ESP8266 Community Forum

- Thu Jul 06, 2017 10:08 pm #68011 What does this mean?

there is no poison after the block. Expected poison address: 0x3fff0194, actual data: 0x0 0x0 0x0 0x0
block start: 3fff00dc

I get this error occasionally on the serial monitor and then the esp8266 resets.

Thanks

- Fri Jul 07, 2017 12:46 pm #68040 Poison (or have you been using Google translate)?
Or do you mean "Position"?

Assumption is the mother of all f*ckups. At least: that's what I'm assuming.

- Fri Jul 07, 2017 1:54 pm #68042 I mean exactly what I wrote, poison.
Subsequently I found some reference to poison here:

https://github.com/esp8266/Arduino/blob ... cfg.h#L129

in relation to memory management. I don't have the a clue what this means.

If anyone can explain would be greatly appreciated.

Thanks

- Sat Jul 08, 2017 8:30 am #68073 This looks like memory allocation boundary check.
This means that some code is overwriting some memory areas that are outside the allocated blocks.
For example, if a 32 bytes block is allocated but code writing into that block writes 40 bytes, it will corrupt the "poison" boundary and when free() been called or some other management functions looking at the sanity of thoses boundaries, the exception will be triggered.