This short application demonstrates a crash in the I2C driver when the I2C bus is unstable.
In most cases, this results in a crash similar to:
Guru Meditation Error: Core 0 panic'ed (Interrupt wdt timeout on CPU0)
Register dump:
PC : 0x40085cbf PS : 0x00060034 A0 : 0x800850cf A1 : 0x3ffb0570
0x40085cbf: vTaskExitCritical at /Users/david/esp32/esp-idf-v3.0/components/freertos/./tasks.c:4571
A2 : 0x00000001 A3 : 0x00000000 A4 : 0x00000002 A5 : 0x3ffb45b0
A6 : 0x00000003 A7 : 0x00060923 A8 : 0x3ffb4784 A9 : 0x000000ff
A10 : 0x00000000 A11 : 0x3ffb05b8 A12 : 0x00000004 A13 : 0x3ffb3450
A14 : 0x00000001 A15 : 0x00000000 SAR : 0x00000017 EXCCAUSE: 0x00000005
EXCVADDR: 0x00000000 LBEG : 0x4000c2e0 LEND : 0x4000c2f6 LCOUNT : 0xffffffff
Backtrace: 0x40085cbf:0x3ffb0570 0x400850cc:0x3ffb0590 0x40083d42:0x3ffb05b0 0x40084149:0x3ffb05e0 0x400822a1:0x3ffb0610 0x400853b1:0x00000000
0x40085cbf: vTaskExitCritical at /Users/david/esp32/esp-idf-v3.0/components/freertos/./tasks.c:4571
0x400850cc: xQueueGenericSendFromISR at /Users/david/esp32/esp-idf-v3.0/components/freertos/./queue.c:2183
0x40083d42: i2c_master_cmd_begin_static at /Users/david/esp32/esp-idf-v3.0/components/driver/./i2c.c:1027
0x40084149: i2c_isr_handler_default at /Users/david/esp32/esp-idf-v3.0/components/driver/./i2c.c:1027
0x400822a1: _xt_lowint1 at /Users/david/esp32/esp-idf-v3.0/components/freertos/./xtensa_vectors.S:1105
0x400853b1: xQueueGenericReceive at /Users/david/esp32/esp-idf-v3.0/components/freertos/./queue.c:2183
Rebooting...
You can use info line *0x40085cbf to get the correct line number. In this case, line 1072 of esp-idf-v3.0/components/freertos/./tasks.c:
configASSERT( xCoreID == tskNO_AFFINITY || xCoreID < portNUM_PROCESSORS);
However, it should be noted that the crash address seems to change each time. It is often within the I2C ISR code path, but typically not in the same place twice in a row. It moves around a lot, and might suggest memory corruption.
It is also interesting to note that many times after the initial crash, the ESP32 will reboot but then crash immediately when calling i2c_driver_install. This occurs over and over until power is removed or the reset button on the board is pressed.
A workaround for this seems to be to add a call to i2c_hw_fsm_reset(i2c_num) at the beginning of i2c_driver_install so that any previous I2C crash caused by a frozen hardware state machine can be cleared. With this workaround in place, the ESP32 does not continually crash after the initial crash, but it will still crash the first time.
Connect 20cm wires to pin 18 (SDA) and pin 19 (SCL). No external pull-ups are required.
Use make menuconfig to set appropriate serial configuration.
Compile this project against ESP-IDF (tested against branch release/v3.0, commit b6b8af498c8cfeb1b209e3035d0531564900952d).
Flash onto an ESP32 board with make flash.
Run the application and use make monitor to observe the log.
Once you see app_main: about to run i2c_master_scan start momentarily but repeatedly shorting SDA and SCL by brushing the ends of your two wires together repeatedly, to simulate a very poor condition I2C bus.
You should see output with random I2C addresses, for example:
I (5161) i2c: detected I2C address on master 0 at address 0x72
I (7011) i2c: detected I2C address on master 0 at address 0x14
I (8091) i2c: detected I2C address on master 0 at address 0x64
I (8231) i2c: detected I2C address on master 0 at address 0x07
...
After some time (up to a minute in my case), you should eventually witness a crash:
I (76531) i2c: detected I2C address on master 0 at address 0x58
I (76681) i2c: detected I2C address on master 0 at address 0x2c
I (76771) i2c: detected I2C address on master 0 at address 0x32
Guru Meditation Error: Core 0 panic'ed (Interrupt wdt timeout on CPU0)
Register dump:
PC : 0x40085cbf PS : 0x00060034 A0 : 0x800850cf A1 : 0x3ffb0570
0x40085cbf: vTaskExitCritical at /Users/david/esp32/esp-idf-v3.0/components/freertos/./tasks.c:4571
A2 : 0x00000001 A3 : 0x00000000 A4 : 0x00000002 A5 : 0x3ffb45b0
A6 : 0x00000003 A7 : 0x00060923 A8 : 0x3ffb4784 A9 : 0x000000ff
A10 : 0x00000000 A11 : 0x3ffb05b8 A12 : 0x00000004 A13 : 0x3ffb3450
A14 : 0x00000001 A15 : 0x00000000 SAR : 0x00000017 EXCCAUSE: 0x00000005
EXCVADDR: 0x00000000 LBEG : 0x4000c2e0 LEND : 0x4000c2f6 LCOUNT : 0xffffffff
Backtrace: 0x40085cbf:0x3ffb0570 0x400850cc:0x3ffb0590 0x40083d42:0x3ffb05b0 0x40084149:0x3ffb05e0 0x400822a1:0x3ffb0610 0x400853b1:0x00000000
0x40085cbf: vTaskExitCritical at /Users/david/esp32/esp-idf-v3.0/components/freertos/./tasks.c:4571
0x400850cc: xQueueGenericSendFromISR at /Users/david/esp32/esp-idf-v3.0/components/freertos/./queue.c:2183
0x40083d42: i2c_master_cmd_begin_static at /Users/david/esp32/esp-idf-v3.0/components/driver/./i2c.c:1027
0x40084149: i2c_isr_handler_default at /Users/david/esp32/esp-idf-v3.0/components/driver/./i2c.c:1027
0x400822a1: _xt_lowint1 at /Users/david/esp32/esp-idf-v3.0/components/freertos/./xtensa_vectors.S:1105
0x400853b1: xQueueGenericReceive at /Users/david/esp32/esp-idf-v3.0/components/freertos/./queue.c:2183
Rebooting...
Without the crash-loop workaround, your ESP32 will probably crash again, and again, at this point:
W (322) app_main: about to run i2c_master_init
D (2332) i2c: i2c_master_init
W (2332) i2c: about to run i2c_param_config
W (2332) i2c: about to run i2c_driver_install
Guru Meditation Error: Core 0 panic'ed (Interrupt wdt timeout on CPU0)
Register dump:
PC : 0x40084137 PS : 0x00060034 A0 : 0x400822a4 A1 : 0x3ffb05e0
0x40084137: i2c_isr_handler_default at /Users/david/esp32/esp-idf-v3.0/components/driver/./i2c.c:1027
0x400822a4: _xt_lowint1 at /Users/david/esp32/esp-idf-v3.0/components/freertos/./xtensa_vectors.S:1105
A2 : 0x3ffb3338 A3 : 0x3ff53000 A4 : 0x00000500 A5 : 0x00000000
A6 : 0x00000000 A7 : 0x3ffb5ea8 A8 : 0x3ffb3338 A9 : 0x00000003
A10 : 0x00000000 A11 : 0x00000001 A12 : 0x80087c68 A13 : 0x3ffb4560
A14 : 0x00000001 A15 : 0x3ffb377d SAR : 0x0000001f EXCCAUSE: 0x00000005
EXCVADDR: 0x00000000 LBEG : 0x4000c2e0 LEND : 0x4000c2f6 LCOUNT : 0xffffffff
Backtrace: 0x40084137:0x3ffb05e0 0x400822a1:0x3ffb0610 0x4000bfed:0x00000000
0x40084137: i2c_isr_handler_default at /Users/david/esp32/esp-idf-v3.0/components/driver/./i2c.c:1027
0x400822a1: _xt_lowint1 at /Users/david/esp32/esp-idf-v3.0/components/freertos/./xtensa_vectors.S:1105
Rebooting...
Note that this crash occurs before the code reaches the "Scan" function, so it is an earlier crash.
I have been able to reproduce this on three ESP32 boards:
- Two "DOIT" boards from AliExpress
- One Wemos LoLin32 Lite from AliExpress
The issue is much easier to reproduce with the DOIT boards, and often occurs within a minute of repeatedly shorting SDA and SCL.
The LoLin32 Lite board seems much more stable in this regard and it is more difficult to get the issue to occur. I found it effective to put a GY-2561 on the end of a 5 metre Ethernet cable (SDA and SCL paired with ground) and attach this to the ESP32's pin 18 and 19 as normal.