(Cross-posted from http://www.intellinuxwireless.org/bugzilla/show_bug.cgi?id=1879) Aireplay-ng's "fakeauth" feature simulates an authentication/association sequence by injecting and sniffing frames on a monitor interface. This works perfectly for 3945ABG and IWL5xxx cards (as well as non-Intel cards with mac80211-based drivers), but fails with a SYSASSERT Microcode SW error on 4965. The error specifically happens after the association request frame is transmitted, the authentication sequence is completed successfully. I am not sure whether the error occurs upon TX of the association request or upon subsequent RX of the association response. However, judging from the information I have so far, the most probable sequence of events is the following: Inject authentication request: success Receive ACK from AP: success Receive authentication response: success Send ACK to AP: success Inject association request: success Receive ACK from AP: success Receive association response: SYSASSERT Send ACK to AP: ucode not ready Steps to reproduce: 1. Load iwlagn with at least debug=0x43fff (to see the complete error dumps). 2. Create at least one monitor interface on the affected PHY (the presence of additional interfaces does not seem to affect the outcome). I'll refer to this interface as mon0. 3. Have a WEP-encrypted AP with no MAC filtering and no shared-key auth ready and in range. Set the card to the same channel as this AP. 4. Run the command "aireplay-ng -1 0 -e <ESSID of AP> -a <BSSID of AP> mon0" (replace mon0 with your monitor interface). Aireplay-ng will retry the auth/assoc sequence a few times, and print a failure. Dmesg shows that a SYSASSERT has occurred. http://pastebin.com/m4c0e5260 is a dmesg dump of the SYSASSERT. The bug occurs the same in both 2.6.28 and the latest wireless-testing, and is distro-agnostic. -- Vista: [V]iruses, [I]ntruders, [S]pyware, [T]rojans and [A]dware. :-) -- To unsubscribe from this list: send the line "unsubscribe linux-wireless" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html