On Sat, Mar 05, 2016 at 07:45:54PM +0200, Jouni Malinen wrote: > > I tried with the current wireless-testing.git snapshot and with the test > case multiple test frames: Great, thanks for the detailed information. I'll try to reproduce the crash here. The non-crashing case sounds like a case of root announcement not being received by addr2 by the time it tries to send frames; IOW changing the original time.sleep(1) to something bigger may work. [I was looking for the part that scheduled the housekeeping timer or queued root announcement after setting gate mode, and haven't found it yet -- if it actually has to wait for the housekeeping timer then it would be up to a minute before the root announcement happens.] > Could you please share the wpas_mesh_gate_forwarding.hwsim0.pcapng file > from a test case run that shows the expected behavior? Sure, find the whole dump here: https://bobcopeland.com/srcs/hwsim-test-logs-2016-03-05.tar.gz Root announcements are at #156 and 173, the proxied frames are 230 and 232. If it makes a difference, I was testing this one case by itself. I'll try a full run shortly. > The crash case looks like this: > [ 8.770098] JKM:mesh_path_send_to_gates:tbl=ffff88001eb48e00 > [ 11.916288] IPv6: ADDRCONF(NETDEV_UP): wlan0: link is not ready > [ 11.931385] IPv6: ADDRCONF(NETDEV_UP): wlan1: link is not ready > [ 11.946013] IPv6: ADDRCONF(NETDEV_UP): wlan2: link is not ready > [ 11.970031] JKM:mesh_path_send_to_gates:tbl= (null) > [ 11.971126] BUG: unable to handle kernel NULL pointer dereference at 0000000000000008 > I.e., the second call to mesh_path_send_to_gates() has > sdata->u.mesh.mesh_paths NULL. Is that broken somewhere else or should > this function check for that NULL case to avoid the crash? So, it seems the test case found a bug at least. This much is new code. :\ mesh_paths is supposed to be not null for the lifetime of the mesh interface. So either there's an initialization race here, or something is clobbering that pointer. Or mpath->sdata is not valid. I'll dig into this. Thanks a lot for testing! -- Bob Copeland %% http://bobcopeland.com/ _______________________________________________ Hostap mailing list Hostap@xxxxxxxxxxxxxxxxxxx http://lists.infradead.org/mailman/listinfo/hostap
