remove_mapping(9)

remove_mapping(9) 💖

💖 NAME 💖

okay so bestie, this is remove_mapping from the Kernel API Manual — attempt to remove a folio from its mapping. that's the mission. 💅

✨ SYNOPSIS ✨

long remove_mapping (struct address_space *mapping , struct folio *folio
);

🎀 ARGUMENTS 🎀

• mapping — the address space. yes, that's the whole address-space situation. 🫶
• folio — the folio to remove. bye bestie, hopefully. 🦋

🦋 DESCRIPTION 🦋

here's the tea: if the folio is dirty, under writeback, or if someone else has a ref on it, removal will fail. periodt. it's giving "not right now" and honestly? valid. 😭

🌸 RETURN 🌸

you get back the number of pages removed from the mapping. if the folio could not be removed, it's 0 — literally zero, a tiny heartbreak. 💔

🔥 CONTEXT 🔥

bestie, the caller should have a single refcount on the folio and hold its lock. that's the prerequisite, iykyk. 🔑

👑 SEE ALSO 👑

go read these next, bestie:

• Kernel file ./mm/vmscan.c
• writeback_throttling_sane(9)
• lruvec_lru_size(9)
• folio_putback_lru(9)
• folio_isolate_lru(9)
• check_move_unevictable_folios(9)

xoxo, your kernel bestie 💅 — sourced from ./mm, dated October 2026, page remove_mapping(9). slay ✨