LPC55S69 dev board parodically requires hard reset to program it

[plong44]

I have an LPCXpresso55S69 dev board hooked up to a Jenkins server and periodically the board gets into a state where it is not flashable with the command line (using crt_emu_cm_redlink).

crt_emu_cm_redlink reports the following error:

PS Z:\> & $MCUX_IDE_BIN/crt_emu_cm_redlink --flash-mass-load-exec "file.bin" -p LPC55S69 --load-base=0x0 --bootromstall 0x50000040 --flash-driver= -x C:/<Project Path>/Debug --flash-dir $MCUX_FLASH_DIR0 --flash-dir $MCUX_FLASH_DIR1 --flash-hashing
Ns: MCUXpresso IDE RedlinkMulti Driver v11.3 (Jan 11 2021 16:41:13 - crt_emu_cm_redlink.exe build 9)
Wc(03). No cache support.
Nc: Found chip XML file in C:/<Project Path>/Debug\LPC55S69.xml
Nc: Restarted LinkServer process (PID 11896).
Ns: redlink server client - failed to receive response (to op 26) from server after 2000ms - timed out (rc 138)
Ns: redlink server client - failed to receive response (to op 26) from server after 2000ms - timed out (rc 138)
Ns: redlink server client - failed to receive response (to op 26) from server after 2000ms - timed out (rc 138)
Nc: Reconnected to existing LinkServer process.
Ns: redlink server client - failed to receive response (to op 26) from server after 2000ms - timed out (rc 138)
Nc: Server OK but no connection to probe 0 core -2(ANY) (after 3 attempts) - Ee(42). Could not connect to core.
Ed:02: Failed on connect: Ee(42). Could not connect to core.
Et:31: No connection to chip's debug port

Any hints or help resolving this issue would be much appreciated!

[converse]

Looks like the background process is having problems - make sure you kill the redlinkserv process before starting.

[plong44]

Thanks for the help Converse. I am not sure if it an issue with the background process on the computer, I do not see redlinkserv in the process list in between flashing attempts. I also just noticed that the issue persists across a reset of the computer, in my mind that points towards something go on with the dev board.

[frank_m]

> I also just noticed that the issue persists across a reset of the computer, in my mind that points towards something go on with the dev board.

I don't think so.

I suspect an issue with the configuration and/or setup. I suppose the redlink server is a separate process, communicating via socket interface. Which means, it could be susceptible to firewall / virus scanner software.

[plong44]

That is an interesting point, I'll check the antivirus logs and whatnot for any evidence of interference. One thing that might not jive with your theory is that I am able to flash the board after I unplug it and plug it back in, if it was an antivirus/firewall issue I would think the problem would persist. Maybe that is a bad assumption though, thanks for help!

[frank_m]

Usually such issues often turn up with remote debug, were debugger and debug server run on different machines.

To be honest, I am not much into MCUxpresso debugger details, but just assume it is similar to GDB. Most Eclipse toolchains run a separate GDB server process on the same machine, comunicating via Unix domain sockets (dunno how you call that on Windows...).

Perhaps you can locate the config/init script for the debug server, and modify it for more verbose output. That might give you a clue.

[plong44]

Thanks again for your help Frank. I ended up putting a relay (controlled by the PC) on the 5v line of a USB cable to allow me to programmatically reset the board. I'll probably try removing my relay board after the next mcuxpresso/windows/etc. update.

[Alice_Yang]

How about test on another PC?