• Locked Locked
  • Replies 3 replies
  • Subscribers 119 subscribers
  • Views 2823 views
  • Users 0 members are here
Options
Related
How was your experience today?
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Interrupt not seen by M0 core

John Mount
Note: This was originally posted on 13th October 2011 at http://forums.arm.com

I am sending an interrupt to the M0 from another processor.   In the NVIC registers I can see the interrupt pending, and I also can see that the interrupt in question is enabled. The firmware never see's the interrupt.   There must be another level of interrupt enable that I have missed.    Any suggestions?    I attached my vector table.   I am asserting general interrupt 0.   
  • Note: This was originally posted on 17th October 2011 at http://forums.arm.com

    Hi Joseph,

    I am using a Cadence Palladium hardware emulator. I verified that the PRIMASK register bit was not set. The firmware is sitting in a loop waiting for something to do.   I have breakpoints in the interrupt code.   Like I said I can see the interrupt pending in the NVIC, but the code never sees the interrupt so it is never serviced?   I have attached in the original post the vector table, "m0_exceptions.c".  I am setting the M0_GeneralInterrupt_0. I am completely baffled at this point!@#$%^

    Thanks,

    John


    Hi John,

    What type of test environment you are using?
    - microcontroller / silicon?
    - FPGA?
    - Verilog simulation?

    A few things to check:
    - value of PRIMASK (if set, it disable all interrupts)
    - value of IPSR (if it is already running another interrupt, then the current priority level might be higher or same as the interrupt).
    - what happen with the program? Does it still running, or does it hangs/crashed when the interrupt take place? Could you check if the vector table is set correctly in the compiled program image?

    regards,
    Joseph
  • Arm Employee Badge
    Note: This was originally posted on 14th October 2011 at http://forums.arm.com

    Hi John,

    What type of test environment you are using?
    - microcontroller / silicon?
    - FPGA?
    - Verilog simulation?

    A few things to check:
    - value of PRIMASK (if set, it disable all interrupts)
    - value of IPSR (if it is already running another interrupt, then the current priority level might be higher or same as the interrupt).
    - what happen with the program? Does it still running, or does it hangs/crashed when the interrupt take place? Could you check if the vector table is set correctly in the compiled program image?

    regards,
    Joseph
  • Arm Employee Badge
    Note: This was originally posted on 17th October 2011 at http://forums.arm.com

    Hi John,

    The most likely cause is that the processor has not been reset after the program memory updated.

    When the emulator starts, the program memory might be invalid and the core enters hardfault, and then lockup if another fault occurred inside hardfault handler.  After the program is updated, you might have changed the PC value and SP value, but not resetting the processor before starting the program execution.  In this case, the priority in the NVIC still think that it is in hard fault (priority level -1) and therefore does not accept interrupts.

    Also, please check the generated program image (disassembled listing / hex / binary).  I know that you have attached the vector table source code, but whether the vector table is placed correctly in the compiled image is a slightly different question.

    Which development tool chain do you use?

    Also, as you are using Cadence Palladium, I guess your company is a ARM customer and maybe you are entitle to ARM support. In such case it might be useful if you can contact ARM support directly. If needed, our FAE can visit you and might help identify issues in your system setup.

    regards,
    Joseph