On Thu, 09 Feb 2012 01:51:51 +0100 Felix Fietkau <nbd@xxxxxxxxxxx> wrote: > Only .get_rate and .tx_status are affected, wireless-testing commit > e1936e9407138b483e6d1332dd944afec8131f30 adds one of the checks, and > my commit adds the other. Maybe John could merge those two to 3.3. I ran the "affected" laptop with ath9k rate control overnight with compat-wireless-2012-02-06 and both patches. As strange as it is, I still got a pair of "-1 events" less than a millisecond apart. It took 18433 seconds (about 4 hours) to reproduce. I wish I put more instrumentation into the ath9k rate control code. I absolutely didn't expect the issue to stay. I tried to find a way to reproduce the "-1 event" quickly, I even put the laptop in a microwave oven while running flood ping to the router. Too bad, nothing interesting happened. The full kernel log is attached. Also, I'm getting "ath: Failed to stop TX DMA, queues=0x001!"; it wasn't there with compat-wireless 3.3. It looks like the issue with ath9k rate control is separate and I have no standing (as lawyers would say) with regard to the already committed patches. -- Regards, Pavel Roskin
Attachment:
dmesg
Description: Binary data