It does not use these fields (1T1R); that's why it went unnoticed.

Yes, this is off-by-one bug. Thanks!

Ok, thanks for the description; I've just misread the code.

the NIC driver might look at this setting and pass it on to the firmware in some way if possible.

In D8262#171838, @adrian wrote:

eg, things like "cancel scan", "is scan finished" methods to query the driver may be useful here.

So, firmware-based scanning will not be implemented as an 'ieee80211_scan_fw' module?

In D8207#170435, @adrian wrote:

Thanks for this!

P.S. is there any need to call memcpy(3) in ieee80211_???_rx_params() ? As I can see, it can return a pointer to ieee80211_rx_params structure instead (NULL if operation fails)

Committed in rS306591

Looks good to me.

In D7979#165342, @adrian wrote:

Hm, don't we have a general scan timeout that should prevent a scan from hanging forever?

Committed in rS306049.

Committed in rS306049.

Works fine when 'hidessid' is not set (however, 'hidessid' bug is not related to this change).

Probably, yes (e.g., 88EU supports 64 MAC ID slots but has only 32 key slots; 12AU/21AU - 128 MAC ID slots, 64 key slots)

That was a bit problematic:

• there are no 40MHz-capable APs without DSSS/CCK support in the range;
• FreeBSD / Linux inserts ERP/11b rates into beacon even when DSSS/CCK support is not advertised.

Obsoleted by https://github.com/s3erios/rtwn.

Looks fine for me (however, I have no device to test with).

P.S. I'm not sure that this code will be called when vap == NULL ...

Sure (ic_macaddr is device's MAC and iv_myaddr may be a) a copy of it b) user-specified via 'ifconfig wlan0 create wlandev <dev> wlanaddr <addr>' c) user-specified via 'ifconfig wlan0 ether <addr>')

In addition to the above:

• switch to ieee80211_iterate_nodes();
• do not assert that node table lock is held, while calling node_age(); that's not really needed (there are no resources, which can be protected by this lock) + this fixes LOR/deadlock between ieee80211_timeout_stations() and ieee80211_set_tim() (easy to reproduce in HOSTAP mode while sending something to an STA with enabled power management).

Wrong for a bit; I'll fix this soon.

Obsolete by rS300433.

Firmware part was committed in r298594, watchdog part in r300242.

ping

Committed in r300241.

Replaced with D6228.

• Apply regdomain/country settings to the MONITOR mode.
• Clenup isocc[] array too; as a result, 'ifconfig wlan0 create wlandev rum0 regdomain FCC4' will fail because of no country.

In D6228#132422, @adrian wrote:

Ah okay. So you're assuming that if a driver has set a regdomain up, it also has setup the relevant channel list. So you're defaulting to no country/no domain as a way for net80211 to realise this isn't the case.

by default - no (vap creation and regdomain/country setup are independent).

Before this can be committed, bsdinstall(8) should be capable to select another regdomain/country.

Rebase.

P.S. About link problem: can you create a bug report for that?
For additional information, you can add
wlandebug_wlan0="state+scan+auth+assoc+crypto"
into rc.conf and attach kernel message log with the moment when it stops working.

Rebase.

Refresh.

ping

In D6030#128349, @adrian wrote:

I'll check with dragonflybsd about the endian conversion routines to make sure they have the same ones in the same place.

rsu(4) is affected too.