xe_gt_sriov_...fig_set_ggtt(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

✨ xe_gt_sriov_pf_config_set_ggtt ✨

Okay bestie, so this is a page from the Kernel API Manual β€” the xe_gt_sriov_pf_config_set_ggtt(9) one β€” and honestly? It's a little baddie of a function that hands out GGTT space to your VFs. Let's get into it πŸ’…


🌸 NAME 🌸

xe_gt_sriov_pf_config_set_ggtt β€” Provision VF with GGTT space.

Like, it's the function that gives a virtual function its own slice of GGTT. That's the whole vibe. πŸŽ€


⭐ SYNOPSIS ⭐

int xe_gt_sriov_pf_config_set_ggtt (struct xe_gt *gt , unsigned int vfid
, u64 size );

πŸ¦‹ ARGUMENTS πŸ¦‹

  • gt β€” the xe_gt (can't be media) πŸ™…β€β™€οΈ like, that's a hard no, it literally cannot be media, okay?
  • vfid β€” the VF identifier, aka which virtual function we're blessing with config ✨
  • size β€” requested GGTT size, the amount of GGTT you're asking for, no cap

πŸ’– DESCRIPTION πŸ’–

So here's the tea πŸ“: if vfid represents PF, then the function will change PF's spare GGTT config. That's it, that's the behavior β€” PF's spare GGTT config gets changed, periodt.

Heads up bestie 🫢 this function can only be called on PF. Not PF-adjacent, not "well technically," no β€” PF only. The vibe check is strict here 😭

πŸ”₯ RETURN πŸ”₯

0 on success or a negative error code on failure. That's the classic Unix deal β€” zero means slay, negative means it's giving problems πŸ’€


πŸ‘‘ SEE ALSO πŸ‘‘

All the siblings, the whole friend group, the entire group chat β€” every single one of them πŸ§šβ€β™€οΈ:


✿ k bye ily, go provision some GGTT ✿

This iconic little number lives in <pink>./drivers/gpu/drm/xe</pink> and is dated <cyan>October 2026</cyan> β€” same date on the tippy-top header and the footer, so it's giving consistency πŸ’œ And yes, both the top and bottom of the original read xe_gt_sriov_...fig_set_ggtt(9)`, which is just the page being extra about itself 😌🫢

πŸ“¦ where this page comes from