1 post / 0 new
sasha
ABRP JS script (iternio ovms-link) and latest stable FW 3.3.006 - Duktape WDT reset / Heap exhaust (umm_malloc)

Hi everyone,

Ran into a reproducible issue with the ABRP plugin (new v2.3.0 branch) on my v3 OVMS module, and wanted to share the crash log and context in case anyone else is hitting this or looking into memory optimization.

What I did:

  1. Installed the latest ABRP plugin and added the CA certificates to /store/trustedca/.

  2. Configured my user token via the OVMS console (config set usr abrp.token ...).

  3. Triggered a single telemetry send to test the setup: script eval abrp.onetime()

What happened: Right after sending the packet, the OVMS module froze, hit a Task Watchdog Reset (WDT tasks: OVMS DukTape), and got stuck in a continuous reboot loop every ~40 seconds until I went into the WebUI and manually disabled the Duktape JS Engine.

Here is the crash log from boot:

This is reset #2 since last power cycle
  Detected boot reason: Crash (12/12)
  Reset reason: Task watchdog (6)
  Crash counters: 1 total; 0 early
  Auto init level: Full
Last crash: abort() was called on core 0
  Current task on core 0: IDLE0; 432 stack bytes free
  Current task on core 1: IDLE1; 440 stack bytes free
  Backtrace:
  0x4008dd2e 0x4008dfc9 0x4010bd00 0x4008418e
  Event: ticker.1@ovms-server-v2 43 secs
  WDT tasks: OVMS DukTape

Has anyone else seen this behavior recently on firmware 3.3.006+?

PS. Previous ABRP version was more stable, AFAIR.