__libeth_xs..._fill_buf_md(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… NAME πŸ’…

okay so bestie, gather round β€” this is __libeth_xsk_xmit_fill_buf_md, and it's literally the internal helper to prepare XSk xmit w/meta. yes, "internal" β€” she's not for public consumption, she's backstage crew only, iykyk. ✨

(this whole page lives in the Kernel API Manual, in case the header at the top felt mysterious πŸŽ€)


πŸ¦‹ SYNOPSIS πŸ¦‹

struct libeth_xdp_tx_desc __libeth_xsk_xmit_fill_buf_md (const struct
xdp_desc *xdesc , const struct libeth_xdpsq *sq , u64 priv );

that's the whole signature, no cap β€” three args in, one struct libeth_xdp_tx_desc out. πŸ’–


🌸 ARGUMENTS 🌸

  • xdesc β€” the xdp_desc that came straight out of the XSk buffer pool, bestie. that's where she's from. 🫢
  • sq β€” the XDPSQ abstraction for the queue. like, the queue's whole vibe wrapped up in one struct. ⭐
  • priv β€” the XSk Tx metadata ops. that's your u64 of metadata goodness. πŸ‘‘

πŸ”₯ DESCRIPTION πŸ”₯

so basically β€” same as __libeth_xsk_xmit_fill_buf, but with a twist. 😭

this one requests a metadata pointer and fills in the extra fields inside libeth_xdp_tx_desc so it can actually ask for the metadata offload. it's giving "same girl, upgraded fit" β€” the base behaviour is the same, she just does more. πŸ’…


🌈 RETURN 🌈

you get back an XDP Tx descriptor carrying the DMA, the metadata request bits, and all the other info you're supposed to hand off to the driver callback. she comes fully loaded, periodt. πŸ”₯

ngl, this is internal β€” it's a helper, not a public API, so treat it like that. πŸ’€

🧚 SEE ALSO 🧚

kernel file ./include/net/libeth/xsk.h plus the whole friend group:

like, the entire extended universe is right there β€” go read them, fr fr. πŸ“šπŸ’œ


that's the whole page, bestie!! πŸ’– source: ./include/net/libeth, dated October 2026 β€” and yeah, she's filed under __libeth_xs..._fill_buf_md(9) in case you need to find her again. mwah πŸŽ€βœ¨

πŸ“¦ where this page comes from