io_uring_enter Linux system call
Submit queued io_uring requests, wait for completions, or do both in one call.
In Linux 7.3, io_uring_enter is defined in the system call table of 12 of the 12 bundled ABIs.
Number on each ABI Linux 7.3
| ABI | No. | Hex | Entry point | Instruction | NR | Argument locations | Kernel table | Interactive view |
|---|---|---|---|---|---|---|---|---|
| x86-64 | 426 | 0x1aa | sys_io_uring_enter | syscall | rax | rdi, rsi, rdx, r10, r8, r9 | syscall_64.tbl:353 | Open |
| i386 | 426 | 0x1aa | sys_io_uring_enter | int 0x80 | eax | ebx, ecx, edx, esi, edi, ebp | syscall_32.tbl:434 | Open |
| AArch64 | 426 | 0x1aa | sys_io_uring_enter | svc #0 | x8 | x0, x1, x2, x3, x4, x5 | syscall.tbl:369 | Open |
| ARM EABI | 426 | 0x1aa | sys_io_uring_enter | svc #0 | r7 | r0, r1, r2, r3, r4, r5 | syscall.tbl:443 | Open |
| RISC-V 64 | 426 | 0x1aa | sys_io_uring_enter | ecall | a7 | a0, a1, a2, a3, a4, a5 | syscall.tbl:369 | Open |
| LoongArch 64 | 426 | 0x1aa | sys_io_uring_enter | syscall 0 | a7 | a0, a1, a2, a3, a4, a5 | syscall.tbl:369 | Open |
| PowerPC 64 | 426 | 0x1aa | sys_io_uring_enter | sc | r0 | r3, r4, r5, r6, r7, r8 | syscall.tbl:519 | Open |
| PowerPC 32 | 426 | 0x1aa | sys_io_uring_enter | sc | r0 | r3, r4, r5, r6, r7, r8 | syscall.tbl:519 | Open |
| IBM Z (s390x) | 426 | 0x1aa | sys_io_uring_enter | svc 0 | r1 | r2, r3, r4, r5, r6, r7 | syscall.tbl:355 | Open |
| MIPS o32 | 4426 | 0x114a | sys_io_uring_enter | syscall | v0 | a0, a1, a2, a3, Stack (5), Stack (6) | syscall_o32.tbl:416 | Open |
| MIPS n32 | 6426 | 0x191a | sys_io_uring_enter | syscall | v0 | a0, a1, a2, a3, a4, a5 | syscall_n32.tbl:367 | Open |
| MIPS n64 | 5426 | 0x1532 | sys_io_uring_enter | syscall | v0 | a0, a1, a2, a3, a4, a5 | syscall_n64.tbl:343 | Open |
Argument locations are the ABI’s calling convention for ARG1–ARG6 in order, not an assignment made for this call. On 32-bit ABIs, 64-bit values may occupy multiple registers and require alignment; assembly wrappers may reorder them. See syscall(2). A table entry does not mean the call exists in every kernel: most optional calls return ENOSYS when the matching configuration option or architecture support is not built.
Kernel declaration
asmlinkage long sys_io_uring_enter(unsigned int fd, u32 to_submit, u32 min_complete, u32 flags, const void __user *argp, size_t argsz);
x86-64, i386, AArch64, ARM EABI, RISC-V 64, LoongArch 64, PowerPC 64, PowerPC 32, IBM Z (s390x), MIPS o32, MIPS n32, MIPS n64 · Declaration: include/linux/syscalls.h:340
Signatures are kernel declarations and may differ from the C library wrappers.
Manual page
Bundled kernel releases
| Release | ABIs bundled | Defined in the table of |
|---|---|---|
| Linux 7.3 | 12 | all 12 |
| Linux 6.18 | 12 | all 12 |
| Linux 6.12 | 12 | all 12 |
| Linux 6.6 | 12 | all 12 |
| Linux 5.19 | 12 | all 12 |
| Linux 4.15 | 5 | none |
| Linux 3.7 | 4 | none |
| Linux 3.0 | 3 | none |
| Linux 2.6.24 | 3 | none |
| Linux 2.6.16 | 3 | none |
| Linux 2.6.12 | 2 | none |
| Linux 1.0 | 1 | none |
Each release is a separate snapshot, and older releases bundle fewer ABIs. This table does not claim the release in which the call first appeared. Numbers are listed where a release differs from Linux 7.3.