Comment 14 for bug 580753

Thanks. Based on the information here so far, this looks likely to be a race condition in either ath9k or mac80211 which would end up in triggering incorrect Michael MIC failure event immediately after the group key is configured (i.e., a broadcast frame being received from the AP at more or less the exact same time when the key is being set).

Quick review of the current implementation did not result in me finding any clear problems in this area. As such, I would like to be able to figure out how to reproduce this without having to use a large number of client devices and to wait for hours.. If that turns out to be too difficult, another option would be in adding some more debugging code into the kernel code to figure out what exactly is the place where this event is generated and under what conditions. Would you be willing to run such a test if needed (i.e., to patch the kernel source code and build your own kernel)?