**** BEGIN LOGGING AT Wed Feb 28 03:00:01 2007 Feb 28 11:44:13 hi.. anyone able to help me with OpenOCD + IXP425 + GW16012? Feb 28 11:45:05 hi hamishm Feb 28 11:45:29 hi Feb 28 11:45:35 i can, but first i'll have to get lunch Feb 28 11:45:42 be back in ~10-15 minutes Feb 28 11:45:57 no prob. thanks:) Feb 28 12:01:47 ok, i've read what you wrote last night (your morning) Feb 28 12:02:21 oh cool Feb 28 12:02:39 i have a large number of changes on top of the xscale svn, mostly merges from trunk, but also some additional debug output Feb 28 12:02:57 haven't worked on that for quite a while, and need to check if it compiles etc. Feb 28 12:03:00 hold on Feb 28 12:03:07 i've tried both ppdev and direct i/o without any difference Feb 28 12:03:47 I've verified the hardware+dongle ok using the DOS software provided by Gateworks Feb 28 12:08:32 anything I can try? Feb 28 12:11:38 I tried cranking the srst/trst delay up to 50000 and didn't notice any change.. even any noticeable delay Feb 28 12:12:18 but that's just random stuff I tried since I don't know what I'm doing ;) Feb 28 12:17:01 the check that fails for you happens before reset is even touched Feb 28 12:17:49 it's very low level communication that fails Feb 28 12:17:50 thought that might be the case given that it had no effect Feb 28 12:18:21 i'm having some problems here right now with my xscale setup, might take a bit to sort this out, hold on Feb 28 12:22:36 hamishm: do you have the board available? Feb 28 12:22:40 yup Feb 28 12:23:29 could you apply this patch: http://mmd.ath.cx/openocd/gw16012_pathmove.patch Feb 28 12:23:52 compile, run with "-f <.cfgfile> -d -l ", and send me the log Feb 28 12:25:21 either dcc, or Dominic.Rath gmx.de Feb 28 12:26:39 either your or my firewall doesn't like this: 14:01 [freenode] DCC SEND from hamishm [0.0.0.199 port 0]: ixp42_gw_debug1.log [3kB] Feb 28 12:27:05 argh possibly firewall/nat at this end Feb 28 12:27:07 I'll email it Feb 28 12:27:19 ok Feb 28 12:28:40 i'll try the DCC again just as a test, if you don't mind. fixed my settings I hope Feb 28 12:29:44 same result Feb 28 12:30:02 ok not to worry. thanks Feb 28 12:30:15 my irc client is running directly on the router, don't think the fault is at my end - i'm receiving a dcc just now from a different source Feb 28 12:30:29 got your log Feb 28 12:31:11 that 0.0.0.199 doesn't sound good. I tweaked my DCC settings here but it obviously didn't make any difference. Feb 28 12:31:32 do you have ppdev and/or lp loaded? Feb 28 12:33:06 originally I had --enable-ppdev and was using that. no luck Feb 28 12:33:16 now I am not using that, though ppdev would still be loaded Feb 28 12:34:03 could you try unloading both modules and run openocd again? Feb 28 12:34:11 currently it fails to communicate with the GW dongle at all Feb 28 12:34:34 could also be a matter of initializing the parport, but at least ppdev should have done that right Feb 28 12:34:40 ok I did that, no change Feb 28 12:35:16 i'm running it as root via sudo to get i/o access Feb 28 12:35:56 that's working fine for me Feb 28 12:36:01 shouldn't be a problem Feb 28 12:36:08 i did notice that there was no difference in the output whether I had the dongle plugged in or not Feb 28 12:37:12 yeah, looks like it doesn't talk to the parport at all Feb 28 12:38:20 hm. 0x378 seems to be correct. that's what the parport module detected (before I unloaded it), and what I used when I ran the gateworks software earlier (under dosemu) Feb 28 12:42:22 config problem or source problem do you think? Feb 28 12:42:30 try adding this http://pastebin.ca/375608 in gw16012.c:gw16012_init after DEBUG("...privileges granted"); Feb 28 12:44:43 hmm parport_port undeclared errors. should it be gw16012_port ? Feb 28 12:44:53 ah, yeah, of course Feb 28 12:45:07 copied that from parport.c Feb 28 12:45:19 and forgot that gw16012 uses a different config variable Feb 28 12:45:55 :( I don't see any change Feb 28 12:46:57 could you try again with ppdev, and send me a log from that, too? Feb 28 12:47:03 ok Feb 28 12:47:10 at least that should make clear if it's an initialization issue Feb 28 12:47:14 do I need to remove that last mod? Feb 28 12:47:18 lp? Feb 28 12:47:22 ah, modification Feb 28 12:47:23 no Feb 28 12:49:06 hmm segfault after "...open" currently Feb 28 12:49:30 Info: openocd.c:83 main(): Open On-Chip Debugger (XScale branch, 2006-11-22 14:00 CEST) Feb 28 12:49:30 Debug: jtag.c:1246 jtag_init(): Feb 28 12:49:30 Debug: gw16012.c:478 gw16012_init(): opening /dev/parport0... Feb 28 12:49:30 Debug: gw16012.c:489 gw16012_init(): ...open Feb 28 12:49:30 Segmentation fault Feb 28 12:50:15 can you run with gdb to see where it segfaults? Feb 28 12:51:08 in gw16012_init, in some assembler code in /usr/include/sys/io.h line 99 Feb 28 12:51:12 99 __asm__ __volatile__ ("outb %b0,%w1": :"a" (value), "Nd" (port)); Feb 28 12:51:49 oh I should've put that last modification inside the #ifndef PPDEV.. oops Feb 28 12:52:53 ok fixed, runs but fails as before Feb 28 12:53:24 the log is identical except it's opening /dev/parport0 instead of requesting privs for 0x378.. nothing different after that Feb 28 13:08:22 i've gotta go.. thanks for your efforts Feb 28 13:09:24 hamishm: i'll look into this once i have my xscale problems sorted out Feb 28 13:09:31 currently it's hard to tell what is going wrong Feb 28 13:09:54 tnx. if you want me to try anything you have my email still? Feb 28 13:10:03 yeah, got it Feb 28 13:10:11 tnx. ok bfn **** ENDING LOGGING AT Thu Mar 01 02:59:57 2007