Threads are dying, I cannot find the issue :(

[russian]

I am trying to reproduce this issue on three computers with three boards. On two of them were I can sit comfortably I run discovery + FT232, on the 3rd one I am using discovery with serial-over-USB. I might be wrong but it looks like all the failures were with real serial port, while the longest successful runs were with serial-over-USB.

Is there a 'serial port attack' test in the testhal bundle?

[Giovanni]

I was thinking to the same thing, maybe there is some error in the serial driver making it loop like we verified. Could you post your serial configuration structures?

Giovanni

[Giovanni]

...I forgot about this: https://sourceforge.net/p/chibios/bugs/558/

Could you try the current 2.6.8 code on the repository? /branches/stable_2.6.x

Giovanni

[russian]

here's the serial state while hanging:

Code: Select all

#define SERIAL_SPEED 115200

static SerialConfig serialConfig = { SERIAL_SPEED, 0, USART_CR2_STOP1_BITS | USART_CR2_LINEN, 0 };

Will try the new code & report

[russian]

22000 cycles/15 hours test passed. looks like this was about https://sourceforge.net/p/chibios/bugs/558/

[Giovanni]

I wish to have realized it was related to the serial port earlier, anyway, case solved

Giovanni