sc_wait_for_packet_egress(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

okay bestie, welcome to the Kernel API Manual entry for sc_wait_for_packet_egress(9) β€” yes, the packet-waiting one. let’s get into it πŸ’–

✨ NAME ✨

sc_wait_for_packet_egress β€” wait for packet. that’s the whole assignment, and it ate πŸ’…

πŸ¦‹ SYNOPSIS πŸ¦‹

void sc_wait_for_packet_egress (struct send_context *sc , int pause );

πŸŽ€ ARGUMENTS πŸŽ€

  • sc β€” valid send context. gotta be valid, bestie, no fake IDs at the function door πŸ™ˆ
  • pause β€” wait for credit return. it’s giving β€œhold up, let the credits come back” ⏳

🌸 DESCRIPTION 🌸

Wait for packet egress, optionally pause for credit return. lowkey iconic behavior.

Egress halt and Context halt are not necessarily the same thing, so check for both. both. not one. both. this is a two-halt situation and you need the full set, no shortcuts πŸ”₯

πŸ’œ NOTE πŸ’œ

The context halt bit may not be set immediately. because of this, it is necessary to check the SW SFC_HALTED bit (set in the IRQ) and the HW context bit to determine if the context is halted.

so yeah β€” SW SFC_HALTED plus HW context bit, both receipts, then you know what’s up. no cap 🫢

🌈 SEE ALSO 🌈


mwah πŸ’‹ β€” this little page comes from ./drivers/infiniband/hw/hfi1, dated October 2026. stay hydrated and check both halt bits, bestie πŸ§šπŸ’–

πŸ“¦ where this page comes from