Loading...
1.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
2
3===============
4bpftool-feature
5===============
6-------------------------------------------------------------------------------
7tool for inspection of eBPF-related parameters for Linux kernel or net device
8-------------------------------------------------------------------------------
9
10:Manual section: 8
11
12.. include:: substitutions.rst
13
14SYNOPSIS
15========
16
17**bpftool** [*OPTIONS*] **feature** *COMMAND*
18
19*OPTIONS* := { |COMMON_OPTIONS| }
20
21*COMMANDS* := { **probe** | **help** }
22
23FEATURE COMMANDS
24================
25
26| **bpftool** **feature probe** [*COMPONENT*] [**full**] [**unprivileged**] [**macros** [**prefix** *PREFIX*]]
27| **bpftool** **feature list_builtins** *GROUP*
28| **bpftool** **feature help**
29|
30| *COMPONENT* := { **kernel** | **dev** *NAME* }
31| *GROUP* := { **prog_types** | **map_types** | **attach_types** | **link_types** | **helpers** }
32
33DESCRIPTION
34===========
35bpftool feature probe [kernel] [full] [macros [prefix *PREFIX*]]
36 Probe the running kernel and dump a number of eBPF-related parameters, such
37 as availability of the **bpf**\ () system call, JIT status, eBPF program
38 types availability, eBPF helper functions availability, and more.
39
40 By default, bpftool **does not run probes** for **bpf_probe_write_user**\
41 () and **bpf_trace_printk**\() helpers which print warnings to kernel logs.
42 To enable them and run all probes, the **full** keyword should be used.
43
44 If the **macros** keyword (but not the **-j** option) is passed, a subset
45 of the output is dumped as a list of **#define** macros that are ready to
46 be included in a C header file, for example. If, additionally, **prefix**
47 is used to define a *PREFIX*, the provided string will be used as a prefix
48 to the names of the macros: this can be used to avoid conflicts on macro
49 names when including the output of this command as a header file.
50
51 Keyword **kernel** can be omitted. If no probe target is specified, probing
52 the kernel is the default behaviour.
53
54 When the **unprivileged** keyword is used, bpftool will dump only the
55 features available to a user who does not have the **CAP_SYS_ADMIN**
56 capability set. The features available in that case usually represent a
57 small subset of the parameters supported by the system. Unprivileged users
58 MUST use the **unprivileged** keyword: This is to avoid misdetection if
59 bpftool is inadvertently run as non-root, for example. This keyword is
60 unavailable if bpftool was compiled without libcap.
61
62bpftool feature probe dev *NAME* [full] [macros [prefix *PREFIX*]]
63 Probe network device for supported eBPF features and dump results to the
64 console.
65
66 The keywords **full**, **macros** and **prefix** have the same role as when
67 probing the kernel.
68
69bpftool feature list_builtins *GROUP*
70 List items known to bpftool. These can be BPF program types
71 (**prog_types**), BPF map types (**map_types**), attach types
72 (**attach_types**), link types (**link_types**), or BPF helper functions
73 (**helpers**). The command does not probe the system, but simply lists the
74 elements that bpftool knows from compilation time, as provided from libbpf
75 (for all object types) or from the BPF UAPI header (list of helpers). This
76 can be used in scripts to iterate over BPF types or helpers.
77
78bpftool feature help
79 Print short help message.
80
81OPTIONS
82=======
83.. include:: common_options.rst
1===============
2bpftool-feature
3===============
4-------------------------------------------------------------------------------
5tool for inspection of eBPF-related parameters for Linux kernel or net device
6-------------------------------------------------------------------------------
7
8:Manual section: 8
9
10SYNOPSIS
11========
12
13 **bpftool** [*OPTIONS*] **feature** *COMMAND*
14
15 *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] }
16
17 *COMMANDS* := { **probe** | **help** }
18
19FEATURE COMMANDS
20================
21
22| **bpftool** **feature probe** [*COMPONENT*] [**full**] [**unprivileged**] [**macros** [**prefix** *PREFIX*]]
23| **bpftool** **feature help**
24|
25| *COMPONENT* := { **kernel** | **dev** *NAME* }
26
27DESCRIPTION
28===========
29 **bpftool feature probe** [**kernel**] [**full**] [**macros** [**prefix** *PREFIX*]]
30 Probe the running kernel and dump a number of eBPF-related
31 parameters, such as availability of the **bpf**\ () system call,
32 JIT status, eBPF program types availability, eBPF helper
33 functions availability, and more.
34
35 By default, bpftool **does not run probes** for
36 **bpf_probe_write_user**\ () and **bpf_trace_printk**\()
37 helpers which print warnings to kernel logs. To enable them
38 and run all probes, the **full** keyword should be used.
39
40 If the **macros** keyword (but not the **-j** option) is
41 passed, a subset of the output is dumped as a list of
42 **#define** macros that are ready to be included in a C
43 header file, for example. If, additionally, **prefix** is
44 used to define a *PREFIX*, the provided string will be used
45 as a prefix to the names of the macros: this can be used to
46 avoid conflicts on macro names when including the output of
47 this command as a header file.
48
49 Keyword **kernel** can be omitted. If no probe target is
50 specified, probing the kernel is the default behaviour.
51
52 When the **unprivileged** keyword is used, bpftool will dump
53 only the features available to a user who does not have the
54 **CAP_SYS_ADMIN** capability set. The features available in
55 that case usually represent a small subset of the parameters
56 supported by the system. Unprivileged users MUST use the
57 **unprivileged** keyword: This is to avoid misdetection if
58 bpftool is inadvertently run as non-root, for example. This
59 keyword is unavailable if bpftool was compiled without
60 libcap.
61
62 **bpftool feature probe dev** *NAME* [**full**] [**macros** [**prefix** *PREFIX*]]
63 Probe network device for supported eBPF features and dump
64 results to the console.
65
66 The keywords **full**, **macros** and **prefix** have the
67 same role as when probing the kernel.
68
69 **bpftool feature help**
70 Print short help message.
71
72OPTIONS
73=======
74 -h, --help
75 Print short generic help message (similar to **bpftool help**).
76
77 -V, --version
78 Print version number (similar to **bpftool version**).
79
80 -j, --json
81 Generate JSON output. For commands that cannot produce JSON, this
82 option has no effect.
83
84 -p, --pretty
85 Generate human-readable JSON output. Implies **-j**.
86
87 -d, --debug
88 Print all logs available from libbpf, including debug-level
89 information.
90
91SEE ALSO
92========
93 **bpf**\ (2),
94 **bpf-helpers**\ (7),
95 **bpftool**\ (8),
96 **bpftool-btf**\ (8),
97 **bpftool-cgroup**\ (8),
98 **bpftool-gen**\ (8),
99 **bpftool-iter**\ (8),
100 **bpftool-link**\ (8),
101 **bpftool-map**\ (8),
102 **bpftool-net**\ (8),
103 **bpftool-perf**\ (8),
104 **bpftool-prog**\ (8),
105 **bpftool-struct_ops**\ (8)