ib_dma_unmap_single(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… ib_dma_unmap_single(9) β€” from the Kernel API Manual, bestie ✨

okay so, hi 🫢 you found the page for ib_dma_unmap_single(9) and honestly? it's a lil cleanup queen. this is the Kernel API Manual's entry, and we're gonna get through it together.


🌸 NAME 🌸

ib_dma_unmap_single β€” the function that destroys a mapping that was created by ib_dma_map_single() πŸ’€βœ¨

like, it's the breakup arc. ib_dma_map_single() set it up, and this one says "we're done here." iconic behavior, no cap.


πŸŽ€ SYNOPSIS πŸŽ€

void ib_dma_unmap_single (struct ib_device *dev , u64 addr , size_t size
, enum dma_data_direction direction );

literally just those four things, in that order, and it returns void so don't go looking for a return value bestie, there isn't one πŸ™ˆ


πŸ’– ARGUMENTS πŸ’–

okay this is the part where we vibe check each param one by one πŸ‘‘

  • dev β€” the device the DMA address was created for πŸ’œ like the specific one. no switching devices on this one, that's not the move.
  • addr β€” the DMA address. yes, the DMA address, that one right there ✨
  • size β€” how big the region is, in bytes. in bytes!! don't be out here guessing πŸ“
  • direction β€” the direction of the DMA 🌈 it's an enum dma_data_direction, and you're gonna tell it which way things were flowing.

πŸ¦‹ SEE ALSO πŸ¦‹

okay this list is long long, but i'm not skipping a single one because every one of these is a main character in their own right, periodt πŸ’… and heads up β€” they all live in the kernel file ./include/rdma/ib_verbs.h:

ngl bestie, that's the whole roster 😭 every single one is listed, so you can go lose yourself in the cross-references for hours. that's the assignment 🧚

🫢 the sign-off 🫢

so yeah! that's the page πŸ’– it comes from ./include/rdma, the version's dated October 2026, and the whole thing answers to the name ib_dma_unmap_single(9). go unmap responsibly, bestie β€” slay πŸ’…βœ¨πŸŒΈ

πŸ“¦ where this page comes from