-fugly-logint ?

[Tobias Burnus]

Tom Roche wrote:
>>        IF (LIMIT_Sg == .TRUE.) THEN
>>            1
>> Error: Logicals at (1) must be compared with .eqv. instead of ==
> Can I fix this error without fixing the code? E.g., can one compile
> with something like -fugly-logint (which I see while googling, but not
> in current docs)?

Looking at the source code of gfortran: No, it does not seem to be 
possible, sorry.

There is actually a reason that Fortran distinguishes .EQV. from ==: 
They have a different precedence (cf. Fortran stanard). Thus, if you 
write "==" should it be higher or lower precedence than, e.g., .NOT.? 
Besides, one rarely needs to do comparisons with logicals. "IF 
(LIMIT_Sg)" is clearer and shortert than "IF (LIMIT_Sg .EQV. .TRUE.)".

> legacy atmospheric-modeling code that has previously been {built with,
> documented for} icc/ifc (except for the most-out-of-date docs, which use
> PGI). Since I'm just building, I'm striving not to edit any sources
> (just build scripts and Makefile's).

It seems to be unavoidable to fix this bug in the atmospheric-modelling 
code as gfortran (at least currently) does not allow it.

Tobias