unlock_sock_fast(9)

💅 unlock_sock_fast — it's literally in the name, bestie

okay so welcome to the Kernel API Manual ✨ section 9 ✨, where we're talking about unlock_sock_fast — the iconic, unmatched, main character complement of lock_sock_fast. you locked it fast, now you unlock it. symmetry, periodt. 🎀

---

🌸 NAME 🌸

unlock_sock_fast — the complement of lock_sock_fast

like... that's the whole vibe. they're a matching set. 💖

---

✨ SYNOPSIS ✨

void unlock_sock_fast (struct sock *sk , bool slow );

returns nothing (void), takes two things. that's it. that's the function. slay. 🫶

---

🦋 ARGUMENTS 🦋

• sk — the socket. the actual socket. our girl. 💜
• slow — slow mode. yeah, it's a bool, and yeah, it matters (see below 👀).

---

🔥 DESCRIPTION 🔥

so this is the fast unlock socket for user context. 🧚

here's the tea: if slow mode is on, we call the regular release_sock. that's it — that's the branch. slow = true and you're taking the regular path, no cap.

lowkey the whole point here: fast unlock is the fast path, but "slow mode" being enabled means it hands off to the normal, regular release_sock instead. so don't forget to pass that flag correctly, bestie — it's literally the difference between the two paths. 💀

---

👑 SEE ALSO 👑

the whole squad, in the original order, no skipping (i would never):

• Kernel file ./include/net/sock.h
• struct sock_common(9)
• struct sock(9)
• sk_user_data_is_nocopy(9)
• __locked_read_sk_user_data_with_flags(9)
• __rcu_dereference_sk_user_data_with_flags(9)
• sk_for_each_entry_offset_rcu(9)
• SOCK_CONNECT_BIND(9)
• lock_sock_fast(9)
• sk_wmem_alloc_get(9)
• sk_rmem_alloc_get(9)
• sk_has_allocations(9)
• skwq_has_sleeper(9)
• sock_poll_wait(9)
• sk_page_frag(9)
• _sock_tx_timestamp(9)
• sk_eat_skb(9)

---

and that's the page, bestie 💅✨ go forth and unlock responsibly 🍓

signed, sealed, delivered from ./include/net 🌈 — dated October 2026 🎀, and yes, this is unlock_sock_fast(9) the whole way through, iykyk 💖