custom hal component failed to compile, undeclared (first use in this function)

No need to write estop components.
The estop_latch component allows you to  build an estop chain in software using muliple instances
linuxcnc.org/docs/2.9/html/man/man9/estop_latch.9.html
There is an example here
forum.linuxcnc.org/47-hal-examples/25861-external-e-stop
I find its easiest to sketch out the estop-latch components in the chain before writing your hal file.
So if you had 1 estop button and say fault sensors for air, lube and a door switch, each fault would be connected to its own estop_latch fault-in and if any one is triggered, it breaks the chain and the machine will stop.
It would be safer to do it this way so you stay inside the Linuxcnc paradigm
 

no I want 1 master estop control, period, writing that as multiples is just to many paths for failure. it's the illusion of safety through excess complexity. in my plan, ANY component sets a estop whether single or parallel ALL estop engages everywhere. your' making a chain of failures there.

Rodw, if you see any obvious errors in loigic or known no no's here with what I have, I am getting close. I did sucessfully operate the tool changer today and I decided to add more logic and fix a few nagging issue. Also included the error states, lots of those, the fatal errors disable the tool turret and set the estop actually. I also expanded the tool table in the turret code. it does work btw. for those who might be interested in how to implement a multi tool per station lathe turret.

edit

edit

more debugging, found a bunch of race conditions, memory leaks and some problems with shared variable, I don't think that one is my fault. anyways rewrote a bunch of logic, brought the code to the middle of the road for coding standards, given a few days to test and adding logic to job, I will release this a complete for broad use.

as always constructive criticism useful

day 6, massive code cleanup, logic improvements, implemented what I think could be considered stair logic, and started creating a whole new program for running tool changers. I am working on creating functional blocks that can be assembled with a gui and assembled into a hal file and compiled with halcompile. my focus is cavemen simple type of code, short, sweet and clean. I have been carefully watching my execution times, looks VERY FAST, everything is implemented inside the component. keeps the hal file cleaner and make life simpler.

this hasn't been tested on real hardware yet, but it does pass every test my simulation machine sets and  I have made a few toolchanges with timeout problems in the last few days. so I changed some logic around and this appears to be working as intended now.

feedback appreciated, but the silence is deafening