Hello,
It seems the same issues seen on the Buffalo WHR-HP-G300N (and documented here: http://kinglee.blogspot.com/2010/07/buffalo-whr-hp-g300n-wireless-drop-on.html) is also happening on my brand new WZR-HP-AG300H. I get the same exact type of errors in my /var/log/messages:
May 9 22:32:02 BuffaloRouter user.warn kernel: [129025.450000] Resetting; Code: 01
May 9 22:32:03 BuffaloRouter user.warn kernel: [129026.320000] Resetting; Code: 01
May 9 22:32:04 BuffaloRouter user.warn kernel: [129027.180000] Resetting; Code: 01
May 9 22:32:05 BuffaloRouter user.warn kernel: [129028.050000] Resetting; Code: 01
Because this is a dualband router and has two wireless interfaces it seems that only one of them drops, the one having been idle for about 12 hours. I've had no issues
copying 300+ GB of data over wifi (which took a couple of days). But if the interface is not in use, it will eventually drop, and the SSID for that interface is no longer broadcast. My current DD-WRT version is what came out of the box (since no other updates are on Buffalo's website as of now): DD-WRT v24SP2-MULTI (12/17/10) std - build 15940
Oddly enough I've had the WZR-HP-G300NH for about a year at my previous address and I've never encountered this issue (haven't had the need to reboot the router for months, it was rock solid) but I think it was because I had a constat flow of WiFi traffic from my IP Camera, which I don't have here, and even if I did it wouldn't be on both channels).
Please, advise if you are aware of the issue, and if there is a fix in the works for it. I don't really feel comfortable upgrading to a development release of dd-wrt, I would prefer to wait for one coming from Buffalo, that is if it doesn't take way too long. At this point, I'm using the Keep Alive feature in DD-WRT (which is really more like kill and resurrect if you ask me) to schedule a reboot daily at 6 AM. This is a dirty, dirty workaround, not a fix, by any definition. I just got off the phone with customer support who directed me to the Forums instead, so they didn't seem to be aware of any issue. Let me know if you need me to provide any more information, logs, etc. I'll be happy to, as long is it leads to a fix.