xe_migrate_resolve(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… xe_migrate_resolve(9) πŸ’…

okay so bestie, welcome to the Kernel API Manual β€” this is the page for xe_migrate_resolve(9) and it is literally the glow-up artist for your buffer objects ✨


🌸 NAME 🌸

xe_migrate_resolve β€” resolve and decompress a buffer object if required. that's it. that's the whole vibe check 🫢


✨ SYNOPSIS ✨

struct dma_fence * xe_migrate_resolve (struct xe_migrate *m , struct
xe_bo *bo , struct ttm_resource *res );

like, here's the exact signature, no cap, copy it character for character πŸ‘‘


πŸŽ€ ARGUMENTS πŸŽ€

  • m β€” the migrate context, aka the main character of this whole operation πŸ’«
  • bo β€” the buffer object to resolve, the one getting the makeover πŸ¦‹
  • res β€” the reservation object, tagging along for the ride 🌈

πŸ’– DESCRIPTION πŸ’–

so what this actually is, bestie: it's a wrapper around __xe_migrate_copy with is_vram_resolve set to true, which is the little switch that triggers decompression if needed πŸ”₯ no extra steps, no drama, it just handles it.

it's giving "delegation queen" honestly β€” all the real work is in __xe_migrate_copy, this function just sets the flag and hands it off πŸ’… periodt.


⭐ RETURN ⭐

you get back a dma_fence that signals on completion β€” like a little "we're done here" text in the group chat 🫢

or, if things went sideways, an ERR_PTR on failure 😭 yeah. that's the bad timeline.


🌸 SEE ALSO 🌸

the whole friend group, icymi β€” kernel file ./drivers/gpu/drm/xe/xe_migrate.c struct xe_migrate(9), plus:


bestie pls note: the synopsis spans two lines the way it does up there β€” that's not me being messy, that's just how it is πŸ™ˆ keep it exact.

okay that's the whole page, you ate it up πŸ’œ this has been xe_migrate_resolve(9) from the Kernel API Manual, living in ./drivers/gpu/drm/xe, dated October 2026 β€” mwah, see you in the next one, bestie πŸŽ€πŸ¦‹πŸ’–

πŸ“¦ where this page comes from