Unreliable debug launching

Another failure mode that I'm getting 100% of the time at the moment, is that I don't even get an error dialog & end up with all processes self-terminating with 

Open On-Chip Debugger (Analog Devices 0.12.0-1.2.0)  OpenOCD 0.12.0 (2025-04-03-09:09)
Licensed under GNU GPL v2
Report bugs to <processor.tools.support@analog.com>
0
adapter speed: 5000 kHz
Info : auto-selecting first available session transport "jtag". To override use 'transport select <transport>'.
Info : xtensa permissive mode is disabled
adapter speed: 5000 kHz
1
Info : Listening on port 6666 for tcl connections
Info : Listening on port 4444 for telnet connections
Info : ICE-1000 firmware version is 1.0.2
Info : clock speed 5000 kHz
Info : JTAG tap: adspsc83x.adjc tap/device found: 0x0282c0cb (mfg: 0x065 (Analog Devices), part: 0x282c, ver: 0x0)
Info : JTAG tap: adspsc83x.cpu enabled
Info : [adspsc83x.apb] Examination succeed
Info : [adspsc83x.axi] Examination succeed
Info : [adspsc83x.ahb] Examination succeed
Info : [adspsc83x.sharcfx] Examination succeed
Info : [adspsc83x.m33] Cortex-M33 r1p0 processor detected
Info : [adspsc83x.m33] target has 8 breakpoints, 4 watchpoints
Info : [adspsc83x.m33] Examination succeed
Info : gdb port disabled
Info : gdb port disabled
Info : gdb port disabled
Info : starting gdb server for adspsc83x.sharcfx on 3333
Info : Listening on port 3333 for gdb connections
Info : starting gdb server for adspsc83x.m33 on 3334
Info : Listening on port 3334 for gdb connections

Hi,

We are looking into this and will get back to you by early next week.

Thanks!

Regards,
Santhakumari.V

I've tried using my ICE-2000 this morning but it does more or less the same : the first launch works but the second doesn't. I've also tried removing the SOM from the carrier board & powering over USB. In that case, on the second launch I get 

Open On-Chip Debugger (Analog Devices 0.12.0-1.2.0)  OpenOCD 0.12.0 (2025-04-03-09:09)
Licensed under GNU GPL v2
Report bugs to <processor.tools.support@analog.com>
0
adapter speed: 9000 kHz
Info : auto-selecting first available session transport "jtag". To override use 'transport select <transport>'.
Info : xtensa permissive mode is disabled
adapter speed: 9000 kHz
Info : Listening on port 6666 for tcl connections
Info : Listening on port 4444 for telnet connections
Info : ICE-2000 firmware version is 1.0.2
Info : ICE-2000 voltage 3.3V
Info : ICE-2000 delay 0
Info : clock speed 9000 kHz
Info : JTAG tap: adspsc83x.adjc tap/device found: 0x0282c0cb (mfg: 0x065 (Analog Devices), part: 0x282c, ver: 0x0)
Info : JTAG tap: adspsc83x.cpu enabled
Info : [adspsc83x.apb] Examination succeed
Info : [adspsc83x.axi] Examination succeed
Info : [adspsc83x.ahb] Examination succeed
Info : [adspsc83x.sharcfx] Examination succeed
Info : [adspsc83x.m33] Cortex-M33 r1p0 processor detected
Info : [adspsc83x.m33] target has 8 breakpoints, 4 watchpoints
Info : [adspsc83x.m33] Examination succeed
Info : gdb port disabled
Info : gdb port disabled
Info : gdb port disabled
Info : starting gdb server for adspsc83x.sharcfx on 3333
Info : Listening on port 3333 for gdb connections
Info : starting gdb server for adspsc83x.m33 on 3334
Info : Listening on port 3334 for gdb connections
Info : accepting 'gdb' connection on tcp/3333
Info : [adspsc83x.sharcfx] Debug controller was reset.
Info : [adspsc83x.sharcfx] Core was reset.
Info : [adspsc83x.m33] external reset detected
system reset complete
Info : New GDB Connection: 1, Target adspsc83x.sharcfx, state: halted
Info : File-I/O: syscall unknown (0), pc=0x20002738
Info : dropped 'gdb' connection
Info : accepting 'telnet' connection on tcp/4444
Info : dropped 'telnet' connection

I've established that with the ICE-2000 at least, I don't actually need to power cycle the board & ICE - just quitting CCES and re-starting is enough to make the subsequent debug session work. After ending that session, I have to quit CCES again otherwise the second gdb port never gets picked up. 

Hi Rob,

A couple things that I am hoping you can try if you don't mind because I was not able to reproduce this. I have been able to connect and disconnect multiple times with no issues. Does this happen with even a very basic application? Does changing the boot mode switch (white rotary switch) to 0 on the SOM change the behavior at all?

Regards,

Chad

I already have the rotary boot switch set to 0. Yeah, my colleague doesn't experience this so it's something specific to my PC I think?

I've just setup a launch group with default settings for SC835W with default preload, and the PerformanceTips_Core1 example loading into Core1 & my Core2 application (I can't find any examples for core2) loading on core 2 and it does the same. 

Note : I don't see the same thing when launching only Core 1, I can launch that multiple times. It's the launch group (loading core1 &2) which fails - I see core 1 briefly load exit.S but then if doesn't hit main & comes up with the timeout dialog & leaves the ICE in a "bad" state, ie LED stuck on. 

First launch : 

Info : starting gdb server for adspsc83x.sharcfx on 3333
Info : Listening on port 3333 for gdb connections
Info : starting gdb server for adspsc83x.m33 on 3334
Info : Listening on port 3334 for gdb connections
Info : accepting 'gdb' connection on tcp/3333
Info : New GDB Connection: 1, Target adspsc83x.sharcfx, state: halted
Info : accepting 'telnet' connection on tcp/4444
Info : dropped 'telnet' connection
Info : accepting 'gdb' connection on tcp/3334
[adspsc83x.m33] halted due to debug-request, current mode: Thread
xPSR: 0x69000000 pc: 0x000074f8 msp: 0x201ff3bc
Info : New GDB Connection: 2, Target adspsc83x.m33, state: halted

Second Launch:
Info : starting gdb server for adspsc83x.sharcfx on 3333
Info : Listening on port 3333 for gdb connections
Info : starting gdb server for adspsc83x.m33 on 3334
Info : Listening on port 3334 for gdb connections
Info : accepting 'gdb' connection on tcp/3333
Info : [adspsc83x.sharcfx] Debug controller was reset.
Info : [adspsc83x.sharcfx] Core was reset.
Info : [adspsc83x.m33] external reset detected
system reset complete
Info : New GDB Connection: 1, Target adspsc83x.sharcfx, state: halted
Info : dropped 'gdb' connection
Info : accepting 'telnet' connection on tcp/4444
Info : dropped 'telnet' connection

Hi Rob,

I can't tell for sure but it appears there is a system reset happening here when there should not be. System Reset should only be selected for preload and neither of the secondary connections. Those should have Skip reset selected as shown below. If you have skip reset selected and a system reset is still happening after preloads then that is the issue and for some reason the IDE is not passing the correct parameters. Can you check this for me?

Thanks,

Chad

The preload is the only one which seems to have System reset enabled

Hi Rob,

As per our Internal team's suggestion, we created a ticket for this observation in private support, and the ticket ID is CS-526484-T9S9Y3

So kindly communicate via this support Ticket ID. 

Best Regards,
Santhakumari.V