cond_synchro...cu_expedited(9)

cond_synchro...cu_expedited(9) 💅

okay so welcome to the Kernel API Manual corner, bestie — this one is short, sweet, and lowkey iconic ✨

---

🌸 NAME 🌸

cond_synchronize_rcu_expedited — conditionally wait for an expedited RCU grace period. That’s the whole vibe 🎀

---

✨ SYNOPSIS ✨

void cond_synchronize_rcu_expedited (unsigned long oldstate );

---

🎀 ARGUMENTS 🎀

• oldstate — this is the value you got from get_state_synchronize_rcu, start_poll_synchronize_rcu, or start_poll_synchronize_rcu_expedited. It’s your little state ticket 🎟️✨

---

💅 DESCRIPTION 💅

okay so here’s the situation: if any type of full RCU grace period has already elapsed since that earlier call to get_state_synchronize_rcu, start_poll_synchronize_rcu, or start_poll_synchronize_rcu_expedited, then we literally just return. No extra waiting, no drama 💖

Otherwise? We invoke synchronize_rcu_expedited to wait for a full grace period. That’s the move 🔥

And yes, bestie, this function does not take counter wrap into account. But counter wrap is harmless — no cap. If the counter wraps, we have waited for more than 2 billion grace periods (and way more on a 64-bit system!), so waiting for a couple of additional grace periods should be just fine. Slay 💀✨

The memory-ordering guarantees? They are the same ones that would be provided by a synchronize_rcu that was invoked at the call to the function that provided oldstate and that returned at the end of this function. Iconic behavior, fr fr 👑

---

🔥 SEE ALSO 🔥

• Kernel file ./kernel/rcu/tree_exp.h: synchronize_rcu_expedited(9), start_poll_synchronize_rcu_expedited(9), start_poll_synchronize_rcu_expedited_full(9), cond_synchronize_rcu_expedited_full(9) 🦋

---

— xoxo, your Unix bestie, signing off from ./kernel/rcu in October 2026 as cond_synchro...cu_expedited(9) 💖✨