18:47 <dormito> any chance anyone is familiar with code warriors (NXP/freescale) TCL scripting? I mentioned previously, but I'm trying to use some of their public docs with openocd. However it looks like they are scanning a 8 bit value into a chips 4-bit IR, and I see nothing in the script that looks (to me)like it would be controlling a jtag router or the likes.

18:49 <PaulFertser> dormito: feel free to pastebin an example

18:50 <dormito> sure, just a moment

18:51 <PaulFertser> Also, results from starting OpenOCD wiith this target for comparison.

18:53 <dormito> it's an lx2160a. which AFAIK dones't quite have opencd upstream support, but I have found an merge request for it (I could probably re-find the link). Anyhow I have what I believe is wa working custome script

18:56 <dormito> so the "authorative" souce of the code warrior TCL: https://docs.nxp.com/bundle/GUID-3FFCCD77-5220-414D-8664-09E6FB1B02C6/page/GUID-161BC9DD-CE91-4951-8858-4DDB91435A75.html under section 3 (lablex LX2160A)

18:57 <dormito> a paste bin with just the TCL from (but otherwise same as above) https://bpa.st/UIFQ

18:58 <dormito> my openocd startup logs (and cli cmd) https://bpa.st/U53Q

18:59 <PaulFertser> If I am to guess "jtag::scan_io 0 8 0x92" means push 0 into IR, then assume DR is 8 bits, push 0x92 there.

19:00 <dormito> I will note that on the webpage, it looks to me like for the ls1028A they first scan out a 4 bit value to IR, then 2 lines later, scan out an 8 bit vlaue

19:06 <PaulFertser> OK, looks like 0 means ir and 1 dr indeed.

19:06 <PaulFertser> And they're really scanning 8-bit value in ir.

19:08 <dormito> there's a thread for openocd support of the lx2160: https://sourceforge.net/p/openocd/mailman/message/37360072/

19:10 <dormito> my cfgs are some what based off that. That uses an irlen of 4, and my config scripts kept that values (and it appears me to be correct).

19:10 <PaulFertser> dormito: what happens if you try 8?

19:11 <dormito> openocd dies on startup (let me make the change and paste the logs)

19:13 <dormito> https://bpa.st/BZDQ

19:17 <PaulFertser> Hm, interesting

19:32 <dormito> I doun't suppose openocd has the ability to "overscan" IR from the TCL/telnet interfaces? it should be fairly simple to to test, I think the "jtag::scan_io 1 64 0x0" line just reads out a value (and since they regocomment is, I'm guessing it's either RO, or safe to load 0x0)

19:47 <PaulFertser> dormito: you can define two taps, one enabled one disabled, then load it and runtime enable the second tap, disable the first, and scan big IR.

19:47 <PaulFertser> dormito: but it's easier to just not define any target for testing

19:48 <dormito> can you do that? I only see the ir/dr cmds associated with a given tap?

19:49 <PaulFertser> dormito: yes, but you can have many taps defined.

19:49 <PaulFertser> dormito: see bcm47xx.cfg for example

19:54 <dormito> by "then load it" do you mean "start openocd the scripts"?

19:55 <dormito> and then via the tcl port, swap the TAPS's enables, and do the scan?

19:56 <PaulFertser> I prefer experimenting via Telnet.

19:56 <PaulFertser> The TCL RPC is for scripting.

19:57 <PaulFertser> Yes, I meant start openocd and then experiment via Telnet.

20:09 <dormito> hmmm. could questions: is there a way to listen on telnet without a working script? I created a 2nd jtag tap (disable) in a tmp script, but I can't seem to toggle the enabled/disabled status from the telenet interface.

20:10 <dormito> s/could questions/couple questions/

20:23 <dormito> hmmm, just started poking around the code (defintely don't fully grok it), but do the taps need enable/disable events defined before the correspondig op will succeed

20:40 <PaulFertser> dormito: no, you should be able to disable and enable taps interactively from telnet.

20:40 <PaulFertser> Using "jtag tapenable ..." command.

21:48 <dormito> PaulFertser: yeah, it was saying enable/disable failed. I was probably overlooking something, however my time to play with it is up for today

21:54 <PaulFertser> dormito: hm, hard to tell without seeing the logs.

21:56 <dormito> I'll post what I found tomorrow. IIRC I endable "--debug=3", and it said rum_command() was returning -4 (but that's from memory)

21:56 <PaulFertser> dormito: feel free to ping me

21:57 <dormito> ok. will do. And thanks for your help so far

21:57 <dormito> just having someone look at the codewarrior cmds and come to the same conclusion is super helpful