Has anyone had any success with JTAG emulation on the pocket beagle yet? I have both Segger and NI debuggers as well as a variety of adaptors for Other demo boards from tag-connect and others, though I can’t get any of them to work, either because the pads are not aligned or because the protocol settings must differ.

I can’t find any documentation about the jtag interface on the pocket, and the documents that do exist (for the black) discuss the 20-pin header and the TI adaptor only.

If I can’t step through using JTAG, is it possible to do so with either of the USB interfaces on the pocket? There isn’t any documentation on that either, so I figured it was prudent to ask.


(Here’s a newbie answer so take that for what it’s worth).

Have only done limited JTAG debugging on the BB-X15 with CCS 7.x - when bringing up the SPL on a custom board with the same Sitara SOC as on the BB-X15.

I THINK I recall that only a hand full of the 20 pins (like 7 maybe?) actually carry data and or clock for JTAG debugging on the 20 pin connector So the subset of JTAG connections on the Pocket Beagle may be identical as on the 20 pin connector??

. But please look this up and correct me if I’m in error!!!

There’s a TI board port series on how to setup CCS to debug the Linux SPL on the BeagleBone Black. It’s old, but TI mentioned on E2E that they’re working on updating this video series.

When debugging the SPL, breakpoints couldn’t be set in CCS. This is a known problem, and I don’t know if TI has a solution yet. Alternatively in the SPL, we were able to click “run until” in the disassembly window to determine the flow of execution. Just that it was in assembly, yuck.

Haven’t tried debugging anything but the SPL using JTAG. I think there have been other discussion threads on the merits of using JTAG vs. GDB and other techniques for debugging LInux. I’ve heard that it can be useful for debugging drivers, but I THINK part of the issue maybe kernel context switching??


From looking at a picture of the JTAG connector on the pocket beagle ( http://uk.farnell.com/beagleboard/bb-pocket/pocket-beagle-board-arm-cortex/dp/2806159 ) , it looks like there are 7 JTAG connectors signals which appear to follow the TI signals as opposed to the ARM JTAG signals.

Here’s a crazy thought…

I’ve often wondered what the boot loader developers (or BeagleBone patch wizards) do when developing SPL/u-boot for a new platform like the Pocket Beagle… Do they resort to JTAG debugging to debug the SPL and u-boot for the new platform or is everything similar enough to other platforms where the console is already up-and-running, so they can spool debug messages out the console…

If the u-boot maintainers routinely use JTAG to debug SPL/u-boot, then is there another discussion forum which discusses JTAG debugging and more detailed u-boot debugging techniques??