Re: diag
Brought to you by:
aeb,
bencollins
From: Clemens L. <cl...@la...> - 2012-10-23 11:45:56
|
Carl Karsten wrote: > A friend is helping me by coding up a diagnostic tool. It is meant to > test firewire hardware. We don't have a baseline to know the quality > of the hardware we are testing, so we are not sure if the results are > due to faulty hardware or faulty code or we just don't know how to > interpret the results. > > Main process: starting receiver > Main: receiver started, pid: 11109; waiting 4 seconds... > Receiver child process 11109 starting... > unexpected packet: 0x7cf2c > unexpected packet: 0x7cf2d > unexpected packet: 0x7cf2e > unexpected packet: 0x7cf2f > unexpected packet: 0x7cf30 > unexpected packet: 0x7cf31 > unexpected packet: 0x7cf32 > unexpected packet: 0x7cf33 > unexpected packet: 0x7cf34 > unexpected packet: 0x7cf35 > unexpected packet: 0x7cf36 > unexpected packet: 0x7cf37 > unexpected packet: 0x7cf38 > unexpected packet: 0x7cf39 > Main process: starting sender This looks as if there are some leftover packets from the last run. In theory, this should not be possible because received packets that are not handled by anybody get thrown away immediately. > And sometimes it gets stuck: > > Sender child process 11517 starting... > send error: 0x10 0x10 is RCODE_SEND_ERROR, and usually means that the sending queue is full. However, the test program takes care to queue less packets than that, so in this case it might indicate that the controller thinks it still has some left-over packets (which might be the same problem as the previous one). It's possible to get such errors when multiple senders are run simultaneously, but assuming that this is not the case, this looks rather like a hardware error. Please name names (i.e., the chip). Regards, Clemens |