xdma_fill_descs(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ¦‹ xdma_fill_descs(9) β€” the kernel API manual tea πŸ’…

okay so bestie, welcome to xdma_fill_descs(9), straight out of the Kernel API Manual. This is the page for xdma_fill_descs πŸ’–


✨ NAME ✨

xdma_fill_descs β€” fill hardware descriptors for one contiguous memory chunk. no cap, that's the whole gig 🫢

and like… plot twist: if the size is bigger than XDMA_DESC_BLEN_MAX, more than one descriptor gets used. she's generous like that ✨ literally one chunk, multiple descs, iconic behaviour.


🌸 SYNOPSIS 🌸

u32 xdma_fill_descs (struct xdma_desc *sw_desc , u64 src_addr , u64
dst_addr , u32 size , u32 filled_descs_num );

that's the signature, copy it exactly, don't be weird about it πŸŽ€


πŸ’œ ARGUMENTS πŸ’œ

  • sw_desc β€” the descriptor container. it's giving "the bag where all the descs live" πŸ‘œ
  • src_addr β€” the first value for the ->src_addr field πŸ‘€ first value, bestie, not the only one
  • dst_addr β€” the first value for the ->dst_addr field 🎯 same energy
  • size β€” size of the contiguous memory block. if this is chunky (bigger than XDMA_DESC_BLEN_MAX), she'll spread across multiple descriptors, periodt πŸ’…
  • filled_descs_num β€” index of the first descriptor to take care of in sw_desc πŸ”’ that's the starting point, main character of the index world

🌈 SEE ALSO 🌈

bestie there is SO much lore here, drink water and read the whole list πŸ“


πŸ’€ no cap, that's every single one of them. i counted. you're welcome.

okay that's the whole page, bestie 🧚 from ./drivers/dma/xilinx, dated October 2026, and this lil manual page is xdma_fill_descs(9) β€” slay, go fill some descriptors πŸ’–βœ¨πŸ‘‘

πŸ“¦ where this page comes from