MultiGeiger icon indicating copy to clipboard operation
MultiGeiger copied to clipboard

Constantly reboot with core dump

Open f0086 opened this issue 1 year ago • 4 comments

I've tried to compile and flash the current release on my multigeiger (which I build on one of the workshops a couple of years ago) but run into the "multiple definition of `hal_init'" problem (https://github.com/ecocurious2/MultiGeiger/issues/461). After usi8ng the current HEAD of the codebase, the problem was gone, but after a successful flash of the firmware, the device constantly crashes and reboots with the following output:

Rebooting...
ets Jun  8 2016 00:22:57

rst:0xc (SW_CPU_RESET),boot:0x17 (SPI_FAST_FLASH_BOOT)
configsip: 0, SPIWP:0xee
clk_drv:0x00,q_drv:0x00,d_drv:0x00,cs0_drv:0x00,hd_drv:0x00,wp_drv:0x00
mode:DIO, clock div:2
load:0x3fff0030,len:1184
load:0x40078000,len:13192
load:0x40080400,len:3028
entry 0x400805e4
GEIGER: 1970-01-01T00:00:06 Logging initialized at level 1.
GEIGER: 1970-01-01T00:00:06 BME_Status: not found
Config version: 015
Config size: 243
Wrong config version. Applying defaults.
[iwcAll]
|-- [iwcSys]
|   |-- 'iwcThingName' with value: 'ESP32-11869500'
|   |-- 'iwcApPassword' with value: <hidden>
|   |-- [iwcWifi0]
|   |   |-- 'iwcWifiSsid' with value: ''
|   |   \-- 'iwcWifiPassword' with value: <hidden>
|   \-- 'iwcApTimeout' with value: '30'
|-- [iwcCustom]
|   |-- [misc]
|   |   |-- 'startSound' with value: 'selected'
|   |   |-- 'speakerTick' with value: ''
|   |   |-- 'ledTick' with value: 'selected'
|   |   \-- 'showDisplay' with value: 'selected'
|   |-- [transmission]
|   |   |-- 'send2Community' with value: 'selected'
|   |   |-- 'send2Madavi' with value: 'selected'
|   |   \-- 'send2ble' with value: ''
|   \-- [alarm]
|       |-- 'soundLocalAlarm' with value: ''
|       |-- 'localAlarmThreshold' with value: '0.50'
|       \-- 'localAlarmFactor' with value: '3'
\-- [hidden]
GEIGER: 1970-01-01T00:00:06 ------------------------------------------------------------------------------------------------------------------------
GEIGER: 1970-01-01T00:00:06 Simple Multi-Geiger, Version V1.17.0-dev
GEIGER: 1970-01-01T00:00:06 ------------------------------------------------------------------------------------------------------------------------
[   840][E][esp32-hal-gpio.c:102] __pinMode(): Invalid pin selected
E (815) gpio: gpio_set_level(226): GPIO output gpio_num error
AP password was not set in configuration
State changing from: 0 to 1
Setting up AP: ESP32-11869500
With default password: <hidden>
AP IP address: 192.168.4.1
AP password was not set.
WiFi SSID was not set.
Will stay in AP mode.
State changed from: 0 to 1
Guru Meditation Error: Core  1 panic'ed (LoadProhibited). Exception was unhandled.

Core  1 register dump:
PC      : 0x40085884  PS      : 0x00050031  A0      : 0x8008167e  A1      : 0x3ffc0f4c  
A2      : 0x00002000  A3      : 0x0001a000  A4      : 0x000637ff  A5      : 0x3ffc0f2c  
A6      : 0x00000008  A7      : 0x3ffbf5a4  A8      : 0x00000000  A9      : 0x00000000  
A10     : 0x000003e8  A11     : 0x003fffff  A12     : 0x80082261  A13     : 0x3ffc0f0c  
A14     : 0x3ffc9998  A15     : 0xfffffbff  SAR     : 0x00000004  EXCCAUSE: 0x0000001c  
EXCVADDR: 0x8008168a  LBEG    : 0x400853a9  LEND    : 0x400853b1  LCOUNT  : 0x00000027  


Backtrace: 0x40085881:0x3ffc0f4c |<-CORRUPTED

I think the relevant line is this one:

E (815) gpio: gpio_set_level(226): GPIO output gpio_num error

I use the Heltec HTIT-WB32 device (with the ESP32-DOWDQ6 chip) and Platform.IO. As board, there is heltec_wireless_stick set. This should be the correct identifier (see here and here)?

f0086 avatar Apr 15 '24 06:04 f0086

#314 again? Backtrace addrs need to be decoded to see where it crashed.

And I don't think it necessarily has to do with that gpio_num error. The crash is quite a bit after that.

ThomasWaldmann avatar Apr 15 '24 15:04 ThomasWaldmann

I think this is due to a new ESP32 Arduino library. It looks like they have changed the interrupt handling. If I comment out the speaker routines (setup_speaker) in the source, it runs through. Unfortunately, I have absolutely no time to investigate this further at the moment. Thomas, perhaps you can investigate this in more detail?

On 15. Apr 2024, at 17:45, TW @.***> wrote:

#314 https://github.com/ecocurious2/MultiGeiger/issues/314 again? Backtrace addrs need to be decoded to see where it crashed.

And I don't think it necessarily has to do with that gpio_num error. The crash is quite a bit after that.

— Reply to this email directly, view it on GitHub https://github.com/ecocurious2/MultiGeiger/issues/474#issuecomment-2057173812, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFSNFB2G5NPEIN4X4FGF3ETY5PYZJAVCNFSM6AAAAABGGVKBZGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDANJXGE3TGOBRGI. You are receiving this because you are subscribed to this thread.

rexfue avatar Apr 15 '24 17:04 rexfue

The speaker works (the device beeps multiple times after it reboots). How can I decode the addresses in the Backtrace? Is there something I can do to provide more information to identify the core problem?

As an alternative (at least for me): Is there some documentation on how to flash the released binary on the device?

f0086 avatar Apr 15 '24 21:04 f0086

Any suggestions how to proceed on this?

f0086 avatar May 03 '24 20:05 f0086