Have input variables from header file be recognized as initialized variables

[Solal_A_]

Hey,

I am currently debugging my code for a runtime error (invalid pointer) but the compiler does not give me valuable information for locating this error even with -traceback. I am therefore make use of all the debug flags I found online with the hope that it may solve the problem:

FLAGS = -O2 -stand f03 -std -check all -traceback -warn all -fstack-protector -assume protect_parens -implicitnone

Using these flags, I actually get a lot more error than the one I had to begin with. One of them is about a variable name which does not have a type:

visc.f90(29): error #6404: This name does not have a type, and must have an explicit type.   [BC_UY]
        CALL FxxSBP_t (d2udy(i,:,k),u(i,:,k),h(2),nptsy, BC_UY)

The variable BC_UY is defined in a header file and loaded in the code with the #INCLUDE statement. Is there a way I can load in the variable from the header files so that the compiler initializes it correctly?

[jimdempseyatthecove]

The #include is a preprocessor statement, not a Fortran statement. Fortran also has an INCLUDE statement. These are not equivalent. Either may be used provided you understand what is happening.

When using the Fortran preprocessor the #include works like it does with the C preprocessor.

The text is inserted into the file (source or other #include) at the position of the #include. *** Any #define that reside within the #include file are defined through the remainder of the compilation unit.

This said, any Fortran variables defined within the #include file are defined within the scope of the Fortran routine where the #include was positioned.

Where are you placing your #include statement?
What are your requirements for BC_UY?
What is the scope of BC_UY? (global, local, module)
You say "so that the compiler initializes it correctly", please clarify.

Jim Dempsey

[Solal_A_]

Hi Jim,

My #INCLUDE statement is placed at the top of the file, before the MODULE is declared. I expect my BC_UY variable to be referred to only inside the module.

Here is my issue: the code I'm using (I only added bits of it, didn't write it all myself) works fine when I do not use the debug flags, but when I do, I get the error showed in my original post. I want to keep using BC_UY from the header files (load in with #INCLUDE) and not get a compile error. This error is not the one I was originally looking for, but since it is fatal, my compiler terminates after it.

Solal Amouyal.

[Lorri_M_Intel]

Fortran name scope is not the same as C/C++ name scope.

If you have included the file that declares BC_UY before the MODULE is opened, then it is NOT defined inside that module.

If you need it declared inside the module, move the #include statement to inside the module.

                       --Lorri

[jimdempseyatthecove]

Also note:

If your module contains an #include file, in which the #include-'d file contains a

#define BC_UY 12345

The macro BC_UY only lives for the duration of the compilation unit from the #define through the end of file (or #undef BC_UY)

Subsequent to compiling the module containing the #define BC_UY 12345 if you were to then use

subroutine foo
  USE YourModuleContaining_BC_UY

Then the macro BC_UY is not defined.

USE is not equivalent to #INCLUDE

Jim Dempsey