ath(4) issues?

[Steve Kargl]

Originally sent to freebsd-current list, but this may be
more appropriate here.


--- Forwarded message from Steve Kargl <***@troutmask.apl.washington.edu> ---

Date: Thu, 22 Nov 2018 10:10:34 -0800
From: Steve Kargl <***@troutmask.apl.washington.edu>
To: freebsd-***@freebsd.org
Subject: Re: ath(4) issues?
User-Agent: Mutt/1.10.1 (2018-07-13)

I have an old D-Link AirPlus G (DWL-630) pccard card
that I have used for years with FreeBSD. Recently,
I see
% dmesg | grep ath
mobile:kargl[201] dmesg | grep ath
[ath_hal] loaded
[ath_dfs] loaded
[ath_rate] loaded
[ath] loaded
ath0: <Atheros 2413> irq 19 at device 0.0 on cardbus0
ath0: AR2413 mac 7.8 RF2413 phy 4.5
ath0: 2GHz radio: 0x0000; 5GHz radio: 0x0056
ath0: ath_rate_tx_complete: ts_rate=27 ts_finaltsi=0, final_rix=0
ath0: bad series0 hwrate 0x1b, tries 1 ts_status 0x0
ath0: ath_rate_tx_complete: ts_rate=27 ts_finaltsi=0, final_rix=0
ath0: bad series0 hwrate 0x1b, tries 1 ts_status 0x0
ath0: bad series0 hwrate 0x1b, tries 2 ts_status 0x1
The "bad series0 hwrate..." message fills syslog. This message
appeared with a month or so old /usr/src and an update to top
of tree (r340736) still produces the message.
So, is this a hardware-about-to-die issue or did someone break
ath(4) with the recent changes to inet?

So, how does one rate limit error messages from the kernel?

% uptime
10:08AM up 1:16, 5 users, load averages: 0.03, 0.10, 0.08
% dmesg | grep "bad series" | wc -l
901

Note, the card appears to work in that I'm using at the moment.

----- End forwarded message -----