ADC3 interrupt handler not working on STM32H743

[tl12]

I have the following problem with my STM32H743:

I want to use the ADC3 for an injected conversion, but the ADC3 interrupt does not call the 

ADC3_IRQHandler, instead the SysTick_Handler is called whenever an ADC3 interrupt happens. 

When I read the ISPR from the SysTick_Handler, I can check which interrupt is active and clear the interrupt in the ADC3 ISR register. (So technically I can get everything to work, but I don't want to handle the ADC3 interrupt in the SysTick_Handler)

I already read the NVIC registers while debugging and the ADC3 interrupt definitely happens, it just isn't handled by the right handler.

I also checked the ISR vector table in the FLASH to see, if the entry for the ADC3 interrupt holds the wrong value but it points to the ADC3_IRQHandler as it should.

I am all out of ideas where to look or what to look for anymore. 

This is what the SysTick_Handler looks like at the moment:

__irq void SysTick_Handler() {
    volatile int ipsr = __get_IPSR();

    if (ipsr == 15) {
        // do the systick thing
    } else if ((ipsr == 143) && (ADC3->ISR & ADC_ISR_JEOS)) {
        ADC3->ISR = ADC_ISR_JEOS;
        // do the ADC thing
    }
}

 Does anyone have an idea, what could cause this?

[TDK]

> ADC3->ISR = ADC_ISR_JEOS;

Put a breakpoint on this line and, when it gets hit, show the value of the SCB->ICSR register in a screenshot, along with SCB->VTOR and the relevant entries in there for the 15th and 143rd interrupt addresses.

[liaifat85]

Use a debugger to step through the code and inspect the exact state when the SysTick_Handler is invoked. Check the state of all registers, particularly IPSR and ICSR, to see what might be causing the jump to the SysTick_Handler instead of ADC3_IRQHandler.

[tl12]

The active exeption at this breakpoint is the ADC3 interrupt and the VTOR value seems to align with the observed vector table offset:

The vector table entries for SysTick_Handler and ADC3_IRQHandler are highlighted.

[TDK · Accepted solution]

143 - 15 = 128. So it's off by 128, I'm sure that's not a coincidence.

I would say that's an issue with SCB->VTOR not being set to a multiple of the vector table length.

If my interpretation of your screenshot is correct, you have SCB->VTOR = 0x08020200. Perhaps this isn't valid and it needs to be a multiple of 0x400. I'm not going to dig into documentation but that seems probable given the symptoms. Try 0x08020000 or 0x08020400.

[tl12]

0x08020400 worked, thanks! :)