HH> I'm leaning towards the volatile specifier on variables, so the compiler
HH> doesn't assume the variable is in a register in mainline code when the
HH> interrupt routine is changing it in RAM. Also, as I recall, the interrupt
HH> was changing a multi-byte variable that was then being evaluated by the
HH> main line code. Perhaps part way through the evaluation in mainline code
HH> the interrupt changes the variable. Perhaps you can disable interrupts
HH> before the evaluation, then enable them after, or disable them, copy to a
HH> local variable, enable interrupts, then evaluate the local.
HH> Harold

I tried with no optimizations even though my C18 version expired and
it is not optimizing as it should. It produces the same effect inside
VSM. As the hours pass i think more and more Proteus has a bug inside
its emulation software. Today it was a bad day for experimenting but i
will try it for sure tommorow with real components and real PIC

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist