From mboxrd@z Thu Jan 1 00:00:00 1970 From: Oleksij Rempel Date: Sat, 07 Dec 2013 18:15:29 +0100 Subject: [ath9k-devel] unwanted client isolation AR9280 @ 5GHz mode In-Reply-To: <1386345616.32924.YahooMailNeo@web125003.mail.ne1.yahoo.com> References: <1385596455.40895.YahooMailNeo@web125002.mail.ne1.yahoo.com> <52971C05.8020205@rempel-privat.de> <1385655109.11438.YahooMailNeo@web125004.mail.ne1.yahoo.com> <1386345616.32924.YahooMailNeo@web125003.mail.ne1.yahoo.com> Message-ID: <52A357B1.4030802@rempel-privat.de> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: ath9k-devel@lists.ath9k.org Am 06.12.2013 17:00, schrieb Gertjan Hofman: > Oleksij, > > I figured I should update the information in this chain. My title should have probably better been, 'No broadcasts on Atheros AR9280'. > > Two things. I put some debugging statements in /kernel/linux-3.11.0/net/mac80211/rx.c. When pinging between two clients, I can see the ARP broadcast being copied back into the buffer: > xmit_skb = skb_copy(skb, GFP_ATOMIC); > never to appear anywhere after that. > > That was with the: Atheros AR9280 Rev:2 usb 1-1: ath9k_htc: Firmware htc_7010.fw. > > The second thing is that I got a new card, a mini-pcie Atheros AR9300 Rev:3, half-mini-PCIe (sold as WPEA-128N). As far as I can tell it uses the same drivers but comes up at 5 GHz fine, no broadcast issues. > > This seems to point back at the firmware. One more frustration is that I am sure I once got broadcast - which I am unable to replicate again. Secondly, a number of times the 9280 comes up without even client-to-master ARP being answered. > > > I may well have a working system now, but I am interested in solving this issue. > If you have any time to look into this, I can send you a AR9280 USB wireless card. Right now it has lower priority, but i'm interested in solving this issue. What kind usb card do you have? Do you able to compile ath9k-htc firmware? And use UART port on this adapter? > In both cases of course, we had to patch the ath driver (regd.c) to ignore the region code issue. > > Cheers > > Gertjan -- Regards, Oleksij