prot_autoc_write_82599(9)

📦 version 7.3.1, in nixpkgs unstable

💅 prot_autoc_write_82599(9) 💅

okay so bestie, we're in the Kernel API Manual and the page is prot_autoc_write_82599(9) — yes, that is the whole name, yes we are pronouncing it with confidence, and yes this is lowkey 82599 driver lore. 🫶

✨ NAME ✨

prot_autoc_write_82599 — Hides MAC differences needed for AUTOC write. periodt. that's the actual NAME line, and honestly it's giving exactly what it says: this is the one that hides MAC differences needed for the AUTOC write. 💅

🌸 SYNOPSIS 🌸

the signature, exactly as the manual serves it:

int prot_autoc_write_82599 (struct ixgbe_hw *hw , u32 autoc , bool
locked );

it returns an int, bestie. that's the type, no cap. ✨

🫶 ARGUMENTS 🫶

  • hw — pointer to hardware structure. literally the hardware structure pointer, the main character of the argument list. 🎀
  • autoc — value to write to AUTOC. so this is the value, and AUTOC is where it goes. 💖
  • locked — a bool to indicate whether the SW/FW lock was already taken by previous prot_autoc_read_82599. it's the “did we already grab the lock?” flag, basically. 🔥

🔥 DESCRIPTION 🔥

okay so this part (82599) may need to hold a the SW/FW lock around all writes to AUTOC. yes, “a the” — that's literally how the page says it, bestie, don't come for me 💀. likewise, after a write we need to do a pipeline reset. so: lock around the AUTOC writes, then pipeline reset after the write. that's the whole description, and it's important. respect the lock, respect the reset. ✨

🦋 SEE ALSO 🦋

Kernel file: ./drivers/net/ethernet/intel/ixgbe/ixgbe_82599.c — the source file, right there.

and now the whole 82599 friend group, each one accounted for:

🎀 source/version: ./drivers/net/ethernet/intel/i... · date: October 2026 · page: prot_autoc_write_82599(9) 🎀
📦 where this page comes from