bpf_kptr_xchg). If not, the verifier fails the
107loading of the BPF program until no lingering references remain in all possible
108explored states of the program.
109
1102.4.2 KF_RET_NULL flag
111----------------------
112
113The KF_RET_NULL flag is used to indicate that the pointer returned by the kfunc
114may be NULL. Hence, it forces the user to do a NULL check on the pointer
115returned from the kfunc before making use of it (dereferencing or passing to
116another helper). This flag is often used in pairing with KF_ACQUIRE flag, but
117both are orthogonal to each other.
118
1192.4.3 KF_RELEASE flag
120---------------------
121
122The KF_RELEASE flag is used to indicate that the kfunc releases the pointer
123passed in to it. There can be only one referenced pointer that can be passed in.
124All copies of the pointer being released are invalidated as a result of invoking
125kfunc with this flag.
126
1272.4.4 KF_KPTR_GET flag
128----------------------
129
130The KF_KPTR_GET flag is used to indicate that the kfunc takes the first argument
131as a pointer to kptr, safely increments the refcount of the object it points to,
132and returns a reference to the user. The rest of the arguments may be normal
133arguments of a kfunc. The KF_KPTR_GET flag should be used in conjunction with
134KF_ACQUIRE and KF_RET_NULL flags.
135
1362.4.5 KF_TRUSTED_ARGS flag
137--------------------------
138
139The KF_TRUSTED_ARGS flag is used for kfuncs taking pointer arguments. It
140indicates that the all pointer arguments will always have a guaranteed lifetime,
141and pointers to kernel objects are always passed to helpers in their unmodified
142form (as obtained from acquire kfuncs).
143
144It can be used to enforce that a pointer to a refcounted object acquired from a
145kfunc or BPF helper is passed as an argument to this kfunc without any
146modifications (e.g. pointer arithmetic) such that it is trusted and points to
147the original object.
148
149Meanwhile, it is also allowed pass pointers to normal memory to such kfuncs,
150but those can have a non-zero offset.
151
152This flag is often used for kfuncs that operate (change some property, perform
153some operation) on an object that was obtained using an acquire kfunc. Such
154kfuncs need an unchanged pointer to ensure the integrity of the operation being
155performed on the expected object.
156
1572.4.6 KF_SLEEPABLE flag
158-----------------------
159
160The KF_SLEEPABLE flag is used for kfuncs that may sleep. Such kfuncs can only
161be called by sleepable BPF programs (BPF_F_SLEEPABLE).
162
1632.4.7 KF_DESTRUCTIVE flag
164--------------------------
165
166The KF_DESTRUCTIVE flag is used to indicate functions calling which is
167destructive to the system. For example such a call can result in system
168rebooting or panicking. Due to this additional restrictions apply to these
169calls. At the moment they only require CAP_SYS_BOOT capability, but more can be
170added later.
171
1722.5 Registering the kfuncs
173--------------------------
174
175Once the kfunc is prepared for use, the final step to making it visible is
176registering it with the BPF subsystem. Registration is done per BPF program
177type. An example is shown below::
178
179        BTF_SET8_START(bpf_task_set)
180        BTF_ID_FLAGS(func, bpf_get_task_pid, KF_ACQUIRE | KF_RET_NULL)
181        BTF_ID_FLAGS(func, bpf_put_pid, KF_RELEASE)
182        BTF_SET8_END(bpf_task_set)
183
184        static const struct btf_kfunc_id_set bpf_task_kfunc_set = {
185                .owner = THIS_MODULE,
186                .set   = &bpf_task_set,
187        };
188
189        static int init_subsystem(void)
190        {
191                return register_btf_kfunc_id_set(BPF_PROG_TYPE_TRACING, &bpf_task_kfunc_set);
192        }
193        late_initcall(init_subsystem);
194