Why declaring a global variable causes the interrupt to be disabed?

Projects Programming
1 
/**
 * Compile options:
 * Board : Generic STM32F1 -> BluePill F103C8 OR BlackPill F103C8
 * USART-> Enabled
 * USBCDC -> Enabled
 * Optimized : Anything without LTO
 * C runtime: Newlib nano
 * Bootloader : HID 2.2
 */

/* Includes */
#include <STM32FreeRTOS.h>
#include <RCSwitch.h>

#define RX433PIN PB6

//String someName = "Ok";

/* Remote transceiver */
RCSwitch remote433 = RCSwitch();

void setup()
{
  //Debug console
  Serial.begin(9600);
  // wait for serial port to connect. Needed for native USB, on LEONARDO, MICRO, YUN, and other 32u4 based boards.
  while (!Serial);

  //config BUILTIN lED for output
  pinMode(LED_BUILTIN , OUTPUT);
  //turns it off
  digitalWrite(LED_BUILTIN, HIGH);

  remote433.enableReceive(RX433PIN);
  Serial.print("Starting...");
}

int i =0;
void loop()
{
  if(millis() - i>1000)
  {
    i = millis();
    digitalToggle(LED_BUILTIN);
  }
  if(remote433.available())
  {
    digitalToggle(LED_BUILTIN);
    Serial.print("Received ");
    Serial.print( remote433.getReceivedValue() );
    Serial.print(" / ");
    Serial.print( remote433.getReceivedBitlength() );
    Serial.print("bit ");
    Serial.print("Protocol: ");
    Serial.println( remote433.getReceivedProtocol() );

    remote433.resetAvailable();
  }
}

I'm using a bluepill STM32 and IDE 2.2.1. in the above code once i uncomment someName the interrupt for the 433 RF won't work anymore. this only happens in STM32. any ideas? (STM32FreeRTOS.h must be included for this to happen)

2

so if you don't include STM32FreeRTOS.h everything works fine?

if you allocate memory differently like

char someName[20] = "Ok";

do you have an issue?

3

Yes if i remove STM32FreeRTOS.h or use std::string the problem is solved.
I've been struggling to find out the problem for days. somehow , something causes the interrupt to not be activated. but how that's related to declaring a variable , i have no Idea!

4

Seems to point at an incompatibility between String and std:string in the context of STM32FreeRTOS.h...
that's weird... Not sure how to debug that... (besides not using the String class)

5

I had a lot of trouble debugging this. but it turns out others are also experiencing the same problem in similar cases.
It seems declaring a Class global Variable sets the basepri register to non-zero value which should normally not happen.

so for now I have to manually set it back to zero

1 Like
6

good fact findings...

7

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.