On Sun, Dec 17, 2006 at 08:41:31AM -0600, Jacob Yocom-Piatt wrote: > i've had this panic occur 6-8 times since i replaced wi cards, one on an > openbsd > client, the other on an openbsd AP running in hostap mode, with ral cards. it > seems to be related to my brother using his winxp laptop to pull >200 KBps > through the ral AP. > > the AP is running a snapshot from Dec 8 that was updated from from a Jul 7 > snapshot. i believe i've applied all the changes needed to keep the machine > up-to-date, but am not certain of this. > > i had to pull the ddb info by typing it myself since this machine doesn't keep > its console on com0 (it says "switching console to com0" then boots to C0), > despite the correct entries being in ttys and boot.conf. > > below is the ddb output followed by dmesg: > > kernel: integer divide fault trap, code = 0 > stopped at rt2661_setup_tx_desc+0xc5: idivl 0x1c(%ebp), %eax > ddb> trace > rt2661_setup_tx_desc(d0a2c000,d67aa070,8,0,5f4) at rt2661_setup_tx_desc+0xc5 > rt2661_tx_data(d0a2c000,d2c48700,d0af3600,0) at rt2661_tx_data+0x3c1 > rt2661_start(d0a2c030,d2c53808,6,23a4bb6) at rt2661_start+0x1b3 > ether_output(d0a2c030,d2c48a00,d077ac6c,d2bf2ec4,d0a537b8) at > ether_output+0x364 > ip_output(d2c48a00,0,d077ac68,1,0,0,44,1) at ip_output+0x98b > ip_forward(d2c48a00,0,0,50,d0a2b05c) at ip_forward+0x1a3 > ipv4_input(d2c48a00,d0a1fc80,0,d08ad000) at ipv4_input+0x258 > ipintr(58,10,10,10,d08ad000) at ipintr+0x64 > Bad frame pointer: 0xd08aee24
This sounds like the rate of 0 problem I thought had been worked around. Try using a fixed rate from the options listed in "ifconfig -m ral0" and see if you can still reproduce it.
