PF2GUC_RELAY_TO_VF(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… PF2GUC_RELAY_TO_VF πŸ’…

okay so bestie, welcome to the PF2GUC_RELAY_TO_VF(9) page β€” it lives in the Kernel API Manual and it's literally just a messenger-girl situation 🧚


πŸŽ€ the vibe (what this thing even is)

This is an H2G message, and the Physical Function (PF) driver uses it to send an embedded VF2PF Relay Message_ over to the VF. that's it, that's the whole job. she's a courier πŸ’Œ

⚠️ important, no cap: this action message must be sent over CTB as a CTB HXG Message_. like, that's the only way she travels. periodt.

πŸ¦‹ the layout (bit-by-bit, she's giving structured ✨)

first, the header dword:

  • bit 31 β€” ORIGIN = GUC_HXG_ORIGIN_HOST_
  • bits 30:28 β€” TYPE = GUC_HXG_TYPE_FAST_REQUEST
  • bits 27:16 β€” MBZ (must be zero, don't be messy πŸ™ˆ)
  • bits 15:0 β€” ACTION = XE_GUC_ACTION_PF2GUC_RELAY_TO_VF = 0x5101 πŸ‘‘

and then the payload dwords, in order:

1     31:0   VFID - target VF identifier
2     31:0   RELAY_ID - VF/PF message ID
3     31:0   RELAY_DATA1
...
n     31:0   RELAY_DATAx

so like β€” dword 1 is VFID, the target VF identifier (who's she mailing πŸ’…), dword 2 is RELAY_ID, the VF/PF message ID, and then dword 3 through dword n are RELAY_DATA1 … RELAY_DATAx, however much data you're hauling. the ... is doing so much work there, she contains multitudes 😭


🌸 SEE ALSO 🌸

go read these next, bestie, i'm obsessed with cross-references honestly 🫢


✨ mwah ✨ that's the whole page. source path is ./drivers/gpu/drm/xe/abi, and she was stamped October 2026 β€” so like, future icon behaviour honestly πŸ’œ take care of her, she's PF2GUC_RELAY_TO_VF(9) and she ate πŸ’…πŸ”₯

πŸ“¦ where this page comes from