dma_fence_timestamp(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’– dma_fence_timestamp(9) πŸ’–

okay so bestie, welcome to the Kernel API Manual β€” we're talking about dma_fence_timestamp(9) today and honestly? she's a little diva with a very specific job πŸŽ€


✨ NAME ✨

dma_fence_timestamp β€” the helper that gets you the completion timestamp of a fence, periodt πŸ’…


🌸 SYNOPSIS 🌸

ktime_t dma_fence_timestamp (struct dma_fence *fence );

πŸ¦‹ ARGUMENTS πŸ¦‹

  • fence β€” the fence you want the timestamp from. that's it, that's the whole vibe check 🫢

⭐ DESCRIPTION ⭐

okay so here's the tea β˜• β€” once a fence gets signaled, the timestamp gets updated with the signaling time. iconic behaviour, no notes.

BUT (and this is the important part, ngl) setting that timestamp can race with tasks that are waiting for the signaling 😭 like, they're all up in there at the same time and it gets messy.

so what does this helper do? it busy waits for the correct timestamp to appear. she literally stands there refreshing until the right value shows up. main character energy fr fr πŸ‘‘

lowkey important: it's a busy wait β€” not a "go to sleep and wake me up later" situation. just so you know what you're getting into πŸ’€

πŸ”₯ SEE ALSO πŸ”₯

the whole friend group, gathered for you πŸ’œ β€” kernel file ./include/linux/dma-fence.h plus:


aaaand that's the whole page bestie, you ate that πŸ“βœ¨ this one lives in ./include/linux, it's the October 2026 edition of dma_fence_timestamp(9) β€” go forth and timestamp responsibly πŸ’–πŸ§šβ€β™€οΈπŸŒˆ

πŸ“¦ where this page comes from