xe_gsc_poison_header(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… xe_gsc_poison_header(9) β€” the Kernel API Manual, bestie edition πŸŽ€

okay so real talk, this is a page from the Kernel API Manual and honestly? it's a whole vibe. we're talking xe_gsc_poison_header β€” the function that goes and poisons the MTL GSC header sitting in memory. yes, poison. dramatic? maybe. necessary? absolutely. ✨


🌸 NAME 🌸

xe_gsc_poison_header β€” poison the MTL GSC header in memory πŸ’€

like, literally that's the job. it takes a header in memory and makes it ~poisoned~. no cap.


πŸ¦‹ SYNOPSIS πŸ¦‹

void xe_gsc_poison_header (struct xe_device *xe , struct iosys_map *map
, u32 offset );

that's the whole signature, copy it exactly bestie, don't get creative with it πŸ™ˆ


✨ ARGUMENTS ✨

three of them. not two, not four. three. let's go:

  • xe β€” the Xe device πŸ–₯️ like, the actual device we're doing this to.
  • map β€” the iosys map to write to ✍️ this is where the writing happens.
  • offset β€” offset from the start of the map at which the header resides πŸ“ so like, where in the map the header is chilling.

πŸ‘‘ SEE ALSO πŸ‘‘

go look at the actual kernel file, it's right here:

Kernel file ./drivers/gpu/drm/xe/xe_gsc_submit.c

and the whole friend group of related functions:

they all live in the same orbit, iykyk 🌈


ngl the whole point is in the name β€” this one poisons the MTL GSC header in memory. don't overthink it, just let it slay 🫢

and that's the page!! πŸŽ€ source path is ./drivers/gpu/drm/xe, dated October 2026, and the function is xe_gsc_poison_header(9) β€” periodt. go poison that header bestie, you got this πŸ’–βœ¨

πŸ“¦ where this page comes from