Intel® Fortran Compiler
Build applications that can scale for the future with optimized code designed for Intel® Xeon® and compatible processors.

Statement functions

wwwmaniac
Beginner
800 Views
Please refer to the following code:
program TEST3
implicit none
integer I
real DUMMY_F_1
real DUMMY_F_2
DUMMY_F_1(I)=real(I)
DUMMY_F_2(I)=DUMMY_F_1(I)
write (6,*) DUMMY_F_2(1) - DUMMY_F_2(0)
end program TEST3

this code compiled with some other compilers (PGF, MipsPro-irix) gives theoutput "1.0" (expected)
compiled with IFC gives "0.0"
can anyone point to the problem? is it a compiler bug or intended behavior?possible solutions (without rewriting the code using contained functions)? any compiler flags that would fix that?
0 Kudos
4 Replies
Steven_L_Intel1
Employee
800 Views
Offhand, this just looks like a compiler bug. I doubt any switches would help. Please report it to Intel Support.
0 Kudos
wwwmaniac
Beginner
800 Views
I use a free unsupported version of IFC, so I am not eligible for Premiere Support from Intel. May be someone who has that would like to submit that to Intel and see what they say.
0 Kudos
Steven_L_Intel1
Employee
800 Views
Yes, you are eligible to use Premier Support. Simply register using the serial number that was sent to you. You can submit problem reports, which will be investigated. Support is limited in that you don't get fixes, but we do encourage you to report problems.
0 Kudos
Steven_L_Intel1
Employee
800 Views
You are eligible for "limited support" through Premier Support. You can report problems so that they will be investigated. Register using the serial number sent to you when you got the compiler.
I tried this program with a recent compiler (8.0.042) and it got the 1.0 result.
0 Kudos
Reply