Hi

When I compile a .for fortran program in my centos 6.7 with gfortran compiler, the error that "Too many arguments in call to **** at (1)" always appear. However, the number of arguments in the subroutine when calling is same with the number when the subroutine is defined, so I am quite confused and do not know how to deal with this problem. Could anyone help me with this?

Regards,

Share the code.

Hi

Thanks so much for your reply.

The attachment is a program by Prof. Brink from WHOI. It is freely available, so I think it is safe to post it here.

When I compile it with gfortran, the problem I described appear.

Kind regards,

Attached Files

BIGLOAD4.FOR-----BIGLOAD4_26113.txt (33.3 KB, 147 views)

Now we also need the precise error message - i.e. which function, and which line.

I think I have a hint. I tried to compile the program as well, with the same error message (the subroutine in question is DTIME).

After uncommenting DTIME from the code, the same error message is issued. The compiler doesn't complain that DTIME is missing.
After replacing the name of this subroutine with blabla, compilation proceeds.

So, your error is most probably provoked by a library function DTIME that "masks" the subroutine DTIME in your source code. How to handle this best, I can't say, but renaming DTIME certainly works.

Note that I get another error message about a missing right parenthesis; perhaps because the parenthesis is too far on the right? I faintly remember that Fortran code lines mustn't be longer than seventysomething characters. Perhaps the GNU Fortran compiler has an option for that.

2 members found this post helpful.

Dear berndbausch

Thank you so much for your kind and detailed reply. When I changed the name DTIME into BALA, it works.

Yeah, I also encounter the second error you mentioned. It seems quite unreasonable.

Many thanks,

Col limit = 72 https://web.stanford.edu/class/me200...03_basics.html for old eg F77 std.
I beleive a modern FORTRAN compiler allows longer lines, but it may be a compile option , rather than the default.

I remember it was so that the last eight positions on a punch card could be used for numbering the cards. In case you fell down the stairs with a pack of cards in your hands, there were machines that could sort them. Not so much of an issue anymore nowadays.

1 members found this post helpful.

Hi,berndbausch

I have made an alternative expression in replace, and it works well.

Many thanks for your kind reply.

Very best wishes

Hi, I'm sorry for the delayed answer. It is exciting find a person who work in the same CTW model!, I'm from Chile, I hope we can share the experience with the model (Or if you know a forum of this topic I'd appreciated that info! ). I run the CTW model too. For this questions it worked for me: "gfortran -ffixed-line-length-0 BIGLOAD4_26113.FOR -o BIGLOAD4_26113.out"
Best Regards!

I know this is from long ago, but I'd say the 'proper' solution is the Fortran EXTERNAL statement. The problem can be compiler dependent (I can reproduce the error with gfortran, but not my favorite compilers).

In the main, perhaps after

CHARACTER*15 FINP,FOT

simply insert

EXTERNAL DTIME

and for at least gfortran the problem is gone. Always a good idea, to be safe, to declare all your own functions/subroutines as EXTERNAL so the compiler knows to use your DTIME (or 'whatever') instead of the compilers DTIME (or 'whatever'). You can name your functions whatever you want because Fortran has no reserved words, but some compilers will get confused (w/o EXTERNAL).

So to the comment: "How to handle this best, I can't say, but renaming DTIME certainly works." [from berndbausch] I think you can conclude that 'best' is 'EXTERNAL', a standard Fortran statement created for this purpose, and then name your routines whatever you like.

Regards