Lahey Fortran compiler
Status: Planning
Brought to you by:
alangenh
System: Suse 7.3 Linux on Intel PIII system
Compiler: Lahey Fortran v5.6
The --dbl flag of the Lahey compiler changes ALL single
precision reals into double precision. This strange
behaviour of this compiler breaks the interfaces using
single precision reals.
In my opinion this is a weakness of the use of compiler
flags for selecting precision.
Is there an alternative for this? The use of specified
types might offer a solution. (I encountered
DOUBLE_KIND as a type in the code).
Logged In: YES
user_id=543744
I have found a similar problem on a NEC SX-5 with the flag
-ew set which essentially does the same thing. My solution
was to define a new preprocessor variable no_4byte and
surround the offending code in the same way
no_8byte_integers is used in mpp_domains.F90