1 .. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
6 -------------------------------------------------------------------------------
7 tool to create BPF iterators
8 -------------------------------------------------------------------------------
12 .. include:: substitutions.rst
17 **bpftool** [*OPTIONS*] **iter** *COMMAND*
19 *OPTIONS* := { |COMMON_OPTIONS| }
21 *COMMANDS* := { **pin** | **help** }
26 | **bpftool** **iter pin** *OBJ* *PATH* [**map** *MAP*]
27 | **bpftool** **iter help**
29 | *OBJ* := /a/file/of/bpf_iter_target.o
30 | *MAP* := { **id** *MAP_ID* | **pinned** *FILE* }
34 **bpftool iter pin** *OBJ* *PATH* [**map** *MAP*]
35 A bpf iterator combines a kernel iterating of
36 particular kernel data (e.g., tasks, bpf_maps, etc.)
37 and a bpf program called for each kernel data object
38 (e.g., one task, one bpf_map, etc.). User space can
39 *read* kernel iterator output through *read()* syscall.
41 The *pin* command creates a bpf iterator from *OBJ*,
42 and pin it to *PATH*. The *PATH* should be located
43 in *bpffs* mount. It must not contain a dot
44 character ('.'), which is reserved for future extensions
47 Map element bpf iterator requires an additional parameter
48 *MAP* so bpf program can iterate over map elements for
49 that map. User can have a bpf program in kernel to run
50 with each map element, do checking, filtering, aggregation,
51 etc. without copying data to user space.
53 User can then *cat PATH* to see the bpf iterator output.
56 Print short help message.
60 .. include:: common_options.rst
64 **# bpftool iter pin bpf_iter_netlink.o /sys/fs/bpf/my_netlink**
68 Create a file-based bpf iterator from bpf_iter_netlink.o and pin it
69 to /sys/fs/bpf/my_netlink
71 **# bpftool iter pin bpf_iter_hashmap.o /sys/fs/bpf/my_hashmap map id 20**
75 Create a file-based bpf iterator from bpf_iter_hashmap.o and map with
76 id 20, and pin it to /sys/fs/bpf/my_hashmap