12:48 <pisquo[m]> Hi everyone, I'm have to debug a CC2544 for a university project using TI's CC-debugger. On the official manual of cc-debugger the official tools listed are IAR Workbench and SmartRF, does anyone know if there's a way to use it with openocd?

12:48 <pisquo[m]> P.s. I tried searching but my google-fu failed me and sorry if it's a dumb question but I'm a beginner in this field.

12:57 <olerem> pisquo[m]: the official TI debugger will probably not work with openocd. You'll need to use some thing different. For example something ftdi based

12:59 <pisquo[m]> olerem: Ok, I'll look into it. Thank you very much!

13:30 <NishanthMenon_> olerem, xds110 works :D there is xds560 that does'nt probably work.

13:31 <olerem> NishanthMenon_: ach, cool!

13:31 <olerem> pisquo[m]: ^

13:32 <NishanthMenon_> btw, I'd like folks' opinion on this little connundrum I ahve: I am trying to get a jtag debugger that works with openOCD.. I'd like to get a level of performance i get from lauterbach (at least some level).. I have been looking at segger j-link.. but curious if there are better options

13:33 <NishanthMenon_> the SoCs i deal with from TI are some complex high bandwidth chips cortex-A based mostly..

13:34 <NishanthMenon_> yes, i have also looked at https://openocd.org/doc/html/Debug-Adapter-Hardware.html :D

13:35 <olerem> NishanthMenon_: hard to say, for example on imx6 i was able to get max speed supported by board which i worked with. It was 15-30MHz JTAG clock speed.

13:35 <olerem> The same clock rate was used on lauterbach for the same board

13:36 <olerem> i used Bus Blaster v3

13:36 <NishanthMenon_> olerem, were you able to get upto 30 on bus blaster?

13:36 <karlp> NishanthMenon_: there's a few ti debuggers that work, but there's no support for the 8051 targets that I'm aware of.

13:37 <NishanthMenon_> karlp, yeah - i am getting no where close to 30MHz with TI xds110... (i get maybe 2MHz or so).

13:37 <olerem> 30MHz is max supported by the ftdi, yes. but the board designe was not able to handle it

13:37 <karlp> I was talking about pisquo's queery, not mega trace :)

13:37 <NishanthMenon_> karlp, aah :D

13:38 <karlp> orbtrace willd be an option for mega trace, but it's still pretty early days.

13:40 <NishanthMenon_> karlp, aah.. https://github.com/orbcode

13:40 <NishanthMenon_> olerem, how fast did you get it to go? just curious..

13:43 <olerem> NishanthMenon_: 15MHz. ftdi has simple devider, there is no way to get anything in between 15 and 30MHz :)

13:43 <NishanthMenon_> olerem, anyday better than what I got :(

13:43 <NishanthMenon_> olerem, thanks..

13:44 <olerem> NishanthMenon_: what do you get?

13:44 <NishanthMenon_> paltry 2MHz :(

13:44 <olerem> not a lot :)

13:44 <NishanthMenon_> i mean, i was using xds110. i will buy a bus blaster and compare the data rates with lauterbach and see.. that should give me a better view.

13:44 <NishanthMenon_> olerem, karlp thanks for the hints..

14:03 <pisquo[m]> <karlp> "NishanthMenon_: there's a few ti..." <- Ok, thank you very much for the help

14:04 <karlp> pisquo[m]: but it _won't_ help you with that cc2544, which is an 8051

14:05 <karlp> you want one of the newer cortex-m radio socs.

14:07 <NishanthMenon_> pisquo[m], openoCD needs to support the debugger interface AND the processor (how to set a breakpoint etc).. he is indicating that even if the board has a compatible debugger, the processor is not supported.

14:37 <PaulFertser> NishanthMenon_: btw, other ft?232H-based adapters should be as fast.

14:37 <PaulFertser> As BBv3

14:38 <NishanthMenon_> PaulFertser, yeah - that is what I ended up concluding in the last hour or so.. https://www.seeedstudio.com/Bus-Blaster-v3.html

14:38 <NishanthMenon_> else i am going to the $1000+ $ range.. which I dont want to do..

14:38 <PaulFertser> NishanthMenon_: I mean there're plenty of ftdi adapters, the cheapest being TUMPA.

14:39 <NishanthMenon_> yup. thanks PaulFertser

14:39 <PaulFertser> NishanthMenon_: and in some cases (3.3 V target Vcc and certain features of hardware line matching) even a plain ft232h breakout board would work.

14:41 <NishanthMenon_> yup. trying to put out a "standard" for my team so that everyone has one.. i think i like BBv3 over all other options i have seen so far..

14:41 <karlp> things like Tumpav2 are meant to be perfectly reasonable too, and not as fragile

14:44 <NishanthMenon_> hmmm.. https://www.diygadget.com/tiao-usb-multi-protocol-adapter-jtag-spi-i2c-serial

14:45 <karlp> they even made changes to it for current revs based on feedback from openocd users :)

14:45 <NishanthMenon_> nice..

14:45 <karlp> PaulFertser: is that the right version there?

14:46 <karlp> you can see the versions here: https://www.tiaowiki.com/w/TIAO_USB_Multi_Protocol_Adapter_User%27s_Manual

14:46 <karlp> that looks like a v2 with the 6pin swd header on the corner

14:49 <NishanthMenon_> uggh.. everything is 14 week lead time.. Sigh the supply chain issues are irritating..

14:50 <NishanthMenon_> karlp, thanks.. i will send both of these to my team.. though i personally like the form factor of bbv3 - easier to packit in a little 3d printed box and swap tagconnect etc in the middle of all the other mess on the desk..

14:52 <PaulFertser> karlp: yes, v2 is with SWD resistor hack integrated, unfortunately, only 3.3 V.

14:52 <PaulFertser> NishanthMenon_: I think TUMPA should be sturdier as it uses a special drive IC rather than a CPLD.

14:53 <NishanthMenon_> aaah.. thanks..

14:53 <PaulFertser> But it has some annoying features too, iirc SRST line isn't open-drain there.

14:54 <PaulFertser> NishanthMenon_: if you need proper SWD then TUMPA is not suitable.

14:54 <PaulFertser> NishanthMenon_: have you considered trying fake jlink for evaluation purposes?

14:55 <NishanthMenon_> like mini-Edu?

14:55 <PaulFertser> There're 10$ clones from China I mean.

14:55 <PaulFertser> Or the full official EDU version for like 60-70 but you are not allowed to use it in commercial setting.

14:56 <NishanthMenon_> not really.. i mean, i am not that bad on budget.. i can afford jlink edu as well.. but i am a bit umm-umm on jlink

14:56 <PaulFertser> Why?

14:56 <PaulFertser> EDU is not suitable for commercial purposes

14:57 <PaulFertser> I mean probably ebay has fake jlink in stock and you can evaluate it fast.

14:57 <NishanthMenon_> somehow i was getting the feel from https://www.segger.com/products/debug-probes/j-link/models/model-overview/ 4MHz vs ~15ish i might get on bbv3 .. vs a different ecosystem..

14:58 <NishanthMenon_> the main reason i am looking is @ 2MHz, folks are complaining on debug speed.. and i want that off the table..

15:02 <karlp> what are they doing that debug speed is the problem?

15:02 <karlp> I mean, is there something else underlying?

15:05 <PaulFertser> karlp: I wasn't able to reproduce Werror in jimtcl, can you please recheck some day?

15:13 <NishanthMenon_> karlp, data refresh rates of memory, etc.. many guys like lauterbach as front end ide view and register views of various peripherals and data structures etc..

15:15 <karlp> PaulFertser: only on fc33 at work, will have to try at home

15:16 <karlp> NishanthMenon_: ah right, people who are expecting all that shit to be updated, allllll the time flickering across their screens.

15:16 <NishanthMenon_> karlp, yup

15:16 <karlp> I've not normally been constrained by "slow" debug, because i'm only having it present what I want, with ITM channels

15:16 <karlp> I'm not trying to have a tab in my editor open that continuously updates with live register contants and a chunk of memory :)

15:17 <karlp> I'm guessing this is rtt related as well

15:17 <karlp> gotta sell those faster dongles I guess :)

15:17 <NishanthMenon_> karlp, https://lore.kernel.org/all/20210903050550.29050-1-nm@ti.com/ was interesting one for example i debugged yesterday night.. i was'nt sure of quite a few conditions.. so i did have variable views, uart reg peripheral view + a few other watchpoints all going in parallel

15:18 <karlp> oh, don'

15:18 <karlp> t get me wrong, I get it happens

15:19 <karlp> it just also happens for people who seem to have underlying issues,

15:19 <NishanthMenon_> hehe... ofcourse..

21:47 <pi0> olerem: is there a cheap way to transmit 1-1000mhz that does not require hackrf

21:48 <pi0> i found this clone from ali https://www.aliexpress.com/item/32949698456.html