Hey all,

I've been investigating some weirdness in a piece of code for a DC based
product, where the front panel access gets locked out in some circumstances.

Anyway - key reading section of the code is pretty simple. there's a
cofunction that runs every 60ms, and this calls a function that reads the
key matrix. So in this case, I guess I can either have a problem with the
cofunction (it's the first cofunction in the code, out of 4, but everything
else I can see via serial port access commands runs fine) or I have a
problem with the keypress reading function, or a problem with the function
that responds to key presses..

looking through the key reading function, there's a persistent (local)
variable that is set to indicate if a particular button (not responding in
fault mode) is pressed on a button read, but nothing is done with this
information until the function runs, and doesn't detect the button pressed,
but sees the persistent variable asserted.. In that case it passes the bit
flag for that button back to the rest of the code.

Anyway. This variable is not declared static. But it is initialised with a
#GLOBAL_INIT section.

This kind of surprises me that a persistent local variable isn't declared
static, and that it could ever work... But then looking through the DC
manual, their own example of using #GLOBAL_INIT shows it being used in a
function to fill variables (a local array of longs) that aren't static
either.

SO - does anyone have much experience with using local variables of
functions in DC as persistent storage? Is it possible this could work
sometimes, but then maybe after running for a while, it might stop working?

The only other thing I can think is if things are getting into a state
where _GLOBAL_INIT() is getting called regularly somehow? The code doesn't
have much in the way of #GLOBAL_INIT declarations, and I'm pretty sure it
could run with no major problems other than what I've seen even if
_GLOBAL_INIT() was going off regularly. So I suppose that's a possibility..

Another option is that memory is getting stomped by something really bad
that I don't know about - but in this case I'd expect a lot more than just
the key access to stop working. and this particular local variable is
sitting in the memory map alongside a bunch of pretty important other
things.

Thanks,

Julian

______________________________
controlSUITE™ software. Comprehensive. Intuitive. Optimized.
Real-world software for real-time control. Details Here!


(You need to be a member of rabbit-semi -- send a blank email to rabbit-semi-subscribe@yahoogroups.com )