STM32Cube V1.9.0 F429 Demo bug

[st140728]

Posted on November 17, 2015 at 21:42

Hi

In the  STM32F429I-Discovery Demo STM32Cube\Repository\STM32Cube_FW_F4_V1.9.0\Projects\STM32F429I-Discovery\DemonstrationsThis demo worked the first couple of times, but then I've had this demo randomly fail when I started to work with it.I tracked it to something happening when   main.c:k_LogInit();   was called.After calling enough times, it randomly starts workingseems like EDIT in the example /EDIT  k_LogInit() is called before k_MemInit() - but it relies on the memory pool to be initialized to 0's which it maybe on powerup.The correct sequence that is working for me is  /* Initialize Joystick, Touch screen and LEDs */  k_BspInit();  /*Initialize memory pools */  k_MemInit();  k_LogInit();      /* Initialize RTC */  k_CalendarBkupInit();   #f429-discovery

[Nickname5522_O]

Posted on November 19, 2015 at 14:28

Hi neilh,

''but then I've had this demo randomly fail when I started to work with it.''What do you mean by the fail? Which module does not work? Could you explain more?-Shahrzad-

[st140728]

Posted on November 20, 2015 at 21:24

It dies. Kaput.With ST-Link I step into it and it never returns. On halting is in an undefined exception vector.

Requires a powered reset to start working.k_LogInit() relies on memory being initialized to 0 - which it is I think is cold power (depending on ram cold initialization), so let the power die and then repower.k_MemInit sets all the buffers to 0, and needs to be done before k_LogInit()Check the code out - I believe it is clear when looking at the code. It hasn't happened since I made the modifications. I use board reset to get the control back to the begining of main