AVR-Chat

NEVER trust a debugger.  :)  Turn off the interrupt and see if it still
happens.  Something else may be taking you to that address and you may be
chasing a red herring.

DLC

> Hi,
>
> This is why it is so frustrating.  I just re-measured the ground
> connection and according to my ohmmeter it is indeed grounded.
>
> And using the debugger it stops on the interrupt:
>
> "interrupt (EXT_INT0) void ext_int0_isr(void)"  The ATtiny.h file
> lists EXT_INT0 as = 2 which is what the datasheet shows.
>
> BTW I'm using the CodeVisionAVR C Compiler.
>
> Thanks,
>
> Richard Cooke
>
>
>
>
> --- In AVR-Chat@yahoogroups.com, "Dennis Clark" <dlc@...> wrote:
>>
>>   It can't fire if it is grounded.  If it IS firing then one of your
>> assumptions is incorrect:
>>
>> A- You are actually attached to ground
>> B- That INT0 is actually what is firing
>>
>>   So, disable the interrupt and see if the problem goes away - If it
> does
>> then assumption A needs to be examined.  If it does not then clearly
>> assumption B needs to be examined.
>>
>> DLC
>
>>
>
>
>
>
>
> Yahoo! Groups Links
>
>
>
>


-- 
Dennis Clark
TTT Enterprises

On Dec 17, 2007 1:59 PM, Dennis Clark <dlc@frii.com> wrote:
> NEVER trust a debugger.  :)  Turn off the interrupt and see if it still
> happens.  Something else may be taking you to that address and you may be
> chasing a red herring.

Check that your int vector table isn't missing any entries, or using
RJMP when it should be using JMP, that the ISRs all end in RETI, and
handle stack properly..  You also might be getting there from a called
routine that exits with an imbalanced stack. (one more or less push,
than pop)