page_vma_mapped_walk(9)

💅✨ page_vma_mapped_walk(9) ✨💅

okay so this is a page from the Kernel API Manual bestie, and it's genuinely important so buckle up 🎀

---

🌸 NAME 🌸

page_vma_mapped_walk — literally just checks if @pvmw->pfn is mapped inside @pvmw->vma at @pvmw->address 💖

⭐ SYNOPSIS ⭐

bool page_vma_mapped_walk (struct page_vma_mapped_walk *pvmw );

🦋 ARGUMENTS 🦋

• pvmw — a pointer to struct page_vma_mapped_walk, no cap. but like, the rules are strict: page, vma, address and flags must be set already, and pmd, pte and ptl must be NULL. it's giving "fill out the form before you walk" 🫶

💜 DESCRIPTION 💜

so it returns true if the page is mapped in the vma, periodt. when that happens, pvmw->pmd and pvmw->pte point to the relevant page table entries, pvmw->ptl is locked, and pvmw->address gets adjusted if that's needed (looking at you, PTE-mapped THPs 👀).

if pvmw->pmd is set but pvmw->pte is not, congrats bestie, you found a PMD-mapped page (usually a THP). but for PTE-mapped THP you gotta run page_vma_mapped_walk in a loop to find all the PTEs that map the THP — yep, a whole loop, she's thorough 💅

for HugeTLB pages, pvmw->pte is set to the relevant page table entry regardless of which page table level the page is mapped at, and pvmw->pmd is NULL. iconic behaviour honestly 🌈

it returns false when there are no more page table entries for the page in the vma — and then pvmw->ptl is unlocked and pvmw->pte is unmapped.

⚠️ and if you need to bail out of the walk before page_vma_mapped_walk returned false, use page_vma_mapped_walk_done — she does the housekeeping for you so you don't leave a mess 😭✨

👑 SEE ALSO 👑

• Kernel file ./mm/page_vma_mapped.c check_pte(9), page_mapped_in_vma(9) 🔥

---

and that's the whole vibe, bestie 🌸 this one's straight from ./mm, dated October 2026 — mwah, go map some pages, ily 💖🎀🦋