On Tue, Jan 26, 2021 at 12:24:37AM +0000, James Cook wrote:
> > > 3) Use the diff below to ignore HT protection. This is cheating! But if
> > > this diff "fixes" the problem we can at least tell that the problem is
> > > related to the use of HT protection.
> >
> > I will try this next.
> >
> > > diff 96e6b9f1cf85b3aeebc30739cfe92746a5b105e0 /usr/src
> > > blob - 2a668d0bd2f3336c39a9d270fe2c2759c442f9e3
> > > file + sys/dev/pci/if_iwx.c
> > > --- sys/dev/pci/if_iwx.c
> > > +++ sys/dev/pci/if_iwx.c
> > > @@ -5270,6 +5270,7 @@ iwx_mac_ctxt_cmd(struct iwx_softc *sc, struct
> > > iwx_node
> > > memset(&cmd, 0, sizeof(cmd));
> > >
> > > iwx_mac_ctxt_cmd_common(sc, in, &cmd, action);
> > > + cmd.protection_flags = 0;
> > >
> > > if (action == IWX_FW_CTXT_ACTION_REMOVE) {
> > > return iwx_send_cmd_pdu(sc, IWX_MAC_CONTEXT_CMD, 0,
>
> With this patch, I saw strange behaviour: when I ran "ping 192.168.0.1"
> (my router) I would see output for just one ping, and then it would
> hang.
Correction: I also see the strange behaviour without the patch.
Changing WiFi channels seems to have mostly gotten rid of the logged
errors, but I'm left with this strange behaviour.
Also, I should say: if I press ctrl-C and run ping again, I again get
one ping in again before ping stops. Sometimes. Sometimes it starts
working normally, as described below.
> A few minutes later, that behaviour is gone (ping operates normally)
> but an ssh connection stops and starts.
>
> I haven't seen any of the "dumping device error log" type errors about
> iwx0 yet.
>
> I'll send you a pcap file for one of the one-shot pings separately.
>
> --
> James
--
James
