[MITgcm-devel] genmake2 problem

[Ed Hill]

On Thu, 2006-04-06 at 10:58 -0400, Patrick Heimbach wrote:
> 
> there's a sublte problem with genmake2.
> We now ommit #define of packages in CPP_OPTIONS.h
> which is good.
> However, a funny thing happens when you have
> a CPP flag that's very similar to a package name,
> but is not a package name.
> For example,  if you add the flag
> #define ALLOW_SEAICEFFF
> in lab_sea/code/CPP_OPTIONS.h
> genmake2 will choke
> 
> Error: In ../code/CPP_OPTIONS.h there is an illegal line: #define
> ALLOW_SEAICEFFF


Hi Patrick,

I can think of at least three ways to fix that and here they are (in my
opinion) ordered from best to worst:

 1) we all agree to use flags of the form ALLOW_${PKG_NAME}
    only for packages -- or perhaps some other pattern such 
    as ENABLE_${PKG_NAME}

 2) within genmake2, turn off the bit of code that tries to 
    dis-allow the manual setting of packages with #define 
    statements

 3) I write a bit of code that is smart enough to enumerate
    all the package names so that non-package ALLOW_* flags
    won't get flagged as being an error -- and keep the 
    now-smarter error checking code [but I think this is a 
    really silly and fragile way to do things]

So, which approach do you and others want to use?

Ed

-- 
Edward H. Hill III, PhD
office:  MIT Dept. of EAPS;  Rm 54-1424;  77 Massachusetts Ave.
             Cambridge, MA 02139-4307
emails:  eh3 at mit.edu                ed at eh3.com
URLs:    http://web.mit.edu/eh3/    http://eh3.com/
phone:   617-253-0098
fax:     617-253-4464