sys_membarrier(9)

💅 sys_membarrier — NAME 💅

okay so bestie, welcome to sys_membarrier(9) from the Kernel API Manual, section 9 🎀 — the syscall that literally issues memory barriers on a set of threads. yes, that one. sit down, hydrate, let's go ✨

---

✨ SYNOPSIS ✨

long sys_membarrier (int cmd , unsigned int flags , int cpu_id );

three args, one long return, zero chill. slay 🫶

---

🎀 ARGUMENTS 🎀

• cmd — takes command values defined in enum membarrier_cmd. that's the whole vibe, it's an enum, it picks what you're doing 👑
• flags — currently needs to be 0 for all commands other than MEMBARRIER_CMD_PRIVATE_EXPEDITED_RSEQ. in that one specific case it can be MEMBARRIER_CMD_FLAG_CPU, which means "hey, cpu_id contains the CPU on which to interrupt (= restart) the RSEQ critical section" 👀
• cpu_id — if flags == MEMBARRIER_CMD_FLAG_CPU, this indicates the cpu on which the RSEQ CS should be interrupted (and cmd must be MEMBARRIER_CMD_PRIVATE_EXPEDITED_RSEQ). no flag, no cpu-picking, that's the rule 💅

---

🦋 DESCRIPTION 🦋

first things first, the error lore 😭

• if this system call is not implemented, -ENOSYS is returned.
• if the command specified does not exist, isn't available on the running kernel, or if the command argument is invalid, this system call returns -EINVAL.
• for a given command, with the flags argument set to 0, if this system call returns -ENOSYS or -EINVAL, it is guaranteed to always return the same value until reboot. like, it's locked in fr fr 🔒
• in addition, it can return -ENOMEM if there is not enough memory available to perform the system call. memory said no 💀

and now the actual promise of the thing, which is honestly kind of iconic:

all memory accesses performed in program order from each targeted thread is guaranteed to be ordered with respect to sys_membarrier. 🧚

if we use the semantic "barrier" to represent a compiler barrier forcing memory accesses to be performed in program order across the barrier, and smp_mb to represent explicit memory barriers forcing full memory ordering across the barrier, we have the following ordering table for each pair of barrier, sys_membarrier and smp_mb:

the pair ordering is detailed as (O: ordered, X: not ordered):

       barrier   smp_mb sys_membarrier barrier          X           X
       O smp_mb           X           O            O sys_membarrier   O
       O            O

ngl bestie, this table arrived looking a lil scrambled from the source — I'm handing it to you character for character exactly as it appears, no fixing, no guessing 💅 the meaning you can trust is the O = ordered / X = not ordered key above 👑

---

🌸 SEE ALSO 🌸

• Kernel file ./kernel/sched/membarrier.c — go read the source, it's giving main character energy 📖✨

---

and that's the page, bestie 🎀 source is ./kernel/sched, dated October 2026, and it signs off as sys_membarrier(9) — periodt 💖🦋🌈