On Sat, 20 Sept 2025 at 15:21, Nuno Teixeira <edua...@freebsd.org> wrote:
> (...) > > btw, the revert conflict is: > > ++<<<<<<< HEAD > + /* Enable seqno offload */ > + ic->ic_flags_ext |= IEEE80211_FEXT_SEQNO_OFFLOAD; > + /* Don't send null data frames; let firmware do it */ > + ic->ic_flags_ext |= IEEE80211_FEXT_NO_NULLDATA; > ++======= > ++>>>>>>> parent of 5bf3c5586b5e (iwx: enable seqno offload) > > Yeah I would've just manually commented out the SEQNO_OFFLOAD line. :-) ok, the locking looks ok. I'm looking at the iwlwifi driver and it does seem to have situations where it /does/ assign a sequence number from the linux wireless stack, but I'm not sure whether it's used in the more recent chipsets. I'll tinker a bit, and I've asked the linux iwlwifi maintainer if the firmware needs sequence numbers from the driver for anything. -adrian
