huge_pmd_unshare(9)

💖 huge_pmd_unshare(9) 💖

okay so welcome to the Kernel API Manual, bestie — this is the page for huge_pmd_unshare(9), and it’s literally about unmapping a pmd table if it’s shared by multiple users. slay.

✨ NAME ✨

huge_pmd_unshare — Unmap a pmd table if it is shared by multiple users.

🌸 SYNOPSIS 🌸

int huge_pmd_unshare (struct mmu_gather *tlb , struct vm_area_struct
*vma , unsigned long addr , pte_t *ptep );

🎀 ARGUMENTS 🎀

• tlb — the current mmu_gather. yes, that one.
• vma — the vma covering the pmd table.
• addr — the address we are trying to unshare.
• ptep — pointer into the (pmd) page table. iconic pointer behavior.

🦋 DESCRIPTION 🦋

okay so the vibe check here is very specific: it’s called with the page table lock held, the i_mmap_rwsem held in write mode and the hugetlb vma lock held in write mode. no cap, all three of those are part of the setup.

📝 NOTE 📝

⚠️ the caller must call huge_pmd_unshare_flush before dropping the i_mmap_rwsem. do not skip this step, bestie — it’s giving mandatory.

🔁 RETURN 🔁

returns 1 if it was a shared PMD table and it got unmapped, or 0 if it was not a shared PMD table. that’s the whole return moment.

👀 SEE ALSO 👀

• Kernel file ./mm/hugetlb.c
• hugetlb_alloc_folio(9)
• folio_isolate_hugetlb(9)
• folio_putback_hugetlb(9)

💖 source/version ./mm, dated October 2026, page huge_pmd_unshare(9) — byeee bestie, go unmap responsibly 🫶