SCX_CMASK_DEFINE_SHARD(9)

Okay bestie, welcome to the Kernel API Manual page for SCX_CMASK_DEFINE_SHARD(9) 💖✨ This one is all about defining an on-stack cmask sized to one shard, and the vibe check is: keep your ranges in order or things get spicy in the bad way. 🫶

---

🌸 NAME 🌸

SCX_CMASK_DEFINE_SHARD — define an on-stack cmask sized to one shard</p> 🎀

---

✨ SYNOPSIS ✨

SCX_CMASK_DEFINE_SHARD (NAME , BASE , NR_CIDS );

That’s the whole call shape, bestie — no cap. 💅

---

🧚 ARGUMENTS 🧚

• NAME — the variable name you’re defining. Name it cute, name it clear, just name it. 💖
• BASE — the first cid of the active range. That’s where the range starts, periodt.
• NR_CIDS — the active range length. It must be <= SCX_CID_SHARD_MAX_CPUS, okay? This is not a drill. 😭

---

💀 DESCRIPTION 💀

Storage is fixed at SCX_CID_SHARD_MAX_CPUS, and the active range is framed by (BASE, NR_CIDS). Think of it like the vibes are set, the storage is set, and the range has to fit inside that whole situation. ✨

🚨 Warning, bestie: passing NR_CIDS > SCX_CID_SHARD_MAX_CPUS leaves the cmask claiming more bits than the storage actually holds. And then subsequent cmask operations will overrun. That is not slay. That is a bug. 💀

---

🦋 SEE ALSO 🦋

Peep the kernel file ./kernel/sched/ext/types.h, plus __SCX_CMASK_DEFINE(9) and SCX_CMASK_DEFINE(9) — iykyk. 🌈

---

Signed off with love from the source/version ./kernel/sched/ext, dated October 2026 🎀💖 Mwah!