iommufd_access_notify_unmap(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… iommufd_access_notify_unmap(9) πŸ’…

okay bestie, welcome to the Kernel API Manual page for iommufd_access_notify_unmap(9) β€” this is the little notification that tells users of an iopt to stop using it, and honestly it’s giving β€œwrap it up” energy πŸŽ€βœ¨.

πŸ¦‹ NAME πŸ¦‹

iommufd_access_notify_unmap β€” notify users of an iopt to stop using it. periodt πŸ’…

✨ SYNOPSIS ✨

void iommufd_access_notify_unmap (struct io_pagetable *iopt , unsigned
long iova , unsigned long length );

πŸŽ€ ARGUMENTS πŸŽ€

  • iopt β€” the iopt to work on, aka the main character of this operation πŸ‘‘.
  • iova β€” starting iova in the iopt πŸ—ΊοΈ.
  • length β€” number of bytes πŸ“.

πŸ’– DESCRIPTION πŸ’–

okay so after this function returns, there should be no users attached to the pages linked to this iopt that intersect with iova,length β€” like, zero, none, nobody πŸ’€. Anyone that has attached a user through iopt_access_pages needs to detach it through iommufd_access_unpin_pages before this function returns. no cap, that’s the whole vibe check πŸ™ˆ.

iommufd_access_destroy will wait for any outstanding unmap callback to complete 🫢. After iommufd_access_destroy, no unmap ops are running or will run in the future.

⚠️ Due to this, a driver must not create locking that prevents unmap to complete while iommufd_access_destroy is running. That’s literally the caveat, don’t skip it πŸ’….

🌈 SEE ALSO 🌈

Kernel file: ./drivers/iommu/iommufd/device.c β€” yes, that’s the spot 🫢.

xoxo, your kernel bestie β€” source ./drivers/iommu/iommufd, October 2026 πŸŽ€πŸ’–

πŸ“¦ where this page comes from