Re: svn commit: r287578 - in head/usr.sbin/wpa: hostapd wpa_supplicant

2015-09-08 Thread Adrian Chadd
The main annoying thing is any migration that hostapd decides to do between 20 and 2040 and 40-only modes. Right now ath just resets the interface and drops all the packets in-flight, which is a big no-no for 11n (we have to save them and restore them.) It's additionally complicated because to do i

Re: svn commit: r287578 - in head/usr.sbin/wpa: hostapd wpa_supplicant

2015-09-08 Thread Jung-uk Kim
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 On 09/08/2015 18:44, Adrian Chadd wrote: > Wait a sec, did anyone review and test that this works out > correctly? I tested hostapd(8) with ath(4) myself and it is working flawlessly for me. I didn't get any review because I thought it wouldn't cha

Re: svn commit: r287578 - in head/usr.sbin/wpa: hostapd wpa_supplicant

2015-09-08 Thread Adrian Chadd
Wait a sec, did anyone review and test that this works out correctly? We already have working 802.11n; it's done inside net80211. It also mostly adds 20/40 coexistence and channel width changing on the fly, which is absolutely going to break absolutely everything as we don't /have/ stable support

svn commit: r287578 - in head/usr.sbin/wpa: hostapd wpa_supplicant

2015-09-08 Thread Jung-uk Kim
Author: jkim Date: Tue Sep 8 21:04:33 2015 New Revision: 287578 URL: https://svnweb.freebsd.org/changeset/base/287578 Log: Enable 802.11n support. PR: 164102 Modified: head/usr.sbin/wpa/hostapd/Makefile head/usr.sbin/wpa/wpa_supplicant/Makefile Modified: head/usr.sbin/wpa/h