1================ 2bpftool-map 3================ 4------------------------------------------------------------------------------- 5tool for inspection and simple manipulation of eBPF maps 6------------------------------------------------------------------------------- 7 8:Manual section: 8 9 10SYNOPSIS 11======== 12 13 **bpftool** [*OPTIONS*] **map** *COMMAND* 14 15 *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-f** | **--bpffs** } } 16 17 *COMMANDS* := 18 { **show** | **list** | **create** | **dump** | **update** | **lookup** | **getnext** 19 | **delete** | **pin** | **help** } 20 21MAP COMMANDS 22============= 23 24| **bpftool** **map { show | list }** [*MAP*] 25| **bpftool** **map create** *FILE* **type** *TYPE* **key** *KEY_SIZE* **value** *VALUE_SIZE* \ 26| **entries** *MAX_ENTRIES* **name** *NAME* [**flags** *FLAGS*] [**dev** *NAME*] 27| **bpftool** **map dump** *MAP* 28| **bpftool** **map update** *MAP* [**key** *DATA*] [**value** *VALUE*] [*UPDATE_FLAGS*] 29| **bpftool** **map lookup** *MAP* [**key** *DATA*] 30| **bpftool** **map getnext** *MAP* [**key** *DATA*] 31| **bpftool** **map delete** *MAP* **key** *DATA* 32| **bpftool** **map pin** *MAP* *FILE* 33| **bpftool** **map event_pipe** *MAP* [**cpu** *N* **index** *M*] 34| **bpftool** **map peek** *MAP* 35| **bpftool** **map push** *MAP* **value** *VALUE* 36| **bpftool** **map pop** *MAP* 37| **bpftool** **map enqueue** *MAP* **value** *VALUE* 38| **bpftool** **map dequeue** *MAP* 39| **bpftool** **map freeze** *MAP* 40| **bpftool** **map help** 41| 42| *MAP* := { **id** *MAP_ID* | **pinned** *FILE* | **name** *MAP_NAME* } 43| *DATA* := { [**hex**] *BYTES* } 44| *PROG* := { **id** *PROG_ID* | **pinned** *FILE* | **tag** *PROG_TAG* | **name** *PROG_NAME* } 45| *VALUE* := { *DATA* | *MAP* | *PROG* } 46| *UPDATE_FLAGS* := { **any** | **exist** | **noexist** } 47| *TYPE* := { **hash** | **array** | **prog_array** | **perf_event_array** | **percpu_hash** 48| | **percpu_array** | **stack_trace** | **cgroup_array** | **lru_hash** 49| | **lru_percpu_hash** | **lpm_trie** | **array_of_maps** | **hash_of_maps** 50| | **devmap** | **devmap_hash** | **sockmap** | **cpumap** | **xskmap** | **sockhash** 51| | **cgroup_storage** | **reuseport_sockarray** | **percpu_cgroup_storage** 52| | **queue** | **stack** } 53 54DESCRIPTION 55=========== 56 **bpftool map { show | list }** [*MAP*] 57 Show information about loaded maps. If *MAP* is specified 58 show information only about given maps, otherwise list all 59 maps currently loaded on the system. In case of **name**, 60 *MAP* may match several maps which will all be shown. 61 62 Output will start with map ID followed by map type and 63 zero or more named attributes (depending on kernel version). 64 65 **bpftool map create** *FILE* **type** *TYPE* **key** *KEY_SIZE* **value** *VALUE_SIZE* **entries** *MAX_ENTRIES* **name** *NAME* [**flags** *FLAGS*] [**dev** *NAME*] 66 Create a new map with given parameters and pin it to *bpffs* 67 as *FILE*. 68 69 **bpftool map dump** *MAP* 70 Dump all entries in a given *MAP*. In case of **name**, 71 *MAP* may match several maps which will all be dumped. 72 73 **bpftool map update** *MAP* [**key** *DATA*] [**value** *VALUE*] [*UPDATE_FLAGS*] 74 Update map entry for a given *KEY*. 75 76 *UPDATE_FLAGS* can be one of: **any** update existing entry 77 or add if doesn't exit; **exist** update only if entry already 78 exists; **noexist** update only if entry doesn't exist. 79 80 If the **hex** keyword is provided in front of the bytes 81 sequence, the bytes are parsed as hexadeximal values, even if 82 no "0x" prefix is added. If the keyword is not provided, then 83 the bytes are parsed as decimal values, unless a "0x" prefix 84 (for hexadecimal) or a "0" prefix (for octal) is provided. 85 86 **bpftool map lookup** *MAP* [**key** *DATA*] 87 Lookup **key** in the map. 88 89 **bpftool map getnext** *MAP* [**key** *DATA*] 90 Get next key. If *key* is not specified, get first key. 91 92 **bpftool map delete** *MAP* **key** *DATA* 93 Remove entry from the map. 94 95 **bpftool map pin** *MAP* *FILE* 96 Pin map *MAP* as *FILE*. 97 98 Note: *FILE* must be located in *bpffs* mount. It must not 99 contain a dot character ('.'), which is reserved for future 100 extensions of *bpffs*. 101 102 **bpftool** **map event_pipe** *MAP* [**cpu** *N* **index** *M*] 103 Read events from a BPF_MAP_TYPE_PERF_EVENT_ARRAY map. 104 105 Install perf rings into a perf event array map and dump 106 output of any bpf_perf_event_output() call in the kernel. 107 By default read the number of CPUs on the system and 108 install perf ring for each CPU in the corresponding index 109 in the array. 110 111 If **cpu** and **index** are specified, install perf ring 112 for given **cpu** at **index** in the array (single ring). 113 114 Note that installing a perf ring into an array will silently 115 replace any existing ring. Any other application will stop 116 receiving events if it installed its rings earlier. 117 118 **bpftool map peek** *MAP* 119 Peek next **value** in the queue or stack. 120 121 **bpftool map push** *MAP* **value** *VALUE* 122 Push **value** onto the stack. 123 124 **bpftool map pop** *MAP* 125 Pop and print **value** from the stack. 126 127 **bpftool map enqueue** *MAP* **value** *VALUE* 128 Enqueue **value** into the queue. 129 130 **bpftool map dequeue** *MAP* 131 Dequeue and print **value** from the queue. 132 133 **bpftool map freeze** *MAP* 134 Freeze the map as read-only from user space. Entries from a 135 frozen map can not longer be updated or deleted with the 136 **bpf\ ()** system call. This operation is not reversible, 137 and the map remains immutable from user space until its 138 destruction. However, read and write permissions for BPF 139 programs to the map remain unchanged. 140 141 **bpftool map help** 142 Print short help message. 143 144OPTIONS 145======= 146 -h, --help 147 Print short generic help message (similar to **bpftool help**). 148 149 -V, --version 150 Print version number (similar to **bpftool version**). 151 152 -j, --json 153 Generate JSON output. For commands that cannot produce JSON, this 154 option has no effect. 155 156 -p, --pretty 157 Generate human-readable JSON output. Implies **-j**. 158 159 -f, --bpffs 160 Show file names of pinned maps. 161 162 -n, --nomount 163 Do not automatically attempt to mount any virtual file system 164 (such as tracefs or BPF virtual file system) when necessary. 165 166 -d, --debug 167 Print all logs available from libbpf, including debug-level 168 information. 169 170EXAMPLES 171======== 172**# bpftool map show** 173:: 174 175 10: hash name some_map flags 0x0 176 key 4B value 8B max_entries 2048 memlock 167936B 177 178The following three commands are equivalent: 179 180| 181| **# bpftool map update id 10 key hex 20 c4 b7 00 value hex 0f ff ff ab 01 02 03 4c** 182| **# bpftool map update id 10 key 0x20 0xc4 0xb7 0x00 value 0x0f 0xff 0xff 0xab 0x01 0x02 0x03 0x4c** 183| **# bpftool map update id 10 key 32 196 183 0 value 15 255 255 171 1 2 3 76** 184 185**# bpftool map lookup id 10 key 0 1 2 3** 186 187:: 188 189 key: 00 01 02 03 value: 00 01 02 03 04 05 06 07 190 191 192**# bpftool map dump id 10** 193:: 194 195 key: 00 01 02 03 value: 00 01 02 03 04 05 06 07 196 key: 0d 00 07 00 value: 02 00 00 00 01 02 03 04 197 Found 2 elements 198 199**# bpftool map getnext id 10 key 0 1 2 3** 200:: 201 202 key: 203 00 01 02 03 204 next key: 205 0d 00 07 00 206 207| 208| **# mount -t bpf none /sys/fs/bpf/** 209| **# bpftool map pin id 10 /sys/fs/bpf/map** 210| **# bpftool map del pinned /sys/fs/bpf/map key 13 00 07 00** 211 212Note that map update can also be used in order to change the program references 213hold by a program array map. This can be used, for example, to change the 214programs used for tail-call jumps at runtime, without having to reload the 215entry-point program. Below is an example for this use case: we load a program 216defining a prog array map, and with a main function that contains a tail call 217to other programs that can be used either to "process" packets or to "debug" 218processing. Note that the prog array map MUST be pinned into the BPF virtual 219file system for the map update to work successfully, as kernel flushes prog 220array maps when they have no more references from user space (and the update 221would be lost as soon as bpftool exits). 222 223| 224| **# bpftool prog loadall tail_calls.o /sys/fs/bpf/foo type xdp** 225| **# bpftool prog --bpffs** 226 227:: 228 229 545: xdp name main_func tag 674b4b5597193dc3 gpl 230 loaded_at 2018-12-12T15:02:58+0000 uid 0 231 xlated 240B jited 257B memlock 4096B map_ids 294 232 pinned /sys/fs/bpf/foo/xdp 233 546: xdp name bpf_func_process tag e369a529024751fc gpl 234 loaded_at 2018-12-12T15:02:58+0000 uid 0 235 xlated 200B jited 164B memlock 4096B 236 pinned /sys/fs/bpf/foo/process 237 547: xdp name bpf_func_debug tag 0b597868bc7f0976 gpl 238 loaded_at 2018-12-12T15:02:58+0000 uid 0 239 xlated 200B jited 164B memlock 4096B 240 pinned /sys/fs/bpf/foo/debug 241 242**# bpftool map** 243 244:: 245 246 294: prog_array name jmp_table flags 0x0 247 key 4B value 4B max_entries 1 memlock 4096B 248 owner_prog_type xdp owner jited 249 250| 251| **# bpftool map pin id 294 /sys/fs/bpf/bar** 252| **# bpftool map dump pinned /sys/fs/bpf/bar** 253 254:: 255 256 Found 0 elements 257 258| 259| **# bpftool map update pinned /sys/fs/bpf/bar key 0 0 0 0 value pinned /sys/fs/bpf/foo/debug** 260| **# bpftool map dump pinned /sys/fs/bpf/bar** 261 262:: 263 264 key: 00 00 00 00 value: 22 02 00 00 265 Found 1 element 266 267SEE ALSO 268======== 269 **bpf**\ (2), 270 **bpf-helpers**\ (7), 271 **bpftool**\ (8), 272 **bpftool-prog**\ (8), 273 **bpftool-cgroup**\ (8), 274 **bpftool-feature**\ (8), 275 **bpftool-net**\ (8), 276 **bpftool-perf**\ (8), 277 **bpftool-btf**\ (8) 278