iommufd_access_pin_pages(9)

📦 version 7.3.1, in nixpkgs unstable

✨ iommufd_access_pin_pages ✨

okay so bestie, welcome to the iommufd_access_pin_pages page from the Kernel API Manual 💖 this one's a lil kernel-level baddie that hands you back a whole list of pages sitting under an iova. no cap, it's giving "let me get those pages for you" energy 🎀

🌸 SYNOPSIS 🌸

int iommufd_access_pin_pages (struct iommufd_access *access , unsigned
long iova , unsigned long length , struct page **out_pages , unsigned
int flags );

literally the whole vibe in one line, iykyk 💅

🦋 ARGUMENTS 🦋

  • access — the IOAS access you're acting on, like, the main character here 👑
  • iova — the starting IOVA, that's where we begin bestie ⭐
  • length — how many bytes you wanna access, tell me how much and I'll go get it 🫶
  • out_pages — the output page list, this is where the goodies land 🍓
  • flags — the IOPMMUFD_ACCESS_RW_* flags, yeah that's exactly how it's spelled in the source, not me being quirky 💀

💜 DESCRIPTION 💜

So what it does: it reads length bytes starting at iova and returns the struct page * pointers. then you (the caller) can kmap them for CPU access. slay 🔥

bestie listen — the caller must perform iommufd_access_unpin_pages when done to balance this. it's literally the rule, don't leave it hanging 😭

This API always requires a page aligned iova. no exceptions, periodt. this happens naturally if the ioas alignment is >= PAGE_SIZE and the iova is PAGE_SIZE aligned — iconic, love that for you ✨

but lowkey, smaller alignments have corner cases where this API can fail on an otherwise aligned iova. so like, heads up 🙈

🌈 SEE ALSO 🌈

Kernel file ./drivers/iommu/iommufd/device.c and honestly the whole friend group:

and that's the whole page bestie — brought to you by ./drivers/iommu/iommufd, dated October 2026, page iommufd_access_pin_pages(9), Kernel API Manual edition 🌸💖🎀 go pin some pages, you got this 🧚💜

📦 where this page comes from