hfi1_may_ecn(9)
β¨ hfi1_may_ecn(9) β¨
okay so bestie, welcome to the hfi1_may_ecn page π
this one lives in the Kernel API Manual and honestly? it's a little vibe check for your packet's congestion bits. let's get into it π«Ά
πΈ NAME πΈ
hfi1_may_ecn β check whether FECN or BECN processing should be done π
π¦ SYNOPSIS π¦
bool hfi1_may_ecn (struct hfi1_packet *pkt );
like, it returns a bool, takes one pointer, that's the whole thing, no cap β¨
π ARGUMENTS π
pktβ the packet to be evaluated π hand it the packet, it'll peer at it
π DESCRIPTION π
So what it does is: it checks whether the FECN or BECN bits in the packet's header are enabled, and that check depends on the packet type π
real talk though β this function only checks for FECN and BECN bits, that's its whole personality π additional checks are done in the slowpath (hfi1_process_ecn_slowpath) so that correct handling actually happens. so like, don't skip the slowpath and expect things to be fine, because they will not be fine π
π₯ SEE ALSO π₯
Kernel file ./drivers/infiniband/hw/hfi1/hfi.h β and here's the whole friend group, iykyk:
rcvhdrq_size(9)hfi1_get_rc_ohdr(9)hfi1_rcd_head(9)hfi1_set_rcd_head(9)hfi1_seq_incr_wrap(9)hfi1_seq_cnt(9)hfi1_set_seq_cnt(9)last_rcv_seq(9)get_hdrqentsize(9)get_hdrq_cnt(9)hfi1_is_slowpath(9)hfi1_is_fastpath(9)hfi1_set_fast(9)sc_to_vlt(9)hfi1_rcvhdrtail_kvaddr(9)hfi1_need_drop(9)
and that's the whole page bestie, go forth and check those bits π
signed, sealed, delivered with love from ./drivers/infiniband/hw/hfi1 β October 2026 π§π (that's the source path and the date, in case you were wondering, and yes i remembered them, you're welcome ππ)
π¦ where this page comes from
- nixpkgs unstable: linux-manual 7.3.1
- π girlified by deepseek-ai/DeepSeek-V4.1-Flash-1017 on : a language model's rewrite of the original, under its terms