11:06 <bq> Should I be concerned if I get a "IR capture error at bit 2, saw 0x05 not 0x...3" error after a successful tap scan?

11:06 <bq> The target is a bit of an odd one for me - it has TDO, TDI, TMS and TCK. No resets

11:06 <bq> But I'm presuming this is normal for FPGA

11:12 <bq> Hmm. Seems to go away with irlen set to 8. Where should this value come from?

11:25 <PaulFertser> bq: hi

11:25 <PaulFertser> bq: your FPGA manual should tell you the IR length, and you should use exactly that in the config.

11:26 <PaulFertser> bq: if you still get capture error with that then probably the target is not compliant and you can use different irmask or ircapture value to ignore that.

11:27 <PaulFertser> bq: you can generate an SVF file for your FPGA with vendors tools and you'll see real IR length there inside (SVF is plain text).

18:19 <diddly> Hello, it seems that the latest release of ChibiOS changed some data structures rendering the openocd -rtos chibios feature broken. Is there a proper place to file a bug, and/or reach out to the original author?

18:23 <PaulFertser> diddly: yes, you can report a bug or (even better) offer a fix to the code.

18:24 <PaulFertser> diddly: the original ChibiOS code was written many years ago, and not changed since then (other than tree-wide cleanups).

18:25 <diddly> Hmm, yea and it seems that was the authors only contribution so not going to be fresh in their mind either

18:26 <PaulFertser> diddly: another thing to try is to use "message:chibios" search on https://review.openocd.org , probably someone has already sent a patch.

18:26 <diddly> I'll see if I can get some info from the ChibiOS author as to what needs to happen and see if I can figure out a way ahead

18:27 <diddly> PaulFertser: hmm nothing on Gerrit unfortunately

18:28 <PaulFertser> diddly: well, you can just sit down and fix it. If you have some specific questions about OpenOCD rtos code I might be able to answer. First thing to do would be to run openocd with -rtos chibios with -d2 or -d3 and note what exactly fails about the chibios code.

18:29 <diddly> as I understand, when he added SMP support the stackup changed and needs some massaging. To do it right though, ideally it should detect which ChibiOS and support both formats

18:30 <diddly> I'll have a deeper look in the next couple weeks and see if I can ask more intelligent questions :)

18:34 <PaulFertser> diddly: sounds sane :)

22:16 <bq> PaulFertser, ah, cool tip about using a produced SVF to see the irlen

22:20 <PaulFertser> bq: glad you liked it :)