ESP3 board core dumps when MDash connects to it

#1

I’m trying to make my ESP32 board work with MDash (it’s actually an ESP32-WROOM-32U). This board is intended to be a commercial product so it needs to be solid. It runs well with Mongoose OS and today I started MDash testing. I’m able to connect to MDash and the board shows as online.

I wrote a very simple Mongoose app that blinks LEDs and flashed it using USB. I included this in my yml:

  # These are needed for MDash and OTA Updates
  - origin: https://github.com/mongoose-os-libs/dash
  - origin: https://github.com/mongoose-os-libs/rpc-service-ota

Sometimes when I try to manage the board I get a core dump. As soon as I click the gear in the MDash dashboard and it switches to the management page. The board then tries to reboot but it appears to be booting corrupted firmware.

I think this is related to an unsuccessful attempt to OTA update it. It looked like the firmware was accepted but then there was an error from my board to the effect that the flash could not be written.

I’m trying to get this to work properly and reliably and would appreciate suggestions, thank you all.

#2

I am seeing “watchpoint 0 triggered” when I used MDash to do an RPC call to reboot the device. Sometimes it happens when I just click the gear on the device tile (on MDash).

Here is some output from the core dump:

[Apr 10 02:39:05.429] mgos_net.c:102          WiFi STA: ready, IP 192.168.1.34, GW 192.168.1.1, DNS 192.168.1.1
[Apr 10 02:39:05.435] mgos_rpc_channel_ws:261 0x3ffbcf40 Connecting to wss://mdash.net/api/v2/rpc, SSL? 1
[Apr 10 02:39:05.642] mgos_mongoose.c:66      New heap free LWM: 224920
[Apr 10 02:39:05.862] mg_ssl_if_mbedtls.c:31  0x3ffc70a8 ciphersuite: TLS-ECDHE-RSA-WITH-AES-128-GCM-SHA256
[Apr 10 02:39:06.715] SW ECDH curve 3
[Apr 10 02:39:07.139] mgos_mongoose.c:66      New heap free LWM: 212484
[Apr 10 02:39:07.382] mgos_rpc_channel_ws.:52 0x3ffbcf40 WS handshake resp 101
[Apr 10 02:39:07.519] mg_rpc.c:293            Sys.GetInfo via WSS_out 148.251.54.236:443
[Apr 10 02:39:08.283] Guru Meditation Error: Core  1 panic'ed (Unhandled debug exception)
[Apr 10 02:39:08.288] Debug exception reason: Watchpoint 0 triggered 
[Apr 10 02:39:08.288] Core 1 register dump:
[Apr 10 02:39:08.294] PC      : 0x4000c357  PS      : 0x00060c36  A0      : 0x8017aa58  A1      : 0x3ffb5cb0  
[Apr 10 02:39:08.299] A2      : 0x3ffc8d00  A3      : 0x3ffc7954  A4      : 0x00000051  A5      : 0x3ffc8d10  
[Apr 10 02:39:08.311] A6      : 0x223a2264  A7      : 0x472e5346  A8      : 0x22746547  A9      : 0x6170222c  
[Apr 10 02:39:08.316] A10     : 0x3ffc71b8  A11     : 0x00000001  A12     : 0x3f429be7  A13     : 0x000011d0  
[Apr 10 02:39:08.322] A14     : 0x3f42a7d8  A15     : 0x3ffc7945  SAR     : 0x00000008  EXCCAUSE: 0x00000001  
[Apr 10 02:39:08.333] EXCVADDR: 0x00000000  LBEG    : 0x4000c349  LEND    : 0x4000c36b  LCOUNT  : 0x00000003  
[Apr 10 02:39:08.339] 
[Apr 10 02:39:08.339] ELF file SHA256: 9a602ebe5071ca7fbdb6cd3fcfcb6d23ac965d717286fb66520fc2eb0e13342f
[Apr 10 02:39:08.344] 
[Apr 10 02:39:08.349] Backtrace: 0x4000c357 0x4017aa55 0x4016cc17 0x4016f29c 0x40170f18 0x40170f47 0x401829f9 0x4016ad51 0x40084406 0x400845cd
[Apr 10 02:39:08.355] 
[Apr 10 02:39:08.355] 
[Apr 10 02:39:08.355] --- BEGIN CORE DUMP ---
[Apr 10 02:39:08.360] mos: catching core dump
[Apr 10 02:39:11.216] ..............
#3

I managed to resolve this. I had the “boards” library included by mistake.