[FLASH-USERS] Eso_map.h inconsistent with setup script

[Klaus Weide]

On Wed, 28 Oct 2020, Slavin, Jonathan wrote:

> I've been using Flash version 4.3 for some time and decided to try to
> transition to 4.6.2. To test it I tried to run a simulation that I just did
> successfully with 4.3. The code compiled without a problem, but crashed
> with the runtime error:
> 
> Error message is [eos_variableMap]: Eos_map.h inconsistent with setup script
> 
> I see that the Eos_map.h that was generated by setup is different between
> version 4.3 and 4.6.2, though I don't know why. Here's the diff (ver 4.3
> and ver 4.6.2):
> 
> 40c40
> < #define EOSMAP_NUM_ROLES 27
> ---
> > #define EOSMAP_NUM_ROLES 28
> 72a73,74
> >
> > #define EOSMAP_CV 28

The number of derived variables that the Eos interface knows about (in its 
'eosData' argument) has apparently increased by one between those 
versions; you should see a similar change in 'Eos.h' files.

There is (unfortunately) a corresponding change in one place in the 
'setup' code, where recognized names of such 'Eos roles' are hardwired. 
With the regular setup method, this can be found in bin/globals.py. Just 
search for one of the recognized names like "GAMC".

> True confessions: I'm still using the unsupported flmake code. I guess that
> I might have to abandon it, though it'd be nice to know what has changed
> that could lead to this problem. Any help would be appreciated.

The 'flmake' code will require a corresponding change. Again look for 
something like "GAMC"; it should be obvious then how to add support for 
"CV".

Klaus