**** BEGIN LOGGING AT Wed May 20 02:59:57 2009
May 20 15:04:10 <vloewis> test
May 20 15:04:56 <vloewis> Hi. Ich könnte die Unterstützung eines openocd-Gurus gebrauchen. Bin ich hier richtig?
May 20 15:07:11 <vloewis> Ok. Ist hier jemand?
May 20 15:29:05 <mib_iq2t65> What does an openocd guru know, that you need ?
May 20 15:31:59 <vloewis> well i need some help with the setup
May 20 15:32:36 <vloewis> i am using a aduc7020, downloaded yagarto installed everything,
May 20 15:32:55 <vloewis> and the debugger doesn't work
May 20 15:33:44 <vloewis> i took me a while to figure out a cfg-file, which at least starts up without any complaints,
May 20 15:34:43 <vloewis> i can set/rest any registers and watch the leds of my demo-board behaving accordingly,
May 20 15:35:14 <vloewis> but when i try to step through code ... zip.
May 20 15:36:38 <vloewis> all examples that i've found so far contain commands,  which cause errors
May 20 15:37:31 <vloewis> now when i try to single step through code using a telnet console i get:
May 20 15:37:48 <vloewis> "should never reach this point"
May 20 15:38:46 <vloewis> "BUG: couldn't calculate PC of next instruction, current opcode was 0x4c000a0d
May 20 15:38:53 <vloewis> "etc. ..."
May 20 15:39:14 <vloewis> does anyone have any ideas what to do about it?
May 20 15:42:45 <vloewis> well just got a call. friend of mine crashed his car. 'll try again later
May 20 20:54:55 <maluta> Hi, is it possible (with openocd) to use parallel JTAG to communicate with AVR32AP7000 and flash my memory?
May 20 20:57:10 <zwelch> maluta: avr support was just recently added, but i think it might be possible
May 20 20:58:20 <maluta> zwelch, I saw a recent thread from February (https://lists.berlios.de/pipermail/openocd-development/2009-February/004707.html)
May 20 20:58:26 <zwelch> if you join the developer mailing list and post a question, i would bet that the developer that added it will provide a definitive answer
May 20 20:58:37 <maluta> zwelch, ok
May 20 22:10:22 <RavenClaw_78> hi, i try to modify a gpio bit using the memory view in eclipse to switch on a led on the LM3S8962_EVAL board. its connected via openocd r1468. But when i set the value to 1, the led flashes for parts of a second and just ends up with everything set to 0. does someone have an idea why this happens ?
May 20 22:25:40 <zwelch> RavenClaw_78: have you tried checking out the latest svn trunk?  also, you don't mention the jtag interface that you're using
May 20 22:29:43 <RavenClaw_78> zwelch: the board is connected via usb over an ftdi chip. I'm using the luminary.cfg interface config and the lm3s6965.cfg target ...its not the exact target but was the closest one and debugging and reading memory works. No, didn't tried the latest from svn
May 20 22:31:19 <zwelch> okay.  i am not experienced enough to give you a meaningful answer, so the best i can recommend would be to join the developer mailing list and post your question there with plenty of details
May 20 22:32:11 <zwelch> still, we will be curious to hear the results from the latest revisions, as there has been a number of changes to the system recently that may improve certain situations and operations
May 20 22:32:35 <zwelch> (though i admit it's still unclear when and which ones ;) )
May 20 22:35:01 <RavenClaw_78> then i'll try to build the latest, hopefully more succesfully than last try...
May 20 22:37:03 <RavenClaw_78> whats still confuses me is if i write a 0x1 then the light flashes ...also with 0x3 ...but not with 0x2 ...->led is on pin0  ...so it must get through somehow but than vanishes again
May 20 22:45:44 <mib_gj3svv> there seems to be a interrupt routine working in the background or something
May 20 22:49:40 <RavenClaw_78> mib_gj3svv: the program i debug is a simple led off, led on loop, working well. so when the program sets the value nothing happens, the light stays on. you mean the write via openocd->jtag might trigger something...hmm I dint set anything up...maybe I should look at the bootstrap code?
May 20 22:51:53 <mib_gj3svv> my standard suggestion is to try using a command window and the telnet interface, then set the momory location using mww (memory write word)  command. This rules out strange Eclipse interactions, they are known to happen :(
May 20 22:57:05 <mib_gj3svv> no I really mean that the write through openocd->jtag work, or you would se no short change, but then something changes it back 
May 20 22:58:26 <RavenClaw_78> great ...that works ...
May 20 22:59:30 <RavenClaw_78> mib_gj3svv: thanx for the tip, never used the telnet interface before ...
May 20 23:08:58 <mib_gj3svv> nice feature with cortex_m3 processors and openocd is you can write to memory and gpio registers while they are running
May 20 23:09:38 <mib_gj3svv> of course then the code on the microprocessor can change it back at any moment.
May 20 23:10:49 <RavenClaw_78> very nice ...i currently write a peripheral register view for the cortex-m3 as part of my bachelor thesis, and modify is the last part missing ...
May 20 23:11:11 <RavenClaw_78> view for eclipse ..
May 20 23:13:43 <mib_gj3svv> ... for me , I use openocd a lot but not eclipse, usually only telnet interface and sometimes raw gdb
May 20 23:14:32 <mib_gj3svv> less strange uncontrollable things happen, and also less project management 
May 20 23:16:58 <mib_gj3svv> so if you want to create an eclipse view, telnet is great to test basic interaction and separate openocd problems from eclipse problems
May 20 23:17:14 <RavenClaw_78> thats right, maybe this view will be used in future university classe ...giving students a easyer first contact with microcontrollers ...
May 20 23:17:55 <RavenClaw_78> everything is working fine so far, just the writing part needs a little "kick" ...*G*
May 20 23:18:17 <mib_gj3svv> nice to hear, good luck :)
May 20 23:18:31 <RavenClaw_78> thanx, and thanx for the talk :-)
**** ENDING LOGGING AT Thu May 21 02:59:57 2009