bpf on x64
perform a command on an extended BPF map or program
sys_bpf(int cmd, union bpf_attr *attr, unsigned int size, struct bpf_common_attr *attr_common, unsigned int size_common)
| number | 321 (0x141) |
|---|---|
| abi | common |
| entry point | sys_bpf |
| since | Linux 3.18 |
| man page | bpf(2) |
| source | elixir.bootlin.com |
Registers
| register | value |
|---|---|
rax | 321, the syscall number |
rdi | int cmd |
rsi | union bpf_attr *attr |
rdx | unsigned int size |
r10 | struct bpf_common_attr *attr_common |
r8 | unsigned int size_common |
rax | the return value |
Then run syscall to enter the kernel.
The syscall instruction clobbers rcx (the return address lands in rcx) and r11 (the flags land in r11).
Since Linux 5.14 the kernel reads the syscall number from eax and ignores the upper 32 bits of rax. Older kernels fail with ENOSYS when those bits are not zero, and mov eax, N clears them, so it works on both.
Same syscall on other architectures
- x86 357
- arm64 280
- arm 386
- riscv64 280
- ia32 357
- x32 1073742145
- loongarch64 280
- ppc64 361
- ppc 361
- s390x 351
- mips64 5315
- mipsn32 6319
- mipso32 4355
- alpha 515
- arc 280
- blackfin 391, removed in 4.17
- cris 358, removed in 4.17
- csky 280
- hexagon 280
- ia64 1341, removed in 6.7
- m68k 354
- metag 280, removed in 4.17
- microblaze 387
- nds32 280, removed in 5.18
- nios2 280
- openrisc 280
- parisc 341
- score 280, removed in 4.17
- sh 375
- sparc 349
- sparc64 349
- tilegx 280, removed in 4.17
- tilepro 280, removed in 4.17
- xtensa 340
Get this as JSON: https://api.syscall.sh/v2/syscalls/x64/bpf