Hi guys

So I was looking at some bootloader source code, with the intention to understand how it works. I think I got the basic idea, but there are a few details that I don't understand and I need some help with these.

The first part that I am not quire understand is how it remap the interrupt to a new locations, here is the source code, where can I find more info about this? I can tell what they do from the comments, but I don't understand them at all.

// *****************************************************************************
// The bootloader code does not use any interrupts.
// However, the application code may use interrupts.
// The interrupt vector on a PIC16F is located at
// address 0x0004.
// The following function will be located
// at the interrupt vector and will contain a jump to
// the new application interrupt vector

    asm ("psect  intentry,global,class=CODE,delta=2");
    asm ("pagesel " str(NEW_INTERRUPT_VECTOR));
    asm ("GOTO " str(NEW_INTERRUPT_VECTOR));

Thanks guys!

PS: using XC8 compiler.

 

hi bug,
Look thru this boot loader tutorial PDF.

E

 

Interesting read. I knew about that option but had been too lazy to look it up, as it does not apply to what I do.

It has a 2010 copyright. I wonder whether MPLabX procedures are the same? Hopefully, this comment is not too OT, but the suggested directive, "Extended Instruction Set Enable bit = Disable" seems counterproductive. I couldn't find it mentioned elsewhere in the document and wonder whether that setting only applies to the bootloader operation* and not the the main firmware?

John

*Might make some sense for firmware that is to be installed regardless of whether the chip is an enhanced version.

 

hi bug,
Look thru this boot loader tutorial PDF.

E

Thanks for the link.

I think that's how they remap the interrupt in their tutorial.

// Since the bootloader isn't going to write the normal reset vector at 0, we have
// to generate our own remapped reset vector at the address specified in the
// bootloader firmware.
#pragma code AppVector = 0x400          // must match bootloader source code definition
void AppVector(void)
{
    _asm GOTO _startup _endasm
}

// For PIC18 devices the high priority interrupt vector is normally positioned at
// address 0008h, but the bootloader resides there. Therefore, the
// bootloader's interrupt vector code is set up to branch to our code at 0408h.
#pragma code AppHighIntVector = 0x408   // must match bootloader source code definition
void AppHighIntVector(void)
{
    _asm GOTO high_isr _endasm          // branch to the high_isr() function to handle priority interrupts.
}

#pragma code AppLowIntVector = 0x418    // must match bootloader source code definition
void low_vector(void)
{
  _asm GOTO low_isr _endasm             // branch to the low_isr() function to handle priority interrupts.
}

#pragma code                            // return to the default code section

If I want to do it in XC8, I guess I can follow the instruction here:
http://microchipdeveloper.com/faq:38
And do something like this?? Am I correct?

// put code in the address 0x008 (in the bootloader)
// and jump to 0x408, high interrupt vector in the application code 
void AppHighIntVector(void) @ 0x008
{
    // asm code: goto 0x408
}

// put code in the address 0x018 (in the bootloader)
// and jump to 0x418, low interrupt vector in the application code 
void low_vector(void) @ 0x018
{
   // asm code: goto 0x418    
}

PS: edited

 

So I did some test, my follow code seem to do the same things, any comments?

#define USE_METHOD_1

#ifdef USE_METHOD_1

// stop xc8 compiler to remove un-used function
asm("GLOBAL _isr_jump");
// put code in interrupt vector
void isr_jump(void) @ 0x08 {
    // jump to interrupt code
    asm("goto _isr_test");
}

#else

// put code in interrupt vector
asm("PSECT intcode");
// jump to interrupt code
asm("goto _isr_test");

#endif