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** }] | { **-d** | **--debug** } | 16 { **-f** | **--bpffs** } | { **-n** | **--nomount** } } 17 18 *COMMANDS* := 19 { **show** | **list** | **create** | **dump** | **update** | **lookup** | **getnext** 20 | **delete** | **pin** | **help** } 21 22MAP COMMANDS 23============= 24 25| **bpftool** **map** { **show** | **list** } [*MAP*] 26| **bpftool** **map create** *FILE* **type** *TYPE* **key** *KEY_SIZE* **value** *VALUE_SIZE* \ 27| **entries** *MAX_ENTRIES* **name** *NAME* [**flags** *FLAGS*] [**inner_map** *MAP*] \ 28| [**dev** *NAME*] 29| **bpftool** **map dump** *MAP* 30| **bpftool** **map update** *MAP* [**key** *DATA*] [**value** *VALUE*] [*UPDATE_FLAGS*] 31| **bpftool** **map lookup** *MAP* [**key** *DATA*] 32| **bpftool** **map getnext** *MAP* [**key** *DATA*] 33| **bpftool** **map delete** *MAP* **key** *DATA* 34| **bpftool** **map pin** *MAP* *FILE* 35| **bpftool** **map event_pipe** *MAP* [**cpu** *N* **index** *M*] 36| **bpftool** **map peek** *MAP* 37| **bpftool** **map push** *MAP* **value** *VALUE* 38| **bpftool** **map pop** *MAP* 39| **bpftool** **map enqueue** *MAP* **value** *VALUE* 40| **bpftool** **map dequeue** *MAP* 41| **bpftool** **map freeze** *MAP* 42| **bpftool** **map help** 43| 44| *MAP* := { **id** *MAP_ID* | **pinned** *FILE* | **name** *MAP_NAME* } 45| *DATA* := { [**hex**] *BYTES* } 46| *PROG* := { **id** *PROG_ID* | **pinned** *FILE* | **tag** *PROG_TAG* | **name** *PROG_NAME* } 47| *VALUE* := { *DATA* | *MAP* | *PROG* } 48| *UPDATE_FLAGS* := { **any** | **exist** | **noexist** } 49| *TYPE* := { **hash** | **array** | **prog_array** | **perf_event_array** | **percpu_hash** 50| | **percpu_array** | **stack_trace** | **cgroup_array** | **lru_hash** 51| | **lru_percpu_hash** | **lpm_trie** | **array_of_maps** | **hash_of_maps** 52| | **devmap** | **devmap_hash** | **sockmap** | **cpumap** | **xskmap** | **sockhash** 53| | **cgroup_storage** | **reuseport_sockarray** | **percpu_cgroup_storage** 54| | **queue** | **stack** | **sk_storage** | **struct_ops** | **ringbuf** | **inode_storage** 55 | **task_storage** } 56 57DESCRIPTION 58=========== 59 **bpftool map { show | list }** [*MAP*] 60 Show information about loaded maps. If *MAP* is specified 61 show information only about given maps, otherwise list all 62 maps currently loaded on the system. In case of **name**, 63 *MAP* may match several maps which will all be shown. 64 65 Output will start with map ID followed by map type and 66 zero or more named attributes (depending on kernel version). 67 68 Since Linux 5.8 bpftool is able to discover information about 69 processes that hold open file descriptors (FDs) against BPF 70 maps. On such kernels bpftool will automatically emit this 71 information as well. 72 73 **bpftool map create** *FILE* **type** *TYPE* **key** *KEY_SIZE* **value** *VALUE_SIZE* **entries** *MAX_ENTRIES* **name** *NAME* [**flags** *FLAGS*] [**inner_map** *MAP*] [**dev** *NAME*] 74 Create a new map with given parameters and pin it to *bpffs* 75 as *FILE*. 76 77 *FLAGS* should be an integer which is the combination of 78 desired flags, e.g. 1024 for **BPF_F_MMAPABLE** (see bpf.h 79 UAPI header for existing flags). 80 81 To create maps of type array-of-maps or hash-of-maps, the 82 **inner_map** keyword must be used to pass an inner map. The 83 kernel needs it to collect metadata related to the inner maps 84 that the new map will work with. 85 86 Keyword **dev** expects a network interface name, and is used 87 to request hardware offload for the map. 88 89 **bpftool map dump** *MAP* 90 Dump all entries in a given *MAP*. In case of **name**, 91 *MAP* may match several maps which will all be dumped. 92 93 **bpftool map update** *MAP* [**key** *DATA*] [**value** *VALUE*] [*UPDATE_FLAGS*] 94 Update map entry for a given *KEY*. 95 96 *UPDATE_FLAGS* can be one of: **any** update existing entry 97 or add if doesn't exit; **exist** update only if entry already 98 exists; **noexist** update only if entry doesn't exist. 99 100 If the **hex** keyword is provided in front of the bytes 101 sequence, the bytes are parsed as hexadecimal values, even if 102 no "0x" prefix is added. If the keyword is not provided, then 103 the bytes are parsed as decimal values, unless a "0x" prefix 104 (for hexadecimal) or a "0" prefix (for octal) is provided. 105 106 **bpftool map lookup** *MAP* [**key** *DATA*] 107 Lookup **key** in the map. 108 109 **bpftool map getnext** *MAP* [**key** *DATA*] 110 Get next key. If *key* is not specified, get first key. 111 112 **bpftool map delete** *MAP* **key** *DATA* 113 Remove entry from the map. 114 115 **bpftool map pin** *MAP* *FILE* 116 Pin map *MAP* as *FILE*. 117 118 Note: *FILE* must be located in *bpffs* mount. It must not 119 contain a dot character ('.'), which is reserved for future 120 extensions of *bpffs*. 121 122 **bpftool** **map event_pipe** *MAP* [**cpu** *N* **index** *M*] 123 Read events from a **BPF_MAP_TYPE_PERF_EVENT_ARRAY** map. 124 125 Install perf rings into a perf event array map and dump 126 output of any **bpf_perf_event_output**\ () call in the kernel. 127 By default read the number of CPUs on the system and 128 install perf ring for each CPU in the corresponding index 129 in the array. 130 131 If **cpu** and **index** are specified, install perf ring 132 for given **cpu** at **index** in the array (single ring). 133 134 Note that installing a perf ring into an array will silently 135 replace any existing ring. Any other application will stop 136 receiving events if it installed its rings earlier. 137 138 **bpftool map peek** *MAP* 139 Peek next value in the queue or stack. 140 141 **bpftool map push** *MAP* **value** *VALUE* 142 Push *VALUE* onto the stack. 143 144 **bpftool map pop** *MAP* 145 Pop and print value from the stack. 146 147 **bpftool map enqueue** *MAP* **value** *VALUE* 148 Enqueue *VALUE* into the queue. 149 150 **bpftool map dequeue** *MAP* 151 Dequeue and print value from the queue. 152 153 **bpftool map freeze** *MAP* 154 Freeze the map as read-only from user space. Entries from a 155 frozen map can not longer be updated or deleted with the 156 **bpf**\ () system call. This operation is not reversible, 157 and the map remains immutable from user space until its 158 destruction. However, read and write permissions for BPF 159 programs to the map remain unchanged. 160 161 **bpftool map help** 162 Print short help message. 163 164OPTIONS 165======= 166 .. include:: common_options.rst 167 168 -f, --bpffs 169 Show file names of pinned maps. 170 171 -n, --nomount 172 Do not automatically attempt to mount any virtual file system 173 (such as tracefs or BPF virtual file system) when necessary. 174 175EXAMPLES 176======== 177**# bpftool map show** 178 179:: 180 181 10: hash name some_map flags 0x0 182 key 4B value 8B max_entries 2048 memlock 167936B 183 pids systemd(1) 184 185The following three commands are equivalent: 186 187| 188| **# bpftool map update id 10 key hex 20 c4 b7 00 value hex 0f ff ff ab 01 02 03 4c** 189| **# bpftool map update id 10 key 0x20 0xc4 0xb7 0x00 value 0x0f 0xff 0xff 0xab 0x01 0x02 0x03 0x4c** 190| **# bpftool map update id 10 key 32 196 183 0 value 15 255 255 171 1 2 3 76** 191 192**# bpftool map lookup id 10 key 0 1 2 3** 193 194:: 195 196 key: 00 01 02 03 value: 00 01 02 03 04 05 06 07 197 198 199**# bpftool map dump id 10** 200 201:: 202 203 key: 00 01 02 03 value: 00 01 02 03 04 05 06 07 204 key: 0d 00 07 00 value: 02 00 00 00 01 02 03 04 205 Found 2 elements 206 207**# bpftool map getnext id 10 key 0 1 2 3** 208 209:: 210 211 key: 212 00 01 02 03 213 next key: 214 0d 00 07 00 215 216| 217| **# mount -t bpf none /sys/fs/bpf/** 218| **# bpftool map pin id 10 /sys/fs/bpf/map** 219| **# bpftool map del pinned /sys/fs/bpf/map key 13 00 07 00** 220 221Note that map update can also be used in order to change the program references 222hold by a program array map. This can be used, for example, to change the 223programs used for tail-call jumps at runtime, without having to reload the 224entry-point program. Below is an example for this use case: we load a program 225defining a prog array map, and with a main function that contains a tail call 226to other programs that can be used either to "process" packets or to "debug" 227processing. Note that the prog array map MUST be pinned into the BPF virtual 228file system for the map update to work successfully, as kernel flushes prog 229array maps when they have no more references from user space (and the update 230would be lost as soon as bpftool exits). 231 232| 233| **# bpftool prog loadall tail_calls.o /sys/fs/bpf/foo type xdp** 234| **# bpftool prog --bpffs** 235 236:: 237 238 545: xdp name main_func tag 674b4b5597193dc3 gpl 239 loaded_at 2018-12-12T15:02:58+0000 uid 0 240 xlated 240B jited 257B memlock 4096B map_ids 294 241 pinned /sys/fs/bpf/foo/xdp 242 546: xdp name bpf_func_process tag e369a529024751fc gpl 243 loaded_at 2018-12-12T15:02:58+0000 uid 0 244 xlated 200B jited 164B memlock 4096B 245 pinned /sys/fs/bpf/foo/process 246 547: xdp name bpf_func_debug tag 0b597868bc7f0976 gpl 247 loaded_at 2018-12-12T15:02:58+0000 uid 0 248 xlated 200B jited 164B memlock 4096B 249 pinned /sys/fs/bpf/foo/debug 250 251**# bpftool map** 252 253:: 254 255 294: prog_array name jmp_table flags 0x0 256 key 4B value 4B max_entries 1 memlock 4096B 257 owner_prog_type xdp owner jited 258 259| 260| **# bpftool map pin id 294 /sys/fs/bpf/bar** 261| **# bpftool map dump pinned /sys/fs/bpf/bar** 262 263:: 264 265 Found 0 elements 266 267| 268| **# bpftool map update pinned /sys/fs/bpf/bar key 0 0 0 0 value pinned /sys/fs/bpf/foo/debug** 269| **# bpftool map dump pinned /sys/fs/bpf/bar** 270 271:: 272 273 key: 00 00 00 00 value: 22 02 00 00 274 Found 1 element 275