Re: [Simulavr-devel] gdbinit for raising interrupts.

[Theodore A. Roth]

On Sun, 2 Nov 2003, Marko Panger wrote:

> Yesterday night I tried the latest CVS version with the rising irq patch 
> included and it works great.
> 
> 1) I just like to ask why the interrupt vector numbering in avr-libc and
> the one in the mega128 datasheet isn't the same.  For example timer0
> output compare interrupt is on vector nr. 10, according to the datasheet
> and if I would like to trigger this interrupt from gdb I have to rise
> irq_9.

I think that avr-libc is more concerned with the the address of the vector 
than number of the interrupt vector. Thus, the RESET vector is at address 
0x0000, etc.

> At the same time the same interrupt is defined on vector 15 avr-gcc
> (__vector_15) ? A little bit confusing or I am just looking at thins from
> a bad angle ?

No, it's most definitely confusing. I should really fix it so that irq_<NN> 
maps into __vector_<NN>. Let me think on that a bit.

> 
> Anyway I am able to trigger interrupts and I am just happy ! Thanks.
> 
> 2) Regarding the timer0 functionality I have made a simple program which 
> stays in a while(2) loop and waits timer0 output comapre function to 
> trigger my interrupt handler but nothing happens.

Can you send me your program? I have a feeling that you might be hitting the 
problem addressed by this patch:

  http://savannah.nongnu.org/patch/?func=detailpatch&patch_id=1764&group_id=997

Ted Roth